Arms Races and Escalation

Abstract

Although major power arms races have been the subject of a great amount of mathematical modelling, there has been little data-based research concerning their impact on international war. This study attempts to determine whether or not these arms races affect the probability that a serious dispute between major powers will escalate to all-out war. To do this, an arms race index is constructed in the following manner: a curve-fitting technique is employed to calculate changes in arms expenditures for each major power as a function of time. The smoothed rates of increase for each of the parties to a dispute are multiplied together, yielding an index whose values will be high only if the two powers have engaged in rapid and simultaneous military expansion prior to the dispute. It was found that disputes preceded by such an arms race escalated to war 23 out of 28 times, while disputes not preceded by an arms race resulted in war only 3 out of 71 times. It was concluded that at the very least, arms races are an important early warning indicator of escalation potential, and may well play a central role in the escalation process. The implications of this finding for the current debate over SALT II were noted.