Abstract
This study presents a theoretical model of asymmetric conflicts that demonstrates that, under fairly general conditions, smaller coalitions (in terms of alliances and economic resources) benefit the most from war. This leads to a paradoxical policy implication, whereby minimizing the likelihood of a new war requires policymakers to eliminate belligerents with a lower incentive for warfare. Using a database on militarized interstate disputes over the last two centuries, I find empirical evidence supporting the hypothesis that the probability of a country attacking another increases in the wealth gap between the attacked country and the attacker.