Doctor of Philosophy, The Ohio State University, 2004, Political Science

Expectations are an important part of our understanding of presidential primaries. Whether we look at expectations in terms of how they drive momentum for some candidates, or as a component of expected utility in the individual decisions of primary voters, the chances of a candidate winning either the party nomination (viability) or the general election (electability) are key variables in understanding their success. Viability and electability are influenced over a long period of time leading up to the primaries, and continue to develop throughout the primary season. By adopting a theory of rational expectations, I look at how expectations change in reaction to information that is made available to voters. Using expectations data from the 2000 National Annenberg Election Study along with data on media coverage and campaign finance in the 2000 presidential nomination process, I show that expectations about the candidates' chances change in response to changes in the information provided by the candidates and the media. In contests involving well-known candidates, voters act rationally by using this outside information to inform their expectations of the candidates' chances of winning. In contests involving only lesser-known candidates, voters do not act rationally, and only base their expectations on past values of their expectations for these candidates. This dissertation provides new insights into how expectations change over the course of a primary campaign, and gives a better understanding of these important variables.

Committee: Herbert Weisberg (Advisor) Subjects: Political Science, General