Abstract
There is a large volume of studies on political participation, including motivations for citizens to become involved in various channels of engagement. Building on this extensive literature, the present study examines factors that affect participation, with particular attention on the linear (conservative vs. progressive) and curvilinear (moderate vs. radical) impact of ideological orientations, and also considers the influence exerted by political interest and partisanship. Utilizing data from Japan spanning nearly three decades, we analyze changes and continuities in patterns of participation in four categories of activities: election turnout; campaigning; system-affirming; and elite-challenging. Empirical analysis reveals that radicals take a more active part in campaigning than moderates, and progressives are more inclined toward elite-challenging activities than conservatives, while no consistent results are found for turnout. The study also discusses possible causes and consequences of a long-term decline in participation in Japan.
Introduction
The participation of ordinary citizens in the political decision-making process has long been lauded for underpinning and enhancing democracy. Scholars have made the normative case linking the quality of democracy with the extent of citizen engagement, positing that mass participation in political activities helps to raise citizens’ sense of political efficacy and trust (Barber, 1984; Pateman, 1976). While turnout in elections, a traditional marker of participation, has decreased in many democracies (Blais et al., 2004; Franklin, 2003), this does not necessarily indicate widespread disillusionment with democracy. Instead, many citizens have turned to less conventional and often more direct channels of interest articulation (e.g. Barnes and Kaase, 1979; Dalton, 2004; Norris, 2002).
A long-standing strand in the voluminous literature on participation focuses on factors that affect individual willingness and ability to become politically involved. 1 This article focuses in particular on the role of ideology as a stimulus for political action. Several studies have identified a relationship between individual ideological orientations and propensity to take part in political activities, particularly protests, and observed that leftists are more likely to be engaged than those on the right (Barnes and Kaase, 1979; Finkel and Opp, 1991; Gundelach, 1995). However, a number of recent works point to rising participation by citizens espousing radical right views (Grande and Kriesi, 2012; Macklin, 2013), which suggests that the impact of ideology may not (only) be pitting left versus right, but (also) moderates versus extremists.
Nearly all the works cited above are based on studies of Western democracies, which raises the question of whether and to what extent their findings are applicable to countries with different historical trajectories and levels of economic and political development. The present study contributes to the literature on political participation, including the link between ideological orientations and different channels of engagement, by examining the case of Japan, an industrialized nation and a consolidated democracy that nonetheless presents differences from its Western counterparts in terms of participatory patterns and ideological understanding. Furthermore, by analyzing data spanning a quarter-century, we are able to trace changes as well as continuities in this relationship against the background of major institutional and party systems changes during the period covered.
This article is organized as follows: The next two sections present some important literature on political participation in general and on the impact of ideology in particular, and are followed by a review of relevant works in Japan. The following section introduces data and variables used in this study, and empirical results are then shown. The last two sections summarize the findings and place them in a broader context.
Who participates?
In line with the assumption that the role of citizens in a representative democracy is to choose between competing elites (Schumpter, 1942), political participation has traditionally been defined as ‘those activities by private citizens that are more or less directly aimed at influencing the selection of governmental personnel and/or the decisions they take’ (Verba and Nie, 1972: 2). Yet this limited role has since been widened to encompass more direct means of interest articulation (e.g. Barber, 1984; Gould, 1988), to include not only ‘conventional’ but also ‘unconventional’ channels of participation (Cain et al., 2003; Clarke et al., 2009). Inglehart and Klingemann (1979: 209) refer to the latter as ‘elite-challenging’ activities, which are ‘likely to take place when one knows how to cope with elites and wants something different from what the elites want’, for example joining boycotts or demonstrations. They contrast with activities that, while seeking changes to certain policies or their implementation, nevertheless implicitly acknowledge the legitimacy of the extant political institutions and processes. Examples of this ‘system-affirming’ category of participation include most forms of community or neighborhood associations, as well as contacts with local or national politicians.
Scholars have employed different classification schemes when analyzing forms of political participation. Verba and Nie (1972) differentiate between voting, campaigning, contacting, and cooperation. Parry et al. (1992) present six dimensions in their study of participation in Britain: voting, campaigning, contacting, collective action, direction action, and violence. Kornberg and Clarke (1992: 236) posit three lines of demarcating various forms of engagement: 1) electoral versus other activities; 2) verbal versus action-oriented; and 3) non-confrontational versus confrontational with respect to protests. Following these studies, we analyze four categories of political participation: election turnout; campaign-related; system-affirming; and elite-challenging.
In a well-known study, Verba, Schlozman, and Brady (1995) listed three reasons why people do not participate politically: because they cannot, they do not want to, or nobody has asked them to. The first cause involves the frequently noted impediments deriving from socio-economic status (Verba et al., 1978; Wolfinger and Rosenstone, 1980). People who are more richly endowed with time, wealth, and social connections can better afford to spend these resources by engaging in individual and group action. Being asked to participate refers to networks and mobilization efforts. Scholars have long noted that whether and how much an individual participates can depend heavily on who he or she interacts with (Giles and Dantico, 1982; Huckfeldt, 1979; Kenny, 1992), as this is associated with patterns of contact and recruitment by churches, labor unions, and various civic associations. Furthermore, political parties have long played a key role in mobilizing voters (Karp et al., 2008), most conspicuously at election time but also as agents in aggregating and expressing opinions supporting or opposing various legislative proposals during inter-election periods.
The second reason listed above, namely that citizens participate when they want to, points to the question of incentives. One of the most well-known equations in political science concerns the calculus of voting (Downs, 1957): an individual decides whether to go to the polling booth based on 1) the amount of benefit she expects to derive from the victory of her preferred party or candidate, multiplied by 2) the likelihood of her vote determining the election outcome, minus 3) the costs of casting a ballot (e.g. gathering information about parties or candidates, queuing at the polling booth). Another term was subsequently added to account for satisfaction from the act of voting itself (Riker and Ordeshook, 1968), due to a sense of civic duty or an affirmation of faith in the political system. This implies that turnout can be motivated by expressive as well as instrumental considerations (Kaase, 1990: 29).
A similar calculus can also apply to other modes of political participation. Compared with voting, these activities – such as contacting one’s representatives, collecting signatures on petitions, volunteering for parties or candidates during election campaigns, or joining protest marches – call for greater commitment of time, effort, and organizational and/or financial resources, so one expects that they require stronger incentives to motivate people to take part. Such incentives can derive from deeply held values and firm policy preferences, which are often encapsulated by ideological orientations. There are numerous examples of issues that divide right from left, but it is also worth considering the possibility of causes that pit radicals toward both poles of the ideological spectrum against those in the center. Both of these propositions are explicated and examined below.
Many studies have highlighted differences between right and left with regard to political participation, particularly protest activities. In so far as conservatives and progressives align themselves on the right and left sides of the ideological spectrum, respectively, it may seem inherent that leftists would have greater incentive to participate because they are more committed to changing aspects of the status quo. Finkel and Opp (1991) posit that left-leaning citizens participate more because their beliefs render protests normatively palatable. Other works draw the link between unconventional channels of participation and libertarian or post-materialist values (Barnes et al., 1979; Gundelach, 1995; Inglehart, 1990), and these values are commonly associated with the left. Furthermore, Hutter and Kriesi (2013) observe a distinguishing feature between the preferred modes of interest articulation for rightists and leftists: the former are disposed toward electoral channels, while the latter resort to protest. A comparative study of 26 European countries by Torcal, Rodon, and Hierro (2016) confirms that this link between leftist orientations and propensity to protest remains significant even after controlling for the ideological distance between individual citizens and their governments. Based on these studies, we hypothesize that:
Citizens espousing leftist orientations, driven by their desire to alter the status quo, are more likely to participate than those with right-wing views.
However, political activism is by no means the domain of leftists. Several studies that have documented the rise of participation by citizens holding radical right views (e.g. Grande and Kriesi, 2012; Van Aelst and Walgrave, 2001) call the assumption of a linear relationship into question, thus implicitly raising the possibility that participation is positively linked with extremism – on both ends of the spectrum. And this is not limited to protests, as a number of older studies also affirm the relationship between extremist views on the one hand, and campaign activities (Miller et al., 1986) and issue advocacy (Scott and Schuman, 1988; Verba and Brody, 1970) on the other. A recent cross-national study provides empirical confirmation that extremists are indeed more prone to participate than moderates in various activities bar voting (Van der Meer et al., 2009: 1439). In addition, normative justification for heterodox channels of participation can be found not only at the left end of the ideological scale but on the opposite extreme as well (Grundy and Weinstein, 1974). All of this leads to the expectation that:
Citizens with radical ideological views, on both ends of the left-right spectrum, are more likely to participate than centrists due to a stronger commitment to their beliefs.
In addition to the possible impact of ideological orientations discussed above, we also test the effects of political interest and partisanship. It is intuitively understandable that citizens who profess interest in politics are more inclined to become engaged, and this should apply to the entire gamut of political activities. While partisanship often partially overlaps with ideology, it is analytically distinct and may be motivated by reasons other than policy preferences. Since parties have the highest stakes in the electoral arena, one expects that partisans are most liable to be mobilized to turn out and help with campaigning activities.
Participation in Japan
Studies of political participation in Japan have mostly concentrated on election turnout (e.g. Arai, 2014; Kabashima, 1988). This is hardly surprising given the relatively low rates of involvement in other activities. Hirano (2012) notes that, other than voting, not only is participation far from prevalent, but one can also observe a downward trend from the 1970s to the 2000s in each type of activity. For example, the proportion of survey respondents reporting that they have contacted politicians or bureaucrats declined from 15 percent to only 6 percent during this period, and those who have joined demonstrations, already low at a mere 8 percent in the 1970s, decreased to less than 1 percent by the late 2000s. This is probably attributable in part to a widespread inclination to avoid political involvement (Nishizawa, 2004). Even activities that do not have explicitly political objectives are often seen as implicitly associated with partisan interests (Sakaiya, 2013; Yamaguchi et al., 2012), thus deterring many ordinary citizens from taking part.
These meager participation rates are not a legacy of the authoritarian culture that persisted in the early postwar decades. In fact, Hirano highlights that today’s elderly who were socialized in that period have continued to be active participants, compared with the declining political engagement among their children and grandchildren. A clear generational effect can be observed: the age groups most active in civic movements consisted of respondents in their 30s in the 1970s, in their 40s in the 1980s, in their 50s in the 1990s, and in their 60s in the 2000s (Hirano, 2012: 145–146). Such evidence showing younger cohorts becoming less involved in all types of participatory modes stands in stark contrast to the portrait of increasing youth engagement in a diverse range of political activities in other advanced democracies (e.g. Dalton, 2016; Zukin et al., 2006).
The question of how ideological orientations affect political participation has seldom been explored. Yamada (2004) reports that whereas partisanship exerts an influence on some modes of participation, ideological orientations per se make no difference. His findings suggest that political participation may be predicated on an electoral winner versus loser dynamic, as voters who identify with several opposition parties are significantly more likely to sign petitions and join demonstrations. Indeed, Ikeda (2007) shows that without controlling for partisanship, ideological progressivism exerts a positive influence on citizens’ non-electoral participation. One should also note that ideology and partisanship do not overlap. Hata (2015) differentiates between ideological moderates and extremists, and finds that extremists are motivated to participate by both the strength of their views and mobilization through organizational networks, whereas only the latter mechanism applies to moderates. This highlights the usefulness of conceptualizing the ideological spectrum in a curvilinear rather than linear fashion, though the said study is limited to only a single point in time.
We try to build on these works by testing both the left versus right and moderates versus extremists hypotheses on different modes of political participation. Furthermore, we provide a longitudinal perspective by expanding the number of data points to cover the span of more than a quarter century.
Data and variables
To test the aforementioned hypotheses, we utilize Japan Elections Study surveys, conducted as panel surveys in 1983 (JES I), 1993–1996 (JES II), 2001–2005 (JES III), and 2007–2010 (JES IV). 2 The surveys we analyze are taken from the 1st wave of JES I (1983), the 2nd wave of JES II (1993), the 4th and 5th waves of JES III (2003), and the 5th and 6th waves of JES IV (2010). All data were collected from face-to-face interviews. These surveys have been widely used by scholars studying various aspects of voter attitudes and behavior in Japan, resulting in a large volume of books and articles.
In order to detect changes in the relationship between ideological orientations and different forms of political participation, data for each of the four points in time we have chosen are analyzed and presented separately. Based on reasons explained in the next section, and in line with common practice in the literature, voting turnout is treated as a category of its own. Another category, labeled ‘campaign’ activities, comprises election campaigning, attendance of political meetings during campaigns, and donations to parties or candidates. Participation in social/civic movements and demonstrations is classified as ‘elite-challenging’ activity, following Inglehart and Klingemann’s (1979: 209) definition. Lastly, non-electoral activities in which participants express their interests through established channels, such as contacting politicians and politically influential individuals, petitioning legislatures and government agencies, and taking part in neighborhood associations to solve local problems, are classified as ‘system-affirming’. 3
As a legacy of the single non-transferable vote (SNTV) electoral system used until 1994, under which candidates from large parties often competed with each other in the same district, district-level party branches, especially those of the LDP, are not well organized (Krauss and Pekkanen, 2011). Instead, many campaign activities rely on candidates’ own personal support organizations called koenkai, which are based on social (at times clientelistic) networks and local interests. While electoral regulations in Japan prohibit door-to-door canvassing during the electoral campaign period, candidates attempt to have political meetings to let voters know who they are and what stances they take. During the campaign, many koenkai members help with campaigning not only as office staff but also on the streets, for example by putting up posters.
While social/civic movements are often not tied to political parties or representatives, insofar as they aim to change existing policies or creating new ones, one can characterize participation in these activities as political. Early examples of such movements in Japan include groups formed to oppose industrial pollution and protect consumer interests; other groups advocate a variety of causes ranging from peace to gender equality. It has been remarked that some social movements in the 1970s have been channeled into backing for progressive local executives (governors, mayors) in opposition to conservative central governments. This accounts in part for the perception of engagement in social movements as ‘elite-challenging activity.’ 4
In contrast to social/civic movements, we categorize participation in neighborhood associations as ‘system-affirming,’ due to the latter’s closer ties with authorities. In a detailed study, Pekkanen et al. (2014: 5–6) state that they are ‘characterized by collaboration with government in the smooth implementation of policy and articulation of local demands’. A long history of government control, as well as their intrinsically limited size and sources, render cooperation rather than confrontation with authorities a logical strategy for neighborhood associations. This is the most common form of citizen participation in Japan, with around 300,000 such groups across the country, yet most members do not engage in activities beyond overseeing cleaning and garbage collection, organizing community festivals, and distributing information about local events.
The distribution of responses for each participation category is shown in Table 1. Over-reporting of turnout is a problem common to most surveys, though it must be noted that the discrepancy between actual and reported turnout exceeds more than 20 percent in all datasets since the 1990s. Yet insofar as inaccurate recall was proportionately distributed across the ideological spectrum, this should not affect hypothesis testing except to reduce variation in the dependent variable. Both campaign-related and elite-challenging activities have decreased over the time span covered by the analysis. Participation in the latter was only 10 percent at the start of this period, and dropped to a paltry 3 percent by 2010, which necessarily raises concern over sample size. Other than reported turnout, system-affirming activities are the only category that has seen a rising proportion of participants. This is largely attributable to a stark increase in neighborhood association engagement. The percentage of respondents who have contacted politicians or other politically influential notables has halved (from over 20 percent to 10 percent), and the rate of petitioning legislatures or government agencies has remained consistently low. 5
Participation in different categories of activities.
Campaign activities: help with campaigns, attend election meetings, donate to campaigns.
System-affirming activities: contact politicians, contact politically influential persons, petition legislatures, join neighborhood associations.
Elite-challenging activities: take part in residents’ movements, join demonstrations.
For each mode of participation, responses are coded dichotomously (have or have not taken part). The nature of the turnout dependent variable requires a logit regression, and for the purposes of methodological consistency and ease of comparison, we employ ordered logit for the remaining three dependent variables, which are constructed by adding the number of activities in which survey respondents reported to have participated. Ideological orientations are measured by respondents’ self-placement on a conservative-progressive scale (corresponding to the conventional left-right scale), while ideological extremism is captured by respondents’ distance from the middle, with higher scores denoting more radical positions. 6 In addition to commonly used socio-demographic characteristics (age, gender, education, income), we also include political interest and partisanship as independent variables. Details on variable coding are shown in the appendix.
Citizens who are interested in political developments and/or feel close to a party are likely to have greater motivation and, in the latter case, stronger organizational support to become politically involved. The positive correlation between age and turnout is well-known, and in the context of a consolidated democracy one expects this to hold true for campaigning and system-affirming activities, since older segments of the population have longer experience with interest articulation and greater stakes in the outcomes of policy making. On the other hand, and again not unexpectedly, younger generations are more prone to engage in elite-challenging activities such as protests. Men have consistently been found to be more active participants than women, and those with longer years of schooling more so than their less educated counterparts. That said, it is worth observing whether and how the gender and education gaps have changed over time.
Empirical analysis
We begin with an analysis of factors affecting election turnout. Logit regression results in Table 2 show no evidence that Japanese citizens on one side of the ideological spectrum are more likely to cast a ballot than those in the opposite camp, but reveal a curious turnaround in the relationship between extremism and turnout. In the early 1980s respondents espousing more radical views were significantly more inclined to make the trip to the polling booth, but two decades later the reverse became true. As overall turnout had decreased markedly over the quarter-century, one might have speculated that it is those eligible voters with more strident opinions who continued to go to the polls, but the result in 2003 suggests otherwise. One hesitates to assert, based on evidence found at a single point in time, that turnout rates are consistently greater among ideological centrists than radicals, but at least it is safe to conclude that declining turnout is not explained by moderates no longer bothering to vote.
Factors affecting election turnout.
**p < 0.01; *p < 0.05; +p < 0.1.
As expected, both partisanship and interest in politics make a significant and substantial difference in boosting the probability of turning out. 7 The positive relationship between age and turnout, observed in many advanced democracies, is confirmed in Japan. Interestingly, on the occasions when education emerges as a significant factor, the coefficients bear opposite signs: in the early 1990s respondents with fewer years of schooling were more likely to make a trip to the polling booth, but this relationship has been reversed since the early 2000s (Sakaiya, 2013). One factor that does not reach statistical significance is also worth noting: while women are less engaged than men in most other modes of political participation, there has not been any marked gender gap with respect to election turnout.
Regarding campaign activities, a category including not only working for candidates or parties but also attending election-related meetings and making campaign donations, Table 3 shows that ideological orientations exert a curvilinear rather than linear impact. Throughout most of the period covered by this study, respondents professing more radical opinions had a significantly higher probability of becoming involved in election campaigns than those placing themselves in the ideological center ground, though this effect disappears in the most recent survey (2010). Since the questionnaires did not probe which specific party or candidate respondents worked for or donated to, and also due to concerns with sample size, we are unable to distinguish between campaign activities in support of different parties. That said, it is possible to draw a link between the decreasing importance of ideological extremism and the de-polarization of the Japanese party system: as distances separating most parties narrowed, that is, as parties which in the past were (perceived to be) located toward either extreme of the spectrum undertook a centripetal shift, the incentive for ideological radicals to take part in election campaigning has probably weakened.
Factors affecting participation in campaign activities.
**p < 0.01; *p < 0.05; +p < 0.1.
Of all the independent variables, partisanship is intuitively the one most directly related to election campaigns, so it comes as no surprise that this variable plays a consistently important role in motivating participation in campaign activities. In contrast to results found for turnout, age is less often significant for campaign involvement, while one now observes a gender gap in most years, with men much more likely to take part in these activities than women. Interestingly, education never makes any difference, whereas income has come to assume greater importance in more recent elections. This hints at which type of resources is most useful for campaign activities.
Concerning the category of activities we label ‘system-affirming’ – contacting politicians and politically influential individuals, taking part in neighborhood associations, and petitioning legislatures and government agencies – Table 4 shows that neither aspect of ideology exerts an impact throughout the period under analysis. Given that legislatures at both national and (especially) local levels have long been dominated by the conservative Liberal Democratic Party (LDP), to the extent that neighborhood associations engage in political activities they more often than not tend to display a conservative bias. This led us to explain a significant positive association between ideological self-placement and the likelihood of participation. Contrary to expectations, this relationship was only confirmed on one occasion, in the early 2000s. 8
Factors affecting participation in system-affirming activities.
**p < 0.01; *p < 0.05; +p < 0.1.
Findings on two other independent variables are worth noting. First, partisanship, which significantly contributed to participation in ‘system-affirming’ activities in the 1980s and 1990s, has lost its explanatory value in recent years, meaning that citizens who do not identify with any parties today – a growing segment of the population – are now just as likely to participate as partisans. Second, the same observation applies to the female dummy variable, meaning that the wide gender difference in participation in earlier decades has since narrowed. We would also note that both of these trends are attributable to changes in pattern of petitioning and neighborhood association involvement; if the dependent variable were confined to the two contacting activities, partisanship never emerges as a significant factor, and the gender gap persists into the 2010s.
Finally, we turn to the ‘elite-challenging’ activities of residents’ movements and demonstrations. The former are often at the forefront of protests against large construction projects such as polluting factories, airports, nuclear facilities, and military bases, and can be understood as NIMBY movements with a mixture of materialist and post-materialist goals. The classification of these movements as ‘elite-challenging’ is justified by their frequent aim of overturning, or at least delaying the implementation of, official policies. Table 5 reveals a long-standing trend that saw left-leaning citizens being significantly more prone to engage in these activities, a finding probably attributable to the fact that a conservative government held the reins of power throughout almost the entire postwar era. That the government had for decades allotted generous budgets for infrastructure building only served to incite opposition from many residents’ movements, while its frequent emphasis on strengthening ties with the US has never ceased to provoke criticism from progressives both during and after the Cold War. Whether the de-linking of progressive views from elite-challenging activities in 2010 represents a passing phenomenon due to a center-left government being in power at the time, or a longer-term trend, warrants further study.
Factors affecting participation in elite-challenging activities.
**p < 0.01; *p < 0.05; +p < 0.1.
In contrast to all other categories of activities discussed above, participation in demonstrations and residents’ movements is not always driven by political interest. More unexpectedly in view of the comparative literature, there is only limited evidence that younger citizens are more likely to take part than their parents and grandparents. Yet perhaps this result is not entirely surprising if one recalls the generational effect mentioned previously: that a cohort who grew up in the 1950s and 1960s has long been the most active in civic movements, while subsequent generations have not followed in their footsteps. Also notable is that partisanship, which once played a highly substantive role, no longer animates these channels of participation. One possible explanation lies in party system de-polarization over the past three decades, so that supporters of different (primarily opposition) parties may be more ready to seek compromises with, rather than confronting, decision-making elites. However, any conclusion drawn from the results in Table 5 must be treated with greater caution than other types of activities due to smaller sample sizes of participants.
Summary
This study has explored factors that influence different modes of political participation in Japan, focusing on ideological orientations and extremism while also looking into the effects of partisanship, political interest, and socio-demographic traits. We conducted a longitudinal analysis of data that spanned a period of important paradigm shifts in Japan, ranging from its geopolitical environment to party system configuration. Empirical results reveal the impact of ideology in two types of political activities throughout most of this period: citizens located toward both ends of the ideological spectrum are more engaged in electoral campaigning – helping parties or candidates, attending meetings, making donations – than those congregated in the center; and progressives are more prone to partake in elite-challenging action – demonstrations and residential movements – than conservatives. Ideological radicalism also exerted some influence on voting turnout, though this effect is only occasional and does not consistently point in the same direction. Neither a linear nor a curvilinear relationship is found with respect to participatory channels predicated on working with or through elites (contacting politicians or notables, petitioning legislatures or government agencies, joining neighborhood associations).
It is worth noting here that results showing the progressive orientation of citizens who have taken part in elite-challenging activities are corroborated by data from Japan in the World Values Survey (WVS). Among the types of activities asked in the WVS questionnaire, participation in boycotts, legal demonstrations, and strikes is significantly correlated with progressive self-placement, though the strength of the relationship has declined over time (especially for boycotts). 9 By contrast, another item in the WVS, signing petitions, has always been more weakly correlated with respondents’ ideological positioning, and ceased to be significant from the mid-1990s onwards.
Our findings suggest that electoral reform has also affected political participation with respect to both turnout and campaigning. The new mixed parallel system, in use since 1996, assigns a plurality of parliamentary seats by first-past-the-post in single member districts. Given a normal distribution in the electorate, the well-known Downsian incentives for ideological depolarization should apply, as parties compete to win the median voter. The rapid decline of the once powerful Japan Socialist Party (JSP) and the emergence of the centrist Democratic Party of Japan (DPJ) as a contender for power both testify to this logic at work.
The centripetal shift by the two major contenders in each district (usually the LDP and the DPJ) made radical voters reluctant to go to the polls in 2003 while still boosting, albeit only slightly, their likelihood of engaging in campaign activities. However, these effects have disappeared by 2010. Part of the reason may be that the labels ‘conservative’ and ‘progressive’, conventionally used in most surveys as well as in academic and press discussions, no longer fully reflect a left-right ideological spectrum, and have thus lost some explanatory power. Another possibility lies in the government alternation in 2009, which saw the long-ruling LDP cede power to the DPJ. Some groups that enjoyed access to and influence over policy makers during long decades of LDP rule found themselves losing their privileged position, while previously marginalized interests could make their voices heard. This might have affected some patterns of (especially non-electoral) participation.
Regarding other predictors of participation, political interest has, understandably, boosted engagement in all modes of interest articulation throughout the entire period. The role of partisanship has followed two divergent trends. Respondents who identify themselves with a party have always had a greater probability of going to the polling booth and being involved in election campaigns, but whereas partisan identification once motivated participation in other types of activities (in the ‘system-affirming’ and ‘elite-challenging’ categories), this is no longer the case. While one may find it an encouraging development that many forms of participation have become the domain of not only those with partisan incentives but also citizens with politically independent views, one may also question whether the persistent decline of partisanship may perpetuate the trend of lower turnout as well as render the electorate more distant from, and disinterested in, election campaigns.
Among socio-demographic factors, age has always been positively related to turnout, but has exerted little influence on other types of activities until the most recent point in time (2010) covered by this study (2010) when similarly positive coefficients are found for other forms of participation. Rather than an elderly population becoming more active, this probably indicates a growing gap between a generation that has maintained its level of engagement and younger cohorts who are more passive. With the exception of turnout, women have long been a significantly lesser presence in various modes of participation than men, though this gender disparity has narrowed markedly for ‘system-affirming’ activities, and also to a smaller extent for electoral campaigning, since the turn of the century. The impact of education is rather sporadic, though its positive relationship with turnout appears to have become established. Perhaps contrary to expectations, longer years of schooling does not help cultivate greater willingness (such as in the form of a sense of civic duty) or ability (through cognitive mobilization) to become active participants.
Discussion
Political practitioners, commentators, and researchers in many advanced democracies have lamented declining election turnout, and Japan is no exception. However, in some of these countries scholars have observed, alongside a detachment from the electoral process, a concomitant rise in engagement in other, more direct channels of interest participation, especially among younger segments of the electorate. The same can hardly be said of Japan. If anything, insofar as one can identify generational differences, it constitutes a widening, and now statistically significant, gap between a still-active older cohort and notably more passive younger strata. The causal process linking higher education, cognitive mobilization, self-expressive values, and propensity for more direct means of participation beyond the ballot box does not seem to have reached its expected endpoint of a more engaged citizenry. If a generational change involving the emergence of more assertive values (Welzel and Dalton, 2014) has indeed taken place in Japan, there may be a growing chasm between beliefs and (in)action among younger cohorts.
It is worth noting that the relatively low levels of participation are not the consequence of satisfaction with how the political system operates. In fact, in a study comparing 12 Asian nations, Park (forthcoming) presents data which show that Japanese survey respondents on average rate political freedom and electoral competitiveness in their country less positively than in a majority of other cases. Such comparison should be taken with a grain of salt, since citizens living in poorer conditions or under authoritarian regimes would probably give high scores so long as their governments meet some minimum standards, whereas the Japanese public employs more exacting measures to judge the quality of governance. Moreover, it is crucial to bear in mind that whatever the causes of discontent with the Japanese government, democratic principles and institutions have not been called into question. Nevertheless, limited participation beyond the polling booth, especially when contrasted with more active publics in many other consolidated and even new democracies, raises concerns about citizens’ willingness and ability to not only voice their demands but also to hold the government accountable.
That said, since this study only covers the period up to 2010 due to lack of comparable data in the past several years, it cannot take into account subsequent events which have affected public participation. First, in 2011 regions in northeastern Japan were devastated by an earthquake, tsunami, and nuclear meltdown. The continued plight of some victims forced to relocate and, more importantly, the risks of nuclear energy exposed by these events, raised awareness and encouraged collective action on the part of many citizens who do not normally pay close attention to politics. Second, the LDP government’s decision in 2015 to ram through parliament bills aimed at expanding the functional and geographical scope of Japan’s exercise of military power provoked strong reactions, with not only elderly anti-war activists fearful of repeating the trauma they experienced during the Second World War, but also college and even high school students playing prominent roles in protests. While these activities arose in response to specific events and would diminish in intensity over time, it is possible that an individual sense of efficacy gained through involvement, and experiences of mobilization and coordination on the part of those who organized these efforts, may have longer-term ramifications for rates and patterns of political engagement. Furthermore, since issues of nuclear power and (especially) collective defense are deeply tied into conventional ideological conflicts in Japan, one may see a revival in the role that ideology plays in mobilization and participation.
According to Welzel and Dalton (forthcoming), several Asian countries such as Japan, South Korea, Taiwan, and Thailand have seen a rise in ‘assertive’ values, as measured by emphasis on social liberalism, gender equality, and giving more say to ordinary citizens. At the same time, the data they present show that these values are less pervasive in Japan than most other countries at similar levels of economic development, and lower than all other established democracies. Also, the prevalence of these orientations has only risen slightly over 30 years. Furthermore, Table 1 implies that what (limited) attitudinal shift one can observe in this period did not always find behavioral manifestation. The causal arrow leading from assertive values to government effectiveness and accountability that Welzel and Dalton convincingly demonstrate is likely based on the assumption that beliefs are translated into action. Whether this represents a missing link in Japan and, if so, how this can be rectified, presents two pertinent topics for future research and important challenges for all those with an interest in enhancing the quality of democratic governance.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.