This book uses both formal theory and quantitative analysis to explore the link among leadership change, domestic institutions, and international cooperation. In addition to explaining how the threat of removal can help democratic leaders commit to deeper and more robust levels of cooperation, it also generates novel predictions over the effects of leadership change itself—not merely its ex ante threat—on the dynamics of cooperation. In particular, Fiona McGillivray and Alastair Smith ask whether cooperation can be harmed by leadership turnover and how leadership change might lead to the restoration of failed cooperation. Taking individual leaders as the unit of analysis, they explain known empirical regularities while predicting new ones and, just as importantly, contribute to the logical extension of the growing body of leader-centric research: Even if leaders act to retain office, what happens once they are deposed?

McGillivray and Smith's leader specific punishment (LSP) theory proceeds from three rather uncontroversial premises. First, national leaders care primarily about staying in office. Second, their citizens wish to maintain cooperative international relations because cooperation enhances their welfare; this can generate incentives to punish (i.e., remove) leaders who forgo international cooperation. Third, domestic political institutions vary in the extent to which they make the removal of national leaders costly.

The authors operationalize the costs of leader removal by drawing on selectorate theory (Bueno de Mesquita et al., The Logic of Political Survival, 2003), which posits that the smaller the winning coalition required to maintain a leader in office, the costlier it is for a citizen to remove that leader, given the risks of being left out of any future winning coalition. In political systems with larger coalitions, supporting a challenger to the incumbent is less risky, as the odds of being left outside the new winning coalition are lower.

Where the cost of leadership removal is low, as in states with large winning coalitions (generally speaking, democracies), the threat of removal at the hands of their citizens induces leaders to cooperate and violate agreements only rarely. On the other hand, where citizens find leader removal costly, as they do in small-coalition (read: less democratic) systems, leaders are freer to cheat in order to maximize short-term gains, secure in the knowledge that while they may forgo international benefits, they will continue to enjoy the benefits of remaining in office.

The use of selectorate theory also facilitates a more important theoretical contribution. Leaders can organize their winning coalitions in a manner of their choosing once in office, and while large-coalition systems face an incumbent with a relatively stable median voter to please with public goods, small coalitions can lead to drastic reorganizations that fundamentally change the nature of the “representative” member of the winning coalition, leading to potentially greater policy volatility from one leader to the next.

This simple extension of their theoretical logic allows McGillivray and Smith to make some novel predictions that can distinguish their explanation for democratic cooperation from its rivals. First, leadership change in small-coalition systems leads to greater policy volatility even during periods of cooperative relations, as the leader shifts private goods to satisfy a new, narrowly defined, winning coalition. Second, during periods of poor relations, leadership change in small-coalition systems should facilitate the renewal of cooperation, as a leader with an unsullied reputation for upholding agreements takes office and is given the chance to start afresh. Finally, leader change in large-coalition systems should have little effect, if any, on cooperative relations, even when those relations are poor, because policy priorities should change only rarely. This is a perhaps unintentionally timely observation, coming as it does with a leadership transition in the United States.

Punishing the Prince provides both a compelling explanation for observed patterns of democratic cooperation and additional empirical content beyond previous explanations of the same phenomenon, including the effects of leadership change itself on patterns of cooperation. In other words, it provides valuable, and all too rare, predictions that can discriminate among competing explanations. The authors test the implications of LSP in both experimental settings and large-n quantitative analyses of international trade and sovereign debt, the former updating previous work and the second a wholly novel test. Moving beyond cooperation, they also relate their predictions to existing evidence about the duration of economic sanctions in more conflictual settings and make policy prescriptions for, first, targeting foreign policies at leaders rather than states and, second, designing international institutions to facilitate compliance through leader-specific mechanisms. If the analysis seems lacking on any point, it is on the absence of original analyses of the effects of LSP on conflictual interactions. However, the decision to focus primarily on cooperation is valuable in its own right, as it provides tests of the theory in multiple, often understudied, realms of international politics.

Finally, like any good work of social science, Punishing the Prince raises fruitful additional questions. For example, LSP assumes that replacement leaders are drawn from a random pool of possible successors (p. 7). This premise supports the prediction of greater policy variability in small-coalition systems, but relaxing it may shed different light on the questions of when and how leader change can lead to an improvement of relations (p. 191). What if, for example, the pool of available successors is undesirable to potential international partners? This may, for instance, lead to rather different patterns of cooperation designed by foreign states to influence the domestic balance at home. Additionally, since the value of cooperating with autocratic leaders is low, they may be particularly attractive targets for cheating, even by democratic leaders, whose citizens are unlikely to punish them for forsaking cooperation of little value (p. 50). This tells us who is likely to get cheated and, perhaps, who should be most reticent to formalize cooperative relations, not because of an inability to cooperate but due to a real fear of being cheated by other states.