Targeting Drug Trafficking Networks: A Comparative Evaluation of SNA-Informed Disruption Strategies

Drug trafficking networks can be targeted using social network analysis (SNA)-informed disruption strategies, however little research has compared interventions across simulated networks generated from empirical networks of varying structure and size. This study compares the relative structural and functional disruption of six sequential actor-removal interventions: high betweenness, degree, and eigenvector centrality; minimum cut sets; and network backbone with random or betweenness-based selection. Eight drug trafficking networks were modelled using exponential random graph models (ERGMs), generating 1,000 simulated networks per dataset, on which the six interventions were tested. Betweenness and degree centrality-based interventions caused the greatest disruption across most networks, followed by eigenvector centrality-based interventions, then backbone-betweenness, while cut sets and backbone-random performed only marginally better than random removal. The findings identify centrality-based targeting, particularly betweenness and degree centrality, as the most reliable disruption strategy for drug trafficking networks, with implications for law enforcement prioritisation of network disruption efforts.