¹ See Putnam, Robert D., “Diplomacy and Domestic Politics: The Logic of Two-Level Games,” International Organization 42 (Summer 1988CrossRefGoogle Scholar); Iida, Keisuke, “When and How Do Domestic Constraints Matter? Two-Level Games with Uncertainty,” Journal of Conflict Resolution 37 (September 1993CrossRefGoogle Scholar); Tsebelis, George, Nested Games: Rational Choice in Comparative Politics (Berkeley: University of California Press, 1990Google Scholar).

² Maoz may be too generous in saying that realist models do not require states' decision latitude to be nil. Kenneth Waltz declares: “The elements of Realpolitik, exhaustively listed, are these: The ruler's, and later the state's, interest provides the spring of action; the necessities of policy arise from the unregulated competition of states; calculation based on these necessities can discover the policies that will best serve a state's interests; success is the ultimate test of policy, and success is defined as preserving and strengthening the state.” Waltz, , Theory of International Politics (Reading, Mass.: Addison Wesley, 1979), 117Google Scholar; emphasis in original.

³ Some such departure is essential to make sense of the realists' favored model of the security dilemma. One condition for a security dilemma is that any state, even if it seeks only security, cannot be sure whether other states are aggressive, rather than just security seekers. But if states' preferences are exogenous—derived from their place in the structure of the international system—their preferences for security/aggression should not differ; or if they do differ, they should be discernible to one another only insofar as the differences derive from their place in the structure.

⁴ For a discussion of this problem in international relations, see Huth, Paul and Russett, Bruce, “Testing Deterrence Theory: Rigor Makes a Difference,” World Politics 47 (July 1990Google Scholar); for a trenchant general challenge to rational choice models on grounds or empirical inapplicability, see Green, Donald and Shapiro, Ian, Pathologies of Rational Choice Theory: A Critique of Applications in Political Science (New Haven: Yale University Press, 1994Google Scholar).

⁵ See, for example, Kahneman, Daniel and Tversky, Amos, “Prospect Theory: An Analysis of Decisions under Risk,” Econometrica 47, no. 2 (1979CrossRefGoogle Scholar); Quattrone, George and Tversky, Amos, “Contrasting Rational and Psychological Analyses of Political Choice,” American Political Science Review 82 (September 1988CrossRefGoogle Scholar).

⁶ Their realpolitik and domestic politics models are of course highly stylized and constrained by their assumptions. Realist theorists might well reject as inaccurate Bueno de Mesquita and Lalman's characterization of their realist assumptions and logic. In that case, however, the burden shifts to the realists to provide a model they consider to be a more accurate representation. Maoz lays out reasoning about how imperfect information in the processes of foreign policy decision making can be an important force in avoiding war, not, as in the most familiar formulations, a factor raising the likelihood of war.

⁷ They do not, however, test this explanation against variants that attribute peace between democracies to other kinds of preferences or perceptions. I have partially incorporated and modified their argument and other arguments as well, in Russett, Bruce, Grasping the Democratic Peace: Principlesfor Post-Cold War World (Princeton: Princeton University Press, 1993Google Scholar), chaps. 2, 4.

⁸ Arrow, Kenneth, Social Choice and Individual Values (New Haven: Yale University Press, 1951Google Scholar); for some broad implications for democracy, see Riker, William, Liberalism against Populism (New York: W. H. Freeman, 1982Google Scholar).

⁹ See, for example, Huth, Paul and Russett, Bruce, “General Deterrence between Enduring Rivals: Testing Three Competing Models,” American Political Science Review 87 (March 1993CrossRefGoogle Scholar); Russett, , Controlling the Sword: The Democratic Governance of National Security (Cambridge: Harvard University Press, 1990CrossRefGoogle Scholar), chap. 2; Alex Mintz, “The Decision to Attack Iraq: A Noncompensatory Theory of Decision Making,” Journal of Conflict Resolution 37 (December 1993).

¹⁰ Limitations are also evident in some of the assumptions, for example, that the domestic costs of initiating use of force rise with the subjective probability of success in a dispute: people do not like to bully weaker partners, so even when they can get their way by force, they prefer to negotiate (p. 46). (The probability of success is positively correlated with the degree to which the relative power balance favors oneself.) Though important to their model, this assumption is not well defended. In the Hempel tradition of seeking “covering laws” to which they appear to subscribe, assumptions should not be unrealistic. See Bruce Bueno de Mesquita, “The War Trap Revisited: A Revised Expected Utility Model,” American Political Science Review 79 (March 1985).

A different kind of problem arises (pp. 166—67) with Bueno de Mesquita and Lalman's claim to explain an apparent anomaly, initially reported by Bueno de Mesquita, that allies are more likely to fight each other than chance would predict. See Mesquita, Bueno de, The War Trap (New Haven: Yale University Press, 1981Google Scholar). Their theory allegedly solves this puzzle, and they show that empirically the general pattern does not apply to democratic allies, who are less likely to fight. But it turns out that in an equation with proper control variables, even nondemocratic allies are less likely to engage in conflict with each other. See Bremer, Stuart A., “Dangerous Dyads: Conditions Affecting the Likelihood of Interstate War,” Journal of Conflict Resolution 36 (June 1992Google Scholar); and Maoz, Zeev and Russett, Bruce, “Normative and Structural Causes of Democratic Peace, 1946–1986,” American Political Science Review 87 (September 1993CrossRefGoogle Scholar).

Maoz notes that “the difficulty of assigning preferences over a multitude of possible situations is perhaps the main deterrent of statistical analyses of game theoretic models in international politics” (p. 405). In their empirical work Bueno de Mesquita and Lalman still use what they acknowledge to be a crude component of utility—a measure of similarity in alliance portfolios. To derive a measure of utilities for alternative outcomes, they must modify this by differences in leaders' risk propensities. The result reduces the impact of system-level calculations on decisions to initiate or escalate conflicts, without employing the context-dependent utilities that Maoz prefers. Maoz did not anticipate (p. 404) “game theoretic predictions” tested with “aggregate empirical data,” as was done by Bueno de Mesquita and Lalman, or by Fearon, James, “Domestic Political Audiences and the Escalation of International Disputes,” American Political Science Review 88 (June 1994CrossRefGoogle Scholar); and in three articles in Journal of Conflict Resolution 38 (June 1994Google Scholar): Fearon, “Signalling vs. the Balance of Power and Interests: An Empirical Test of a Crisis Bargaining Model”; John Conybeare, “Arms vs. Alliances: The Capital Structure of Military Enterprise”; and Ido Oren, “The Indo-Pakistani Arms Competition: A Deductive and Statistical Analysis.”

¹¹ Recent, very different efforts to treat preferences as variable across time and/or space include Wendt, Alexander, “Collective Identity Formation and the International State,” American Political Science Review 88 (June 1994CrossRefGoogle Scholar); Lumsdaine, David, Moral Vision in International Politics: The Foreign Aid Regime, 1949–1989 (Princeton: Princeton University Press, 1992Google Scholar); and the literature on the “democratic peace,” to which Maoz was an early contributor. See Maoz, Zeev and Abdolalli, Nasrin, “Regime Type and International Conflict,” Journal of Conflict Resolution 33 (March 1989Google Scholar).

¹² This suggests an empirical test: do Bueno de Mesquita and Lalman's empirical results fit their model's expectations more closely at lower (presumably less stressful) levels of conflict than at the high levels of war and peace decision? Another important challenge to expected utility theory is mounted by Mintz (fn. 9), who argues that even though a particular action might offer the greatest expected utility by holistic criteria, a decision maker may reject it because it fails to satisfice on a single critical dimension, such as protecting one's position in domestic politics.

¹³ Further grounds for rejection of that notion are given in recent work showing that democracies are more likely than are authoritarian regimes to win the wars on which they embark (80% of the time—not bad considering the odds of a coin toss are 50–50) and more likely to hold accountable those leaders who embark on wars they lose. See Lake, David, “Powerful Pacifists: Democratic States and Wai',” American Political Science Review 86 (March 1992CrossRefGoogle Scholar); Bruce Bueno de Mesquita, Siverson, Randolph, and Waller, Gary, “War and the Fate of Regimes: A Comparative Analysis,” American Political Science Review 86 (September 1992), 639Google Scholar–46; Bueno de Mesquita and Siverson, “War and the Survival of Political Leaders: A Comparative Analysis” (Paper presented at the annual meeting of the American Political Science Association, Washington, D.C., September 1993). Maoz later remarks, however, that people with rigid, authoritarian, cognitively simple personality structures who work in hierarchical or centralized regimes are more likely to take a strategic perspective (p. 530). (Take that, game theorists!)

¹⁴ Walter Isard makes this point more generally, that a decision maker “may wish to minimize his losses, maximize his gains, minimize his expected losses, or maximize his expected gains. Or he may want to minimize his regret.” Isard, See, Understanding Conflict and the Science of Peace (Cambridge: Blackwell, 1992), 30Google Scholar. Isard also notes the goal may be satisficing or maximizing expected utility. See also the reservations expressed by Nicholson, Michael, Rationality and the Analysis of International Conflict (Cambridge: Cambridge University Press, 1992CrossRefGoogle Scholar), esp. chaps. 3, 6.

¹⁵ Leng, Russell, Interstate Crisis Behavior, 1816–1988: Realism vs. Reciprocity (Cambridge: Cambridge University Press, 1993CrossRefGoogle Scholar); Huth and Russett (fn. 4; fn. 9).

¹⁶ Notably in the special issue on the rational deterrence debate, World Politics 41 (January 1989Google Scholar). An example of the weakness in Vasquez's approach emerges in his remark that “alliances do not prevent war or promote peace; instead they are associated with war … although they are probably not a cause of war” (p. 159). A formal model helps make much more sense out of ambiguous empirical results on this matter, as shown by Alastair Smith, “Alliance Formation and War” (Paper presented at the annual meeting of the American Political Science Association, Washington, D.C., September 1993).

¹⁷ Bueno de Mesquita and Lalman certainly do not agree. Rather, they take pains to show that their analysis is not limited to particular kinds of wars, that they consider it self-evident that it applies to rivalries (p. 243), and that they have had empirical success with all wars, not only big ones.

¹⁸ I develop these points in Russett, , “The Alleged Role of Nuclear Weapons in Controlling the Soviet-American 'Enduring Rivalry,' and in the Future,” in Gelstad, Jorn and Norstad, Olav, eds., Nuclear Technology and International Politics (London: Sage, 1994Google Scholar). The fact that the United States and the Soviet Union were not involved in any direct territorial conflict may have been an important mitigating factor. Vasquez pays great attention to territorial disputes, suggesting that “the error in realist thinking is to assume from this experience that there is a constant struggle for power operating, when in fact there may be only a struggle for contiguous territory” (p. 295). Kalevi J. Holsti contends that about half of all wars between Westphalia and World War I were over territory. Holsti, , Peace and War: Armed Conflicts and International Order, 1648–1989 (Cambridge: Cambridge University Press, 1991CrossRefGoogle Scholar).

¹⁹ Exemplars of this point of view are Kugler, Jacek and Organski, A. F. K., “The Power Transition: A Retrospective and Prospective Evaluation,” in Midlarsky, Manus, ed., Handbook of War Studies (Boston: Unwin Hyman, 1989Google Scholar); and Doran, Charles, Systems in Crisis: New Imperatives of High Politics at Century's End (Cambridge: Cambridge University Press, 1991CrossRefGoogle Scholar).

²⁰ Oddly, Vasquez also shows little interest in game theory despite its extensive applications to the realm of dyadic relations. “In the end, probably more will be learned by explaining deviations from rational behavior then by formal models themselves” (p. 219). It is hard to know quite what this means. Presumably Bueno de Mesquita and Lalman would contend that while other models of decision making, like prospect theory, can help explain anomalies, their expected utility model should be used to identify central tendencies. Maoz might say that the usefulness of different models varies across different contexts and therefore prevents any one from being “more” useful.

²¹ Maoz and Leng (fn. 13) would largely agree, particularly in the context of crisis bargaining, during which psychological and domestic political variables may override “rational” interstate power calculations.

²² Bueno de Mesquita and Lalman make the point that game theory can account for such changes, including the end of the cold war (pp. 248–50). Changes in equilibria can follow from small variations in variables. Contra John Lewis Gaddis, chaos theory or catastrophe theory is not required. See Gad-dis, “International Relations Theory and the End of the Cold War,” International Security 17 (Winter 1992–93).

²³ And democracies are readier to use those mechanisms. See Dixon, William, “Democracy and the Management of International Conflict,” Journal of Conflict Resolution 37 (March 1993CrossRefGoogle Scholar); idem, “Democracy and the Peaceful Settlement of International Conflict,” American Political Science Review 88 (March 1994Google Scholar); and Raymond, Gregory A., “Democracies, Disputes, and Third-Party Intermediaries, “Journal of Conflict Resolution 39 (March 1994Google Scholar). On the role of new norms and rules for the system, which the great powers consciously established for the Concert of Europe when they and their statesman learned that the old ones did not work, see Schroeder, Paul, The Transformation of European Politics, 1763—1848 (Oxford: Oxford University Press, 1994Google Scholar). See also Schroeder's scathing review of the historical accuracy of realists' generalizations; Schroeder, “Neo-Realist Theory and International History: An Historian's View,” International Security 19 (Summer 1994Google Scholar).