An earlier version of this article was presented at the annual meetings of the American Political Science Association, Washington, D.C., 2–5 September 1993. The article draws in part on chapter 1 of James D. Fearon, “Threats to Use Force: Costly Signals and Bargaining in International Crises,” Ph.D. diss., University of California, Berkeley, 1992. Financial support of the Institute on Global Conflict and Cooperation of the University of California is gratefully acknowledged. For valuable comments I thank Eddie Dekel, Eric Gartzke, Atsushi Ishida, Andrew Kydd, David Laitin, Andrew Moravcsik, James Morrow, Randolph Siverson, Daniel Verdier, Stephen Walt and especially Charles Glaser and Jack Levy.

^1. Of course, arguments of the second sort may and often do presume rational behavior by individual leaders; that is, war may be rational for civilian or military leaders if they will enjoy various benefits of war without suffering costs imposed on the population.While I believe that “second-image” mechanisms of this sort are very important empirically, I do not explore them here. A more accurate label for the subject of the article might be “rational unitary-actor explanations,” but this is cumbersome.

^2. For the founding work of neorealism, see Waltz, Kenneth, Theory of International Politics (Reading, Mass.: Addison-Wesley, 1979)Google Scholar. For examples of theorizing along these lines, see Jervis, Robert, “Cooperation Under the Security Dilemma,” World Politics 30 (01 1978), pp. 167–214CrossRefGoogle Scholar; Walt, Stephen, The Origins of Alliances (Ithaca, N.Y.: Cornell University Press, 1987)Google Scholar; Mearsheimer, John J., “Back to the Future: Instability in Europe After the Cold War,” International Security 15 (Summer 1990), pp. 5–56CrossRefGoogle Scholar; and Glaser, Charles, “Realists as Optimists: Cooperation as Self-Help,” International Security 19 (Winter 1994/1995), pp. 50–90CrossRefGoogle Scholar.

^3. The sense of “mechanism” is similar to that proposed by Elster, although somewhat broader. See Elster, Jon, Political Psychology (Cambridge: Cambridge University Press, 1993), pp. 1–7CrossRefGoogle Scholar; and Elster, Jon, Nuts and Bolts for the Social Sciences (Cambridge: Cambridge University Press, 1989)CrossRefGoogle Scholar, chap. 1.

^4. For an influential example of this common assumption see Snyder, Glenn and Diesing, Paul, Conflict Among Nations (Princeton, N.J.: Princeton University Press, 1977)Google Scholar.

^5. See, for examples, Blainey, Geoffry, The Causes of War (New York: Free Press, 1973)CrossRefGoogle Scholar; Howard, Michael, The Causes of Wars (Cambridge, Mass.: Harvard University Press, 1983)Google Scholar, especially chap. 1; and Stein, Arthur, Why Nations Cooperate: Circumstance and Choice in International Relations (Ithaca, N.Y.: Cornell University Press, 1990), pp. 60–64Google Scholar. Even the case of World War I is contested; an important historical school argues that this was a wanted war. See Fisher, Fritz, Germany's Aims in the First World War (New York: Norton, 1967)Google Scholar.

^6. The quotation is drawn from Waltz, Kenneth, Man, the State, and War: A Theoretical Analysis (New York: Columbia University Press, 1959), p. 188Google Scholar.

^7. For a careful analysis and critique of this standard argument on the difference between the international and domestic arenas, see Wagner, R. Harrison, “The Causes of Peace,” in Licklider, Roy A., ed., Stopping the Killing: How Civil Wars End (New York: New York University Press, 1993), pp. 235–68Google Scholar and especially pp. 251–57.

^8. See Herz, John H., “Idealist Internationalism and the Security Dilemma,” World Politics 2 (01 1950), pp. 157–80CrossRefGoogle Scholar; and Jervis, “Cooperation Under the Security Dilemma.” Anarchy is implicated in the security dilemma externality by the following logic: but for anarchy, states could commit to use weapons only for nonthreatening, defensive purposes.

^9. Jervis, “Cooperation Under the Security Dilemma.”

^10. For an analysis of the security dilemma that takes into account signaling, see Kydd, Andrew, “The Security Dilemma, Game Theory, and World War I,” paper presented at the annual meeting of the American Political Science Association, Washington, D.C., 2–5 09 1993Google Scholar.

^11. The most developed exception I know of is found in Van Evera, Stephen, “Causes of War,” Ph.D. diss., University of California, Berkeley, 1984, pp. 61–64Google Scholar.

^12. See de Mesquita, Bruce Bueno, The War Trap (New Haven, Conn.: Yale University Press, 1981)Google Scholar, and “The War Trap Revisited: A Revised Expected Utility Model,” American Political Science Review 79 (03 1985), pp. 157–76Google Scholar. For a generalization that introduces the idea of a bargaining range, see Morrow, James D., “A Continuous-Outcome Expected Utility Theory of War,” Journal of Conflict Resolution 29 (09 1985), pp. 473–502CrossRefGoogle Scholar. Informal versions of the expected utility argument are everywhere. For example, Waltz's statement that “A state will use force to attain its goals if, after assessing the prospects for success, it values those goals more than it values the pleasures of peace” appears in different ways in a great many works on war. See Waltz, , Man, the State, and War, p. 60Google Scholar.

^13. See, for classic examples, Thucydides, , The Peloponnesian War (New York: Modern Library, 1951), pp. 45 and 48Google Scholar; and von Clausewitz, Carl, On War (Princeton, N.J.: Princeton University Press, 1984), p. 85Google Scholar.

^14. Bueno de Mesquita, The War Trap.

^15. A proof is given in the Appendix.

^16. On the theory of bargaining with outside options, see Osborne, Martin J. and Rubinstein, Ariel, Bargaining and Markets (New York: Academic Press, 1990)Google Scholar, chap. 3; Perry, Motty, “An Example of Price Formation in Bilateral Situations,” Econometrica 50 (03 1986), pp. 313–21CrossRefGoogle Scholar; and Powell, Robert, “Bargaining in the Shadow of Power” (University of California, Berkeley, 1993, mimeographed)Google Scholar. See also the analyses in Wagner, R. Harrison, “Peace, War, and the Balance of Power,” American Political Science Review 88 (09 1994), pp. 593–607CrossRefGoogle Scholar; and Wagner, “The Causes of Peace.”

^17. See, for example, Luard, Evan, War in International Society (New Haven, Conn.: Yale University Press, 1992), p. 191Google Scholar. Schroeder notes that “patronage, bribes, and corruption” were “a major element” of eighteenth-century international relations. See Schroeder, Paul, The Transformation of European Politics, 1763–1848 (Oxford: Oxford University Press, 1994), p. 579Google Scholar.

^18. On Cuba, see May, Ernest, Imperial Democracy (New York: Harper and Row, 1961), pp. 149–50Google Scholar. On the Louisiana purchase, military threats raised in the U.S. Senate apparently made Napoleon more eager to negotiate the sale. See Lyon, E. Wilson, Louisiana in French Diplomacy (Norman: University of Oklahoma Press, 1934), pp. 179 and 214ffGoogle Scholar.

^19. Gambetta, Diego, The Sicilian Mafia: The Business of Private Protection (Cambridge, Mass.: Harvard University Press, 1993), p. 214Google Scholar.

^20. In one of the only articles on this problem, Morrow proposes a private information explanation for states’ failures to link issues in many disputes. See Morrow, James D., “Signaling Difficulties with Linkage in Crisis Bargaining,” International Studies Quarterly 36 (06 1992), pp. 153–72CrossRefGoogle Scholar.

^21. See Levy, Jack, “The Causes of War: A Review of Theories and Evidence,” in Tetlock, Philip E. et al. , eds., Behavior, Society, and Nuclear War, vol. 1 (Oxford: Oxford University Press, 1989), pp. 209–333Google Scholar. Recent work using limited-information game theory to analyze crisis bargaining places the strategic consequences of private information at the center of the analysis. See, for examples, de Mesquita, Bruce Bueno and Lahnan, David, War and Reason (New Haven, Conn.: Yale University Press, 1992)Google Scholar; Fearon, James D., “Domestic Political Audiences and the Escalation of International Disputes,” American Political Science Review 88 (09 1994), pp. 577–92CrossRefGoogle Scholar; Morrow, James D., “Capabilities, Uncertainty, and Resolve: A Limited Information Model of Crisis Bargaining,” American Journal of Political Science 33 (11 1989), pp. 941–72CrossRefGoogle Scholar; Nalebuff, Barry, “Brinksmanship and Nuclear Deterrence: The Neutrality of Escalation,” Conflict Management and Peace Science 9 (Spring 1986), pp. 19–30CrossRefGoogle Scholar; and Powell, Robert, Nuclear Deterrence Theory: The Problem of Credibility (Cambridge: Cambridge University Press, 1990)CrossRefGoogle Scholar.

^22. Blainey, , The Causes of War, p. 246Google Scholar.

^23. Ibid., p. 54. Blainey also blames patriotic and nationalistic fervor, leaders’ (irrational) tendency to surround themselves with yes-men, and crowd psychology.

^24. See White, Ralph K., Nobody Wanted War: Misperception in Vietnam and Other Wars (New York: Doubleday/Anchor)Google Scholar, chap. 7; Blainey, , The Causes of War, p. 54Google Scholar; and Lebow, Richard Ned, Between Peace and War: The Nature of International Crises (Baltimore, Md.: Johns Hopkins University Press, 1981), p.247Google Scholar.

^25. Harsanyi, John C., “Games with Incomplete Information Played By ‘Bayesian’ Players, Part III,” Management Science 14 (03 1968), pp. 486–502CrossRefGoogle Scholar.

^26. Aumann observed an interesting implication of this doctrine: genuinely rational agents cannot “agree to disagree,” in the sense that it cannot be commonly known that they are rational and that they hold different estimates of the likelihood of some uncertain event. See Aumann, Robert, “Agreeing to Disagree,” The Annals of Statistics 4 (11 1976), pp. 1236–39CrossRefGoogle Scholar Emerson Niou, Peter Ordeshook, and Gregory Rose note that this implies that rational states cannot agree to disagree about the probability that one or the other would win in a war in The Balance of Power: Stability in the International System (Cambridge: Cambridge University Press, 1989), p. 59Google Scholar.

^27. On bounded rationality, see Simon, Herbert A., “A Behavioral Model of Rational Choice,” Quarterly Journal of Economics 69 (02 1955), pp. 99–118CrossRefGoogle Scholar.

^28. This analysis runs exactly parallel to work in law and economics on pretrial bargaining in legal disputes. Early studies explained costly litigation as resulting from divergent expectations about the likely trial outcome, while in more recent work such expectations derive from private information about the strength of one's case. For a review and references, see Cooter, Robert D. and Rubinfeld, Daniel L., “Economic Analysis of Legal Disputes and Their Resolution,” Journal of Economic Literature 27 (09 1989), pp. 1067–97Google Scholar.

^29. Blainey, The Causes of War.

^30. This take-it-or-leave-it model of international bargaining is proposed and analyzed under conditions of both complete and incomplete information in Fearon, James D., “Threats to Use Force: The Role of Costly Signals in International Crises,” Ph.D. diss., University of California, Berkeley, 1992Google Scholar, chap. 1. Similar results for more elaborate bargaining structures are given in my own work in progress. See Fearon, James D., “Game-Theoretic Models of International Bargaining: An Overview,” University of Chicago, 1995Google Scholar. Powell has analyzed an alternative model in which both sides must agree if the status quo is to be revised. See Powell, “Bargaining in the Shadow of Power.”

^31. See Claim 2 in the Appendix.

^32. For examples and discussion on this point, see Fearon, “Threats to Use Force,” chap. 3.

^33. See the Appendix for proofs of these claims. Cheap talk announcements can affect outcomes in some bargaining contexts. For an example from economics, see Farrell, Joseph and Gibbons, Robert, “Cheap Talk Can Matter in Bargaining,” Journal of Economic Theory 48 (06 1989), pp. 221–37CrossRefGoogle Scholar. These authors show how cheap talk might credibly signal a willingness to negotiate seriously that then affects subsequent terms of trade. For an example from international relations, see Morrow, James D., “Modeling the Forms of International Cooperation: Distribution Versus Information,” International Organization 48 (Summer 1994), pp. 387–423CrossRefGoogle Scholar.

^34. The conclusion is likewise altered if the possibility of repeated interactions in sufficiently similar contexts is great enough that reputation building can be supported.

^35. On signaling costs in crises and audience costs in particular, see Fearon, “Threats to Use Force,” and “Domestic Political Audiences and the Escalation of International Disputes.” For an excellent analysis of international signaling in general, see Jervis, Robert, The Logic of Images in International Relations (Princeton, N.J.: Princeton University Press, 1970)Google Scholar.

^36. For developed models that make this point, see James Fearon, “Deterrence and the Spiral Model: The Role of Costly Signals in Crisis Bargaining,” paper presented at the annual meeting of the American Political Science Association, 30 August-2 September 1990, San Francisco, Calif.; Fearon, “Domestic Political Audiences and the Escalation of International Disputes”; Morrow, “Capabilities, Uncertainty, and Resolve”; Nalebuff, "Brinkmanship and Nuclear Deterrence”; and Powell, Nuclear Deterrence Theory.

^37. Albertini, Luigi, The Origins of the War of 1914, vol. 2 (London: Oxford University Press, 1953), pp. 183–87Google Scholar.

^38. Ibid., p. 187.

^39. Ibid., p. 158. For the full text of the cable, see Kautsky, Karl, comp., German Documents Relating to the Outbreak of the World War (New York: Oxford University Press, 1924), doc. no. 71, p. 130Google Scholar.

^40. Jarausch, Konrad, “The Illusion of Limited War: Chancellor Bethmann Hollweg's Calculated Risk,” Central European History 2 (03 1969), pp. 48–76CrossRefGoogle Scholar. The quotation is drawn from p. 65.

^41. See Turner, L. C. F., Origins of the First World War (New York: Norton, 1970), p. 101Google Scholar; and Jarausch, , “The Illusion of Limited War”, p. 63Google Scholar. Trachtenberg writes that “one of Bethmann's basic goals was for Germany to avoid coming across as the aggressor.” See Trachtenberg, Marc, History and Strategy (Princeton, N.J.: Princeton University Press, 1991), p. 90Google Scholar.

^42. Albertini concludes that “on the evening of the 27th all the Chancellor sought to do was to throw dust in the eyes of Grey and lead him to believe that Berlin was seriously trying to avert a conflict, that if war broke out it would be Russia's fault and that England could therefore remain neutral.” See Albertini, , The Origins of the War of 1914, vol. 1, pp. 444–45Google Scholar. See also Turner, , Origins of the First World War, p. 99Google Scholar.

^43. See White, J. A., The Diplomacy of the Russo-Japanese War (Princeton, N.J.: Princeton University Press, 1964), pp. 142–43CrossRefGoogle Scholar; and Nish, Ian, The Origins of the Russo-Japanese War (London: Longman, 1985), pp. 241–42Google Scholar.

^44. Westwood, J. N., Russia Against Japan, 1904–5: A New Look at the Russo–Japanese War (Albany: State University of New York Press, 1986), p. 22CrossRefGoogle Scholar. Estimates varied within the Japanese leadership, but with the exception of junior-level officers, few seem to have been highly confident of victory. For example, as the decision for war was taken the Japanese navy requested a two-week delay to allow it to even the odds at sea. See Nish, , The Origins of the Russo–Japanese War, pp. 197–200 and 206–7Google Scholar.

^45. See, for example, Walder, David, The Short Victorious War: The Russo–Japanese Conflict, 1904–5 (London: Hutchinson, 1973), pp. 53–56Google Scholar; and Nish, , The Origins of the Russo–Japanese War, p. 253Google Scholar.

^46. See White, The Diplomacy of the Russo–Japanese War, chaps. 6–8; Nish, , The Origins of the Russo–Japanese War, p. 241Google Scholar; and Lebow, , Between Peace and War, pp. 244–46Google Scholar.

^47. White, , The Diplomacy of the Russo–Japanese War, p. 139Google Scholar. Nish writes that “many Russians certainly took a view of [the Japanese military] which was derisory in comparison with themselves. It may be that this derived from a deliberate policy of secrecy and concealment which the Japanese army applied because of the historic coolness between the two countries.” See Nish, , The Origins of the Russo–Japanese War, p. 241Google Scholar.

^48. The British were the major exception, who as recent allies of Japan had better knowledge of its capabilities and level of organization. See Nish, , The Origins of the Russo–Japanese War, p. 241Google Scholar.

^49. See, for example, Langer, William, “The Origins of the Russo–Japanese War,” in Schorske, Carl and Schorske, Elizabeth, eds., Explorations in Crisis (Cambridge: Harvard University Press, 1969), p. 44CrossRefGoogle Scholar.

^50. For some examples, see Fearon, “Threats to Use Force,” chap. 3. For a formal version of reputational dynamics due to private information, see Nalebuff, Barry, “Rational Deterrence in an Imperfect World,” World Politics 43 (04 1991), pp. 313–35CrossRefGoogle Scholar.

^51. See, for examples, Jervis, Robert, Perception and Misperception in International Politics (Princeton, N.J.: Princeton University Press, 1976), pp. 62–67Google Scholar; Posen, Barry, The Sources of Military Doctrine (Ithaca, N.Y.: Cornell University Press, 1984), pp. 16–17Google Scholar; and Glaser, Charles, “The Political Consequences of Military Strategy,” World Politics 44 (07 1992), p. 506CrossRefGoogle Scholar.

^52. For dynamic game models that demonstrate this, see Powell, Robert, “Absolute and Relative Gains in International Relations Theory,” American Political Science Review 85 (12 1991), pp. 1303–20CrossRefGoogle Scholar; and Fearon, James D., “Cooperation and Bargaining Under Anarchy,” (University of Chicago, 1994, mimeographed)Google Scholar. On tit-for-tat and the impact of the shadow of the future, see Axelrod, Robert, The Evolution of Cooperation (New York: Basic Books, 1984)Google Scholar; and Oye, Kenneth, ed., Cooperation Under Anarchy (Princeton, N.J.: Princeton University Press, 1986)Google Scholar.

^53. This argument about military variance runs counter to the usual hypothesis that offensive advantages foster war. For a discussion and an empirical assessment, see Fearon, James D., “Offensive Advantages and War since 1648,” paper presented at the annual meeting of the International Studies Association, 21–25 02 1995Google Scholar. On the offense-defense balance and war, see Jervis, “Cooperation Under the Security Dilemma”; and Van Evera, “Causes of War,” chap. 3.

^54. For the argument about leaders' views on first-strike advantages in 1914, see Van Evera, Stephen, “The Cult of the Offensive and the Origins of the First World War,” International Security 9 (Summer 1984), pp. 58–107CrossRefGoogle Scholar.

^55. See, for example, Trachtenberg, History and Strategy, p. 90.

^56. This is suggested by results in Myerson, Roger and Satterthwaite, Mark, “Efficient Mechanisms for Bilateral Trading,” Journal of Economic Theory 29 (04 1983), pp. 265–81CrossRefGoogle Scholar.

^57. Schelling suggested that efficient coordination in stag hunt-like preemption problems might be prevented by a rational dynamic of “reciprocal fear of surprise attack.” See Schelling, Thomas, The Strategy of Conflict (Cambridge, Mass.: Harvard University Press, 1960)Google Scholar, chap. 9. Powell has argued that no such dynamic exists between rational adversaries. See Powell, Robert, “Crisis Stability in the Nuclear Age,” American Political Science Review 83 (03 1989), pp. 61–76CrossRefGoogle Scholar.

^58. Taylor, , The Struggle for Mastery in Europe, 1848–1918 (London: Oxford University Press, 1954), p. 166Google Scholar. Carr held a similar view: “The most serious wars are fought in order to make one's own country militarily stronger or, more often, to prevent another country from becoming militarily stronger.” See Carr, E. H., The Twenty Years' Crisis, 1919–1939 (New York: Harper and Row, 1964), pp. 111–12Google Scholar.

^59. To my knowledge, Van Evera is the only scholar whose treatment of preventive war analyzes at some length how issues of credible commitment intervene. The issue is raised by both Snyder and Levy. See Evera, Van, “Causes of War,” pp. 62–64Google Scholar; Snyder, Jack, “Perceptions of the Security Dilemma in 1914,” in Jervis, Robert, Lebow, Richard Ned, and Stein, Janice Gross, eds., Psychology and Deterrence (Baltimore, Md.: Johns Hopkins University Press, 1985), p. 160Google Scholar; and Levy, Jack, “Declining Power and the Preventive Motivation for War,” World Politics 40 (10 1987), p. 96CrossRefGoogle Scholar.

^60. According to Hiro, President Bush’s main concern at the first National Security Council meeting following the invasion of Kuwait was the potential increase in Iraq's economic leverage and its likely influence on an “already gloomy” U.S. economy. See Hiro, Dilip, Desert Shield to Desert Storm: The Second Gulf War (London: Harper-Collins, 1992), p. 108Google Scholar.

^61. On compensation, see Gulick, Edward V., Europe's Classical Balance of Power (New York: Norton, 1955), pp. 70–72Google Scholar; and Schroeder, Paul W., The Transformation of European Politics, 1763–1848, pp. 6–7Google Scholar.

^62. See Trachtenberg, , History and Strategy, pp. 56–59Google Scholar; Albertini, , The Origins of the War of 1914, vol. 2, pp. 129–30Google Scholar; Turner, Origins of the First World War, chap. 4; Joll, James, The Origins of the First World War (London: Longman, 1984), p. 87Google Scholar; and Evera, Van, “The Cult of the Offensive and the Origins of the First World War,” pp. 79–85Google Scholar.

^63. Williamson, Samuel, “The Origins of World War I,” Journal of Interdisciplinary History (Spring 1988), pp. 795–818 and pp. 797–805CrossRefGoogle Scholar in particular; and Lieven, D. C. B., Russia and the Origins of the First World War (New York: St. Martins, 1983), pp. 38–49CrossRefGoogle Scholar.

^64. Levy, Jack S., “Preferences, Constraints, and Choices in July 1914,” International Security 15 (Winter 1990/1991), pp. 234–36CrossRefGoogle Scholar.

^65. Levy argues that preventive considerations are rarely themselves sufficient to cause war. See Levy, “Declining Power and the Preventive Motivation for War.”

^66. See for example Holsti, Kalevi J., Peace and War: Armed Conflicts and International Order 1648–1989 (Cambridge: Cambridge University Press, 1991)CrossRefGoogle Scholar; and Vasquez, John, The War Puzzle (Cambridge: Cambridge University Press, 1993)CrossRefGoogle Scholar.

^67. The argument is formalized in work in progress by the author, where it is shown that the conditions under which war will occur are restrictive: the states must be unable to continuously adjust the odds of victory by dividing up and trading the land. In other words, the smallest feasible territorial transfer must produce a discontinuously large change in a state's military chances for war to be possible. See also Wagner, , “Peace, War, and the Balance of Power,” p. 598Google Scholar, on this commitment problem.

^68. See Jakobson, Max, The Diplomacy of the Winter War: An Account of the Russo-Finnish Conflict, 1939–1940 (Cambridge, Mass.: Harvard University Press, 1961), pp. 135–39Google Scholar; and Evera, Van, “Causes of War,” p. 63Google Scholar. Private information and incentives to misrepresent also caused problems in the bargaining here. See Fearon, “Threats to Use Force,” chap. 3.

^69. See Fudenberg, Drew and Tirole, Jean, Game Theory (Cambridge, Mass.: MIT Press, 1991)Google Scholar, chap. 8.

^70. This condition is satisfied for a broad range of distributions. See Fudenberg, and Tirole, , Game Theory, p. 267Google Scholar.

^71. The assumption that type c_B=x - p chooses not to fight is immaterial.