Building a Functional One-Party Democracy

Why Japanese Voters Make Decisions Differently from Americans, and Why Different Isn’t Necessarily Odd

By Jennifer Dodgson and Ts’ang Chung-shu

I think there are many people with the same perspective as me. The most important thing is to vote for parties and politicians that will realistically improve your life, rather than based on ideals.

- Japanese voter, 2018

Japanese politics is different from US politics. And not in the quirky, hand-wagging way that European politics is different. Japanese politics is not US politics expressed using a different idiom, but rather a world of its own.

When considering our own (mundane, unexotic) political institutions the default approach is to understand all behviour as a response to a self-correcting system of incentive structures. In the case of a political world that is not immediately familiar, however, there is a strong temptation to fall back on culture as a catch-all explanation for its many and disparate phenomena. From business manuals explaining the bushido spirit, to scholarly analyses of neo-Confucian corporatism, Japanese exceptionalism has become a literary genre of its own, as popular in Japan as elsewhere. No explanation is required for any cultural phenomenon, because culture itself is the explanation for everything. The result is that intelligent people with letters after their names will readily accept culture as the reason behind the LDP’s dominance of Japanese politics or the proliferation of second- and third-generation politicians in the Diet, when they wouldn’t dream of using it to explain away a Rand Paul filibuster or a government shut-down.

Part of the problem lies in the fact that while a large and interesting body of work exists to analyse Japanese politics in purely empirical terms, it is a) written almost entirely in Japanese and b) published in exceptionally dull psephological journals. Someone with a lucrative career as a State Department panjandrum/television talking-head can hardly be expected to jump upon every new edition of Senkyo Kenkyū to keep up to date on whether SUR methods are out-performing OLS in predicting economic voting patterns, and his audience would find the information tedious even if he did. Far better for all concerned to evoke the unfathomable inscrutability of the mystic East instead…

Nevertheless, while individual papers may be somewhat heavy-going for a non-specialist reader, the picture that they build up is an interesting one. Rather than depicting a world dominated by a ingrained adhesion to Confucian conformity, it shows instead a fully-functional system of checks and balances, albeit one whose mechanisms are entirely different from those familiar in the US and Europe. The world it reveals is driven not by faith in ancient cultural norms, but by clear and calculated self-interest.

So how does this world operate?

Firstly, and most-importantly, Japanese voters are largely non-ideological. While they may have strong opinions on political subjects, these opinions often have little or no correlation with one another. It is relatively easy to predict a US or a European voter’s position on public health if you already know his position on law and order, but this is seldom the case in Japan.[1] Moreover, opinions themselves are not necessarily a good predictor of voting patterns. In US electoral analysis a voter’s approval of a policy’s execution is generally dependent upon his approval of the policy itself; in other words, voters who approve of the way Obamacare was carried out are considered as a subset of those who approve of Obamacare in general.[2] Japanese voters are far more likely to take the position that a policy that they themselves dislike can nevertheless be admired for being well-executed. In other words, it’s not uncommon for voters to dislike what Shinzo Abe does, but like the way he does it.[3]

So if Japanese voters do not vote ideologically, how do they decide how to allocate their ballots? This is a much trickier problem to solve. Research using statistical and econometric methods tends to suggest that social and physical proximity to the candidates plays an important role, as does the state of the economy (see the notes at the end of this article for more detail). Nevertheless, any situation in which dozens of factors go into a making a single decision — such as voting — is always going to make it difficult for statisticians to decide where the deal-maker and deal-breaker issues lie.

Our research team blundered into this problem in 2018. We are average statisticians at best, and certainly incapable of designing anything like the complex econometric models that are currently popular in election research. This route being blocked to us, we decided that if we wanted to find out how voters make decisions, the simplest way would be to ask them.

In 2018 we published a large-scale open-ended survey on the Japanese freelancing platform, collecting responses from 500 citizens of voting age (reduced to 451 to weed out nonsense answers, duplicate IP addresses and so on). Participants were paid 100 yen for their time, and asked two questions:

1. What do you take into consideration when voting in general elections? (For example: policies, party image, local issues, the economic situation, the personality of the leaders etc.) How do you make up your mind?

2. From your point of view, do you think that other voters decide in the same way that you do? What do other voters take into consideration? (For example: policies, party image, local issues, the economic situation, the personality of the leaders etc.)

(The aim in asking the second question was to identify incidences of social desirability bias among the answers to question one; people are generally happier to attribute behaviour of which society is likely to disapprove to others than to themselves.)

Conventional wisdom tends to hold that there is little point doing open-ended questionnaires, given the short attention span and low intelligence of the average respondent (though researchers tend to phrase things rather more diplomatically), and we were warned repeatedly prior to commencement that our survey would turn out to be an expensive failure. In fact, the responses were not just useful from a research perspective, they were also fun to read. Indeed, part of the reason we wrote this article was to argue for more use of open-ended survey research. (For interested readers, we have made the raw data available via the National University of Singapore’s website. A detailed methodological explanation has also been included at the end of this article.)

What We Discovered

While most responses mentioned a similar set of factors (the economy, social services, education), these factors were not always framed in the same way. So, while two different individuals may have mentioned economic policy as contributing to their voting decisions, one may have framed these issues in terms of managerial competence, another may have focused on the personal advantages he expected to gain. There were five main trends:

A) Managerial framing: when people talk about making decisions based on their evaluation of parties’ or politicians’ competence and ability to carry out their planned policies, or describe politics in terms of responding to contingent problems. (Examples: “I consider whether politicians have concrete policies or just talk; I think it’s important for parties to be able to implement their policies exactly”, “whether, looking at current economic and local problems, they’ll be able to fix what’s broken without undoing the good aspects”.[4])

B) Advantage framing: when people describe decisions being made on the basis of individual benefits that are likely to result, either in the form of pork-barrel projects or of policies that advantage a group of which one is a member. (Examples: “taking my child into consideration, I’d like to see free education implemented, so I think I decided based on that policy”, “I guess maybe it’s tied up with money; whoever’s done favours for your own local area”.)

C) Ideological framing: when people mention beliefs, opinions, philosophies or religion as a factor in decision-making. (Examples: “people whose beliefs are relatively left-wing will probably go for the Communist Party”, “since Soka Gakkai and New Komeito get support, it seems like decisions might be made along religious lines”.)[5]

D) Social framing: when people describe themselves or others as having been influenced by the opinions of those in their social circle, work-related block voting campaigns, local loyalties, or public opinion in the general sense. (Examples: “if you went to school with a candidate or a Diet Member, even if you don’t know them directly, or they’re just a friend of a friend, I think you’ll feel like you’ve got a connection and support them”, “I don’t really know about politics so I just vote for whoever the people around me favour”.)

E) Media framing: when people describe themselves or others as having been influenced by the media. (Examples: “there are many people who are very image-oriented and are influenced by television and mass communications”, “I think people are influenced by the images of parties that they get from television, isn’t that why it’s so easy for celebrity Diet Members to win?”)

When respondents were describing their own preferences, by far the most frequent position was the managerial one. Almost 50% of respondents described themselves assessing parties or politicians based on their perceived levels of competence. By contrast, fewer than 10% mentioned ideology as a factor in their own decisions, and (as we shall see below) barely any of these respondents expressed any partisan attachment. When describing others’ voting decisions participants were more likely to mention the advantages that they hoped to gain from backing a particular party or candidate, and to mention social influences or the media as factors.

The Dog that Didn’t Bark: Non-Ideological Voters

Unsurprisingly, given the existing research mentioned above, our respondents were almost entirely uninterested in ideology and grand political principles. Eleven people mentioned “conservatism” (保守, hoshu), but only one of these responses could be considered a declaration of partisan affiliation. The others were more nuanced, often implying small-C conservatism and describing it in relatively neutral terms:

“The countryside is conservative and the LDP’s power base is strong, so I vote for anyone but the LDP, choosing based on party image and whether it’s riding the wave of the times.”

“I decide based on gut feeling. Even though the LDP is conservative and uncompromising, over time it has always provided peace of mind.”

“Because there is no final answer in politics, I always try to vote for whichever of the opposition parties seems to be doing best at the time. The LDP is just as conservative as it used to be, but its quality has declined. Practically, it’s to do with the quality of their human resources.”

Seven people mentioned broadly conservative opinions, mainly concerning foreign relations, but just one explicitly linked this to partisan support. Only one respondent self-identified as liberal, though twelve mentioned holding broadly progressive opinions (mainly concerning redistribution and pacifism). Left-right alignments were only mentioned six times, with just two respondents saying that they affected their own voting choices. While twelve respondents mentioned beliefs (思想, shisō), only three said that they influenced their own decisions. One respondent actively disapproved of their intrusion into the political process. Similarly, eleven mentioned ideals (理想, risō), a similar split concerning whether they constituted a good basis for electoral decisions.

“I think there are many people with the same perspective as me. The most important thing is to vote for parties and politicians that will realistically improve your life, rather than based on ideals.”

“Whether policy is weak or not is my top priority, but I also pay attention to the ideological image of a party, whether it’s more left or right wing.”

“Probably what I take into consideration isn’t that different from what other people consider. But I don’t really understand people who are fixated on particular beliefs.”

“Most people vote based on personal profit and loss, rather than on ideals. Or based on the hot air produced by the mass media. People who are busy with work and personal business can only give a few minutes to paying attention to politics.”

Linked to this relative lack of interest in ideological combat, participants also showed an almost total absence of party affiliation.

Loyalty without Party Loyalty

While many Japanese voters regularly vote for the same party across all elections, existing research shows that party loyalty, insofar as it is understood in the US and Europe as a sense of identification with a “team”, is almost unknown in Japan.[6] The Japanese World Values Survey team even went so far as to drop the standard question on party loyalty from their poll as it was “not relevant in Japan”.[7]

Our data tended to confirm this trend. When respondents mentioned a party or an alignment, it was usually in the context of denying any affiliation. So, when participants mentioned the LDP it was usually in reference to the LDP as a default option or as a force to be hedged against.

“In Japan the current ruling party, the LDP, will always win. However, since life under the LDP’s politics is getting hard, I don’t support them. I try to vote for whatever opposition party is likely to beat the LDP.”

“I try to balance things so that one party doesn’t get too prominent. (For example, voting for the Communist Party so that the LDP doesn’t become like a one-party dictatorship.)”

“Mainly it’s to do with policies and party image, but since I think that probably the only decent party that exists in Japan is the LDP, I more or less just vote for them without thinking about it in elections. (It’s not that I support the LDP, just that the other parties are too terrible.)”

However, ideas of loyalty, identification and attachment were far from absent. Many respondents described local connections and personal proximity as being a factor in their voting decisions, though people were more likely to describe others voting in this way, rather than say that they did it themselves.

“In the provinces, people vote for someone who’s powerful locally, or who’s had a close relationship with that area for a long time. In large cities, like ordinance-designated cities, people just vote based on gut feelings as a result of listening to street speeches and suchlike.”

“[I vote based on candidates’] parties’ policies, their local relationships, which companies support their activities, their previous work, whether they’re a second-generation politician or not, whether someone from my local area is running or not.”

“Regardless of policy content and party image, I think there are many people who vote for the candidate whose message has been highlighted by the company where they work, or who they’ve had a personal acquaintance with for a long time.”

“I don’t know much about other voters. Even among the people I know, we don’t talk about which party they vote for. However, I feel as though people vote for the party that their company supports because they want to fit in, or if the people around them have supported a party for a long time. It’s not a matter of individual volition.”

“I guess political parties’ images aren’t very important. I get the impression that the further you go into rural areas, the stronger group consciousness gets, and people vote for the same fixed parties and candidates.”

“(…) If you’re working for a company that supports a political party, I think that if you’re not interested in politics you’ll vote according to the company’s instructions. If people have politicians among their friends and acquaintances I think they’ll vote for them so that they can receive benefits themselves.”

In other words, many voters prefer to back a person to whom they have some sort of personal attachment, regardless of the party that he or she has chosen to join. When a politician’s party did enter into the calculation,it was either on the basis that a politician belonging to a strong party is more likely to have the power and connections to implement his or her promises, or related to the idea that a politician can be judged based on the company he keeps.

“In local politics, local government is dependent upon strong politicians, and it’s nothing to do with parties. Because it’s to do with things that are close to people’s everyday lives, parties don’t matter and you just rely on a strong local representative. Even if they’re from the opposition party, they can still be effective. But in national politics the LDP is still better.”

“Regarding candidates, I emphasise character. Because of this I watch political shows as much as possible and try to go to street speeches as well. In addition to this I pay close attention to the philosophy, achievements and policies of the political party to which the person belongs.”

The absence of large ideological and party-identification voting blocs is linked to the way in which political issues are perceived. The majority of participants did not see social problems as having a single, immutably correct answer, as tends to be the case among ideological groups. Instead, they tended to view politics as a problem-solving exercise demanding managerial skills.

Politics as Management

The economy was by far the most-cited policy issue among respondents (mentioned 123 times). In contrast with U.S. and European economic debate, however, not a single person chose to describe economic issues in ideological left/right terms. Instead, perspectives were generally framed as reflections on the government’s economic management skills or on the respondents’ own economic interests. The same thing can be said of tax policy, which was regularly mentioned in reference to personal circumstances but never cited as a matter of ideological principle.

Other policies that were frequently mentioned were those affecting specific groups: young people, the parents of young children, and senior citizens. As in the case of economic policy, these were described as ways in which a government’s ability and decisions were likely to influence ordinary people’s daily lives, not as arenas for ideological combat.

Only three policies that are traditionally viewed as ideological issues were mentioned: foreign relations and defence(mentioned by eleven people), constitutional reform (mentioned by ten people), and nuclear power (mentioned by six people). However, even these issues were not always described in ideological terms. Only three people framed their comments on foreign policy from an explicitly ideological perspective, with two doing so on the constitutional reform issue, and none on the nuclear issue. The majority either framed them neutrally (just mentioning them as being important), or described them in terms of managing contingent phenomena:

“[I vote based on] foreign relations, defence, the economy, personality and ability to take action. After that, someone who has a good grasp of the current domestic and international situation, and the capability to explain what is most important.”

“I vote based on policy. If possible, I’d like to get rid of nuclear power plants, so I’ll vote for whoever will do that.”

“[Other people vote based on] economic policy, wages, taxes, whether a party uses money cleverly to improve people’s lives. A party that’s skillful in dealing with international relations.”

Many people described politics from a purely administrative point of view, as being the business of steering the country towards optimal outcomes under conditions of uncertainty. Rather than being tethered to ideological principles, a good leader was expected to be able to develop pragmatic policies base in response to present circumstances, and have the persuasive abilities and strength of character necessary to assure their adoption and implementation by the Diet, the bureaucracy, and ordinary citizens. In many cases, people said explicitly that having the ability and capacity to carry policies out was more important than the policies themselves.

“I think party image is crucial. Of course, policies are important as well, but I definitely think that party image is important because from a party’s image and its past performance you can tell whether what it’s proposing is feasible or not.”[8]

“I think it’s decided by considering whether candidates and parties demonstrate political attitudes (manifestos, policies) that are well-adapted to the current situation of the country or the region.”

“Firstly I consider what goals they’re setting, then I vote based on whether or not they can be realised.”

“The most important thing to consider is how realistic the policies are. After that I look at the person’s character. People who are too theoretical are no good, so they have to have good charisma.”

“It’s the ability to implement policy and leadership strengths of the representative. If a representative proposes policies without the strength to carry them out that’s just opportunism and they won’t deal with the problems that really need solving.”

This idea of politics as the art of the possible was related to two other trends in the data — the importance accorded to past success in judging politicians, and the need for leaders to be in tune with the Zeitgeist.

Nothing Succeeds like Success: Extrapolating Future Performance from Past Achievements

Evaluation of past performance is known to feature as a significant component in voting decisions across all electoral systems, and was a recurring theme among our responses:[9]

“First of all I consider whether or not the policies of the party currently in power have been effective. How was their economic policy? There are various issues in Japan at the moment, but the economic uncertainty is the main one.”

“While policies are very important, I take into consideration how long the party’s spent in power. At the end of the day, I consider the ability to implement policies and whether or not the feasibility of their political activities is guaranteed.”

“I focus on the current government’s economic results. If they’re unsatisfactory, I look at other parties’ economic policies. I won’t vote for parties that just criticise the current administration and don’t have a clear vision.”

“My first priority is policies. In addition to this I judge based on what the party has done so far, and whether its policies are really feasible.”

However, this was not merely a case of punishing poor performance or evaluating competence based on results. There was also a clear trend among the responses to consider past successes as providing a practical foundation for future projects. A politician who, by his actions, has succeeded in building up a local base, uniting a network of supporters, or climbing the party ladder, is seen as having accumulated resources that will enable him to accomplish even greater things.

“Whether the policy being presented is really going to be carried out. In a political situation where a party that hasn’t been able to acquire a good image finds itself ruling over the people, without popular support it can’t achieve anything. Whether the people around me have a good image of the party.”

“Policies, whether or not past election pledges have been fulfilled, candidates’ qualities as politicians, whether they’ve been recognised by a supporting party or not.”[10]

“I check carefully to see how deeply rooted the candidates are in the local community, and whether they’re carrying out policies to benefit the region. In addition to this, I look at the party image, the organisational connections and the consciousness of the people standing for election. Because if they’re an independent, they can’t contribute to the revitalisation of the area.”

This, in turn, was linked to the idea that being a member of a successful group confers advantages upon its members, which in turn attracts more people to the group, making it yet more successful and allowing it to bestow even greater benefits upon its members — a systemic explanation for the LDP’s self-perpetuating dominance of Japanese post-war politics.

“As far as local politics are concerned, if you live in an LDP fiefdom, they’ve got the upper hand and if you want to profit from politics you’ve got to be on their side. From a long term viewpoint that’s how voting works, and in such circumstances people who do otherwise are just left swinging in the wind.”

“(…) I think that in this country there are many people who want to be in the ‘winning group’ (especially old people). Even in elections, because there are ‘winners’ and ‘losers’, people chose to go for the ‘winning side’…”

This seems to be a reversal of the tendency among US and European electorates to want to keep politicians on their toes by “kicking the rascals out” every so often. While a US voter who judges that a particular party or politician has grown too comfortable in office is likely to see this as a reason to vote against them, a Japanese voter is more likely to lean in the opposite direction, and see this as a reason for keeping them in place.

This is connected to another noteworthy phenomenon present among the responses: the idea that power exists outside of and independently from the legal frameworks that grant it to office-holders.

Power Detached from Institutions

Respondents frequently described political power as distinct from the various laws and offices that confer it upon individuals. If anything, the perception of power as being the product of informal social contacts or personal charisma was more prevalent than the idea of power as being dependent upon bureaucratic authority.

Politicians with strong networks or grassroots support groups were frequently described as being powerful whether they won or lost any given election (with their pre-existing power also being seen as an aid to victory). Similarly, perceptions of parties’ power did not necessarily depend upon their having won the previous election.

“I consider policies, party leadership, political power etc. First of all I ask about policies. On top of that, I judge based on what kind of party leadership and political power they have. Even if a party has a coherent viewpoint, if it doesn’t have popularity and power nothing will change.”

“I’m different, but it’s common for the majority of people in the Japanese countryside to support someone who seems like a strong person. Concerning national politics, it tends to be centred on urban areas; in local politics there’s a tendency to avoid risks, so people aren’t really keen to support opposition parties.”

“I think that having a strong grassroots support base is more important than policies or the economic situation, because in the area where I live the same candidates often get re-elected year-on-year. I think that other voters generally vote for the parties they prefer, rather than actual policies.”

“I think people look to celebrities or people that have a strong local power-base, not people who’ll leave their mark and take care of citizens.”

“Familiarity with the parties. In the countryside precedence goes to your own feeling about which party has a strong local power base, rather than your opinions.”

Power was also regularly described as an inherent attribute of certain individuals, being seen as a product of charisma, strength of character and intellect.

“I look at the personality and temperament of the candidates. Image is also an important thing to think about. Then there’s whether or not they’ve got the power to take action. On the campaign trail I guess it can be hard to tell, it’s more of an intuitive thing, isn’t it?”

“Generally I think it’s the party image. Looking at the party high-ups’ statements and actions, you can easily tell who has a strong will to benefit the people, and get a feeling about who has the capacity to take action — that sort of thing.”

“Considering the political philosophy, achievements, and behaviour of individual politicians and parties up to now. Choosing parties and politicians that have the power to maintain domestic well-being and establish appropriate relations with other countries.”

“Of course it’s policy [that is important], but even if a person has been criticised, if they have the power to carry the population with them I think that’s important, so I consider the qualities of the party leaders and their achievements.”

“I think many people choose people with power, or who are well-known — that sort of thing. Some people respond to personal requests for their vote, some people just vote for the same person every time.”

A similar trend appeared in comments regarding popularity, which was often described as force in and of itself, separate from the social or institutional mechanisms that create and convey it.

Riding the Wave: Zeitgeist and Popularity as a Source of Influence

Popularity was seen as a factor liable to help ensure the implementation of a party’s or a politician’s policies, as well as proof of managerial competence. It was seen as a reliable indicator that a politician had the charisma and the aura of success necessary to attract the support of others, which in turn would ensure the success of his proposed policies.

“The leader’s charisma and career history. The true nature, goals and meaning of the policies. In addition, the visuals and their way of speaking, and whether they’re capable of drawing people in. If you can really believe what they say, and they’re not just talking big. A refined person.”

“I think it’s probably a person with a strong will, who works for others and not just themselves, someone who exerts all their efforts on behalf of the country, someone who can really attract the adhesion of those around them and create enthusiasm — that kind of person.”

Popularity was not mentioned only in the context of leader-follower interactions, but also in more abstract terms, as a reflection of the Zeitgeist. When discussing trends in popularity, respondents often used variations on the phrases “流れに乗る”(nagare ni noru, riding the current) and “風に乗る” (kaze ni noru, riding the wind). Respondents used these phrases to imply both the voters’ tendency to favour already-popular politician, and an ability on the part of a candidate to sense the direction in which public opinion is running and to adapt his or her policies and presentation accordingly:

“I vote mainly according to parties’ policies and the flow of public opinion. Before the elections I get information from news reports etc. It bothers me if there are too many scandals.”

“I think that at the end of the day people just go with the flow of the situation. Everyone’s voting for this one person, and I don’t have any particular opinions, so I’ll just vote for them out of conformity.”

“There are too many people voting based on ‘what’s in the air’ without applying clear criteria or indicators. If that happens I think there’s a risk of populism and authoritarian leadership.”

“It’s mainly to do with what’s in the media and news reports, it feels like a popularity-based vote, going with the flow of the times — that sort of feeling. Anyway, there’s not much interest in politics, it seems like.”

The same terms were also used to describe the extent to which a politician or a party could be said to fit in with the spirit of the age, with the implication that a good “fit” will produce good leadership. A politician who corresponds with the Zeitgeist, it was implied, would achieve popularity and success naturally. This idea would appear to connect with the conception of politics as a contingent business: a politician with his finger on the pulse of the current age will be better able to work out practical solutions to its problems.

“My number one priority is the parties’ policies. I see whether the content is appropriate to the times. On top of this, I consider leaders’ personalities.”

“It was based on the image of the party when I was deciding. When the LDP was declining and the DPJ had gained momentum, I voted for the DPJ. Looking back now, I think I made a mistake in voting for them.”

“I think that many people vote based on the image of the party just before the election. For people who don’t support a particular party and haven’t had time to listen to their policies, voting based on the parties’ images according to public opinion is the safe option.”

“You get a sort of hunch about who’s doing well at a particular time. For example, lately it seems like it’s the Party of Hope. I think about the good and the bad aspects, and the main think is whether or not the party leader looks like someone who could defeat the LDP.”

“I decide primarily based on political stability, I think. You should judge based on a holistic perspective, including policy and leadership qualities. However, sometimes you ride the wind of new politics. At such times, it’s not rational, but you now feel that the continuation of the current administration is the riskier option.”

Conclusion: Elections as Performance Reviews

Only a small proportion of our participants saw elections in terms of ideology or principles. Instead, they tended to regard the electoral process as being something akin to the annual performance reviews to which civil servants and company employees are regularly subjected. Policies were not generally judged on whether they were ideologically right or wrong, but whether they were appropriate to the time and place. Similarly, politicians’ ability to carry out their plans was regarded as being at least as important as having good plans in the first place. Previous successes were considered not merely as a proof of political skill, but also as furnishing social and economic capital to be deployed in the advancement of future projects. As such, popular support was, in itself, considered to be a reasonable indicator of future performance: an adept politician who is in tune with the Zeitgeist will be able to build up a strong personal network, and thus find it comparatively easy to implement his or her policies. In this way, popularity becomes a self-reinforcing cycle, with followers being attracted exponentially as a politician’s support increases. In short, the average respondent appeared to share Otto von Bismarck’s view of politics: that “the statesman’s task is to hear God’s footsteps marching through history, and to try and catch on to His coattails as He marches past.”

While many participants were not shy to admit that they voted based on their own interests, the more socially acceptable position seemed to be that electing a politician who is a skillful manager is in the interests of the nation as a whole. Such a person would be able to run the country so as to ensure its — and by extension its inhabitants’ — success. The over-riding impression given by the responses was that voters hope to be able to group together behind a skillful manager and enjoy increased collective prosperity as a result, rather than banding together in partisan groups to defend their turf and beliefs against internal threats and competitors.

This vision of politics as a search for the best club to join helps, in turn, to explain how Japan can function as a democracy while almost always returning the same party to power. A successful party has the human and financial resources to improve the lives of its followers; by doing this it attracts more followers, and thereby gains more resources to redistribute — a self-reinforcing cycle. It is the moment at which a party is no longer able to provide economic benefits that its hold on government fails: the LDP has only fallen out of power following economic downturns, when it was unable to make good on its promises to ensure the continual improvement of citizens’ daily lives. Outside of economic crises, minor parties tend to find themselves locked out. Not only do the highest-skilled candidates and the richest donors tend to gravitate towards their better-established rivals, but the only way to succeed politically is by redistributing resources, while the only way to gain resources for redistribution is through political success.

In US and European electoral systems, the opposite mechanism tends to apply. Voters habitually judge any group that is perceived as having become too powerful as a potential threat to the rights and interests of those who do not belong to it, and vote for whoever seems best placed to undermine its dominance. Small parties and maverick candidates can thus sweep to power unexpectedly if they successfully present themselves as being on the side of the underdogs against the elites.

However, while the Japanese system is certainly different from the US or European ones, it is not so different that the only way to explain it is by recourse to aged cultural tropes. Indeed, from a strictly objective point of view, the Japanese political system can often seem much simpler to understand than its US and European counterparts, which rely on the counter-intuitive use of institutionalised conflict to prevent the overwheening dominance of any particular group. Commentators should beware of the natural tendency to treat the familiar as the default, and anything that differs from it as an exotic cultural quirk. It is, after all, not difficult to imagine a Japanese-dominated alternate history in which silver-haired academic NHK pundits are busy explaining to domestic audiences the inscrutable inwardness and eccentric rites that govern American politics.

To contact the authors, please email, or visit

Annex I: Existing Literature

The first noteworthy aspect of Japanese voting behaviour is that ideological voters make up only a small part of the electorate. Hirofumi Miwa (2014) found that fewer than two in five Japanese voters cherished any beliefs that could be described as ideological, and only half of these held positions that appeared to reflect a conventional progressive-conservative schema. Furthermore, a substantial proportion of respondents appeared merely to be endorsing what they saw as majority opinions on the issues mentioned, rather than expressing profoundly-held individual beliefs (Hirofumi Miwa, 2014).[11] By contrast the Pew Research Centre found around that around 70% of Americans hold an ideological viewpoint that places them somewhere on the conventional conservative-progressive continuum (Dimock, Kiley et al., 2014).[12] Yoshihiko Takenaka found that rather than heralding a switch towards a U.S.-style bipartisan ideological competition, the electoral reform of the 1990s was, in fact, followed by a decline in levels of both ideological belief and ideological influence on voting decisions (Takenaka, 2008).[13] These findings can be linked to another phenomenon that has been noted by various observers: the fact that the LDP has moved right in recent years, while voter positions appear not to have changed (Kago, 2015).[14] Takuya Murakawa demonstrated that the rightward shift was the product of existing Diet Members becoming more conservative, rather than the result of voters replacing moderate Diet Members with more right-wing ones (Murakawa 2014).[15] This coincides with findings by Yoshihiko Takenaka, Masahisa Endo and Willy Jou suggesting that as Parliament has grown more ideological — with many more LDP members openly expressing right wing positions — voters have becoming more materialistic, interested more in outcomes and a less in matters of principle (Takenaka, Endo and Jou, 2015).[16] The fact that Diet members feel themselves to be at liberty take up explicit ideological positions in the face of popular disinterest without fearing electoral sanction appears to indicate that voters’ priorities lie elsewhere.

These trends are compounded by the fact that for large segments of the electorate — and particularly for LDP supporters — foreign and military policy is simply not a deal-breaker issue (Kimura, 2000).[17] When Ichiro Miyake looked at the comparative importance of various issues to Japanese voters in 1999, only 1% mentioned “foreign relations”. Defence and constitutional reform were either not mentioned or mentioned by so few respondents as to be statistically insignificant (Miyake, 1999).[18] Similarly, while Sadafumi Kawato found defence and security to be the subjects upon which public opinion was most divided, only a small proportion of the electorate cited them as being important factors in their voting choices. It is also interesting that while they had a significant effect on vote choice among opposition supporters, none was observed among LDP supporters and independents (Kawato, 1988).[19]

It would appear, then, that actively disagreeing with the LDP on defence issues does not necessarily prevent people from voting for their candidates. Kimura searched through several decades’ worth of electoral data for “deal breaker” issues. He found that in only one of the elections studied (1972) was defence raised among the principal issues affecting vote choice, and even then, it only prompted defection among one group of voters: those already dissatisfied with politics in general (Kimura, 2000).[20] This can be compared with findings published by Yoshiaki Kobayashi indicating that among independent voters, 30.69% of those who opposed the LDP’s defence policy voted for the party nevertheless. Among committed LDP supporters, this rose to 63.44% (Kobayashi, 1987).[21] In other words, a substantial tranche of voters approves of the competence with which the LDP implements its foreign and defence policies, despite disapproving of the policies themselves (Miyake, 2000).[22] This should be contrasted with U.S. models, which tend to see evaluation of a policy as being dependent upon approval of the same. Under such a model, voters who — for example — positively evaluate the conduct of the Iraq war, will be a subset of the voters who approve of the war in principle. Or, as Merrill Shanks and Warren Miller put it, “position variables (concerning preferred policy direction) should be regarded as causally prior to valence variables (or performance evaluations)” (Shanks and Miller, 1990).[23]

Even more interestingly, it would seem that not only do LDP supporters vote differently from opposition supporters, they also use a different approach in deciding whom to vote for. Studies have shown that LDP supporters, and particularly those belonging to the party’s ko’enkai, are relatively uninterested in policy issues, as compared to supporters of other parties (Watanuki and Miyake, 1997), a finding confirmed by Aiji Tanaka and Sherry Martin (Tanaka and Martin, 2003).[24] Masahiro Yamada, studied swing voters in the 2009 election and concluded that LDP loyalists were interested more in candidate’s personalities and the benefits that they could bring to the local area, while opposition and swing voters had a greater interest in policy (Yamada, 2012).[25] In 2016, a similar study carried out using in-depth survey data on loyalists and swing voters in Kanazawa Prefecture came to the same conclusion (Fujii, 2016).[26] In fact, the effect is so strong that when Yoshiaki Kobayashi produced a series of models that measure the influence of voter opinions on specific policies across multiple elections from 2001 to 2004, it showed that the effect of policy opinions was so inconsequential among LDP supporters that a large proportion continued to back the LDP, despite the fact that other parties’ manifesto positions were closer to their own personal views (Kobayashi, 2006).[27]

While the existence of pocketbook voting is notoriously difficult to prove in any electoral system, a substantial tranche of circumstantial evidence exists in its favour in the Japanese case. Firstly, there is the frequently-noted fact that the LDP is strongest in rural areas (Chiavacci, 2010), which do disproportionately well from national-level economic redistribution and individual candidates’ infrastructure projects (De Wit and Steinmo, 2002).[28] A study by Yusaku Horiuchi and Jun Saito showing that over-represented districts (predominantly in rural areas) receive more subsidy money per-capita, implying that Diet Members function, at least in part, as vectors for channelling public investment towards their constituencies (Horiuchi and Saito, 2003).[29] In a particularly interesting study, Takashi Fukuchi used aggregate data comparing regional public infrastructure investment and voting trends across Japan. They uncovered a cyclical relationship between public infrastructure investment and voting: when investment dropped, votes were lost, and the government responded by increasing public investment in order to win them back (Fukuchi, 1981).[30] Though, by contrast, Jun Saito argues in favour of an equal-but-opposite phenomenon of “reverse accountability”. Suggesting that the LDP holds out promises of future infrastructure investment, to be rescinded if the region in question fails to show sufficient levels of support at the ballot box (Saito, 2010).[31] Takeshi Iida used a similar argument to explain the large number of votes gained by the LDP and the Communist Party (JCP) in the 2014 election called by Shinzo Abe. He suggests that in light of Japan’s slow growth rate, the LDP has been incapable of satisfying all constituencies and thus been obliged to privilege certain groups. As a result, the “winners” tend to support the LDP, while many of the “losers” have responded by switching their allegiance to the JCP (Iida, 2016).[32] Similarly, Tanaka and Martin showed that groups who tend not to benefit personally from LDP policies (housewives, students, urban residents) show a higher-than-average tendency to back opposition parties and vote based on policies rather than personal advantage (Tanaka and Martin, 2003).[33] Ko Maeda, meanwhile, looked at the difference between rural and urban campaign strategies and voting trends, finding that urban voters were less likely to belong to political relationship networks than rural ones and thus more likely to vote based on parties and policies rather than local candidates and financial benefits (Maeda, 2009).[34]

In a similar vein, Tomoyuki Ide argued that the declining support for progressive parties in post-war Japan is better explained using an economic model than a sociological or a policy-distance one. As the economy prospered and voters’ quality of life improved, it was no longer in their interests to support aggressive redistribution policies. He concludes that the Japanese electorate is closer to Downs’ rational choice model of voting (whereby voters privilege their own risk-adjusted interests) than the Michigan or Columbia models (according to which voters decide primarily based on sociological and partisan identification factors) (Ide, 2000).[35] Similarly, the lack of economic debate during the years when the economy was growing fast enough to satisfy all groups, followed by the rejection of the LDP after the bubble of the 1980s burst, is frequently explained as being the reflection of a pragmatic public that makes retrospective judgements of governments’ managerial capabilities based on economic results (Sakaiya, 2013).[36]

In addition to this, there is some evidence that the LDP’s dominance is self-perpetuating. Tanaka and Martin, for example, argue that the LDP’s monopolistic control over local networks effectively prevents other parties from developing patronage structures of their own and thereby constituting an effective threat to LDP dominance over the political system (Tanaka and Martin, 2003).[37] After its years in power the LDP has come to be seen as the default option by a substantial proportion of the electorate, meaning that elections often function as “referenda on LDP government” (Reed and Thies, 2003).[38] Conversely, voting for an alternative (and possibly untested) party, is seen as a risk (Hata, 2017).[39]

Annex II: Methodology

Participants were recruited via (a Japanese equivalent of Amazon’s Mechanical Turk) and offered a fee of 100 yen each for their cooperation. The participants were presented with a questionnaire soliciting information regarding their age, sex, place of residence and level of interest in local, national and international politics. This was followed by two open-ended questions:

1. What do you take into consideration when voting in general elections? (For example: policies, party image, local issues, the economic situation, the personality of the leaders etc.) How do you make up your mind?

2. From your point of view, do you think that other voters decide in the same way that you do? What do other voters take into consideration? (For example: policies, party image, local issues, the economic situation, the personality of the leaders etc.)

Two rounds of pre-testing were conducted; one using 10 Japanese citizens working in Singapore, and a second online, among 25 volunteers recruited using the Crowdworks platform. The questions were refined and simplified as a result of this pretesting process, and the “for example” section added in to clarify the request. A total of 501 participants were selected, reduced to 451 after duplicate identities and unusable/unreadable responses were removed. Of the retained respondents, 53.88% were female and 46.12% were male. 52.55% had a university education (around 5% higher than the national rate), 28.82% had a high school education, and 1.55% had a primary-level education, with the rest declining to state. Concerning their place of residence, 45.23% lived in large cities, 45.45% in towns, and 10.02% in villages or rural areas (the nationwide urbanisation rate is around 93.4%). Concerning age groups, 23.95% were between 18 and 33, 37.69% were between 31 and 40, 29.05% were between 40 and 50, 6.87% were between 50 and 60, and 2.22% were over 60. This sample thus skews younger than the Japanese electorate, a factor which must be borne in mind, as younger groups tend to contain a higher proportion of floating voters, and to base their decisions on national policy to a greater degree. Crowdworks does not allow for IP tracking, but participants’ IDs were checked to prevent the submission of multiple responses by single individuals.

For participants self-reported levels of interest in politics, see Table 1 below, though these figures should be considered in light of the tendency to provide socially desirable answers in surveys regarding citizen participation (Holtgraves, 2004).[40] It is possible that actual interest in politics at all levels is significantly lower than reported here.

The social desirability problem was also a key factor in designing the two survey questions. Evidence suggests that while people will try to vote according to socially approved criteria (policy, principle, the national interest), their choice is nevertheless influenced by a wide variety of less socially acceptable factors (candidate appearance, party loyalty, self-interest).[41] In an attempt to control for social desirability effects, we asked participants to describe their own decision-making processes, but also asked them to comment on how they believed their fellow-citizens to make voting choices, on the basis that people are often more willing to ascribe negative behaviour to others than to themselves.[42] It should be noted that the aim of this exercise was not to use one set of results to invalidate the other, but rather minimise gaps in the data and build up an image of what is considered to be “typical” voting behaviour, in order to inform analysis of individuals’ self-reported decision-making strategies.

In the event, there were several factors that were more frequently cited in relation to respondents’ own decision-making processes than to others’ and vice versa.

Word tables for both sets of data were created using the EKWords software package in conjunction with algorithms developed in-house to isolate major themes, before the responses were hand-coded according to these themes and others that emerged during the coding process.[43]

[1] Miwa Hirofumi (2014) ‘Gendai nihon ni okeru sōten taido no ideorogī-teki ikkan-sei to seijiteki senren’ (Ideological consistency and political sophistication in contemporary Japanese issue attitudes), Nenpō Seijigaku (Political Science Annual Report) Vol. 65, №1, pp. 148–174.

[2] J. Merrill Shanks, and Warren E. Miller (1990) ‘Policy direction and performance evaluation: Complementary explanations of the Reagan elections,’ British Journal of Political Science Vol. 20, №2, pp. 143–235.

[3] Kobayashi Yoshiaki (2006) ‘Manifesuto senkyo ikō no sōten taido tōhyō’ (Attitude voting following the “manifesto election”), Senkyo kenkyū (Electoral Research) Vol. 21, pp. 7–38.

[4] All quotations translated by the author.

[5] Soka Gakkai is a Buddhist sect founded in the 1930s. New Komeito is its affiliated political party.

[6] Maeda Yukio (2006) “Trends in party support rates as seen in news polls (1989–2004)” (“Jiji seronchōsa ni miru seitō shiji-ritsu no suii (1989–2004)”), Central Research Services Report (Chūō chōsa-hō).

[7] Yamazaki Seiko (2015) WV5 Results: Japan 2005 Technical Record, World Values Survey.

[8] Note: These quotes have not been cropped from more wide-ranging answers. Each constitutes the respondent’s entire answer to the question.

[9] Shanks and Miller (1990)

[10] For many years a quirk of the Japanese electoral system meant that candidates frequently stood for election without the endorsement of any particular party. This was conferred upon them only after they had proved themselves capable of winning. This trend declined somewhat following the electoral reforms of the 1990s.

[11] Hirofumi (2014)

[12] Dimock, Michael, Jocelyn Kiley et al. (2014) Beyond Red vs. Blue: The Political Typology, Washington D.C.: Pew Research Center.

[13] Takenaka Yoshihiko (2008) “Gendai nihonjin no ideorogī saikō: tairitsu kōsokuryoku kiteiryoku” (“Reconsidering modern Japanese ideology: confrontation, constraint and predictive power”), Ronsō gendai bunka, kōkyō seisaku (Discussing Contemporary Culture and Public Policy) Vol. 7, pp. 25–63.

[14] Kago Rieko (2015) “Nihon ni okeru ‘sayū tairitsu’ no genzai” (“The current state of ‘right-left conflict’ in Japan”), Leviathan Vol. 57, pp. 146–169.

[15] Murakawa Takuya (2014) “Jimintō habatsu no seisaku ichi henka toso yōin” (“The causes of policy position change among LDP factions”) in Gendai Nihon seiji-ron enshū I, II (Exercises in contemporary Japanese political theory I, II), Tokyo: Tokyo University Faculty of Law.

[16] Takenaka, Yoshihiko, Endo Masahisa and Willy Jou (2015) “‘Yūkensha no datsu ideorogī to Abe seiji” (“Abe’s politics and the decline of ideology among Japanese voters”), Leviathan Vol. 57, pp. 25–46.

[17] Kimura Takahiro (2000) “‘Taishutsu’ ni kansuru sōten no eikyō no bunseki” (“An analysis of the influence of ‘exit’ issues”), Seisaku kagaku (Policy Science) Vol. 8, №1, pp. 89–97.

[18] Miyake Ichiro (1999) “Chūtohanpa ni owatta seisaku tōhyō” (“Slouching towards policy voting”), Senkyo kenkyū (Electoral Research) Vol. 14, pp. 50–62.

[19] Kawato Sadafumi (1988) “Shūsan dōjitsu senkyo to Nakasone ninki” (“The ‘same day election’ and Nakasone’s popularity”), Hokudai hōgaku ronshū (Hokkaido University Law Review) Vol. 39, №2, pp. 238–180.

[20] Kimura, 2000.

[21] Kobayashi Yoshiaki (1987) “Tōhyō kōdō to seiji ishiki ni kansuru keiryō bunseki” (“An econometric analysis of voting behaviour and political consciousness”), Senkyo kenkyū (Electoral Research) Vol. 2, pp. 26–63.

[22] Miyake Ichiro (2000) “Seitō no seisaku ni taisuru manzoku-do” (“Rates of satisfaction with parties’ policies”), Senkyo kenkyū (Electoral Research) Vol. 15, pp. 96–108.

[23] Shanks, Merrill J. and Warren E. Miller (1990) “Policy direction and performance evaluation: Complementary explanations of the Reagan elections,” British Journal of Political Science Vol. 20, №2, pp.143–235.

[24] For LDP supporters’ relative lack of interest in policy issues, see: Watanuki Joji and Miyake Ichiro (1997) Kankyō hendō to taido hen’yō (Environmental and Attitude Changes), Vol. 2, Tokyo: Bokutakusha. For the differences in decision-making strategies between LDP supporters and independents, see: Tanaka Aiji and Sherry Martin (2003) “The new independent voter and the evolving Japanese party system,” Asian Perspective, pp. 21–51.

[25] Yamada Masahiro (2012) “2009-nen Shūinsen ni okeru suuingu vuōtā no seijiteki ninchi to seijiteki jōhō kankyō” (“The political environment and cognition of swing voters in the 2009 House of Representatives election”), Seisaku kagaku (Policy Science) Vol. 19, №3, pp. 165–178.

[26] Fujii Ayaka (2016) “Senkyoku senkyo ni okeru ryūdō tōhyō to kōhosha tōhyō no kankei” (“The relationship between voter behaviour and constituency voting trends”), Ningen shakai kankyō kenkyū (Human and socio-environmental studies) Vol. 32, pp. 7–20.

[27] Kobayashi Yoshiaki (2006) “Manifesuto senkyo ikō no sōten taido tōhyō” (“Attitude voting following the ‘manifesto election’”), Senkyo kenkyū (Electoral Research) Vol. 21, pp. 7–38.

[28] For the LDP’s strong showing in rural regions, see: Chiavacci, David (2010) “Divided society model and social cleavages in Japanese politics: No alignment by social class, but dealignment of rural-urban division,” Contemporary Japan Vol. 22, №1–2, pp. 47–74. For economic redistribution favouring rural areas, see: DeWit, Andrew and Sven Steinmo (2002) “The political economy of taxes and redistribution in Japan,” Social Science Japan Journal Vol. 5, №2, pp. 159–178.

[29] Horiuchi Yusaku and Saito Jun (2003) “Reapportionment and redistribution: Consequences of electoral reform in Japan,” American Journal of Political Science Vol. 47, №4, pp. 669–682.

[30] Fukuchi Takashi (1981) “Tōhyō kōdō no keiryō-keisaigakuteki bunseki — kōdo seichō ki ni okeru Jimintō shijiritsu teika no bunseki” (“Econometric analysis of voting behaviour: an analysis of falling LDP support rates during a period of high-growth”), Kikan riron keizaigaku (Theoretical Economics Quarterly) Vol. 32, №1, pp. 29–44.

[31] Saito Jun (2010) Jimintō chōki seiken no seiji keizaigaku: Rieki yūdō seiji no jiko mujun (The political economy of long-run LDP dominance: the contradictions inherent in pork-barrel politics), Tokyo: Keiso Shobo.

[32] Iida Takeshi (2016) “Surging progressives in the conservative mood: The conditional effects of income and urbanism on vote choice in the 2014 Japanese Lower House election,” Asian Journal of Comparative Politics Vol. 1, №1, pp. 6–24.

[33] Tanaka and Martin, 2003.

[34] Maeda Ko (2009) “Has the Electoral System Reform Made Japanese Elections Party-Centered?” in Political Change in Japan: Electoral Behavior, Party Realignment, and the Koizumi Reforms, New York, Walter H. Shorenstein Asia-Pacific Research Center, pp. 47–66.

[35] Ide Tomoyuki (2000) “Sengo Nihon no kakushin seitō shiji-ritsu teika” (“The decline in support for progressive parties in post-war Japan”), Shakaigaku Hyōron (Sociological Review) Vol. 51, №3, pp. 298–313.

[36] Sakaiya Shiro (2013) “Posuto 55-nen taisei ki ni okeru seisaku-teki tairitsu kōzō” (“Structuring policy conflicts after the end of the 55 system”), Retrieved 24 August 2017:

[37] Tanaka and Martin, 2003.

[38] Reed and Thies, 2003.

[39] Hata Masaki (2017) “Higashi Nihon daishinsai to seiji ishiki ― sonzai kyōi kanri riron ni motozuku hoshu-ka genshō no kenshō” (“Political consciousness and the Tohoku earthquake: verifying conservative-shifts based on existential threat-management theory”), 2016-nendo sankasha kōbo-gata niji bunseki kenkyūkai gendai nihonjin no seiji ishiki to tōhyō kōdō ni kansuru, dēta no niji bunseki, kenkyū seika hōkoku-sho (Participants in the 2016 secondary analysis research group on modern Japanese political consciousness and voting behaviour, secondary analysis of data, research report), Tokyo, University of Tokyo Centre for Social Research and Data Archives.

[40] Holtgraves, Thomas (2004) “Social desirability and self-reports: Testing models of socially desirable responding,” Personality and Social Psychology Bulletin Vol. 30, №2, pp. 161–172.

[41] For the struggle to vote for socially appropriate reasons, see: Brennan, Geoffrey and Alan Hamlin (2000) Democratic devices and desires, Cambridge, Cambridge University Press. For the influence of other factors, see: Olivola, Christopher Y. and Alexander Todorov (2010) “Elected in 100 milliseconds: Appearance-based trait inferences and voting,” Journal of Nonverbal Behavior Vol. 34, №2, pp. 83–110.

[42] Hoorens, Vera (1993) “Self-enhancement and superiority biases in social comparison”, European review of social psychology Vol. 4, №1, pp. 113–139.

[43] DJSoft: EKWords

Doing research on public opinion.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store