New Publication in the Journal of Peace Research

There is scant empirical research on the role of the external acquisition of coercive technologies, and even fewer studies explore the respective causal mechanisms of their consequences. The authors argue that the existing literature has not adequately considered the potential simultaneity between conflict initiation and arms purchases. In contrast, their study explicitly takes into account that weapon inflows may not only causally induce conflicts but may themselves be caused by conflict anticipation.

Following a review of applicable theoretical models to derive their empirical expectations, they offer two innovative approaches to surmount this serious endogeneity problem. First, the authors employ a simultaneous equations model that allows them to estimate the concurrent effects of both arms imports on conflict onsets and conflict onsets on imports. Second, they are the first to use an instrumental variable approach that uses the import of weapon types not suitable for intrastate conflict as instruments for weapon imports that are relevant for fighting in civil wars.

Relying on arms transfer data provided by the Stockholm International Peace Research Institute for the period 1949-2013, they provide estimates for the effect of imports on civil war onset. The authors' empirical results clearly show that while arms imports are not a genuine cause of intrastate conflicts, they significantly increase the probability of an onset in countries where conditions are notoriously conducive to conflict. In such situations, arms are not an effective deterrent but rather spark conflict escalation.