Portal de Periódicos da CAPES

The Electoral Consequences of Candidate Appearances on Soft News Programs

2011; Taylor & Francis; Volume: 28; Issue: 1 Linguagem: Inglês

10.1080/10584609.2010.540304

1091-7675

Masaki Taniguchi,

Media Influence and Politics

Abstract The shift in political communication toward soft news has been accelerated by politicians who appear on "soft news" shows, but the literature on the electoral effects of such soft news appearances is limited, especially outside the United State. This article will demonstrate, using a national survey of voters, a survey of politicians, and an Internet-based experiment on Japan's 2007 House of Councillors election campaign, that candidates' appearances on soft news programs have a positive effect on their election results. Candidates accrue votes from appearing on traditional news shows, but they accrue more votes from appearing on soft news shows. Moreover, politicians who make frequent appearances on soft news programs tend to gain more candidate-oriented, rather than party-oriented, votes. However, if the candidate's political message is weak, then the positive electoral effects diminish. In order to maximize their votes, candidates would be well advised to appear both in soft and traditional media and pay attention to their political, as well as entertainment, message. Keywords: soft newsinfotainmentelectionJapan Acknowledgments An earlier version of this article was presented at the annual meeting of the Japanese Political Science Association, Tokyo, October 2007. I appreciate the helpful comments and support of Ikuo Kabashima, Yoshiaki Kobayashi, Satoshi Machidori, Samuel Popkin, Crystal Pryor, Steven Reed, Shiro Sakaiya, Gill Steel, Naoko Taniguchi, Hidenori Tsutsumi, and the anonymous reviewers. I also acknowledge the financial support of the 21st Century Center of Excellence program and a grant-in-aid for scientific research from the Ministry of Education, Culture, Sports, Science and Technology. Notes 1. In communication studies, "soft news" refers to certain types of articles and feature stories within the traditional news media (CitationEveland, 2006). According to CitationCurran, Iyengar, Lund, and Salovaara-Moring (2009), hard news is defined as reports about politics, public administration, the economy, science, technology, and related topics, while soft news consists of reports about celebrities, human interest, sports, and other entertainment-centered stories. This is similar to the definition employed for the experiments herein; however, depending on the degree of soft/hard news (based on the above definition), a TV program itself may be termed "soft (hard) news," as seen in Baum's research. The analysis of politician and voter surveys here uses the same categorization of TV news as Baum. 2. CitationBaumgartner and Morris (2006) claim the opposite effect: People exposed to jokes about candidates in The Daily Show rate candidates more negatively. 3. The recent trend of Japanese political news leaning toward entertainment is also pointed out in several studies written in English, for example CitationPharr and Krauss (1996), CitationKrauss (2000), and CitationKabashima and Steel (2007). 4. This does not include so-called "talent politicians" (celebrities who ran for office and were TV personalities, comedians, hosts, or actors/singers). 5. The original data were recoded so that higher figures reflect increased frequency. The control variables were similarly recoded. 6. When there were two LDP (or DPJ) candidates in the same prefecture, I used the average frequency of the two candidates. For other surveys using the consumption indicator and media appearance frequency (or frequency of TV campaign ads of each candidate), see CitationFreedman and Goldstein (1999). 7. Average value minus standard deviation is less than 0. 8. I am thankful to the reviewers for their suggestion on this point. 9. Even more advantageous for this analysis is the fact that both LDP and DPJ leaders of the House of Councillors were not up for reelection in the 2007 election and both the speaker (LDP) and vice-speaker (DPJ) retired, which means that many of the prominent figures of the House were not included in the data. 10. The average value of the dependent variable was 1.04 (SD = 0.19). 11. The dependent variable was adjusted for Hokkaido. In the Hokkaido constituency (two seats) the DPJ nominated Katsuya Ogawa and an independent candidate, Kaori Tahara. Accordingly, the DPJ decided to allow three of the 12 DPJ House of Representative district branches in Hokkaido to support Tahara instead of Ogawa. Based on this, I allotted 75% of the total votes earned by DPJ in Hokkaido under the proportional representation system (1,217,315 votes in total) to Ogawa. 12. According to wording in the UTA Survey, an appearance can only be counted if it includes "going to the studio, accepting coverage, or otherwise appearing by your own choice." For example, if a newscaster indirectly covers the politician's activities during a traditional news program, this does not count as an appearance, but if the politician makes a comment directly during the news story, that qualifies as an appearance. 13. According to CitationDruckman (2005), the difference between TV news and newspapers is quantitative, and few qualitative differences, or differences in terms of content, are observed. Here I follow Druckman's analysis and include the print media in deriving the principal component score. 14. This is defined as follows: [(number of total valid votes)/(number of members to be elected in an ordinary election in the electoral district)] × 1/6. If this is higher than the number of votes obtained, the candidate will not be accepted as elected, even if the ranking order was ostensibly high enough to obtain a seat. 15. For examples of survey experiments done in order to evaluate the political effects of soft news consumption, see CitationNabi and Hendriks (2003) and CitationSlater, Rouner, and Long (2006). 16. The voting age in Japan is 20. 17. The effective responses from each group were 498 for Treatment Group 1 (Video 1), 502 for Treatment Group 2 (Video 2), and 860 for the control group (no videos). Response rates for the treatment groups may be lower than those for the control group because of a nonresponse bias: Assignments to each group were made as the questionnaires were being sent. In cases where a certain attribute leads to a nonresponse bias that impacts participation in a survey, adjustment is needed (CitationHoriuchi, Imai, & Taniguchi, 2007). However, I did not use any adjustment in this experiment as there was no significant difference between the treatment groups and the control group in any of the major factors (demographics, interest in politics, interest in educational issues, frequency of newspaper and news consumption, and DPJ support). 18. The two videos used in this experiment can be viewed at the author's Web site (http://www.masaki.jou-tokyo.ac.jp/PC). I would like to thank Kan Suzuki for generously allowing the use of his video materials for the survey experiment. 19. Respondents were given a debriefing at the end of the experiment. 20. This consisted of the principal component score on four variables (interest in politics, interest in diplomatic issues/international relations, frequency of newspaper consumption, and frequency of TV news consumption).