Public Perceptions of Wartime Atrocities: Evidence from a Conjoint Experiment

When is the public more likely to react to wartime abuse? Existing research suggests that governments, especially democracies with strong human rights lobbies, face a domestic cost for supporting states engaged in human rights abuses during conflict. But what types of abuses are more likely to damage a state's reputation? We theorize that the type and severity of violence, the identity of victims and perpetrators, the location of the conflict, and the strategic motivations for the conflict likely shape public perceptions of reputational harm. However, we argue that the effects of the type of violence depend on both the identity of the perpetrator and the victim. Using a conjoint experiment with a sample of about 1,500 Americans, we show how perceptions of harm vary by the type and severity of violence, identity of victims and perpetrators, conflict location, and strategic motivations. Interestingly, people react most powerfully to wartime torture by police and when aid workers and civilians are targeted. Our findings make important contributions to the study of public opinion, conflict, contentious politics, and human rights.