The Insignificance of the Liberal Peace

the ~iberal Peace Doyle’s 1983 essay pointed out that no liberal democracy has ever fought a war with another liberal democracy,’ scholars have treated pacifism between democracies as ”the closest thing we have to a law in international politics,”2 and the field is nearly at the point where this ”law” is accepted without q~estion.~ The purpose of this article is to raise critical questions about this newly rediscovered law of peace among democracies. The primary question I address is whether the statistic that democracies never (or rarely) fight wars with each other is significant. Doyle supported his argument by showing that since 1816 no nations he considered to be liberal had fought wars with each other. Although he used a computerized database on interstate wars, he did not perform any probability analyses to show that zero is statistically ~ignificant.~ Zero is a powerful statistic, which David E. Spiro is an Assistant Professor of Political Science at the University of Arizona in Tucson. He is the author of Hegemony Unbound: Petrodollar Recyclingand the De-Legitimationof American Power (Zthaca: Cornell University Press, forthcoming 2995). The author thanks Edward D. Mansfield for the many discussions that motivated this study; and he is grateful for comments from Robert Art, Michael Doyle, John Matthews, John Owen, and Randall Schweller. This article was written partly under the auspices of the Institute for International Economic Studies, Tokyo, Japan. 1. Michael Doyle, “Kant, Liberal Legacies, and Foreign Affairs,” Parts I and 11, Philosophy and Public Affairs, Vol. 12, No. 3 (Summer 1983), pp. 205-235; and No. 4 (Fall 1983), pp. 323-353. The argument is summarized in Michael Doyle, ”Liberalism and World Politics,” American Political Science Review, Vol. 80, No. 4 (December 1986), pp. 1151-1169. The absence of wars among liberal democracies was noted before Doyle’s article, but it was not seen as confirmation of any particular theory. See Peter Wallenstein, Structure and War:On lnternational Relations 28201968 (Stockholm: Raben and Sjogren, 1973);and Melvin Small and J. David Singer, “The WarProneness of Democratic Regimes,” The /erusalem /ournu2 of International Relations, Vol. 1, No. 1 (Summer 1976), pp. 50-69. 2. Jack S. Levy, “Domestic Politics and War,” journal of Interdisciplinary History, Vol. 18, No. 4 3. See, for example, David Lake, ”Powerful Pacifists: Democratic States and War,” America Political Science Review, Vol. 86, No. 1(March 1992),pp. 24-37. One exception, not yet published, is Joanne Gowa, ”Democracies, Autocracies, and Foreign Policy,” mimeo, Princeton University, 1993. See also John Owen, ”How Liberalism Produces Democratic Peace,” International Security, Vol. 19, No. 2 (Fall 1994), pp. 87-125; and Christopher Layne, ”Kant or Cant: The Myth of the Democratic Peace,” ibid., pp. 5-49. 4. The computerized data set is from the Inter-university Consortium for Political and Social (1988), pp. 653-673. InternationalSecurity, Vol. 19, No. 2 (Fall 1994), pp. 50-86 0 1994 by the President and Fellows of Harvard College and the Massachusetts Institute of Technology. 50 The Insignificance of the Liberal Peace I 51 seems beyond debate or refutation, but as John Mearsheimer observed, ”democracies have been few in number over the past two centuries, and thus there have not been many cases where two democracies were in a position to fight each ~ t h e r . ” ~ As an example of how zero can be statistically insignificant, consider that people win million-dollarlotteries in the United States every day of the week, but not one single member of my immediate family has ever won one. Something can happen all of the time, and still the fact that it never happens to a certain group of individuals does not mean anything. No one needs to explore what it is about the nature of my family that prevents winning lotteries, because zero is not a significant result. It is predicted by random chance. If the absence of wars between democraciesis predicted by random chance, this is not to say that chance is a good explanation for war. It does mean that an explanation we know to be untrue-random chance-predicts the absence of war between democracies better than liberal theories of international relations, and therefore the absence of wars...