Capability Distribution between Allies, Geographical Proximity and Alliance Duration

This study explores how geographical proximity and power distribution between allies interact and affect alliance duration. Morrow (1991), based on the security-autonomy trade-off model, claims that asymmetric alliances are likely to last longer than symmetric ones. This paper, however, argues that the effect of the relative power distribution between allies on alliance duration can be conditional upon the distance between allies because geographical proximity between allies may serve to aggravate or dampen the autonomy concerns of allied states—for example, one ally may be concerned about the mounting influence of its alliance partner and this concern could be severe in case of unequal alliances because of presumably the overwhelming influence of the former on the latter, but this kind of concern may be alleviated if the two allies are geographically remote. By contrast, in case of symmetric alliances, the equal power status between allies may relieve such autonomy concerns but geographical remoteness between them may impede the effective and efficient operation of alliances. Hence, this study hypothesizes that geographically remote and unequal alliances are more likely to last longer than geographically close and unequal alliances and that geographically close and equal alliances are more likely to last longer than geographically remote and equal alliances. These hypotheses are put to test against the allied-dyad data covering the period of 1816-2002 and I find strong support for these hypotheses. However, the findings in the robustness check suggest that these findings be accepted with some reservation especially in the case of bilateral alliances.