Abstract
Levy et al. (2015) propose that inconsistency audience costs are caused in two ways: backing out of military commitment or backing into foreign conflicts. We replicate their experiment in July-August, 2021. Like many other studies, we find evidence for audience costs caused by backing out. However, our findings indicate that, unlike Levy et al., citizens are no less supportive of a leader who backs into a military conflict despite an initial commitment to stay out than one who behaves consistently. This study has significant theoretical and methodological implications. Theoretically, the importance of inconsistency in audience cost literature may be overstated or inconsistency audience costs may be bounded by temporal domains or contexts. Methodologically, our study emphasizes the necessity of replications because major findings can become conventional wisdom without such additional analysis.