Ethnic war as a commitment problem

TitleEthnic war as a commitment problem
Publication TypeConference Paper
AuthorsFearon, JD


Increased ethnic violence in Eastern Europe may be the most striking immediate consequence of the end of the Cold War and the break up of the Soviet Union. The standard explanation for this outcome,  found universally in newspapers and widely in the opinions of the experts they cite, is that the new ethnic violence is the product of age-old, primordial ethnic hatreds that were "suppressed" in the communist era.  Even where the outcome has not been violence or massacre but simply increased ethnic strife, as in the Baltic states, Czechoslovakia, and Hungary, the explanation given is basically the same: age-old hatreds, resentments, and feuds that can now be openly expressed.
On slightly closer examination, however, the "age-old hatreds" explanation does not hold up.
This paper offers of an alternative explanation, and briefly considers evidence from the Yugoslav case to provide a (very partial) empirical evaluation. The alternative explains the surge of ethnic violence as the result of a commitment problem that arises when two political communities nd themselves without a third party that can guarantee agreements between them. The problem is that in post-Soviet Eastern Europe, ethnic majorities are unable to commit themselves not to exploit ethnic minorities in a new state. Ethnic minorities, such asthe Serbs in Croatia, Armenians in Azerbaijan, and possibly Ossetians in Georgia, anticipate that regardless of what the ethnic majority's leaders agree to now, there is no solid guarantee that the leaders will not renege in the future due to the play of majority politics in the new state to come. Given this anticipation, fighting now in hopes of winning secession from a weak, barely formed state may appear the superior alternative. However, the ethnic war that ensues may leave both majority and minority worse off than if the majority could make a credible commitment not to abuse the minority in the new state.
Put dierently, I am arguing that ethnic violence might be protably understood as a species of preventive war, and that the real problem of preventive war is the inability to make commitments in an anarchic environment. For many international relations scholars, it is an article of faith that anarchy and the "security dilemma" that is said to ensue are root causes of international conflict.
Remarkably, however, the field has not identied specic mechanisms by which anarchy produces violence -- we lack arguments that directly link the
absence of a power that can guarantee agreements to war. Under anarchy, nothing stops states from using force if they wish. But if using force necessarily entails costs, then it is not clear why or how anarchy should prevent states from locating peaceful bargains that would avoid the price of ghting.
This paper offers a specic mechanism explaining how anarchy can make it impossible for states to negotiate an agreement that would avoid the costs of war. Anarchy implies that a rising state is unable to commit itself not to exploit the greater leverage it will have in future bargaining. This prospect of a worse peace in the future can make it reasonable for a declining state to fight now, even though both states would prefer a peaceful (but incredible) bargain in which the rising state agreed to restrain its demands in the future. The mechanism is essentially the same in the case of ethnic violence. Ethnic majorities cannot credibly commit themselves not to exploit the greater bargaining leverage they will have against ethnic minorities once the new state has consolidated. From this perspective, ethnic war does not appear as something entirely distinct from international war.
The paper has three major sections. The first briefly considers a preliminary methodological question: What is the point or value of thinking about ethnic violence from a rationalist perspective? The second section develops a simple game model of the commitment problem faced by ethnic groups in plural societies, showing exactly how the logic sketched above operates. The model also generates a set of hypotheses on when the commitment problem is more or less likely to be resolved (that is, what makes ethnic war more or less likely). The key factors prove to be (1) the size of the expected change in relative military power between groups that would result from formation of a new state; (2) the relative size of the ethnic minority; (3) whether majority and minority groups costs for fighting are low, as may occur if they are more rural than urban and if they are not strongly interdependent in the economic terms; and (4) whether institutions can be created that give minority groups political power that is at least proportional to their numbers.
The third section of the paper considers how the 1992 war in Croatia came about, arguing that the commitment problem was a principal cause of the conflict. A few other Eastern European cases and some policy implications are considered in the conclusion.