This book challenges the rational choice assumptions that undergird much of the extant international relations literature on deterrence, bargaining, cooperation, economic behavior, and the exercise of power. In an ambitious extension of his previous work, Jeffrey D. Berejikian attempts to construct “a new set of theoretical propositions about international politics securely anchored to empirical research in cognitive psychology” (p. 2). In the tradition of Graham T. Allison's (1971) Essence of Decision, Berejikian meticulously demonstrates the degree to which behavioral expectations regarding strategic interaction are tied to underlying conceptual models of decision making. He substitutes prospect theory's empirically grounded observations regarding framing effects and loss aversion for rational choice assumptions regarding invariance and net asset valuation. The result is a comprehensive cognitive model of state interaction that is capable of explaining preference reversals, risk acceptance, and nonmaximizing choice. The boldness of this enterprise more than compensates for minor flaws in execution.

The publication of International Relations Under Risk marks the culmination of more than a decade of work attempting to integrate prospect theory into international relations research. The book includes portions of articles published in the American Political Science Review and the Journal of Peace Research, as well as a significant amount of new material that molds the chapters into a very coherent whole. The books flows smoothly—from an introduction and defense of prospect theory in Chapters 1 and 2; to the modification of basic expectations regarding strategic interaction in Chapters 3, 4, and 5; to an empirical investigation of the cognitive models in Chapters 6 and 7. The empirical chapters focus on European Community (EC) and United States decision making involving the negotiations that eventually produced the Montreal Protocol restricting chlorofluorocarbon production. The author concludes that observed shifts in negotiating frames and risk-acceptant cooperation can only be explained by cognitive models of choice.

Berejikian's introduction and defense of prospect theory is relatively brief. The uninitiated reader would find a better review in Chapter 2 of Rose McDermott's (1998) Risk-taking in International Politics (as he readily acknowledges, p. 123, n. 6). The author does, however, introduce the relevant concepts—reference points, aspiration levels, the s-shaped value function, framing effects, the endowment effect, and loss aversion. The credibility of prospect theory as an alternative to rational choice theory is then carefully established. Berejikian reviews the massive body of literature utilizing prospect theory in a range of disciplines, emphasizing its empirical success and noting the Nobel committee's recent endorsement of behavioral economics (p. 2). The review of the application of prospect theory to international relations puzzles is less persuasive. Much of the first generation work that he cites has been heavily criticized, and some more recent publications in the area are oddly omitted (most notably, James W. Davis's Threats and Promises, 2000). Berejikian's responses to critiques regarding differences between the “real world” and laboratory setting, the identification of reference points, the data requirements of the theory, and collective versus individual choice are spirited, but not completely convincing. Too often the default defense is that a “similar critique applies equally to rational choice” (p. 17) or that the explanatory power of prospect theory is “an empirical question” (p. 21).

The strength of the book and the author's most significant contribution to the literature can be found in Chapters 3, 4, and 5. In each chapter, Berejikian initially “presents the logic and behavioral expectations of a rational model … then modifies or extends the analysis by integrating prospect theory” (p. 2). Over the course of a mere 50 pages, he undermines rationalist understandings of military deterrence, economic threats, negotiation, cooperation, collective action problems, and the relative versus absolute gains debate. In several instances, the integration of prospect theory's descriptive observations produces counterintuitive expectations regarding strategic interaction. Where prospect theory and rational choice models offer similar predictions, Berejikian contends that prospect theory more accurately characterizes the process of decision making (see pp. 5–6). He constructs cognitive models that reveal how credible deterrent threats can induce a “losses frame” and produce a risk-acceptant challenge (p. 41); how losses frames can produce both “new cooperative agreements” and “defection within established cooperation” (p. 57); how framing effects illuminate the different aspects of collective action problems that “require governments to make a contribution from existing endowments” versus those that “restrict consumption of an existing resource” (p. 61); and why “states in a gains frame pursue absolute gains and are risk averse, whereas states in a losses frame chase relative gains and are risk acceptant” (p. 76).

Armed with this new set of expectations (derived from prospect theory) regarding strategic behavior, Berejikian embarks on a study of the EC and U.S. negotiating strategies that eventually resulted in their acceptance of the Montreal Protocol. Chapters 6 and 7 reveal the extent to which shifts in problem frames favored risk-acceptant strategies to secure relative advantage by the EC and the United States. The interaction of these strategies then produced unintended cooperation on a major environmental issue (see pp. 96–97). The author demonstrates a significant depth of knowledge about the cases—identifying the dynamic positions of relevant government, multinational, and domestic political actors. He does fall prey to some of the problems associated with identifying problem frames (too often suggesting that actors are “in” a frame, instead of focusing on how alternatives are framed) and measuring the subjective probabilities and utilities associated with outcomes under consideration (relying on ordinal estimates of utility and verbal representations of probability), but the cases are clearly intended to illustrate and not fully test the comprehensive cognitive model.

In the end, Berejikian clearly succeeds in his effort to demonstrate the value of integrating empirical observations from cognitive psychology into our models of state interaction. Descriptive knowledge of how individuals and groups actually make decisions increases the predictive and explanatory value of our conceptual models. For Berejikian and other scholars impressed by contemporary cognitive research, the “as if” defense of the rationalist enterprise is no longer adequate (pp. 5–6).