Does criminal violence spread? Contagion and counter-contagion mechanisms of piracy

Research shows that political and criminal violence cluster spatially but neglects the wide range of mechanisms driving contagion and, more importantly, the role of counter-contagion efforts. After identifying permissive conditions for piracy, I hypothesize that piracy clusters in locations conducive to successful attacks. Pirates engage in risk-reducing behaviour: they return to areas where they have been previously successful but also adapt this learning-based decision to constraints imposed by EU counter-piracy. The analysis relies on uniquely detailed data on piracy and counter-piracy in monthly grid-cells off Somalia (2005-2013). Results show that although successful attacks foster more attacks and contagion, EU counter-piracy reduces contagion. Even within most successful locations, rescue operations reduce incidence of piracy by 89% in the following month. The article contributes to existing contagion/diffusion literature by identifying specific channels of contagion (contiguity and learning) and by factoring in containment policies that can limit and reduce criminal and political violence.