The Trade-Off between Equal Representation and Electoral Participation: The Effect of Redistricting on Voter Turnout

ABSTRACT

Equal representation and electoral participation lie at the core of democracy. However, the two are sometimes contradictory. When redistricting is used to correct malapportionment, a typical example of unequal representation, it can discourage citizens from voting by increasing their anxiety about whether their interests are represented and increasing their information cost. The effect of redistricting on electoral participation has not been accurately estimated due to difficulty isolating the effect from past redistricting and other factors. Japan’s upper house conducted its first redistricting in 2016, providing an ideal opportunity to identify and isolate the effect of redistricting on electoral participation by avoiding the usual methodological problems. Using an original dataset on Japan’s upper house elections from 2001 to 2019 and employing a differences-in-differences design, this study reveals that redistricting reduced voter turnout by 10.3 percentage points and that the effect lasted until the second election after redistricting.

KEYWORDS:

• Redistricting
• equal representation
• electoral participation
• voter turnout
• Japan

Acknowledgements

The earlier versions of this article were presented at the Japanese Consortium for Political Science Workshop in 2017 and the European Consortium for Political Research General Conference in 2017. I would like to thank Masahisa Endo, Masaaki Higashijima, Jun'ichi Hirano, Takeshi Iida, Hideki Kido, Petter Y. Lindgren, Ryo Nakai, Etsuhiro Nakamura, Hironori Sasada, Sohei Shigemura, Naoki Shimizu, Daniel M. Smith, Chris Winkler, Masahiro Yamada, Yuki Yanai, Masahiro Zenkyo, anonymous reviewers, and the journal editor for their helpful comments.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Notes

1 Hayes and McKee (2012) state, ‘[w]e assume that most individuals do not decide whether or not to turn out on the basis of the U.S. House race (although probably some do). Instead, it is the higher-profile contests (i.e., presidential, gubernatorial, and senatorial elections) that primarily drive turnout rates’ (p. 120). If so, this assumption raises new questions: do voters carefully consider the benefits and costs of voting in deciding whether or not to vote in the House election of such low salience, and thus, can an analysis of roll-off accurately estimate the effect of redistricting on electoral participation? While this study acknowledges the usefulness of roll-off, its potential for development as a method to measure electoral participation should be examined in future research.

2 Abramowitz, Alexander, and Gunning (2006) find little effect of partisan gerrymandering on competitiveness. Gelman and King (1994) find a positive effect.

3 The upper house has strong veto power over the lower house on the enactment of bills, although the Constitution provides for the supremacy of the lower house. If the upper house rejects a bill passed in the lower house, the lower house may pass it again by a two-thirds majority of present members in order to enact it (Article 59). In fact, all recent governments that lacked a majority of seats in the upper house, such as the cabinets of Keizo Obuchi (1998–2000), Yasuo Fukuda (2007–2008), Taro Aso (2008–2009), Naoto Kan (2010–2011), and Yoshihiko Noda (2011–2012), had difficulty passing bills and maintaining power. Furthermore, the cabinet of Ryutaro Hashimoto (1996–1998) resigned to take responsibility for defeat in an upper house election. The cabinets of Yoshiro Mori (2000–2001) and Yukio Hatoyama (2009–2010) were forced to resign by ruling parties right before an upper house election mainly because their unpopularity was expected to have an adverse effect on election outcomes.

4 See Komatsu (2015) for the legislative process of the electoral reform of Japan’s upper house.

5 Data on voter turnout rates were obtained from the website of the Ministry of Internal Affairs and Communications: http://www.soumu.go.jp/senkyo/senkyo_s/data/sangiin/ichiran.html.

6 To test the effect of redistricting on the change in voter turnout from the previous to the current election, this study runs a model with ΔVoter Turnout_it, the change in voter turnout from election t-1 to election t, as the alternative dependent variable instead of Voter Turnout_it (see Appendix 2). In Column A2 in Table A2, Redistricting_it, is statistically significant and positive for ΔVoter Turnout_it, which indicates that redistricting caused a decrease in voter turnout from the previous election.

7 ‘Omitted variable bias occurs when two conditions are true: (1) the omitted variable is correlated with the included regressor; and (2) the omitted variable is a determinant of the dependent variable’ (Stock & Watson, 2014, p. 180). Electoral competitiveness, the number of candidates, and economic conditions are determinants of voter turnout. However, there is little theoretical reason to expect that these variables affect the implementation of redistricting. Therefore, the exclusion of these variables does not create omitted variable bias. Rather, they are the outcome variables of the implementation of redistricting, so their inclusion would create post-treatment variable bias. For further confirmation, this study runs a model with additional control variables for electoral competitiveness, the number of candidates, and economic conditions (see Appendix 2). Column A2 in Table A2 shows that even with these additional control variables, the effects and statistical significance of Redistricting_it remains almost unchanged.

8 The demographic data of prefectures used in this study were obtained from the website of the Ministry of Internal Affairs and Communications: http://www.stat.go.jp/data/k-sugata/

Additional information

Funding

This work was partly supported by a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science [research project number: [18K01413] and the Research Fund from the Quantitative Methods for International Studies Program at Kobe University.

Notes on contributors

Naofumi Fujimura

Naofumi Fuijmura is Professor of Political Science at the Graduate School of Law, Kobe University in Japan. He received his PhD from Kyoto University. His research interests include electoral systems, legislatures, and political parties. His work has appeared in Election Law Journal, European Political Science Review, Journal of Legislative Studies, and Party Politics. E-mail: [email protected]