Alliance formation and the nature of alliance relationships have been frequent subjects of study in international politics. The questions of whether or not states primarily balance or bandwagon, why states choose one ally over another, and the manner in which alliances shape state behavior are important not only for historical reasons but also because of the role that major alliances continue to play today. For example, the continued existence of NATO—and its activities in humanitarian intervention—suggests that something beyond the creation of an alliance to confront a common enemy has occurred. Likewise, the special relationship between Israel and the United States, in spite of policy differences and the possibility that the relationship complicates other relations for both states, has been the subject of a vigorous and, at times, controversial scholarly and policy debate.

Marco Cesa's Allies yet Rivals provides an interesting contribution to this field of study. Cesa argues that states form alliances primarily to decrease uncertainty regarding an ally's expected actions. States exchange some freedom of maneuver in order to reduce this uncertainty. The primary contribution of the work, however, is Cesa's relatively simple typology of alliances based primarily on two factors: the power relationship between allies and the congruity of the allies' interests. This typology allows Cesa to make some predictions regarding the freedom of movement of alliance partners, as well as the degree to which allies will work together to achieve their interests.

Cesa acknowledges that both the relative power between allies and the congruence of their interests lie on a continuum, but he uses ideal types in order to classify alliances. Alliances between states with similar capabilities are symmetric, while those between states with different capabilities are asymmetric. When state interests are largely congruent (such as in a defensive alliance with one clearly defined common enemy), the alliance is homogenous. When states have limited interests in common and several objectives that are not shared, the alliance is heterogeneous. Combining the two sets of classifications yields four types of alliances: 1) aggregation alliance (symmetrical/homogenous); 2) guarantee alliance (asymmetrical/homogenous); 3) deadlock alliance (symmetrical/heterogeneous); 4) hegemonic alliance (asymmetrical/heterogeneous).

In analyzing the nature of these different types of alliances, Cesa is informed by Glenn Snyder's Alliance Politics (1997), particularly the alliance security dilemma: the simultaneous fear of both abandonment and entrapment by an ally. In this dilemma, an ally needs to be capable of and willing to help a state achieve its interest, but not so powerful that it can dictate the terms of an alliance. Each of the alliance types will be marked by different levels of fear of abandonment and relative bargaining power, as well as the lasting power of the alliance. In an aggregation alliance, bargaining is marked by an equal exchange of services that will normally end once the two states have achieved their basic objective, whether defensive or offensive. Fear of abandonment is real, but is tempered by the common cause. In a guarantee alliance, the more powerful state is less dependent on the weaker member and so less fearful of abandonment, while the opposite holds for the less powerful state. The more powerful state may also try to strengthen its partner in order to attain more benefit from the alliance and to lessen the fear of free riding. Deadlocked alliances are characterized by mistrust and fear of abandonment and entrapment. Bargaining is more likely to be coercive and may be marked not by a positive exchange of services but, rather, by threats of withholding services, as the two states are fearful not only of losing their ally but also of being pulled into a conflict in which they have little interest. Finally, in the hegemonic alliance, a weaker state is dependent on a more powerful actor that may largely dictate the relationship. The weaker state's need for the larger state's support may require it to adopt positions it would otherwise oppose. Bandwagoning relationships like those described by Stephen Walt, in Origins of Alliances (1987), may fall into this category.

While the typologies are interesting, their utility is somewhat limited in that the author does not provide specific, testable hypotheses regarding state behaviors that one should expect in each type of case, but rather provides broad descriptions of the nature of the relationships. Neither do the typologies provide sufficient insight into when state alliances are most likely to form and between which ones. In general, states prefer to be the more powerful partner and prefer allies that are able to both render services and share the objectives of the more powerful state. However, such opportunities may not present themselves frequently. Outside of the typology, ally availability appears to be a major factor affecting alliance choice—a view that the case studies support. In an examination of the types of alliances, it is not clear why deadlocked alliances form—the author notes that the alliance may be “disadvantageous to both” (p. 78). Finally, the author notes that neither the power relationship nor the congruence of interests between allies is strongly predictive of the staying power of the alliance (p. 227). Nonetheless, the typologies do provide a fairly straightforward framework that can serve as a useful way to categorize alliances.

To illustrate the theoretical typology, Cesa examines four alliances from the eighteenth century: British alliances with the United Provinces (guarantee alliance), France (hegemonic alliance) and Prussia (aggregation alliance), and France's alliance with Austria (deadlock alliance). The case studies themselves illustrate the difficulty of applying broad ideal types to history. In particular, the British alliance with the United Provinces evolved over time as interests began to diverge between the two parties, and the external threat (France) began to recede. It is interesting to note that during the course of that alliance (1702–56), Britain also formed an alliance with France (1716–31), the purported object of the alliance with the United Provinces. Steady conflict across the continent during the time period does provide for frequent shifts in alliances and opportunities to study bargaining, but also makes the consistent application of the typologies difficult.

To conclude, Allies yet Rivals does provide a useful typology that could be made stronger with further refinement and effort to produce predictions regarding intraalliance behavior. The case studies are an interesting addition in that they cover an era not frequently used in the study of international relations. The book would have greatly benefited from applications of some of the lessons from the case studies and the typologies to more recent alliances, at least briefly in the conclusion. For example, one area not considered by the theory, but which plays a significant role in the case studies, is the impact that shifts in leadership (the death of monarchs, including wars over succession) had on alliance formation, a topic that would be particularly interesting given recent changes in the Middle East. I do recommend the book to scholars interested in the dynamics of intraalliance relationships as the basic typology provides an elegant way to think about different types of alliances.