Having linked dialogue to democratic legitimacy, I have identified the forces that discourage candidates from adopting dialogueenhancing strategies. These theoretical understandings of the candidates' incentive structures were developed into a two-player game that models campaigns in mass elections. The game focuses on candidates' strategic choices over what to discuss in campaigns. In the last chapter, a case study of the 1994 California gubernatorial race tested this model. Taken together, the model and the case study show that the choice to dialogue will likely lead to electoral defeat. Now attention can be devoted to a broader study of dialogue in U.S. Senate campaigns. In the next chapter, I will refine the model in order to develop and test predictions concerning the appearance of dialogue in these races. This second step will advance the ultimate goal of promoting better campaign discourse. This chapter takes on the first step in the Senate study – an examination of campaigning as it naturally occurs.

Senate elections present a superior set of situations to further study the dynamics underlying dialogue. In the first place, they present a larger number of cases, in a given time, than any comparable race. At least thirty-three senate elections happen in every cycle. In contrast, the United States has undertaken the process of electing a president only seventeen times since World War II. At the same time, unlike elections to the House of Representatives, where the majority of elections are effectively decided prior to the campaign, senate elections generally have two carefully orchestrated and reasonably financed campaigns.