Bannister et al. (2015) found growing inequalities in exposure to crime due to crime rates falling more in some areas than others. This raises a number of questions. How do crime fads and trends emerge? How do they spread to other neighbourhoods? Why are some neighbourhoods impervious to particular fads and trends, but susceptible to others? Where do particular fads and trends start and finish? To what extent is the crime linkage non-spatial? Are neighbourhoods linked through movements in crime linked in other ways? And how fragile or robust is the network of crime links? To understand inequalities in exposure to crime and why they shift over time, we need to understand the wider mechanism for crime transmission.
In this paper we develop a theory of "Crime Cascades". Drawing on the theory of "Information Cascades" developed in the economics literature, we develop a social network conceptualisation of how dynamic links between neighbourhoods in crime patterns emerge. We then apply this theory to data on local crime trends in England and Wales, complemented by a more detailed analysis of crime data in South Yorkshire, to develop a statistical network analysis of the geography of crime dynamics. We use this model explore the extent to which crime networks are shaped by or transcend ethnic, social and administrative frontiers and boundaries.
For more information about this research, please contact Professor Gwilym Pryce.