Learning Curves: Three Studies on Political Information Acquisition

Abstract

What are the effects of political information on public opinion, political participation, and electoral outcomes? In this dissertation, I address these questions and investigate the ways that people acquire and incorporate information based on their levels of political knowledge and attentiveness. I examine the effects of political information among three groups whom we would expect to learn differently: those people with little knowledge or interest in politics; the potentially interested who possess some, but not much, knowledge; and the attentive experts. In my first chapter, I look at the effects of information on people with little or no knowledge of politics by asking, "Do candidate visits affect voting decisions and candidate evaluations?" I link survey data with the location and topics of all speeches given by George W. Bush and John Kerry in 2004 to empirically test the conventional wisdom that candidate appearances change electoral outcomes. I find that candidate visits do provide information to voters and that those effects are conditioned on consumption of local media. In my second chapter, I look at people with some knowledge of politics: college students. I ask, "How does the information that students 'incidentally' encounter in electronic social networks like Facebook.com shape their knowledge of current political events and their participation?" To answer these questions, I conducted a survey with an embedded experiment. I find that students do learn from Facebook, though the effects are small and vary across groups. My third chapter investigates the ways that the politically attentive incorporate information by asking, "What campaign information matters? Which campaign events are actually informative?" I develop a new measure of information flow using data from a political prediction market and a Bayesian estimation technique that adapts models from the economics literature. This measure offers a reliable way to describe the importance of campaign events that does not suffer from either post hoc judgments or reports from the principals involved in the campaign. Together, these projects address the consequences of political information in contemporary politics.