Doctoral Dissertation Research in Political Science: Running on the Brand

University Of California-Berkeley, Berkeley CA

Investigators

Abstract

This project investigates the choices that US legislative candidates make in their campaigns as strategic members of a collective party organization. The central question motivating the project is: When do congressional candidates run with or away from their party at election time? In addressing this question, the project aims to fill three major gaps in extant political science research. First, due to the lack of historical advertising data, we know relatively little about the kinds of issues and statements candidates typically raise in their campaigns, especially for elections prior to 2000. Second, we have only a limited understanding of the role that parties play in these elections and of change over time in that role. Finally, the most widely held view of parties in congressional organization is based on the largely untested assumption that the electoral prospects of party candidates are strongly linked to the fates of the parties? reputations in Congress. According to this view, candidates have a restricted ability to develop campaign strategies to deal with good and bad partisan election swings. Yet no systematic analysis is available to appraise this fundamental assumption or to explore the messages actually issued by candidates. This project will examine the strategic statements communicated directly by candidates in their advertisements over the last forty years. The primary research task will be to code and analyze the content of the over 80,000 radio and television congressional ads housed at the Julian P. Kanter Political Commercial Archive, at the University of Oklahoma. The ads will be examined through a combination of traditional and computer-automated approaches, in order to assess the kinds of issues the candidates raise, the sorts of images and references they make about the parties and party leaders, and the frequency with which they make clear, ambiguous, or inconsistent statements over time and across districts. The research will also evaluate the level of coordination that is exhibited within parties on similar election messages as each party grows more ideologically extreme and unified. The project will also conduct survey experiments aimed at addressing the behavioral foundations of partisan candidate strategies and the ways that those strategies might be linked to the information available to voters. The project will make several broader contributions. The campaign data collected in the project will be of interest to scholars in multiple areas, including media and communication, social and public policy, economics, and political science. The computer-based coding method developed and used in this project could find additional application in such fields as linguistics and legal scholarship. The datasets and findings from the project will shed light on a range of important aspects of contemporary US politics, and may prove to be a resource for teachers, journalists, and policy-makers.