The Empirical Promise of Game Theory
This is an advance summary of a forthcoming article in the Oxford Research Encyclopedia of Politics. Please check back later for the full article.
Over the last 20 years, game-theoretic approaches in international relations have coalesced around the bargaining model of war. This model suggests that peaceful bargaining should be allowed as an alternative to war, thus highlighting a fundamental puzzle: why does war occur, even though it is costly and destructive? The bargaining model of war is a significant advance over previous game-theoretic approaches and has proved to be a simple and fruitful foundation for future studies.
Yet, from the beginning, critics have questioned the empirical promise of the bargaining model. The fundamental frictions that it identified as causes of war—information and commitment problems—are always present. Thus, the argument goes, the bargaining model cannot explain why war erupted in some cases but not others.
These critiques are incorrect. To the contrary, the bargaining model of war, and game theory more generally, holds much empirical promise. By spelling out how strategic interactions may lead to conflict, game theory provides valuable insights for understanding the causal mechanisms leading to war. In turn, the fact that the bargaining model highlights fundamental frictions of the international system is one of its virtues.
An important step for game-theoretic approaches is to specify conditions under which information and commitment problems are more or less severe, and thus more or less likely to engender war. Already, game theory has provided such empirical predictions in two substantive areas—the effect of domestic politics on interstate war, and the relationship between the balance of power and conflict. Looking ahead, game theory can offer additional predictions in other substantive areas and holds much empirical promise.