de Mesquita: The War Trap Revisited

de Mesquita, B., 1985. The War Trap Revisited: A Revised Expected Utility Model.”. The American Political Science Review, 79(1), 156-177.

“An expected utility approach to the study of international politics offers both the opportunity to deduce propositions about international conflict, and, through the application of admittedly crude indicators, to evaluate the usefulness of those propositions as explanations of actual behavior” (156).

“My main objective here is to reconstruct the model so that it reflects risk through the introduction of concavity or convexity into the utility functions. In doing so, it is imperative that the model give each actor the opportunity to have a differently shaped utility function, with the extremity of the function’s curvature embodying the extremity of the decision maker’s willingness (or reluctance) to take chances” (156). The second goal of this piece deals with the relative subjectivity of actor’s perceived utility: de Mesquita attempts to build the model to avoid interpersonal comparisons.

This addition to The War Trap was then explored through game theory rationalist models using both the convex and concave curves for either risk averse or risk acceptant actors to determine the relative utility different actors put in potentially conflictual contexts.

Bennett and Stam: A Universal Test of an Expected Utility Theory of War

Bennett, DS, and AC Stam. 2000. “A Universal Test of an Expected Utility Theory of War.” International Studies Quarterly 44:451-480.

This account explores the expected utility theory of war, a rationalist approach to modeling the interaction of states that stems from the work of Bueno de Mesquita and Lalman. In War and Reason, these authors explore the International Interaction Game (IIG), but only on 707 dyads and only for Europe. These authos attempt to expand upon that approach.

“Rational choice applications to war initiation begin with the assumption that states can be modeled as rational actors who make choices about war and peace by assessing the costs and benefits of alternative actions” (452).

“The IIG posits a series of interactive decision paths that lead to a set of eight different possible outcomes for any dyadic relationship at any given moment. Given adequate data, a prediction can be made of the expected outcome in equilibrium from any given interstate dyadic interaction” (452). It is assumed that this model “holds up” against space and time.

“Overall, our results support the prior findings of Bueno de Mesquita and Lalman that the game-theoretic predictions of the IIG and the empirical measures developed to test them successfully predict behavior” (477).