Peacekeeping and Civilian Protection in Civil Conflicts: A Response to Kocher's Re-Analysis

Abstract

In a re-analysis of data on civilian deaths and UN peacekeeping personnel commitments, Mathew Kocher argues that peacekeeping is not associated violence against civilians. However, his analyses suffer from several problems. First, his dependent variable does not measure the phenomenon we sought to study in our article. Also, his dependent variable makes it nearly impossible to draw inferences about the determinants of civilian deaths, given the highly variable nature of one-sided violence. Second, his analyses suffer a sample selection problem, because he only analyzes civilian deaths in wars with peacekeeping operations. Finally, Kocher’s visual interpretations of violence graphs are tenuous, at best. Using a more appropriate sample and a number of robustness checks, we find that compared to wars without UN operations, civilian deaths are fewer in conflicts with larger amounts of UN military troops and police. We appreciate Kocher’s contribution for furthering the discussion of how to assess UN missions, and we hope to see more work on the theoretical mechanisms by which the UN operates in the future.