Personalized Campaign Behavior in New Democracies: The Case of South Korea (Pre-Recorded)

• In Event: Parties & Legislatures

Abstract

This project is about the personalization of politics on the constituency candidate level. More specifically, this paper explores the causes and consequences of the personalization of constituency candidates' electoral/campaign behavior in South Korea. This study asks under what conditions candidates take personalization strategy over party-oriented strategy in their campaigns and how their strategy is rewarded by voters in Korean general (National Assembly) elections? With the new Korean Candidate Behavior Dataset created for this project, the paper aims to understand constituency candidate behavior which is often left as a black box in an electoral behavioral cycle. Mainly this paper argues that candidate behaviors do not occur in a contextual vacuum: country and district political and socio-economic context impacts candidates' strategic choice of campaign behavior, and candidate's individual factor also interacts with the connection between the context and candidate behavior.
The research question is answered by analyzing the new Korean Candidate Dataset within the mixed-method tradition – first, use qualitative content analysis of Korean Candidate Advertisements/Manifestos to create the dataset and then specify a quantitative multilevel statistical model to test the argument. The Korean Candidate dataset is constructed based on Candidate Advertisements/Manifestos and information from Korean Statistical Information Service (KOSIS) and Korean National Election Commission (KNEC). From the Candidate Advertisements/Manifestos that every candidate is required to publish to run an election, we can collect candidates' personal and public information, such as candidate's party, age, education, career, experience, policy, personal characteristics, personal/political connections to various groups. Since candidates can construct their advertisement as they want within a 12-page limitation, candidates can strategically highlight any factors/information they want. Hence, the Candidate Advertisements/Manifestos provide invaluable information about candidate behavior/campaign strategy. From KOSIS and KNEC, we can collect country and district level socio-economic information (e.g., economic prosperity, unemployment rate, education, age) and political information (e.g., incumbent/open seat, partisan balance, election outcome of t-1 and t). The Korean Candidate Dataset, combining contextual setting and candidate behavior information, enables us to understand under what conditions candidates choose personalization strategy and how that affects election outcome.
This study about the personalization of politics in constituency candidate behavior in South Korea preserves theoretical and methodological importance. Theoretically, focusing on the constituency candidate level, this study will enhance our understanding of the cause and consequences of personalization of politics at the district level. Most of the previous personalization of politics or populism studies focused on the national leader level (president or chancellor) or party level (Karvonen 2010, McAllister 2007, Garzia and De Agenlis 2014; Rahat and Sheafer 2007) left us with a lack of understanding of the district level personalization of politics. Although some exceptional district-level works increased our understanding (Zittel, 2015; Gschwend and Zittel, 2015; Bogglid and Pederson, 2018; Cross and Young 2015, Eder, Jenny, and Muller 2015; Selb and Lutz 2015), the studied cases are mostly advanced democracies.
Secondly, by focusing on the general election in Korea, which is one of the most politically dynamic new democracies in Asia, this study will enhance our understanding of the personalization of politics in the new democracies. This study will be one of the rare studies that attempt to theoretically understand constituency candidates' personalization of electoral behavior in Korea. Starting with Korea's case, this project can expand to the case of Taiwan and Japan, which share the same style of Candidate Advertisements/Manifestos and thus, can add more comparative aspects of analysis.
Lastly, introducing the new dataset of Korea's candidate behavior, this study offers methodological contribution in understanding candidate behavior with the more objective term. Most of the previous studies on candidate behavior relied on candidate surveys, such as Comparative Candidate Survey (CCS). Although survey research provides excellent value, it can be suffered by subjectivity or selection bias because it depends on retrospective subjective evaluation of candidates' campaign behavior. On the other hand, by focusing on every candidate's Candidate Advertisement in the 19th and 20th general election, this study will reduce the selection bias and offer an objective measure of candidate behavior.

Author

• Leeann H Youn, University of Kansas