Defense Pacts and Deterrence: Caveat Emptor

In a previous study, we provided a novel empirical test that indicated that alliance formation is often a poor means of deterring militarized disputes or preventing war. In this issue, Leeds and Johnson respond, arguing that our results differ from their previous work because we analyze the time surrounding alliance formation, which they claim is inappropriate when testing deterrence theory. Here we outline the reasons why examining alliance formation is critical to the analysis of deterrence. We also demonstrate that the empirical concerns raised by Leeds and Johnson are largely inconsequential by reproducing our results using the same design implemented in existing studies that uncover evidence of deterrence success. We find additional support for our primary conclusion that alliance formation often fails to deter adversaries and instead increases likelihood of militarized conflict.