Arms Races and the Outbreak of War, 1816-1980.

Abstract

This study investigates the impact that arms races have on the probability of major power rivals going to war. Twenty-two enduring rivalries in the years 1816-1980 are the focus of the study. Correlates of War Project data are used to construct an index measuring the intensity of an arms race. Controls for relative capabilities, military allocations, and contiguity are instituted in order to isolate the impact that military buildups have on war. Military buildups were found to exercise little direct impact on the likelihood that a militarized dispute between the rivals would escalate to war. Wars broke out in rivalries after at least two prior militarized disputes, regardless of the presence of an arms race or any other condition. Military buildups were critical, however, in affecting the relative capabilities of the contending nations. Parity was the escalatory condition for rivalries in the 19th century, while preponderance precipitated war in the 20th century. Arms races and unilateral buildups were dangerous when they caused a shift to the century's escalatory condition or inhibited a shift away from that condition. The escalatory capability distributions were necessary, but not sufficient, conditions for war. War resulted from a militarized dispute in a rivalry only if the dispute site was contiguous to one of the rivals and the military allocation ratio of the dispute target was high. The requirement for contiguity indicates that the salience of the dispute issue was important in determining peace or war. The analysis of military allocations demonstrated that one of the rivals had to be adequately prepared (often as a result of a military buildup) to fight before war broke out.