Understanding why the European major powers found themselves embroiled in “The War to End all Wars” in the late summer of 1914 has been an ambition of political scientists and historians for almost a century. Given all this attention, just about every possible logical explanation for its onset has been proposed at one point or another. In his well-researched recent book, Frank C. Zagare does not aim to bring additional explanations to the table. Instead, he uses an existing, formal, theoretic framework to identify a logically consistent and plausible explanation for the outbreak of World War I.
Zagare's framework for analyzing the July Crisis is perfect deterrence theory, which he developed with D. Marc Kilgour in a series of articles and a previous book (Perfect Deterrence, 2000). Perfect deterrence theory is a set of related game-theoretic models in which actors have incomplete information about each other's credibility. In many ways, the current book can largely be seen as an exploration of the empirical validity of the theory. To do this, Zagare pursues a methodological approach similar to the analytic narratives used by Robert Bates and his coauthors (Analytic Narratives, 1998), in which the insights from a game-theoretic model are used to identify a causal explanation for a particular historical event.
In the first three chapters of the book, Zagare outlines his methodological approach, considers alternative theories, and summarizes the basic components of perfect deterrence theory. To a great extent, the author presents his theory in contrast to two versions of classical deterrence theory, which he labels “structural” and “decision-theoretic.” (Liberal explanations for the war are also considered, but they are quickly dismissed.) Given that perfect deterrence theory was developed in response to classical deterrence theory, this focus is understandable. However, the analysis would have been further enriched if the author had provided more discussion as to why a deterrence framework is the most appropriate way to model this particular situation, rather than other game-theoretic frameworks commonly found in the literature, such as bargaining models or leader-centric models (e.g., selectorate theory).
The centerpiece of the book is an evaluation of strategic decisions made at four key points leading up to the outbreak of the war: Germany's decision to enter into a defensive alliance with Austria-Hungary in 1879, Germany's “blank check” to Austria-Hungary in July of 1914, the escalation of a local conflict between Austria-Hungary and Serbia into an expanded war among the continental European powers, and the failure of Britain to deter a German attack on France. For each case, Zagare presents and analyzes a game-theoretic model of deterrence or extended deterrence. The author does a nice job of presenting and describing the theoretical models in a manner that will be accessible to a general audience. After analyzing each model, he explains how the implications of the theory can be used to identify an explanation for the historical events in question.
In Chapters 4 and 5, the author examines the strategic relationship between a defender (Germany) and its protégé (Austria-Hungary) before the outbreak of the war. In both cases, he finds that Germany likely pursued an action favorable to its protégé (the defensive alliance of 1879 and the blank check of 1914) because Austria-Hungary had a credible exit option of aligning with another state if Germany demurred. Chapters 6 and 7 turn the attention to the decisions of the European powers to escalate a local war into a world war. According to Zagare, the Great War was a case of failed deterrence. First, Russia failed to deter Germany and Austria-Hungary because these states mistakenly believed that Russia would stand aside if Austria-Hungary attacked Serbia. Similarly mixed signals due to its “saddle strategy” prevented Britain from convincing Germany that it would respond to an attack on France.
The author's account of the causal process behind the events leading up to World War I is plausible and largely persuasive. However, the analysis does illustrate some of the issues that all scholars face when they attempt to use an abstract theoretical model to explain a specific empirical event. First, existing models must often be amended to fit the specific facts of a case, or assumptions about the specific case must be made to fit the structure of the model. Second, complex models of incomplete information often have multiple equilibria, which are defined primarily by the players' prior beliefs about each other's preferences. Given that it is difficult to precisely identify actors' beliefs empirically, one cannot often identify a priori which equilibrium a model predicts in a given situation.
To illustrate these tensions, consider Zagare's analysis of the onset of the continental war in Chapter 6. To fit the situation with his Asymmetric Escalation Game, one must assume that Germany and Austria-Hungary act as one actor. However, the evidence indicates that the leaders of these states had different beliefs and perhaps different preferences, which calls into question the plausibility of this assumption. The author argues that prior to the July Crisis, both Germany and Austria-Hungary believed that Russia would not respond militarily to an attack by Austria-Hungary on Serbia. In an attempt to signal its resolve, the Russians pursued partial mobilization against Austria-Hungary. In response to these events, Zagare claims, the Germans received the message and began to push for a political settlement, while the Austrians continued to believe that the Russian threat was not credible. Since the leaders of both states received the same signal, their varied responses must have been the result of a stronger prior belief by the Austrians than by the Germans that Russia was not willing to intervene in an Austro-Serbian conflict.
Therefore, the equilibrium prediction of the model depends not only on whether Germany or Austria-Hungary believed that Russia would fight but also on the strength of that belief. While the former may not be too difficult to determine empirically, the latter is much more challenging to decipher. Thus, if we simply knew that a Central European power initially believed that Russia was not resolved and then Russia engaged in a partial mobilization, both a negotiated settlement and the outbreak of continental war would be consistent with the predictions of the theoretical model. To be sure, Zagare recognizes many of the issues that arise when creating an analytic narrative, and he believes that the upside of his methodological bet outweighs the downside (pp. 12–13). Readers will be able to make their own judgment on the method's merits.
A common critique of game-theoretic work in international relations is that it is often not connected to the empirical world that it is meant to explain. By combining a sophisticated model with detailed historical analysis, The Games of July does a nice job of showing that this does not need to be the case. No doubt this book will be of interest to security studies scholars and students of World War I, and it would be a nice addition to graduate course syllabi as an illustration of the successful combination of formal theoretic and historical analyses.
