In this study, we look at the relationship of arms races to war, with appropriate consideration of rivalries. Are arms races more common in rivalries than in lesser competitions? Are they merely a consequence of rivalry competitions? How do the patterns of arms races map with those of war in rivalries? We explore these concerns with an empirical examination of rivalry and non-rivalry populations in the 1816-2000 period. In brief, we find that: arms races occur most frequently in the context of enduring rivalries; arms races are more likely in the middle and later stages of rivalry; the frequency of arms races is higher in rivalries with war than rivalries that do not experience war; and only when arms races occur in the later phases of rivalries is there an increased chance of war. Our study narrows the scope of the arms race-war relationship relative to past studies, demonstrating that the arms race-war relationship is conditional on rivalry processes.
- Arms races
- Heckman selection model