To Remain or to Terminate? The Quest for Strategic Autonomy: An Autonomy Cost Model of Alliance Duration

Alliances are agreements among independent states to cooperate militarily in the face of potential or realized military conflict. However, not all alliances endure for decades. Existing empirical evidence suggests that there is a variance in the lifespan of alliances. I argue that alliances are more likely to be dissolved when their treaty obligations undermine autonomy in a state’s core security or military strategy. In this vein, I also insist that ‘strategic autonomy’ is a key factor in the termination of alliance and alliance duration. Alliances with strong ‘strategic autonomy constraints’ would not last longer than others. I use the ATOP data and principal components analysis (PCA) technique to test this argument to identify a strategic autonomy variable. The extracted principal components (PCs) are used as explanatory variables in regression. This paper finds that alliances with strategic autonomy constraints do not last longer than alliances that are not constrained, and the empirical results are consistent with the hypotheses.