9308139 Myerson The goal of this project is to develop a better understanding of how the design of political institutions can affect political behavior, with hope that such understanding may be valuable for guiding decisions about constitutional design in new and reforming democracies. In particular, the researcher will continue his research on game-theoretic comparison of political institutions by looking into several extensions. First, electoral systems will be compared by analyzing large finite voting games in which the numbers of voters of different types are independent random variables. Then the effectiveness of different electoral systems for reducing government corruption by considering single transferable vote and other electoral systems will be investigated. Second, the tendency of plurality voting elections to become two-party contests is known as Duverger's law. The investigator will work to derive Duverger's law from game-theoretic analysis. Third, a variety of game-theoretic models that may account for observed differences between parliamentary and presidential systems will be studied. This question has been an important area of study in the empirical literature on comparative politics, but there is a need for game-theoretic models to guide and complement empirical work. Finally, the investigator's long term plan is to integrate the results of his research, together with other related results in formal political theory and social choice theory, into a general monograph on game-theoretic models for comparing political institutions. ***
