Powell: War as a Commitment Problem

Powell, Robert. 2006. War as a Commitment Problem. International Organization.

In Brief

Powell argues that war is either a commitment problem or an informational problem (indivisibility problems don't exist). But commitment problems are often poorly defined and categorized, particularly in the case of prolonged conflict. Powell discusses the types of commitment problems and some models of conflict based on them.

Place in the Literature

Powell builds upon Fearon's ^{vague citation} work on conflict as a failure of a bargaining game.

Limitations of Informational Explanations

1. Asymetric information may explain the outbreak of conflict, but it is not very good at explaining prolonged conflict
2. A satisfied state ALWAYS prefers appeasing a dissatisfied adversary state no matter how large the concession it takes to satisfy the dissatisfied state. This creates odd implications and interpretations of clift (eg the German decision to attack Britain was based precisely on the idea that Britain was going to resist). Thus in certain cases, fighting ensues when the resolution of uncertainty reveals that a state is facing an adversary it would rather fight than accommodate. This is not well modeled in the standard informational account.

Bargaining Indivisibility

Bargaining indivisibility does not adequately explain war since even if an issue is indivisible there are still agreements both sides prefer to fighting. The problem is that states cannot commit to these agreements. Both sides to a conflict merely need to find a bargaining game that provides them with similar odds to a warfare game, which would then avoid the significant non-recoverable costs of war (eg mediation). The problem is that it is extraordinarily to commit to such a mechanism.

Preventive War

Powell argues that the three general types of preventive war--preventive war caused by anticipated shift in power distribution, preventive war caused by first-strike advantaged, and preventive ward caused by concessions which shift the military balance--are all part of the same general strategic problem, in that all the bargainers are trying to divide a flow of benefits in a situation in which (1) it is impossible to commit to a future division of benefits, (2) each actor has the option of locking in a share of the flow at time t, (3) the use of power is inefficient in that it destroys some of the flow, and (4) the distribution of power, and thus the amounts actors can lock in, shifts over time.

1. Fearon has shown that in an infinitely repeating game in which changes in capabilities are constant, there is a non-war equilibrium. Powell argues, in contrast, that when there is a rapid shift in capabilities (or an anticipated one anyway) the equilibrium is still war absent commitment.
2. the costs associated with appeasing domestic coalitions can change the game significantly (essentially turning war bargaining into a two stage game in which there is some positive probability of losing power, cf Bueno de Mesquita ^{vague citation}).
3. Costs of preserving the status quo, e.g. political costs of deterring USSR may be such that we have a dictatorship, can be significant.