Association for Public Policy Analysis & Management

Understanding terrorist targets of choice has been crucial in counterterrorism policy-making as it provides important guidelines for allocating security resources. Scholars have intensively investigated the determinants of terrorist target selection and the target categorizing and examined the spatial clustering of terrorist targets. Yet systematic research remains limited on the group clustering of terrorist targets.

This study conducts a spatial econometric analysis of terrorist target choices to further understand the clustering pattern of terrorist attacks from a target group lens. This study analyzes the panel data drawn from the Global Terrorism Database (GTD), which includes terrorist attacks in North America and Western Europe from 1970 to 2015.The target data had been sorted into four groups based on the attributes of each target category: Government-Designated, Government-Undesignated, Private-Designated, and Private-Undesignated. Utilizing spatial autocorrelation analysis, this study assigns the similarity of target categories as the spatial distance, then sets weight coefficients of different target category based on if the target category are in the same target group or not. Global Moran’s I statistic indicates there is a significant clustering phenomenon of terrorist attacks. Local Moran’s I analysis identifies four clusters of terrorist targets. Also, the analysis of diffusion pathway shows that the clustering was weakening in the Government-Designated group while strengthening in Private-Undesignated. Lastly, this study explains the clustering pattern of terrorist attack by path-dependence theory and contagion theory.

This study is one of the early attempts to apply concepts and tools from spatial econometrics to studying group clustering of terrorist attacks. Findings from this study can shed light on allocating counterterrorism resources and hardening potential terrorist targets.