The intersection of domestic politics and international conflict is one of the most important areas of inquiry in international relations, if not the field of political science. Matthew A. Baum and Philip B.K. Potter’s War and Democratic Constraint makes a novel and significant contribution to this literature, with new theoretical insights and in-depth, multimethod empirics. The book focuses on the mechanisms through which the general public influences foreign policy, enabling and constraining national foreign policy decisions concerning war and peace. The authors persuasively argue that the interaction between the media and domestic political institutions could represent the missing link in academic understanding of how countries make foreign policy choices.

Baum and Potter’s fundamental claim is that as the number of political parties in a system increases and the media in a country becomes more open, the give-and-take of domestic politics is more likely to place constraints on leaders who want to take their countries to war. Moreover, neither media access nor the number of political parties, on their own, are enough to significantly influence the militarized behavior of a country. The United States, for example, has free media but a small number of political parties. Therefore, the United States is less militarily constrained than a country like Germany, which has free media and a larger number of political parties.

The authors’ theory makes several important predictions, validated with in-depth research, that should shape scholarship on domestic political institutions and international conflict from this point forward. In contrast with prior explanations of democratic military behavior focused on the inherent qualities of democratic institutions, Baum and Potter show that democratic countries with larger numbers of political parties and a free press are systematically more constrained than other democracies. This prediction differs from a simple distinction between presidential and parliamentary systems due to the role of media access in their theory, providing a way to further distinguish between types of democracies. Additionally, when it comes to conflict initiation, their results suggest that previous null findings on democratic conflict initiation (no relationship between democracy and military conflict initiation) reflect a failure to differentiate between types of democracies. In particular, democracies with open media access but a smaller number of political parties appear to be more conflict prone, while other democracies are less so.

Baum and Potter’s findings on conflict reciprocation—the probability that once a country initiates a military dispute against a target, the target will give in rather than join or escalate the dispute—are particularly interesting. It seems plausible that countries that are less likely to initiate military disputes are more credible when they do initiate such actions because they are more selective when it comes to the use of force. The authors show empirically that the same characteristics that make countries less likely to initiate military disputes—open media and a larger number of political parties—make the disputes those countries do initiate less likely to be reciprocated.

Given the ability of opposition parties with messages amplified by the media to publicly exert pressure, leaders will only initiate disputes when they are very committed because they would have to pay a high political price if they back down. The targets of disputes will recognize this dynamic, meaning that the target leadership recognizes that the aggressor leader is resolved, thus making the target more likely to back down. This provides new microfoundations for the logic of audience costs (James Fearon, “Domestic Political Audiences and the Escalation of International Dispute,” American Political Science Review 88 [1994]: 577–592; Kenneth Schultz, “Domestic Opposition and Signaling in International Crises,” American Political Science Review 92 [1998]: 829–844).

The ability to differentiate between types of democracies along the dimensions of media openness and political parties, as Baum and Potter do, therefore provides two key insights relevant for the audience costs literature on domestic politics and conflict reciprocation. First, they more coherently link the probability of conflict initiation to the probability of conflict reciprocation than previous research. This can help sync up varied research agendas in international conflict. Second, by explaining the logic of democratic constraint and how it may only apply to certain types of democracies, Baum and Potter’s research and theory could break a logjam in research on regime type and audience costs. Over the past several years, competing researchers have made the case that audience costs do not exist, that democracies lack unique advantages in generating such costs (and thus lack credibility) when they initiate conflicts, and more generally have pushed back at research suggesting that democracies have more credibility when they initiate military disputes due to the ability of the public to punish democratic leaders if they back down. The authors provide a way forward by highlighting specific mechanisms that make regimes better able to generate audience costs, something testable by other researchers in the future.

Other micro-level research conducted by Baum and Potter is similarly impressive. In Chapter 5, they explain how media access and the number of parties explain the decisions by states to initially join the U.S.-led coalitions in Afghanistan and Iraq despite public opposition in many cases, especially in the Iraq case. Using content analysis of news sources, Chapter 6 shows that media coverage in democracies with open media access and a larger number of political parties differs in fundamental ways from other types of democracies. Chapter 7 then provides case studies of key European countries, such as Germany, showing how their behavior in the Iraq War case is consistent with the authors’ theory.

A book as broad as War and Democratic Constraint, precisely because of the claims it makes, does raise some important questions. For their macro claims about conflict initiation and reciprocation, the authors’ models do, of course, suffer from the inherent inferential limitations of cross-national, observational time-series analysis, especially when using data like the Militarized Interstate Disputes and International Crisis Behavior data sets. The large number of robustness tests are reassuring, as are the more micro-level tests in Chapters 6 and 7, in particular.

Even in path-breaking books there is also always room to long for more: More here means the case research in some places. Take Germany, for example, which is discussed in Chapter 7, among other places. With a large number of political parties and open media, Baum and Potter’s theory predicts German opposition to the U.S.-led wars in Afghanistan and Iraq and Germany's failure to join the U.S.-led coalition in Iraq. Yet Germany deployed more than 3,000 troops to the U.S.-led coalition in Afghanistan, the International Security Assistance Force (ISAF). Why one conflict but not the other? One answer is that the invasion of Iraq was less popular with the German public than the invasion of Afghanistan, meaning that German political leaders had more flexibility to join the Afghanistan coalition. Yet that suggests an interesting model of public opinion. Rather than viewing public opinion as elite driven, with the public consuming media from chosen sources, Baum and Potter could be suggesting that public opinion operates, at the outset, more independently from elite preferences. Alternatively, they could be arguing that German opposition parties did not oppose German intervention in Afghanistan, which trickled down to the public as well. This begs the question: Why would German opposition parties not have the same incentive, political differentiation, to criticize potential German intervention in Afghanistan that they did in Iraq? If one believes that the public takes cues from elites in developing preferences on issues, opposition parties could have scored political points by drumming up opposition to German intervention in Afghanistan as well.

More broadly, if Baum and Potter are right that democracies with more political parties and greater media access are less likely to initiate militarized disputes, it might also mean that they are less able to confront real external threats that exist. Essentially, with a free press able to magnify the commentary of opposition parties, and many opposition parties with political incentives to oppose the governing coalition, regardless of the facts on the ground, might this mean that those states are actually less able to face down real external threats? It might be harder for the ruling coalition to persuade the voting public that a real threat exists, especially if some of the relevant information was secret due to classification.

One might respond that in an existential threat scenario, if politics faded to the background, one would expect those high media access/high party countries to unify and thus be able to confront a significant external challenge. However, one might still expect those states to be slower to recognize those larger threats because, again, parties with an incentive to criticize the ruling coalition and the ability to get their message out in the media could constrain the ruling coalition to a relative extent. This strikes the reviewers as plausible, and not inconsistent with Baum and Potter’s story.

Additionally, these countries might also be less able to successfully protect national interests once a conflict starts. The inherent features of the domestic environment—a large number of parties and free media—would still exist once a country gets into a conflict. This is what makes adversaries less likely to reciprocate; for example, adversaries become more likely to give up, once one of these states does start a conflict (Chapter 4). However, if reciprocation occurs and a conflict escalates, these countries should be less able to stay the course. Multiple parties with political incentives to take down the ruling coalition and free media could lead to more casualty sensitivity, for example, on the part of these kinds of democracies.

Another interesting facet of Baum and Potter’s argument is that, as in much of the international relations literature on the relationship between domestic institutions and international conflict, leaders do not play a role. The way a national leader behaves in a given situation (conflict initiation), and whether the behavior of the leader is considered credible (conflict reciprocation) depends entirely on factors related to the security environment and then the interaction of parties and the media environment. This means factors such as the individual persuasiveness of the leader, for example, do not make a difference. Some leaders are great orators who use the bully pulpit of leadership to persuade the nation to take a particular direction, such as Ronald Reagan in the United States (though the United States is, of course, a state with a relatively low number of parties but a high level of media access).

Similarly, leaders may have different inherent risk propensities. For example, leaders might interpret the threat environment differently and thus might have a different inclination toward initiating a military conflict in a given situation. Yet variation in leaders does not play a role in this story—it is just about how variations in the domestic environment condition the behavior of the leader. Future research might seek to integrate this story about the domestic institutional environment, which significantly advances our understanding of how democratic regimes behave, with our story about leader backgrounds and how they shape leader behavior.

None of these issues should distract from the fact that War and Democratic Constraint is a tremendous and persuasive piece of scholarship. It should definitely shape how scholars research the relationship between domestic political institutions and international conflict moving forward. It also provides crucial insights relevant for other fields, ranging from communications (due to the focus on media access) to history and others.