![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
Abstract
This paper examines whether candidates in Japanese local elections are more likely to declare affiliation with parties as the population size of districts increases, and thus the importance of party reputations and resources for mobilization increases. To test this hypothesis, we use data from Japanese municipal legislative elections between 1999 and 2010 where the vast majority of candidates run as independents. We exploit variations in the population size of districts before and after massive municipal mergers. We find that the percentage of candidates affiliated with political parties increased when the number of votes required for winning a seat increased as a result of municipal mergers. Our analysis also finds that candidates in municipalities that merged were more likely to join parties, compared to those in municipalities that did not experience mergers.
Notes
1 The information is obtained from Local Authority Election Statics 2010 published by the New Zealand Department of Internal Affairs.
2 For example, in Victoria, Australia, almost every candidate for local council ran as independent in 2008 (Economou, Citation2010). In Ireland, independent members constitute about 8% of the total seats in its national parliament, the Dail, and play a critical role in government formation (Bolleyer & Weeks, Citation2009). In Australia, electoral support for independent candidates has been steadily increasing in federal elections during the past several decades (Costar & Gurtin, Citation2004; Rodrigues & Brenton, Citation2010)
3 The information on municipal assemblies is based on the report by the Ministry of Internal Affairs and Communication in 2006.
4 As we discuss in the subsequent section, being elected with a few hundred votes is possible in small towns and villages in Japanese municipal elections.
5 However, the Yokohama city assembly has multiple electoral districts. The largest at-large district has 52 members elected from a single district (Setagaya city assembly election in 2007).
6 Our supplementary analysis confirmed that the fiscal condition of municipalities had a crucial effect on the decision to merge. Municipalities with the larger amount of deficit were more likely to merge with others. Another potential explanation is that municipal mergers occurred when parties predicted that they could gain in size under the merged entity. In other words, there might be more mergers among municipalities where parties are either strong or have a potential to grow. Our supplementary analysis checked this possibility by examining the relationship between the LDP vote share of municipalities in the 2000 Lower House election and the decision of municipalities to merge. We used the LDP vote share in the 2000 Lower House election as the measure of underlying partisan strength in municipalities because LDP was the governing party until 2009 and thus could exercise a strategic influence on the decision of local governments to merge. The results indicated that the LDP vote share did not have any statistically meaningful impact on the decision to merge, implying that this political explanation did not account for the decision of mergers. The details and results of the supplementary analysis are presented in Online Appendix 1 at http://dx.doi.org/10.1080/17457289.2014.933837.
7 These numbers are based on the official report by the Ministry of Internal Affairs and Communications.
8 The maximum size of increase is 1,380,000.
9 Online Appendix 2 is found at http://dx.doi.org/10.1080/17457289.2014.933837.
10 Some municipalities even decreased the size of the assembly after mergers in order to reduce the cost associated with legislative activities. After the merger, municipalities have an option to elect a new set of legislators with the new municipal boundaries or to let all the incumbents who are elected before the merger temporarily serve in a new legislature for two years after the merger. In the latter case, municipalities should hold a new election under a new boundary once the two-year term is over.
11 Online Appendix 3 is available at http://dx.doi.org/10.1080/17457289.2014.933837. Municipalities in large metropolitan areas, such as Osaka city and Nagoya city, are divided into multiple districts, yet we treat those municipalities as a single district because they experienced no major merger with neighboring municipalities. This means that we include those large municipalities as a control group in our analysis. The results reported below hold even if we exclude the municipalities in metropolitan areas. Furthermore, some municipalities held their first election using their pre-merger municipal boundaries as districts. These municipalities have no impact on the results.
12 We excluded only five municipalities for this reason. In addition, we excluded the data of elections held just before the merger if all the merging municipalities did not hold elections. For example, if difference in the election cycle led one of two merging municipalities to hold an election before the merger but not the other, we could not create the outcome variable for the area because of the lack of information on the second municipality. We therefore dropped the election data for the first municipality before the merger. This does not affect our results because the newly merged municipality must hold an election immediately after the merger.
13 When information is not available in the Asahi Shimbun database, we consult databases of other newspapers, such as Yomiuri Shimbun and local newspapers.
14 Elections in Okinawa were excluded due to a large amount of missing data.
15 The information is available at http://www.soumu.go.jp/gapei/gapei.html.
16 In 1999, the percentages of candidates belonging to the LDP, DPJ, Komei, and JCP were 2.6%, 1.0%, 4.5%, and 7.1%, respectively.
17 Online Appendix 4 is found at http://dx.doi.org/10.1080/17457289.2014.933837.
18 We have also used the number of citizens over 15 years of age as a numerator, yet our analysis shows the substantively the same results.
19 Although it is better to use the actual number of votes casted in elections as the numerator, we employ the size of the population instead because the actual number of votes is unavailable when elections are uncontested. For the same reason, we do not use “Houtei Tokuhyo Suu” (the minimum number of votes necessary to win a seat) as an indicator.
20 We did not choose the election in 2003 because most municipalities merged in 2005.
21 In contrast, [Population]jt before the merger equals the average population size per seat in multiple municipalities that merged into a new municipality.