Stott, C., Adang, O., Livingstone, A . and Schreiber, M. (2007) ‘Variability in the collective behaviour of England fans at Euro 2004: “Hooliganism”, public order policing and social change’, in European Journal of Social Psychology, 37: 75—100.

This is a test of a model of crowd behaviour developed earlier. This model (ESIM) apparently shows links between personal and social factors affecting social identity, in this case an identity that needs to hooligan behaviour on the one hand, or a peaceable fan behaviour on the other.

English football fans have long been associated with hooliganism, and this team are interested in the social psychological mechanisms that seemed to produce disorder in some circumstances and not in others. In this particular tournament, there was no fan violence in Lisbon, but there was in Albufeira. The study gathered some data by observation and semi structured interviews [this is called phenomenological analysis in the study]. [I already have some reservations about whether this data is going to test ESIM or confirm it]. ESIM stands for ‘elaborated social identity model of crowd behaviour ’ (76). The model assumes that social identity emerges as a result of intergroup relations, especially influenced by group members with legitimacy and power. Interaction can affect whole social categories, and produce prototypes, which further influence crowd participants. However, participants can also change intergroup relations. This helps us spell out particular processes [and seems capable of explaining just about anything—when things go as predicted, the group members are clearly influencing interaction, then when they don’t, other participants are resisting these group pressures]. An example is the emergence of rioting behaviour in a political protest  against the poll tax. Participants initially understood themselves to be doing nonviolent protest, but the police intervened with force, and this legitimated conflict with the police, and violence became more widespread. The police in turn escalated their interactions and a riot developed. [the whole article seems obsessed with this one policy implication, that heavy handed policing causes crowd hostility].

The model really needs to be tested to explain the non appearance of disorder [a strange idea of a test here—that one model should explain everything! I think the intention is to install this model was capable of informing any kind of policy analysis in any kind of football fans situation. It certainly looks pretty banal. Thus there was no violence by Scotland fans because ‘they understood their social relations with police and other groups as non confrontational and characterized by legitimate behaviour on all sides’ (77). What it all seems to boil down to is whether the fans C police intervention as legitimate]. The authors evidently feel that if the model can explain a move away from violence it has been validated in some way.

They seemed to have got quite a bit of support from the UK Home Office and the Police Academy of the Netherlands. They did some previous research observing occasions where different police strategies were used and various levels of disorder emerged. Overall, disorder seemed associated with policing levels that the fans perceived as disproportionate and illegitimate, usually high profile policing. In these situations, the fans engaging in conflict appeared to be more prototypical, more central to the group itself. The team were able to recommend good practice to a number of European police forces. Portugal’s main police force was included, but not their local forces.

The team used an ethnographic method, modified for crowd behaviour. They describe this as ‘semi structured observations’ (79) [the structure being provided by the model, presumably]. First they observed events, fan behaviour, fan group interactions, police deployment, and police interactions with fans. They used photos and video when possible. They also interviewed fans using a semi structured schedule. Obviously, interviews were not possible during periods of disorder. Interviews were ‘still driven by specific theoretical concerns and made relevant to the, often rapidly developing, surrounding events’ (80). There were also meetings with fans, and invitations to reply to website questionnaires..

The researchers first described the incidents that they observed, and attempted to triangulate these descriptions with data provided by interviews with the police and other observers on a parallel project. They also used photographs, video, and TV coverage. Then they analyzed fans' accounts, using a thematic analysis based on the concept of identities and how they evolved. ‘Our initial approach to the state or was therefore informed by our theoretical framework and linked to the analytical questions’ (81). [This raises all the doubts about circularity for me]. The data were sorted according to expectations, and the accounts of what happened, although fans were also invited to comment on ‘who counted as “in group” and who as “out group”’ (81). The intention was to achieve the best fit between the analysis and the data, and extracts are provided to illustrate analytic points. The team are aware that there might be a problem using participants' accounts as evidence of cognitive representations, but consider that they have produced a ‘parsimonious theoretical explanation’ (81).

[A series of descriptions of various encounters follows. For me, some of the most interesting episode seem to fall outside the model, however, and they include the curious beliefs about the English as a belligerent 'race' who will take no insults, and the strange self pitying nature of English fans who believe they are misunderstood as patriots. The accounts seem to set up the main theoretical findings pretty well—that excessive police intervention causes trouble. Some of the notes are quite odd, as in ‘a group of four males who, as a field notes record, did not appear to be English’ --how could you tell from appearances? (84). They claim that they have observed occasions where both hooligan and calm fans were present, and that the main variable appears to have been police intervention styles. In some circumstances, English fans were clearly willing and able to defuse situations and prevent confrontation.]

The analysis of the fans' views indicate that some fans found the police in Lisbon to be non confrontational and helpful. Fans saw themselves as similar to fans from other nations. Overall, there was a ‘social identity among England fans that was defined in terms of positive social relations with the [police] positive relations with the other national fan groups and differentiation from “hooligan” activity’ (88). This enabled some fans even to do self policing. Generally, they confirm resentments when faced with more aggressive policing which they see as unjustified, and described one way in which violence escalates following rumours of police brutality in Albufeira. These descriptions were linked to expectations that they would be treated as hooligans. The accounts show that some non violent fans became violent as a result. [This is talked up rather as ‘the emergence of a social identity among England fans… that was defined in terms of negative social relations with the [police] and increasing appropriateness of and identification with “hooligans”’ (90).

Overall, the authors think that their model has been shown to be useful [and their policy and recommendations to the police]. This is addressed in rather psychological terms about prototypes, [although there is little relation between these data and the psychological terms as such]. They admit thats football banning orders might also have been responsible for the lack of violence at other venues, but insists that policing is the main variable. There are some interesting policy findings, especially that fans can be persuaded to self police as long as the police are not too heavy handed. The previous study of Scottish fans showed that self policing was even prototypical.

The study concludes with some self criticisms, especially about the data. They admit that they know very little about the actual social psychological mechanisms involved, and advocate a deeper analysis. However, policy implications are the most important ones. They would seek additional quantifiable data, including on how fans labelled themselves. They also agree that it would have been nice to follow the same fans throughout the tournament, although they argue that their cross sectional sampling is an valid way to assess differences and change, largely because they claim ‘there is no evidence of or reason to assume a priori differences between the fans in these two contexts, [which] suggests that the differences emerged as a function of contextual variation’ (93). The implications for policing are confirmed. High profile tactics only sustain expectations of violence in the future, and justify an argument that English fans are seen as troublemakers. The authors flirt with more general implications for social identity and how particular social relations inside a group can confirm particular identities. At its most general,  this addresses theoretical concerns and social psychology about the causes of violence and the contextual elements.

back to key concepts