This study investigates candidate behavior in American electoral campaigns. It centers on a question of equal importance to citizens and scholars: how can we produce better political campaigns? The project takes an innovative approach to answering this question by bringing together critical and empirical methods as well as game theory in a sophisticated yet readable text. The answer comes in four parts. First, Simon develops the idea of dialogue as a standard for evaluating political campaigns. Second, he shows that candidates' self-interest in winning leads them to avoid dialogue, that is substantive campaign discourse. Third, he demonstrates the beneficial effects produced by the little dialogue that actually occurs. Fourth, he pinpoints the forces responsible for these rare occurrences. The major lesson of this work is that campaign reform under its present guise will not bring about the more substantive campaigns that the public desires.