Essays on Voter Preference and Political Advertising

Chapter 1: This chapter studies the impact of TV advertising in the 2008 presidential election on the decision of individuals on whether to vote and for whom to vote. I use survey data from the American National Election Studies and advertising data from the Wisconsin Advertising Project. The survey asks individuals to report their vote intention each month for several months before November and I exploit this panel structure to examine the channels through which advertising affects an individual's vote choice. I show that advertising for a candidate significantly increases the probability of individuals voting for that candidate, particularly among those who expressed an interest in politics and who were younger. Advertising mainly operates through the channels of market expansion (persuade individuals to switch from not voting to voting) and reinforcement (lower the probability to abstain for those who have decided to vote for the candidates). Business-stealing effect (switching between candidates) is weak, especially at the end of the campaign.;Chapter 2: This chapter investigates the extent to which vote choices can be rationalized by ideology. In a spatial voting framework, if voters vote ideologically, they choose candidates whose positions are closest to their own. Rationalizability increases with dimension of ideological space so higher dimensional space makes it difficult to identify voters' preferences. I tackle this problem by employing comprehensive data from the 2008 U.S. presidential election survey conducted by the American National Election Studies. In the study, individuals reported their vote choices, their own ideological positions as well as candidates' perceived positions in eight policy issues. Respondents also reported how important each policy issue was to them. This information provides meaningful restrictions on ideological distance weight assignment. I find a small but significant (about 8.8%) set of responses to be inconsistent with ideological voting. Results suggest individuals make use of a wide range of issues to decide their votes. Candidate quality may compensate for differences in ideological distance but only for a small number of voters.