«by Dyron Keith Dabney A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Political Science) ...»
The Type A campaign strategy is typified by organizational tactics by using local politicians and interest groups (senkyo keiretsu, and reiki dantai keiretsu). Type A oriented candidates rely on a broad web of support from influential, local notables, namely local politicians and associational leaders of enterprise-based labor unions, professional organizations, neighborhood associations and trade groups. Such intermediaries or “vote-brokers” try to amass a “gathered vote” on behalf of the candidate.
A Type A candidate will often try to build an “electoral keiretsu,” when local politicians and a Diet candidate have determined a mutual benefit from a political alliance. In general, local politicians need pork and political legitimacy provided by national assembly members, while national assembly members need the gathered vote and local-level information provided by local assembly members. Pork more or less tends to be the bottom line of the relationship in the case of rural-based local politicians than urban-based local politicians.
The Type B campaign strategy is typified by tactics that appeal directly to voters, including face-to-face encounters and the mass media. Most importantly, candidates pay careful attention to how they will be perceived by voters, including crafting their personal image—young and energetic in the case of our candidates—and enunciating policy positions seen as attractive in their districts. Type B candidates tend to emphasize their attachment to a political party. They take every chance they can to appear on television, although such opportunities are limited particularly in the formal campaign period.
outdoor greetings, rallies and assemblies (gaitō enzetsu, enzetsukai and shūkai), sound truck canvassing (gaisensha no yūsei), door-to-door canvassing and mail and telephone solicitation. Public speeches and political rallies are occasions for candidates to present their qualifications for political office and their legislative goals to district constituents, and to distribute political and campaign literature.1 Type B-oriented campaigns also pay attention to the design and placement of campaign posters, and the content delivered by the sound truck that tour neighborhoods supplemented with live and pre-recorded political messages.
The most enigmatic aspect of Japanese election campaign tactics, one with both Type A and Type B characteristics, is the kouenkai, literally “support association.2” When Curtis (1971) did his field work in Kyuushuu in 1967, he found kouenkai to be a relatively new phenomenon; a way for candidates to gather reasonably stable votes at a time when old social solidarities of villages and local notables were gradually dissolving.
Those had been the principle vote mobilization mechanism in most of Japan in the prewar period, and into the postwar, but over time they could gather a smaller and smaller proportion of the votes needed to win a seat.
Two sorts of explanations have been advanced for why the kouenkai evolved to perform this function. One is cultural, that Japanese respond best to a sense of personal connection (or kankei) rather than joining mass organizations or movements, or 1 Candidates deliver these speeches to large and small audiences in intimate settings such as neighborhood or community halls, or large, formal settings, such as a town/city auditorium. The dominant settings for public speeches and rallies, however, are train station entrances and shopping centers in which candidates can take advantage of ample pedestrian traffic.
2 A comprehensive description of kouenkai is offered by Kabashima and Yamada, in “Kouenkai to Nihon no Seiji” (Kouenkai and politics in Japan) in The Annuals of the Japanese Political Science Association, 1994 (Tokyo: Japan Political Science Association, 1995).
structural. In Western electoral politics, the organizational vacuum left by declining traditional social ties tended to be filled by political party organizations. In Japan, however, the multi-member district (MMD) electoral system, which required a party to average at least two seats in every district to secure a majority of the legislative seats, meant that party organization was useless to candidates for getting elected. The only party in Japan that consistently ran multiple candidates in all election districts was the LDP, but neither its national party nor local party branches could support one candidate over another, so they were left to fend for themselves through non-party mobilization methods. Of course, both these explanations could well be valid.
Kouenkai can be characterized as Type B to the extent that people simply sign up for it autonomously, that is, without the directive of any person of socio-political authority, because they are impressed with the candidate for their own reasons, and they would like to show their support, meet like-minded fellow voters, and help get their choice elected. An American equivalent would be the “Citizens for X” groups, independent from party organizations, that are ubiquitous in American elections. Such kouenkai are often more mailing lists than groups—as pure Type B as can be imagined.
However, kouenkai are Type A to the extent that people are brought into them through other social relationships (i.e., organizations and associations) in which local notables and elected politicians are involved, and that they take on a hierarchical structure.
Following a kouenkai strategy most often means efforts to get members more and more attached to the group, such as by sponsoring trips and get-togethers (to the point where the benefits seem to require some reciprocity). A telling difference with “Citizens for X”
organization and attach people to it, rather than upward as campaign contributions. The funds were used for gift-giving to its members for weddings, funerals, and Japanese holidays, and for group excursions (some to Tokyo to visit the legislature, some to hot springs). This gift-giving could easily be interpreted as simple voting buying on one level, but these were common practices in Japanese society. Another telling difference between these two models of grassroots citizen support is that kouenkai typically lasted beyond the campaign season and endured as long as the politician continued to run for office.
It is important to keep in mind the dual nature, or ambiguity of kouenkai when investigating Japanese election campaigning. In our study, we found that kouenkai were generally thought of as more Type A than Type B among the candidates interviewed, and empirically kouenkai based tactics most often fell under the Type A rubric.
Political candidates tend to gravitate toward the Type A or Type B strategy when running for office, but the two strategic approaches are hardly mutually exclusive. In fact, all Japanese candidates typically use both, and for that matter people running for election in all countries use some combination of direct and indirect campaign strategies that more or less correspond to our distinction between Types A and B strategy.
Nonetheless, the conventional strategy prescription among inexperienced challengers is the Type B strategy, while the strategy prescription for experienced incumbents (or experienced challengers) is the Type A strategy. The case studies presented in chapters three and four will examine campaign strategy under pre-reform and post-reform
Type A and Type B strategies.
Reform and its Effects We turn now to providing the historical and institutional context for our study.
Election campaign behavior for Japan’s Lower House, the more powerful chamber of the National Diet, was the source of considerable public inquiry and criticism over the last few decades. Fueling this inquiry and criticism of campaign behavior was the multimember district (MMD), single non-transferable vote (SNTV) election system adopted in 1925 which shaped candidates’ campaign behavior, and made for a distinctive election campaign.
Under the MMD election system, 511 Lower House seats were apportioned across 129 two-to six-member districts.3 A feature of the MMD election system was that it forced candidates from the same party seeking to gain a majority of the votes to compete for seats in the same district. To illustrate this point, we need only look at the history of one political party, the Liberal Democratic Party (LDP), which was particularly singleminded about securing a majority seats in the Lower House. A long-standing indictment of this election system feature was that it facilitated one-party dominance by the LDP 3 The MMD election system was in effect prior to 1925, but the key difference was that the district magnitude consisted of single-members districts and multi-member districts with two to three members.
Initially, the district magnitude consisted of 122 three- to five-member election districts—53 three-member districts, 38 four-member districts and 31 five-member districts. The three to five-member MMD election system was adopted because it enabled dominant conservative groups to obtain a majority of seat in the 1925 House of Representatives elections (Hirano, 2006). Both the number of districts and the district magnitude have been subject to mandatory adjustments over the years in according with election laws.
candidate in election districts.4 Such intra-party competition encouraged candidates to define themselves by personality (i.e. personal image), and what Reed and Theis (2003) call “hyperpersonalistic” patron-client relationships. The MMD electoral system tended to play up competition over personality and personal relationships (particularly among co-partisan candidates) instead of policy issues and political party platforms. In other words, the electoral system promoted competition for votes using the Type A rather than Type B campaigning.5 Almost universally, academics, journalists, and even politicians (when they were speaking to the general public rather than trying to fight their own campaigns), deplored the Type A strategy (because of its link to money politics produced by intraparty competition) and called for more of the Type B strategy. The motivation behind the growing demand for more Type B campaigning, rather, an electoral system that promoted more Type B campaigning, was due to the dissatisfying levels of political accountability in Japanese electoral politics; a hallmark of democratically-held elections. By political accountability we mean candidates’ 1) responsiveness to and representation of members of the electorate to get elected to office; 2) commitment to drawing voters into the political and electoral processes; 3) campaigns and subsequent actions in office based on explicit policies that reflect voter preferences; and 4) transparent use of campaign contributions for political and campaign activities.
4 LDP’s dominance in the Lower House was also tied to its failure as the ruling party to reapportion legislative district seats according to population changes. Rural election districts, where the LDP received significant voter support, were over represented by a 4:1 ratio. In this way, the LDP must assume some culpability in the indictment of the MMD election system.
5 A long-acknowledged curiosity among scholars is that the MMD electoral system easily could have driven candidates (namely those fielded from the same political party) to differentiate themselves on substantive policy issues, but instead, the system drove them to distinguish themselves on personality.
accountability; candidates were not accountable to the policy preferences of voters, at least a significant percentage of voters, and the electoral system gave little power to voters to oust candidates who failed to be accountable to them. Under the MMD electoral system, candidates could get elected to office—and stay in office—with as little as 10 percent of the vote (in a five-candidate district-- 20 percent is needed to guarantee victory but most winners got much less than that). In other words, the MMD system gave candidates an incentive to cater to the policy interests and expectations of a small slice of the electorate to get elected. Indeed, this slice of the electorate often was “vertical,” concentrated in one geographic area, or otherwise homogeneous in composition.6 The logic of electoral reform was that if the machinery of politics shifted from the MMD to the MMM electoral system (namely the SMD portion), candidates would have to gain a much larger percentage of the vote, requiring an appeal to a cross-section of the electorate that represents the “median voter.” That is best managed through the Type B strategy, where the Type A strategy is most effectively aimed at small niches of voters accessible through political “pork” and constituency service. Appealing to the median voter favors Type B strategy via policy issue appeals and policy records.
The aspect of campaigning that stimulated the most calls for reform was campaign money. Both Type A and Type B strategies require substantial funds, of 6 While political turnover regularly occurred under the MMD electoral system (mostly among weaker incumbents, or incumbents tied to scandal), it mostly applied to candidates (and political parties) who had not mastered the Type A strategy well-regarded for generating a reliable percentage of voters. One consequence of the SMD competition is the higher level of electoral safety it is known to produce among incumbent candidates (i.e., incumbency advantage). Short of strong challengers from which they must defend their seat, incumbent candidates have less incentive to expand their support base beyond voters who put them in office. While accountability problems still exist under the MMM electoral system, the SMD seat competition demands that candidates are accountable to a larger share of the electorate than the MMD seat competition.