Why do some wars drag on for years, while others are quickly resolved after a few battlefield clashes? Sarah Croco’s excellent new book explores the conditions under which wars are terminated. She argues that the answer has much to do with domestic politics. Her core insight is the importance of the culpability of leaders, of those individuals who led the state when the war began, regardless of the eventual war outcome (so culpability does not always mean “guilty of a defeat”). She finds that the culpable leaders are far less likely to want to end wars, whereas nonculpable leaders—those who came to power after a war began—are more likely to accept necessary compromises to end them. The argument has an intuitive appeal at a time when America’s experience in Iraq, and the different approaches of Presidents George W. Bush and Barack Obama, are likely to loom large in readers’ minds.

The core logic of the argument is straightforward. Voters are more likely to politically punish a culpable leader for a bad war outcome than a nonculpable leader. This gives a culpable leader greater incentive to continue a war in the hopes of achieving a better outcome or even just delaying the inevitable defeat. Consequently, Croco’s theory predicts, and her statistical analysis confirms, that culpable leaders tend to have relatively bimodal war outcomes: They either win big (when the gamble pays off) or they lose big (and face the wrath of their domestic audience). Nonculpable leaders, on the other hand, tend to have more mediocre outcomes, with fewer outright wins or losses, and more negotiated settlements. Her argument that this logic applies not just to democracies but also to nondemocracies (though not quite as strongly for the latter) is an interesting one. Although the domestic audience in autocracies cannot punish culpable leaders via elections, Croco argues that elites in autocracies often find other ways to punish culpable leaders who lose wars. In this sense, she builds upon Jessica Weeks’s findings on the similarities between democracies and some forms of autocracies (in Dictators at War and Peace, 2014).

Peace at What Price? has a conventional structure. After the introduction, there is a theory chapter and three empirical chapters, followed by a conclusion. Each of the empirical chapters focuses primarily on a statistical analysis, though there are some illustrative historical examples sprinkled throughout. The first empirical chapter, Chapter 3, tests and finds support for the book’s hypotheses about leader tenure: Culpable leaders are indeed more likely to be punished (compelled to exit office) if they lose a war than are nonculpable leaders. The next chapter tests the implications for war outcomes. As expected, culpable leaders tend to have a relatively high “win” rate, whereas nonculpable leaders are relatively more likely to end a war in a “draw.” Chapter 5 then extends the analysis to legislative leaders as opposed to the executive leaders studied in Chapter 3. Chapter 5 focuses only on the U.S. context in contrast to previous chapters. Here, however, the analysis is somewhat less convincing. Her findings (pp. 142–45) suggest that the effects of simple partisanship tend to be far more important than culpability, and she does not conduct any statistical test on whether voters punish culpable legislators as they do culpable executives. Indeed, she finds that voters punished Republicans in 2006 “regardless of the Republican incumbent’s position on the war” (p. 148).

Croco’s theoretical focus positions her research squarely in a growing body of scholarly work on leaders and elites in international relations. This corpus includes Leaders at War by Elizabeth Saunders (2011); Why Leaders Fight by Michael Horowitz, Allan Stam, and Cali Ellis (2015); and Leaders and International Conflict by Giacomo Chiozza and Hein Goemans (2011), among others. Leaders and elites are more difficult to study in some ways than masses, whose preferences are more amenable to the survey experiments that have come into vogue in IR. Yet the growing body of insightful leader-centric research suggests that the explanatory payoff to studying elites is well worth the effort.

Although Croco is not eager to challenge the rationalist bargaining model of war (p. 48), her book is the latest to point to the shortcomings of using it as the dominant model for understanding war. Her analysis is essentially monadic, paying little attention to adversaries’ strategic reaction to culpability. If her argument is correct, it raises the question why adversaries (who can observe culpability as well as voters, presumably) do not adjust their bargaining demands downward when facing culpable leaders and upward when facing nonculpable leaders. If they did, one would expect convergence in the rate of “wins” and “draws” among culpable and nonculpable leaders, albeit with different substantive settlements to the war. Yet that is not what Croco finds; instead, her findings suggest little strategic adjustment on the part of opponents. She suggests that culpability might even cut in the other direction: “An adversary may not trust a culpable foe to commit to unfavorable terms because he knows the leader will face repercussions from his citizens if he does not win. Given such a scenario, the adversary may feel he has no choice but to continue the war” (pp. 47–48). That logic suggests exactly the opposite: Adversaries will adjust their bargaining demands upwards when facing culpable leaders. Croco might be right, but her work points to an unfortunate indeterminacy in the underlying theoretical framework. As is so frequently the case with the bargaining model, it is possible to construct a rationalist story that fits any possible empirical pattern.

It is not the job of Peace at What Price? to defend the bargaining model, however. If one is looking to criticize the book itself, the absence of any attention to the domestic political effects of war victories is more notable. Croco focuses almost entirely on the consequences of war losses. Yet if every war is a calculated gamble, and voters know that, should they not reward victorious leaders for the same reasons they punish culpable defeated leaders? This would seem to follow from the author’s logic, but some obvious counterarguments leap to mind: Winston Churchill’s electoral defeat after World War II or George H. W. Bush’s loss in the 1992 election after victory in the Persian Gulf War. Perhaps Croco’s argument is asymmetric and does not apply to war victories, or perhaps these two examples are outliers and the broader trend does support the idea that voters reward incumbent leaders who are victorious in war. The book, however, stays silent on the topic.

All in all, Croco provides a tight, focused argument supported by a robust empirical analysis. The book is a significant contribution to the work on state leaders, one that many instructors will choose to teach in the years to come.