Generalizing power transitions as a cause of war

Abstract

In this thesis, I ask three questions about the nature of power transition theory. First, I ask whether power transition theory can be generalized beyond identification of great powers or regional hierarchies. Lemke and Werner introduce the concept of a multiple hierarchical order, in which mutually relevant regional powers can go to war over dissatisfaction with a regional status quo. I submit that this concept can be generalized into a continuous concept to include all states within the umbrella of the theory. Second, I ask how often status quo states initiate war in power transition cases. Jack Levy explains that status quo states have a motive to launch a preemptive war against a revisionist state, before it becomes too powerful to defeat. I submit that these motivations lead to a high incidence of status quo actor-initiated war in power transitions. Finally, I ask whether the rate of change of relative power matters during a transition period. I hypothesize that quick changes in the relative difference in power between two states would create a fast-closing window of opportunity. This closing window creates a crisis and motivates leaders to move quickly, leading to a higher probability of avoidable war. Incorporation of rate of power transition could explain war in power transition cases yet to achieve true parity, or even explain peace in a period of parity and revisionism. To test these questions, I create a large, inclusive (571,000+ N) dataset of nearly all dyads between 1821 and 2001, using the Correlates of War Composite Index of National Capabilities as the basis of power independent variables, and a composite of distance and power measurements to determine the relevance independent variable. I run a number of regressions of the power and relevance independent variables against the onset of war. I reach decisive conclusions about the nature of power dynamics in the international system, and propose their incorporation into the power transition literature. Generalized, continuous measurements of relevance, parity, and rate of change of power transition increase the explanatory power of the model; the revisionist state does not always or even usually provoke power transition war; finally, higher rates of power transition lead to a higher probability of war. The thesis ends with a number of shortfalls with the model I propose, and a number of further revisions and expansions of power transition theory.