Co-convictions and specialisation in criminal careers - a modest proposal

The importance of co-convictions has rarely been recognised. Using a dataset (n=67,052) of all those convicted in England andWales (1979-2001) for arson, blackmail, kidnapping or threats to murder, we consider how much information on co-convictions is normally overlooked, and the extent towhich knowledge of co-convictions contributes to predicting
serious recidivism. We identify that co-convictions are pervasive, ranging from47%in total for arson through to 77% for kidnapping. Specialisation in crime is a contentious issue and Sullivan et al. (2006) has identified the need for short termmeasures of specialisation in criminal careers research.We claim that that co-convictions act as a measure of short-term specialization. However, while specialisation (as measured through co-convictions) is an important additional predictor of recidivism, the evidence suggests that its effect can vary dramatically according to which type of recidivist activity is under study.