| Abstract |
The current research tests two hypotheses regarding the structure and evolution of dynamic terrorist networks. First, some argue that terrorist networks exhibit stringlike structures of sparse connections as an effort to maximize operational security. Others maintain that terrorist networks tend to contain a few highly-connected individuals who play crucial roles in connecting the larger network. Using Separable Temporal Exponential Random Graph Models (STERGMs), we test both of these hypotheses on several multi-wave terrorism networks from the John Jay & ARTIS Transnational Terrorism Database (JJATT). Our dataset includes networks from prominent attacks and bombings that occurred during the last three decades(e.g., the 2002 Bali Bombings), where nodes represent individual terrorists and ties represent social relationships. In contrast to theories of maximizing operational security, results suggest that there is a tendency for terrorist networks to become increasingly well-connected as they prepare for an attack. We also find evidence that highly central nodes develop even more ties in the years preceding an attack, signifying that terrorist networks tend to be structured around a few key actors. Our findings have the potential to inform strategies of counterterrorism efforts by suggesting which actors in the network make the most influential targets for law enforcement. |