Arms Races and War Initiation

Abstract

Intriligator and Brito argue that certain arms races can lead to peace and certain disarming races can lead to war. They support this position with a formal model of the relationship between weapons levels and war initiation. This article evaluates and criticizes the Intriligator-Brito model. The results of their model are shown to depend upon questionable assumptions both about the conditions under which nuclear attack might be initiated or deterred and about the nature of nuclear war itself. Two sets of alternatives to the Intriligator-Brito (I-B) model are developed. In contrast to the Intriligator-Brito analysis, most of the topologies generated by the alternative models suggest that arms races are unlikely to yield stable peace. The issue of disarmament is more complicated. If security is identified with the ability to inflict painful retribution on the enemy, then situations of near complete disarmament are necessarily hazardous. However, the main implication of the alternative models is that mutual disarmament—at least mutual disarmament to moderate weapons levels—will not increase the chance of war and may well diminish it. At least one of the alternative models points out that no situation of mutual deterrence may be possible. Arms races may not always lead to war nor disarmament to peace, but considerable circumspection is needed when drawing policy inferences from war initiation models of the Intriligator-Brito genre because the results are highly sensitive to the underlying assumptions.