Candidates’ Characteristics, Electoral Performance, and Legislative Outcomes: (Quasi-) Experiments

Abstract

Explaining who citizens elect is central to political science. It is now more than ever a rapidly developing field with the ongoing revolution of research methods (e.g., experimental designs and image recognition technology). Recent literature has investigated how the characteristics of candidates, such as incumbency or physical appearance, determine electoral performance (e.g., Lee 2008; Todorov et al. 2005), but results are mixed. One reason for the mixed results is that the literature tends to ignore institutional contexts that can mediate the relationship between candidates’ characteristics and electoral outcomes. Using experimental and quasi-experimental designs, this panel sheds new light on how the political or physical characteristics of candidates affect their electoral performance and legislative influence. We seek to solve problems in the existing literature by examining how the effects of candidates’ characteristics on electoral results are conditioned by electoral environments such as district magnitude and electoral timing. In addition, we extend our investigative scope to their role in the policymaking process and seek to explain the determinants of their legislative influence. Our panel contributes to a better understanding of the effects of candidate quality on electoral and legislative outcomes as well as to the development of research methods in social science. This panel consists of four papers: First, Tsukiyama explores how electoral systems shape incumbency advantage. Focusing on elections for prefectural assemblies with district magnitude ranging from 1 to 17 and using a regression discontinuity design, he finds that incumbency advantage decreases as district magnitude increases. Second, Shimizu examines how the relationship between candidates’ characteristics and electoral performance is conditioned by electoral environments. He focuses on election timing and estimates the effect of public support for parties and economic performance on candidates’ electoral performance. Third, Fujimura looks at the determinants of legislators' influences in the policymaking process. Using a differences-in-difference design, he shows that legislators' influence on legislative outcomes increases when they become a committee chair. Finally, Nakamura seeks to explain the mechanism behind the effect of candidates’ physical appearance on voters’ decisions. Using picture-based conjoint analysis and eye tracking technology, he analyzes how the visual cues of candidates change the relative importance of factors in voters’ decisions. Hijino and Kido will serve as discussants. Hijino is a specialist in local politics and Kido is a specialist in party politics and quantitative research. Both will offer insightful comments on the four papers.