The question of how normative form changes during a riot, and thus how collective behaviour spreads to different targets and locations, has been neglected in previous research, despite its theoretical and practical importance. We begin to address this limitation through a detailed analysis of the rioting in the London borough of Haringey in 2011. A triangulated analysis of multiple sources of data (including police reports, media accounts, and videos) finds a pattern of behaviour shifting from collective attacks on police targets to looting. A thematic analysis of 41 interview accounts with participants gathered shortly after the events suggests that a shared anti‐police identity allowed local postcode rivalries to be overcome, forming the basis of empowered action not only against the police but to address more long‐standing grievances and desires. It is argued that collective psychological empowerment operated in a ‘positive feedback loop’, whereby one form of collective self‐objectification (and perceived inability of police to respond) formed the basis of further action. This analysis of the development of new targets in an empowered crowd both confirms and extends the elaborated social identity model as an explanation for conflictual intergroup dynamics.

Previous research has shown that riots spread across multiple locations, but has not explained underlying psychological processes. We examined rioting in three locations during the August 2011 disorders in England to test a social identity model of riot diffusion. We triangulated multiple sources to construct a narrative of events; and we analysed interviews with 68 participants to examine experiences. In line with the model, we found evidence for two pathways of influence: “cognitive” and “strategic”. For some participants, previous rioting was highly self‐relevant, and shared identity was the basis of their subsequent involvement. For others, previous rioting was empowering because it demonstrated the vulnerability of a common enemy (the police). In each location, interaction dynamics mediated the link between initial perceptions and collective action. The utility of this social identity approach is that it is able to account for both the boundaries and the sequence of urban riot diffusion.

In August 2011, over four days, rioting spread across several cities in England. Previous accounts of these riots have indicated the roles of police racism, class disadvantage, and spatial affordance. However what remains unclear is how these structural factors interacted with crowd processes spatially over time to govern the precise patterns of spread. The present paper provides a micro-historical analysis of the patterns and sequences of collective behaviour as the 2011 riots spread across North London, drawing upon multiple data-sets (archive, interview, video, official report, news coverage). The analysis suggests that initial stages of escalation in the broader proliferation were the result of protagonists deliberately converging from areas of relative deprivation in order to create conflict, but that they did so as a meaningful social identity-based expression of power. We show how over time these motivations and patterns of collective action changed within the riot as a function of intergroup interactions and emergent affordances. On this basis we provide support for the argument that political, social and economic geography were key determining factors involved in the pattern of spread of the 2011 riots. However, we also suggest that an adequate explanation must correspondingly take into account the interplay between social identity, the dynamics of intergroup interaction, and empowerment process that develop during riots themselves.