Stability and the Distribution of Power

T relation between the distribution of power and the probability of war has been at the center of a long and important debate in international relations theory. Is an even distribution of power more or less stable than a preponderance of power? The balance-of-power school generally argues that an even distribution of power is more stable, whereas the preponderance-of-power school generally argues the opposite, that a preponderance of power is more stable.1 Empirical efforts to resolve this debate have yielded equivocal results. Siverson and Tennefoss, for example, find an even distribution of power to be more stable, whereas Kim and Moul find a preponderance of power to be more stable. Singer, Bremer, and Stuckey find an even distribution to be more stable than a preponderance of power during the nineteenth century and the opposite during the twentieth century. Maoz and Bueno de Mesquita and Lalman, by contrast, find no significant relation between the probability of war and the distribution of power. Finally, Mansfield finds evidence of a nonlinear, quadratic relation in which the probability of war is smallest when there is an even distribution and a preponderance of power.2