We welcome the opportunity to engage with Matthew Baum and Philip Potter about our book, Why Leaders Fight, along with their great book, War and Democratic Constraint. Both of our books highlight how the reality of foreign policy making, especially concerning the use of military force, depends on a number of domestic political, international, and leader-centric factors. We agree with many of the issues they raise in their very fair review, and believe they constitute a great example of critical research questions on leaders and international politics that we hope scholars in the field pursue in the years ahead.

To begin, Baum and Potter raise the possibility that the microfoundations of some of our arguments about leader risk are under-informed when applied to particular types of leader experiences. There is undoubtedly some truth here. Given that the Leader Experience and Attributes Dataset (LEAD) covers over thirty leader attributes, we had to pick and choose where to focus for the text itself. Moreover, as we argue in the book, it is difficult to make finer-grained predictions based on the interactions of particular leader attributes, e.g., point predictions for a leader with military service and combat experience, but who is young, and who has a given level of education and family situation. The interaction of different leader attributes, as well as the interactions between leaders, actually represents an extremely promising avenue for future research. In the area of military service, however, which they raise, we believe we are on firm ground. Our expectations about the effect of military service with combat, in particular, compared to other types of military service, comes out of classic research in civil-military relations by Huntington, Janowitz, and others. Similarly, our arguments about age flow from research in psychology on time horizons, as Horowitz and Stam explored in a previous paper with Rose McDermott.

The issue that Baum and Potter suggest about censored data is a good one. It is certainly the case that we can only observe those leaders who come into office. Their suggestion to look at a random sample of political figures who competed to become head of state but did not gain power makes sense. Gathering data on this sample of leaders might be difficult, but it is worth pursuing. One possibility might be focusing on the American case. Gathering data on the major party candidates for office (e.g. Trump and Clinton in 2016) is entirely possible, and could help future researchers model, at least for the American case, the particular expected impact of the president who was elected in the United States compared to the president who could have been elected.

We also agree that building more data on the relationships between leaders would be valuable. For example, it is possible that leaders facing off against potential adversaries with similar experiences might understand those potential adversaries better, helping them come to agreements easier than when the leaders have had very different experiences prior to entering office. Alternatively, leaders with similar experiences might feel like they understand the weaknesses of their counterpart, making them more prone to aggression. Finally, Baum and Potter are correct to point out the potential role of outlier leaders in our analysis. While we attempted to control for that statistically, it is entirely possible that there is something about a Saddam Hussein or Muammar Qaddafi that goes beyond what our leader-experience data capture. Our book focuses on the interaction between leader attributes and traditional political science variables such as regime type and material power. What their comment suggests, however, is that ongoing research should consider the role of nature and genetics, perhaps, to more fully flesh out the whole leader, in addition to background experiences.

Once again, we want to thank Professors Baum and Potter for engaging with us on our respective books. The interaction between our work, in fact, could be a promising area for scholars moving forward. Bringing together the attributes of leaders with the characteristics of the media could help explain why some leaders are better able to leverage the bully pulpit than others, for example, further filling out our models of the relationships between leaders, the domestic institutions they work in, and the international security environments that shape them and that they shape.