Soka Gakkai voting

The lay Buddhist organization Soka Gakkai remains a gap in our knowledge of Japanese politics. We can see its presence in Komeito, the political party it founded in 1967, and speculate about its influence over the Liberal Democratic Party (Yamada, 2003; Sado, Reference Sado2005) but its political activity remains unclear. Researchers (Metraux, Reference Metraux1988, 2003; Seager, Reference Seager2006) have shared their interviews with top leaders in the Gakkai and Komeito, but there has been no systematic study of rank-and-file Soka Gakkai members in Japan since the late 1960s (Thomsen, Reference Thomsen1963; McFarland, Reference McFarland1967; White, Reference White1970; Dator, Reference Dator1969; Azumi, Reference Azumi1971), in spite of Komeito's pivotal role in the balance of parliamentary power.Footnote ¹

This article is the product of rare fieldwork in the Soka Gakkai, and makes two claims redefining how we think about Komeito voters. This study's first claim is that the Gakkai relies on policy claims more than social identity to justify members' support for Komeito. Second, Komeito partisans in the Gakkai repeat the same policy appeals to recruit floating voters, rather than relying on social ties to motivate independent voters.

Currently, scholars assert that Gakkai members support Komeito solely because of their social identity (Baerwald, Reference Baerwald1986; Curtis, Reference Curtis1999; Mulgan, Reference Mulgan2000; Hrebenar, Reference Hrebenar2000). Some qualify that assertion (Stockwin, Reference Stockwin, Ishida and Krauss1989; Richardson, Reference Richardson1998), but they all explain the party's support by linking political activity to Gakkai member identity. Put simply, Gakkai members vote Komeito because they think that's what Gakkai members do, and independents vote Komeito to please a friend or relative in the Gakkai. The standard portrayal of Gakkai members – typically citing survey data from the 1960s – is that members join the Gakkai in search of a social identity (and its attendant connections) that they lost moving to urban areas for low wage jobs. This new identity ties them to Komeito through the figure of (former President) Daisaku Ikeda, spiritual leader of the Gakkai and founder of Komeito. When I talked to Japanese non-members about my research, for example, I heard a single common response, from professors to politicians, ‘They vote that way because Ikeda tells them to.’

This model of Komeito partisanship closely parallels the literature on partisanship in America. In The American Voter, Campbell et al. (Reference Campbell, Converse, Miller and Stokes1960) postulate that voters' partisanship stems from their social identity, not policy preferences, the way Southern conservatives voted Democrat because that was part of what it meant to be a white Southerner. More recently, Green et al. (Reference Green, Palmquist and Shickler2002) have renewed this view of American voters.

In The Japanese Voter, Flanagan et al. (Reference Flanagan, Kohei, Miyake and Richardson1991) find that this theory held true in Japan, too. Voters were not concerned about political issues – they became partisans as part of a social group. Reed (Reference Reed2003) updates this theory and specifically applies it to the Soka Gakkai, implying that members vote for Komeito as an expression of their membership in the Gakkai, not out of support for Komeito's policies.

While the conventional wisdom about Komeito may have been true in the 1960s, I find that its supporters have changed along with the rest of Japan's voters. Past research does show that this social identity effect dominated post-war partisanship, but de-alignment among Japanese voters has forced scholars to confront the weakening role of social identity in partisanship (Richardson, Reference Richardson1998). Based on direct observation of Gakkai practice, I argue that today, policy appeals are central to the organization's political justification and campaign outreach. Social ties remain important, but they function as enablers – membership brings people to meetings, where they hear policy messages, and social networks offer pathways for partisans to transmit policy messages to a larger audience. Ultimately, though, the messages are about policy.

In support of these arguments, I offer direct evidence – ethnographic reporting and new primary sources from Gakkai socialization efforts – and indirect evidence, testing the model's implications. In this case, the conventional model implies that: (a) Komeito and the Gakkai organization will not waste limited time at meetings justifying their policy positions to members, since members will vote Komeito regardless, and (b) references to Komeito that occur within the Gakkai context will emphasize social identity, not policy. By contrast, my model implies the opposite: the two organizations will spend time justifying their policy positions, and limit social identity-based appeals.

Since forming a ruling coalition with the LDP, Komeito has come under fire for pursuing policies contrary to its stated ideology (e.g. Yamada, 2003; Eto and Hichiri, Reference Eto and Hichiri2003; Sado, Reference Sado2005). If the current model of social identity voting is correct, then Komeito's voter support should remain unaffected, leading those authors to assert that Komeito's leadership has nearly unlimited freedom to change policies. On the other hand, if Komeito partisans justify their support by agreeing with Komeito policies, then contradictory policies are likely to create dissonance among voters and reduce the party's electoral support.

This produces a second testable implication of the article's theory. The conventional model implies that Komeito support should remain constant (paralleling Gakkai membership). By contrast, my model suggests that Komeito support should be high when the party supported a government that echoed the themes in Gakkai policy messages – during the Koizumi era – and decline when the coalition moved away from those policies – during the Abe era.

The bulk of this article, coming after a description of my research methodology, is divided into testing these two implications, looking at the content of political messages first, and the timing of electoral support second.

Fieldwork in the Soka Gakkai

The first phase of my research investigated why Gakkai members become Komeito partisans. As Koellner (Reference Koellner2002) notes, we know that organizational membership has a significant affect on Japanese voting, but we have few empirical descriptions of why. The ethnographic technique used here is more useful for that than large-N statistical analysis because it focuses on process rather than outcomes (Bayard de Volo and Schatz, Reference Bayard de Volo and Schatz2004).Footnote ² It involves close observation of social practices, giving them meaning by putting them in context, especially in situations like this where little is known about the subject (Fenno, Reference Fenno1990; Leblanc, Reference Leblanc1999; Silverman, Reference Silverman2006).

I conducted fieldwork in Japan during the second half of 2007. After securing the cooperation of the Gakkai public relations office, I attended approximately 30 regular meetings as an observer, usually with an escort from the organization. I supplemented this with individual discussions with organization officials, local district leaders, and rank-and-file members (without organizational escorts). We chose these meetings to present a diverse cross-section of the membership: weekly division (men's and women's) meetings of 10–30 people, zadankai monthly meetings were the divisions come together with 20–50 people, and occasional special events. Geographically, they were divided among three primary sites: Moriguchi, Hirakata, and Tanba. The first two are industrial suburbs north of Osaka. Moriguchi is home to Panasonic's headquarters and the meetings I visited were in blue-collar areas; Hirakata is less dense and the meetings I attended suggested a higher income level. Tamba is in Hyogo prefecture, divided into dormitory towns at the southern end to small farming communities at the northern end. Gakkai membership is not uncommon in these areas – Hirakata even hosts one of the Gakkai's two elementary schools. To provide contrast, I also attended meetings in eastern Gifu prefecture, where membership is said to be less common.Footnote ³ This wide range of meetings, I believe, provides enough leverage to make generalizations about the Gakkai membership.

In this article, I analyze messages that appeared in district-level Gakkai leaders' (chikubucho, shibucho) speeches, Komeito political videos, and organizational publications, comparing the relative prominence of group membership and policy preference. Messages that emphasize the personal connections between Gakkai and Komeito, or the breadth of Komeito support among Gakkai members, are group membership oriented. Possible examples of this include: insistence that ‘we [Gakkai members] are all in this together’, reminding listeners that Komeito politicians are Gakkai members, mentioning the Komeito in Gakkai publications, or juxtaposing symbols of party and religion. On the other hand, policy oriented messages are likely to: present information current political concerns, stress the Komeito's role in making policy, or deliberately avoid Gakkai symbolism. The relative frequency of these message categories provides evidence for their relative importance in member mobilization, telling us what moves Gakkai members to become Komeito partisans.

The second phase of my research investigated outreach to floating voters. Ideally I would have liked to observe outreach directly, but I was not allowed to attend meetings during an actual election campaign. When I pressed officials on the subject, they said it was a precaution to ‘prevent misunderstandings’, which I suspect is a legacy of the election law violation trials of the 1960s. I was able to attend meetings immediately after the July 2007 Upper House campaign ended, meetings during the run-up to 2008 campaigning, and meetings where members discussed outreach for elections in neighboring cities' elections, and I supplemented these observations with a written questionnaire in December of 2007. The instrument asked them to describe the arguments they used in voter outreach, coding their answers into three categories: (1) expressions connecting organization and party, which would only make sense if addressed to other Gakkai members, (2) references to their own support for Komeito, which suggests they are appealing to people who will be swayed by ties of friendship or blood, and (3) references to policy issues, which appeal independently of Gakkai membership or personal ties.

While this research challenges the conventional wisdom about Komeito and the Soka Gakkai, it is important to acknowledge the handicaps that scholars have labored under. The Gakkai's reputation for secrecy is deserved (Kawanishi, Reference Kawanishi1996; Shimada, Reference Shimada2007); my own research began after more than two months of negotiations with organizational representatives in Central and Western Japan before I finally secured permission to attend regular meetings, and even then I was excluded during election campaigns. Personally, I believe that this secrecy is a response to the Japanese tabloid press's relentless criticism, and I hope that my research will encourage the Gakkai and other scholars to cooperate in producing a more accurate picture of the organization and its members.

Justifying partisanship

Gakkai members' tendency to support Komeito is well established. In my own survey of 82 active members, every single one admitted to being a Komeito partisan. In the Japan Election Study (JES III) dataset, there is no question about Gakkai membership, but we can infer it from respondents' answer to the question of whether they belong to a religious group whose members support Komeito – every one who answered ‘yes’ to that question also supported Komeito. Whatever measure one cares to use, the power of Gakkai membership is clear. The reason for that partisanship, however, is not.

The Soka Gakkai has three primary channels for justifying Komeito partisanship to active members: direct appeals from local Gakkai leaders, Komeito outreach at Gakkai meetings, and printed publications.Footnote ⁴ During my fieldwork I was able to observe all three, and each message I witnessed followed a similar pattern: (1) present a social problem, (2) describe where different parties stand in relation to that problem, and (3) explain why the Komeito's stand on the problem is best – for the country, not for the Gakkai. In this section, I discuss each of these channels in turn.

Outreach and the floating voter

This analysis would be incomplete without examining how Komeito partisans in the Gakkai reach out to independent voters. In spite of how many authors reiterate the 1960s-era image of Komeito as a party of Gakkai members only (e.g. Baerwald, Reference Baerwald1986; Stockwin, Reference Stockwin, Ishida and Krauss1989; Richardson, Reference Richardson1998; Curtis, Reference Curtis1999; Yoshikawa, Reference Yoshikawa1999; Sado, Reference Sado2005), scholars who look more closely at the Gakkai and Komeito know that the popular image of the party's exclusivity is wrong. Komeito partisans only account for about half of the party's electoral support.

In 1992 – before the Komeito began Upper House electoral cooperation with other parties – it won 6.4 million votes, but a survey conducted by the Yomiuri Shinbun on 1 May 1991 found that only 3% of approximately 100 million voters are Komeito partisans (this figure also includes those who lean toward Komeito rather than supporting it).Footnote ⁷ Data from the 2000 JEDS study suggest that today 3% of voters identify themselves as Komeito partisans, but the party received 7.6 million votes in the 2000 upper house election. Past research documents the existence of Komeito support among non-partisan voters as well: White (Reference White1970) included similar data from the 1960s in an appendix, Richardson (Reference Richardson1974) alludes to them in a footnote, and Watanuki (Reference Watanuki and Flanagan1991) and Reed (Reference Reed2003) also cite them. Subtracting the 2 million or so ballots that we can attribute to strategic voting by LDP members, that leaves more than 2 million ‘friends’, who vote Komeito without declaring themselves partisans.

Among Gakkai members, the act of recruiting a Komeito voter is referred to as f-tori, which literally means ‘gathering friends’, and in practice means they contact acquaintances and urge them to vote Komeito. Unfortunately, while this practice is well known, little research has been done on it, though we do know that members of the women's division are the most vigorous campaigners (Soka Gakkai Problem Research Center, 2001), and that the outreach does not target inactive Gakkai members (Harada and Murayama, Reference Harada, Murayama and Igami1995).Footnote ⁸

Here again, we see social ties playing an enabler role. Members do not cold-call random voters, they reach out to people with whom they have prior relationships. Superficially, it is easy to imagine that these relationships motivate independent voters to support Komeito. For example, I heard some female members fret that they were the only believer in the family, and a husband in a labor union might decide that it is easier to lie to his boss about voting Socialist than it is to lie to his Gakkai wife about voting Komeito. Alternatively, an employee might vote Komeito if the company owner makes his Gakkai affiliation public.

Nevertheless, we must recognize that this practice does not confirm that independent voters choose Komeito because of their social ties. Consider the common refrain from non-believers that ‘they [Gakkai members] only become my friends at election time’, or that a member of one group I attended saved his high school yearbook so he could remember who to call. A report on the women's division campaign practices includes a photocopy of a social network diagram, designed to help members remember all the possible connections they might have, from pre-school classmates to workers at the local dry cleaners (Soka Gakkai Problem Research Center, 2001: 83). These are thin ties at best, and it stretches credulity to imagine they can carry the weight of political partisanship.

The policy-driven model in this paper suggests an alternative hypothesis. Social ties are the message carrier, policy appeals provide the message content. This is similar to the partisan socialization process discussed in the previous section: policy messages provide the content for converting members, social ties provide the impetus to carry that message among the electorate. To see how this works, consider this account from a woman who talked about reaching out to an elderly co-worker during the July 2007 election. Her testimonial centered on listening to him, and being there to help when his children were far away – politics only entered as an aside, a brief comment that the man appreciated her claims that the Komeito works to help people in his position. She did not say ‘vote Komeito because I would appreciate it’, she pointed out how Komeito's plan to lower insurance payments for elderly citizens would benefit him. Her social tie to the man gave her an opportunity to present a Komeito policy message, to which he responded positively. This is just one anecdote, but it echoes the results from a larger study of Komeito outreach.

To gather data, my questionnaire asked respondents what they say when a potential target asks: ‘Why should I vote Komeito?’ This was an open-ended question, and the replies were coded into categories, with the results shown in Table 1, where each cell indicates the percentage of respondents who used to that category of argument.Footnote ⁹ The first category (column 1) is policy-based appeals, like that of the woman mentioned above. The second category is claims about group identity, which might be used to persuade inactive Gakkai members to vote Komeito. The third category of responses relies on personal ties, which the conventional wisdom suggests should predominate.

Table 1.

Arguments used to recruit floating voters

Notes: Total respondents: 82.

The percentages add up to more than one because respondents provided up to three arguments apiece.

Contrary to the conventional wisdom, these survey results provide strong support for a model of policy-based electoral support instead of one based on personal ties. Almost all respondents make policy-based appeals. The most common pattern was to insist that Komeito works for the common people first (49%), as suggested by this quote from an older man:

After describing how the party uses its position in government to bring real change, a young member sounds the same theme:

Others flesh that claim out with assertions about specific Komeito policies (combined to 50%), like this response from a college student, addressing his peers more pressing concern:

An older women emphasizes issues that her peers are likely to be more interested in:

As these few examples suggest, Komeito appeal is focused on the party's concrete policy successes. It is apparent how closely the respondents' outreach strategies match the Gakkai's partisan socialization strategies. Members learn to support Komeito by hearing about the party's work on behalf of regular people, and the specific policies it proposes. Once they internalize that message, they repeat it to a larger audience of non-partisans during political campaigns.

The third most common response was to point out how hard Komeito representatives work. On occasion, these respondents fall back on a generic ‘they're working hard, so please support them (gambatteimasukara, yoroshiku onegaishimasu)’ strategy, but more commonly they say they'd tell listeners about the party's ‘achievements’ (jisseki).

On the other hand, very few people relied on personal ties. For example, only 7% of respondents used personal ties to justify asking friends to vote Komeito, just six out of 82 people. These few respondents did made it clear that they were making a personal request of their listeners, like this college student division leader does:

One more person (seven out of 159) based their appeals on group membership, confirming my anecdotal evidence that Harada and Murayama (1995) are right about campaigning not being aimed at inactive Gakkai members.

Earlier in the section, I described how members use social ties as opportunities rather than as justifications, and one final quote from the survey shows this in action. The respondent, an unmarried woman, says:

She reconfirms that the listener has enough of a personal connection to care what she thinks, and then she uses that opportunity to present the Komeito's policy appeal.

Electoral support

Without access to evidence of this emphasis on policy, contemporary authors overestimate Komeito leaders' freedom to choose their policy agenda, even labeling Gakkai members as mind controlled (Eto and Hichiri, Reference Eto and Hichiri2003), ensorcelled (Yamada, Reference Yamada2004), and deceived (Sado, Reference Sado2005), incapable of making independent political choices. The normally staid Japan Times says that Komeito is ‘addicted to ruling power’ and unwilling to pursue its promised agenda in the face of LDP opposition.Footnote ¹⁰ Even among academics, the belief that Komeito voters will continue to support the party regardless of its policies, that the party is ‘unfettered’ by policy constraints (Reed, Reference Reed2003) is a fixture in the literature; it appears in my own earlier work as well (Ehrhardt, Reference Ehrhardt2008). Only Metraux (2003) and Shimada (Reference Shimada2007) suggest otherwise, and they limit themselves to pointing out that small-scale dissension within the Gakkai may potentially pose constraints.

Mitchell Killian's (Reference Killian2007) work on abortion policy in the United States proposes that the strength of policy constraints depends on how a party mobilizes supporters. Parties that rely on social appeals are likely to have more leeway, because policy divergence doesn't impact members' central beliefs. Parties that rely on policy appeals to gather votes are more constrained, since hypocrisy is readily apparent and strikes directly at members' expressed rationale for voting.

Applying this to the Japanese context predicts that Komeito leaders operate under constraint, given that their electoral support comes from policy appeals. To test this prediction, I show how it helps us understand the fluctuation in Komeito votes over the last decade, which increased from 2001 to 2005, then dropped in 2007.