I am grateful to Joshua Kertzer for his thoughtful review, which does an excellent job of summarizing the key contributions of my book and also highlights important issues for future research and reflection. I will use this space to focus on two of the issues that Kertzer raises.

First, Kertzer is correct to note how the relationship between statements of resolve and resolve itself presents a methodological challenge. If, as I argue, statements of resolve are costly signals, then we would expect leaders who make more resolved statements to be more genuinely resolved on average. If it is also true that greater genuine resolve causes leaders to send additional nonverbal signals of resolve, to fight harder, and/or to behave otherwise differently from less resolved leaders in disputes, then it may be difficult to disentangle the effect of resolved statements from other effects of resolve itself. On pages 79–83 of the book, I address this methodological challenge by controlling for other signals of resolve (admittedly with rougher measures); by controlling for factors that might influence resolve, including salience as measured by New York Times reporting; and by using a matching technique. However, Kertzer is correct that survey experiments are another tool that can shed further light on this topic, and this type of work should be encouraged. As noted in my review of Kertzer’s book, his own method of estimating true levels of resolve also has potential to help us further disentangle the effect of resolved statements from the effect of genuine resolve.

Second, I want to engage with Kertzer’s points about the application of my book’s arguments to the Trump administration. Kertzer is correct that the book argues that extremely resolved and even provocative language, such as Trump uses, can be an effective tool in international coercive bargaining when the necessary conditions for credibility are present. I also agree that low popularity may not impede Trump’s credibility as much as it has impeded the credibility of previous presidents because Trump seems less concerned with maintaining a wide support base. On the other hand, even though Republicans control both Congress and the executive branch, many members of Congress and Trump’s own administration are less hawkish on issues such as Iran and North Korea, which may create an impediment to following through and weaken Trump’s credibility. However, the biggest impediment to Trump’s credibility is likely to be his reputation. While my book focuses primarily on the ability to follow through on statements, I also find that a reputation for backing down reduces the effectiveness of statements (pp. 96–97). Indeed, a reputation for backing down can weaken the effectiveness of statements so much that whether a president has the ability to follow through no longer matters (pp.119–21). This suggests that threats of “fire and fury” without following through will diminish Trump’s ability to make credible statements in the long term.

Kertzer also points out several other fruitful directions for future research. I do not necessarily agree that the ability of negative characterizations to function as effective conflict bargaining tools poses a major challenge to traditional bargaining models of war, since an implicit threat can often be inferred from a negative characterization. However, I agree that my findings regarding negative characterizations suggest the opportunity for productive dialogue with other research traditions. I hope that discussion of these issues continues.