Partisanship and democratization.

Shin, Doh Chull ; Tusalem, Rollin F.

How do attachments to political parties among the mass publics of East Asia affect the process of democratization in the region? Analyses of the East Asia Barometer surveys reveal that partisanship motivates East Asians to endorse the democratic performance of their political system and embrace democracy as the best possible system of government. These findings accord, by and large, with the socialization, cognitive dissonance, and rational choice theories of partisanship.

KEYWORDS: cognitive dissonance, democratization, East Asians, partisanship, rational choice, and socialization

**********

Political parties perform a variety of functions for "achieving, maintaining, and improving the quality of democracy." (1) How effectively parties can perform such democratic functions, however, depends largely on whether they form an institutionalized party system that has taken root deeply in society. (2) One indicator of such a viable party system concerns the extent to which voters identify with a party or have a party preference. (3) Without widespread citizen attachment to political parties, the electoral process becomes highly volatile and the legislative process becomes highly unpredictable in its formulation of major policies. Such partisan attachment, known as partisanship, is, therefore, essential to the institutional process of democratization. (4)

The democratization effects of mass partisanship go well beyond the institutional dimension of democratization. (5) As a political predisposition often called "a prime mover," partisan attachment powerfully shapes many other political attitudes and beliefs. (6) As discussed in earlier articles, it motivates ordinary citizens to engage in electoral and other political activities. Furthermore, as a long-term affective orientation, it motivates them to perceive and react positively to the political world in which they live. (7) By promoting favorable attitudes toward the democratic system in which citizens live, partisanship can contribute to the process of legitimizing democratic rule in the minds of the masses. (8)

To date, however, few studies have sought to explore the consequences shifting partisanship has on the cultural dimension of legitimizing democratic rule by directly examining partisanship's link to support for democracy. (9) Research on this link has dealt mostly with old democracies in Europe and North America. As a result, little is known about how orientations to political parties shape commitment to democracy among the citizens of nascent democracies in other regions.

In Eastern Europe, support for new democratic regimes is 13 percentage points higher among those identified with a party than those without a party identification, and opposition to undemocratic alternatives is 9 percentage points higher among the former than the latter. (10) In Africa, partisan affiliations also significantly shape popular reactions to democracy. (11) Among those identified with the winning party, for example, 74 percent expressed satisfaction with democracy. The corresponding figure for those identified with losing parties was only 47 percent. Clearly, identification with a winning political party motivates Africans to support democracy and evaluate its performance positively to a greater extent. In Asia, however, no empirical study has yet investigated how partisanship and its status affect popular reactions to democracy.

Is partisan attachment facilitating the democratization process in East Asia? Does partisanship motivate East Asians to accept democratic politics, while encouraging them to reject its authoritarian alternatives? Does it also motivate them to perceive the performance of the existing democratic political system in a positive light? Do those attached to the ruling and opposition parties perceive it differently? If so, why do partisans and nonpartisans react differently to the democratization process? This article explores these questions by analyzing the first wave of the East Asia Barometer (EAB hereafter) surveys conducted in Japan, Korea, Mongolia, the Philippines, Taiwan, and Thailand. (12) Unlike the Comparative Study of Electoral Systems and World Values surveys used in other articles, the EAB surveys tapped democratic support from multidimensional and multilevel perspectives. Japan, the only established democracy in East Asia, is included in our analysis to explore whether partisanship plays a different role during the course of democratization.

This article is organized into several sections. In the section following immediately, we review recent scholarship on the role of partisanship in democratization and propose a number of theories and hypotheses linking partisanship with prodemocratic and antiauthoritarian orientations. In the second section, we explicate the meanings of partisanship and democratic support, the two key concepts underlying these hypotheses. We also introduce indicators of these concepts. We then move to the largest part of the article, which analyzes the relationships between partisan attachments and reactions to institutional and cultural democratization among East Asians. In the final section, we summarize key findings and discuss their implications for the democratization of mass citizenries in East Asia.

Theoretical Considerations

The above review of the literature reveals that a sense of identification with a political party is a "super attitude," that is, one that influences other political attitudes and beliefs. It also reveals that psychological attachment to a political party promotes citizen support for the democratic political system in which the party exists. Why does partisan attachment give rise to support for democratic political order among the citizens who live in a democracy? The existing literature on partisanship does not offer any single unified or general theory explaining, either deductively or inductively, why popular support for political parties translates to popular support for democracy. Instead, the literature offers a number of specific theoretical propositions linking an affective orientation to a political party with various components of democratic support.

One proposition emphasizes the socialization role that partisanship plays in political life. (13) Specifically, it posits that attachment to political parties exposes citizens to the issue positions and policy programs their favored parties advocate and stimulates them to become psychologically and behaviorally involved in the political process. Over time, repeated involvement in the political process integrates partisans into the party-based democratic system, and consequently leads them to develop favorable attitudes toward the system itself. Familiarity with the democratic process breeds contentment or satisfaction with it. Political involvement also encourages citizens to endorse the idea that the democratic political system needs parties to function correctly. Thus, this theory of political learning and socialization holds that partisans will be more supportive of democracy than nonpartisans when comparing democracy with its alternatives.

A second theoretical proposition is derived from the psychological theory of cognitive dissonance. (14) According to this theory, individuals seek to maintain consistency among their attitudes and beliefs by avoiding those that contradict them. Those favorably attached to political parties, therefore, seek to avoid cognitive dissonance by orienting themselves favorably toward the democratic political system populated with political parties. As much as they embrace their own political parties, therefore, they favor the political system in which those parties exist. To maintain consistently positive attitudes toward democratic politics, we expect partisans, as compared to nonpartisans, to be more favorably oriented toward the ideals and practices of democracy.

A third theoretical proposition treats partisanship as a predisposition that can mediate the impact of electoral winning and losing on support for the political system in which people live. (15) Taking the perspective of rational choice or prospective regime performance, this proposition emphasizes the expected benefit from the outcome of an electoral contest. Of the political parties taking part in the contest, only the winning party is likely to implement its favorite policies and thus benefit its supporters. Therefore, those identified with the winning party are expected to be more satisfied with and supportive of the political system than those who are identified with the losing parties. Thus, this theory differs from the socialization and cognitive dissonance theories, which claim all partisans, winning and losing, will be more supportive than nonpartisans. (16)

Conceptualization and Measurement

From the three theories, we derived a number of hypotheses linking partisanship with positive reactions to institutional and cultural democratization. To test these hypotheses, we need to define and measure the key concepts of partisanship and democratization. This section explicates each of these two key concepts, and outlines a strategy for measuring them with responses to the EAB surveys conducted in six East Asian countries during the 2001-2003 period.

Partisanship

What constitutes partisanship? Since the publication of The American Voter in 1960, a large body of literature has accumulated to address that question. (17) The authors of these studies often equate partisanship with an affective or emotional attachment to a particular political party. Emulating their research on political consequences of partisanship, our study focuses on affective identification with a particular party. It should be noted, however, that this conception of partisanship appears narrow when compared with another conception in which partisanship represents a frame of mind toward political parties in general. (18)

To measure affective partisan orientations among East Asians, we selected one of the two questions the EAB surveys asked to tap partisan attachments. The EAB asked respondents to name the one political party to which they felt closest on the list of all the parties in their country. Those who did not name any political party were recognized as nonpartisans. Those who named the governing party and those who named an opposition party were recognized, respectively, as winning and losing party identifiers. (19)

Democratization

The existing literature on third-wave democracies generally agrees that democratization takes place in individual citizens and their political regime. (20) To examine the democratic changes at the regime level, (21) we selected a pair of questions from the EAB surveys. The first question asked respondents to rate the extent to which their current regime was operating as a democracy or dictatorship on a 10-point scale. The second question asked them to rate on a 10-point scale the extent to which they were satisfied or dissatisfied with the way the existing regime was performing as a democracy. By considering these two assessments together, we estimated the extent to which respondents recognized the current political system as a well-functioning democracy.

To examine the democratic changes taking place at the level of institutions, we selected from the EAB surveys a set of four questions dealing with four key institutions of representative democracy: the national government, political parties, parliament, and local government. To determine how responsibly and accountably these institutions have performed in the eyes of citizens, the surveys asked respondents how much trust they had in each institution. We converted their responses into an index of institutional trust. Values of this index range from a low of 0 to a high of 4.

A third level is the democratization of cultural beliefs and values. It is a multilayered or multilevel phenomenon because citizens simultaneously comprehend democracy not only as an abstract idea but also as a political system-in-practice. (22) It is also a multidimensional phenomenon because it involves the acceptance of democracy as well as the rejection of its alternatives.

Embracing democracy as an abstract idea is not the same as embracing democracy as a working regime. Democracy as an abstract idea represents certain political principles, values, and ideals while democracy-in-practice represents the actual workings of the democratic institutions that govern citizens' daily lives. (23) Ordinary citizens who have lived most of their lives under authoritarian rule may have an idealized vision of what democracy can be, and yet not be fully supportive of the democracy they see at work in their country. Consequently, popular support for democracy must be differentiated into two broad categories: normative and practical.

To measure normative democratic support, we selected an item that asked respondents to express their desire to live in a democracy or dictatorship on a 10-point scale. Those who scored 6 or higher on this scale were considered supporters of democracy-in-principle. Thus, calculating their percentage provides a measure of normative democratic support.

To measure support for democracy-in-practice, we selected a set of three items from the EAB surveys. The first item asked respondents to appraise the suitability of democracy for their country. The second item asked them whether they agreed with the view that democracy is always preferable to any other kind of government. The third item asked them whether they believed democracy would be capable of solving the problems of their society. Those who endorsed democracy as a suitable, preferable, and efficacious system of governance were considered supporters of democracy-in-practice. Thus, calculating their percentage provides a measure of practical democratic support.

In addition, citizens with little experience and limited sophistication about democratic politics may be uncertain whether democracy or dictatorship offers satisfying solutions to the problems facing their societies. Under such uncertainty, citizens often embrace both democratic and authoritarian political propensities concurrently. (24) Consequently, the acceptance of democracy does not necessarily cause the rejection of authoritarianism or vice versa. We selected three questions to tap orientations toward or against the practices of authoritarian rule: whether respondents favored the return of the military in the country; whether they would approve of getting rid of parliament and elections and having a strong leader decide everything; and whether they would approve of one-party rule. Calculating the percentage of respondents who reject all three of these forms of undemocratic rule reveals the extent of the citizens' dissociation from the virtues of authoritarian rule.

Findings from the EAB Surveys: The Contours of Partisanship

We begin the analyses by categorizing East Asian citizens into nonpartisans, losing partisans, and winning partisans. Evidently, as shown in Table 1, the proportions of nonpartisans and partisans vary a great deal across the six countries in East Asia. In Thailand, for instance, partisans constitute a minority (41 percent) of the electorate. In Mongolia, on the other hand, partisans constitute an overwhelming majority (90 percent). In the other four countries, partisans outnumber nonpartisans (53 percent in Japan, 57 percent in Taiwan, 54 percent in the Philippines, and 73 percent in Korea).

Among partisans, winning partisans outnumber losing partisans in three of the six countries: Japan, Korea, and Thailand. In the other three countries, Mongolia, the Philippines, and Taiwan, losing partisans outnumber winning partisans. More notable is that winning partisans do not constitute a majority in any of the six East Asian countries surveyed. Even in consolidated democracies like Japan (32 percent) and Korea (38 percent), there is only a mere plurality of respondents that identify with the ruling party. This finding indicates that the attachments of most East Asians to political parties are highly uneven from one country to another. Within each of the six countries surveyed, moreover, partisan attachments are not concentrated in one or two parties; instead, they are highly dispersed among several parties, including the winning party.

Like those in other third-wave democracies, party systems in East Asia are neither institutionalized nor based on class distinctions. Working- or labor-class activism, which supposedly solidifies partisanship identification in the West, (25) is lacking in East Asia. As a result, most parties are formed on the basis of elite interests and personal ties, and fail to represent the voices and policy demands of the various sectors of society. (26) Accordingly, mass attachments to these parties remain tentative and fleeting from one election to another. This may be the reason why none of the ruling parties in East Asia holds support from a majority of its electorate.

Facilitating Institutional Democratization

Democratization always involves the democratic transformation of authoritarian political institutions and procedures. How does attachment to political parties affect the process of such institutional democratization in East Asia? To address this question, we first compare the extent to which nonpartisans, losing partisans, and winning partisans perceive their current political system as a democracy that performs to their satisfaction.

Regime Performance

For each of the six East Asian countries, Table 2 reports the proportions of partisans and nonpartisans who rated their current regime as a well-functioning democracy. Of these countries, Thailand stands out as the only country where large majorities of both partisans and nonpartisans endorsed the current regime. Why is it that Thais, unlike those of other nationalities, were uniformly and unusually high in their endorsement of their nascent democratic rule? It appears that Thaksin Shinawatra's winning the largest popular mandate ever in the January 2001 National Assembly elections, and subsequently forming the first majority democratic government in their country, encouraged respondents to the EAB survey conducted later in that year to become overly enthusiastic about democracy as a political ideal as well as political practice without dissociating themselves from authoritarianism. (27) Even among their nonpartisans, more than four-fifths (84 percent) were supportive of their government as a well-functioning democracy. This figure is from 36 to 46 percentage points higher than what was observed in five other countries.

Table 2 also shows that, in every country, partisans, either losing or winning, are more satisfied with their democratic experience than their nonpartisan counterparts. In all six East Asian countries, nonpartisans are the least positive about the way their political system performs as a democracy. With five of the six East Asian democracies, majorities of partisans endorse the current level of institutional democratization, while majorities of nonpartisans refuse to endorse it. Thailand is the only country where a majority of ordinary citizens support their democracy regardless of their partisanship. This finding that partisans, whether winning or losing, endorse the current democratic regime confirms the socialization hypothesis that partisanship familiarizes ordinary citizens with the democratic political process, and this experience breeds a greater sense of satisfaction with the performance of their political system as a democracy. It also confirms the cognitive dissonance hypothesis that the citizens who are favorably oriented toward political parties are likely to embrace the political system in which the parties operate.

Table 2 also reports that, in every country, winning partisans are more likely than losing partisans to appraise their political system in a positive light. Regardless of the authoritarian political system in which they once lived or the level of socioeconomic modernization their country has achieved, East Asians who identify with their ruling party are the most positively oriented to the democratic performance of the existing political systems. Once again, Thailand is the only country where an overwhelming majority (90 percent) of winning partisans support the current regime as a well-functioning democracy. In other East Asian democracies, including Japan, the oldest democracy in the region, a relatively small majority or large minority of winning partisans endorse it: Japan's Liberal Democratic Party (57 percent); Korea's Uri Party (66 percent); Mongolia's People's Revolutionary Party (59 percent); the Philippines' governing coalition of Gloria Arroyo Party, Lakas Ng Tao, Liberal Party, and People Power Coalition (48 percent); (28) and Taiwan's Democratic Progressive Party (56 percent).

This finding that winning partisans are more favorably oriented to the existing democratic regime than losing partisans is statistically significant in all East Asian countries with the exception of Mongolia. It is, therefore, in general agreement with the rational choice or prospective regime performance proposition that supporters of the winning party are more satisfied with and supportive of the existing political system than those of losing parties, as the former expect greater benefits from the political system than the latter. (29)

Institutional Trust

We also compared the levels of trust East Asians place in four major democratic institutions--the national government, the parliament, political parties, and the local government--across the three categories of nonpartisans, losing partisans, and winning partisans. For each partisanship category, we calculated the mean number of the institutions respondents reported they were trusting and placed them on a 5-point scale ranging from a low of 0 to a high of 4. Figure 1 presents these means.

[FIGURE 1 OMITTED]

In Figure 1, we see that partisans are not always more trusting of democratic institutions than are nonpartisans. In four of the six East Asian countries--Japan, Mongolia, the Philippines, and Taiwan--the former tend to trust these institutions more, but not always significantly more, than do the latter. In Korea, Mongolia, and Taiwan, for instance, the differences between their mean index values range from 0.01 in Mongolia to 0.14 in Japan on the 5-point scale ranging from 0 to 4. This finding clearly shows that mere attachment to a political party does not promote institutional trust either consistently or powerfully. It also suggests that neither the socialization nor cognitive dissonance theory of partisanship holds true at the subsystem level of particular institutions.

Concerning winning and losing partisans, Figure 1 shows that the former always report significantly higher mean levels of institutional trust than do the latter. Winning partisans register the highest level of trust for democratic institutions in every East Asian country, from Japan, the oldest established democracy in the region, to the third-wave democracies of Korea, Mongolia, the Philippines, Thailand, and Taiwan. This finding accords with the rational choice model of partisanship, which holds that winning partisans will have greater support for institutions of representative democracy than will losing partisans or nonpartisans. Moreover, the impact of winning partisanship on institutional trust appears to be more pronounced than its impact on the appraisal of the existing political system as a well-functioning democracy. The eta coefficients reported in Table 2 and Figure 1 indicate that winning partisanship is more strongly associated with institutional trust than perceptions of a well-functioning democracy in four of the six East Asian countries--Japan, Mongolia, Taiwan, and Thailand.

Fostering Cultural Democratization

To comprehend the role partisanship plays in the entire process of democratization, we need to examine whether partisanship also matters in cultural democratization, that is, orienting East Asians toward democracy and away from authoritarianism. We differentiate those orientations into two categories, normative and practical support for democracy.

Normative Support for Democracy

For each East Asian country, Figure 2 reports the percentages of nonpartisans, losing partisans, and winning partisans who expressed a democratic desire by choosing 6 or higher on a 10-point scale in which a score of 1 means complete dictatorship and 10 means complete democracy. Most notable, all East Asian countries embrace democracy as an ideal political system. Regardless of their partisanship, large majorities ranging from 84 percent to 99 percent support democracy-inprinciple. In Japan, Korea, Mongolia, and Thailand, more than 9 out of 10 people expressed the desire to live in a democracy rather than in a dictatorship. Only in the Philippines (11 percent) and Taiwan (12 percent) did more than 1 out of 10 people express the desire to live in a nondemocratic regime.

[FIGURE 2 OMITTED]

Another notable feature of Figure 2 is that partisanship is not very instrumental in shaping the citizens' desire to live in a democracy. Specifically, partisanship does not always mean higher levels of support for democracy-in-principle. For instance, in the two countries of Japan and Thailand, nonpartisans are more supportive of democracy as an ideal political system than partisans. Only in one of the four other countries, Taiwan, do partisans report significantly higher levels of normative democratic support than do nonpartisans. In none of the six East Asian countries, moreover, do winning partisans report a significantly higher level of such democratic support than do losing partisans. These findings, when considered together, make it clear that partisanship, either losing or winning, does not distinctively shape normative support for democracy.

Practical Support for Democracy

What effect does partisanship have on support for democracy-in-practice among East Asians? We used the index described above to reveal the strength of practical support for democracy. Citizens who respond positively to all three items in the index are fully supportive of democracy as it is practiced within their polity. Table 3 reports percentages expressing such full practical support for democracy among those in three categories of partisanship.

Between nonpartisans and partisans, the latter tend to support democratic practices more than the former, although they do not always do so. In Japan, the Philippines, Taiwan, and Thailand, levels of practical democratic support are higher among partisans than nonpartisans, but in Korea and Mongolia, nonpartisans report higher levels of such support than do partisans (37 percent vs. 32 percent and 40 percent vs. 37 percent, respectively). In addition, between winning and losing partisans, the former tend to report greater support for democratic practices than do the latter, although not always. In five of the six countries, winning partisans are more supportive of democratic practices than losing partisans, but in Mongolia, the percentage of losing partisans who support democracy-in-practice is significantly greater than the percentage of winning partisans who do (58 percent vs. 37 percent). Considering these findings together reveals that in East Asia today, partisans are more likely to support democratic practices than nonpartisans, and that winning partisans are more likely to do so than losing partisans.

Opposition to Authoritarian Rule

People in third-wave democracies lived many years under authoritarian regimes; consequently, some may remain attached to its virtues even as they embrace democracy as the best political system. To examine fully the role of partisanship in cultural democratization, we also considered popular detachment from authoritarianism. Respondents who rejected all three authoritarian regimes discussed above are considered true authoritarian opponents. Table 4 compares percentages of authoritarian opponents across three categories of partisanship.

Of the six countries reported in the table, Japan, Korea, and Taiwan are the three countries where majorities of both partisans and nonpartisans are detached from the virtues of authoritarian rule. In Mongolia, only partisans--winning and losing--are detached from them. In the Philippines and Thailand, on the other hand, majorities of nonpartisans as well as partisans remain attached to authoritarianism. This finding suggests that the rejection of authoritarian rule among the mass publics of East Asia has to do with the types of political regime they experienced in the past and with the extent to which they are socio-economically modernized. The more oppressive their past regime was, the more quickly they dissociate themselves from it. The more modernized they are, the more quickly they reject it.

Table 4 shows that partisans are not detached any more from authoritarianism than nonpartisans. This runs counter to the socialization, cognitive dissonance, and rational choice hypotheses, which link partisanship to democratization. In fact, nonpartisans are more detached from authoritarianism than partisans of either a losing or winning party in five of the six countries. In Korea, for instance, more than three-quarters (77 percent) of nonpartisans are fully detached from the virtues of authoritarian rule, whereas only two-thirds (68 percent) of winning partisans are. Taiwan is the only country where opponents of authoritarianism are more numerous among partisans than nonpartisans.

Concerning partisans only, winning partisans are not any less authoritarian than losing partisans. Only in the Philippines are the former significantly less authoritarian than the latter. In the other five countries, it is losing partisans, not winning partisans, who reject authoritarianism in a greater proportion. In Mongolia, for instance, 52 percent of losing partisans reject all three alternatives to democracy, while only 33 percent of winning partisans do the same. In Taiwan, 69 percent of the former are opponents of authoritarian rule, while only 60 percent of the latter are. Our findings in Table 4, when considered all together, offer little evidence to support the hypotheses claiming that partisanship contributes to cultural democratization by dissociating the masses from the virtues of authoritarian rule.

Determining Independent Influence

The final step of our analysis isolates partisanship from other variables known to influence democratization and determines whether partisanship independently contributes to democratization. (30) To address this question, we first selected five demographic characteristics--gender, age, educational attainment, class standing (subjective), and community type--as control variables. (31) We then employed Multiple Classification Analysis to estimate the beta statistics for partisanship and five other control variables. (32) Table 5 reports the beta coefficients for the relationships of the partisanship variable with two institutional and three cultural domains of democratization.

Being analogous to standardized regression coefficients, the beta statistics reported in the table indicate the relative power of each predictor. According to these coefficients in the table, there is a great deal of variation from one country to another in the way partisanship shapes citizen reactions to democratization. In Korea and Mongolia, for instance, it matters significantly in all five domains, including normative support for democracy. In the Philippines, on the other hand, it matters only in two domains. From one domain to another domain of democratization, partisanship also plays a significantly different role. In the domain of democratic regime performance, it matters significantly in all six countries. In the domains of normative and practical democratic support, it matters similarly only in three of the six countries. Why does partisanship motivate citizens to react to democracy differently across its domains and across countries within the same region? This is an important question that needs further research.

Nonetheless, Table 5 shows that partisanship matters significantly for all six East Asian countries in at least one institutional domain (the regime as a whole and/or particular institutions) and at least one cultural domain (democracy-in-principle, democracy-in-action, and/or antiauthoritarianism) of their democratization even after holding other factors constant. In three of the six countries--Korea, Mongolia, and Taiwan--it shapes most or all of their democratization domains significantly. In four countries--Japan, Korea, Mongolia, and Taiwan--partisanship influences the process more powerfully than do any of the five other demographic variables, including educational attainment. In short, partisan ties do shape the cultural and institutional dynamics of democratization in East Asia.

Comparisons of the beta coefficients across countries suggest a contrasting pattern of the role partisanship plays over the course of democratization. In the two domains of institutional democratization, the magnitude of these coefficients is higher for Japan, an established democracy, than most of the five new East Asian democracies. In all three domains of cultural democratization, however, their magnitude is lower for the former than the latter. Over the course of democratization, it appears that partisanship becomes stronger in motivating ordinary people to embrace democratic institutions, but it weakens in orienting them toward democracy and away from authoritarianism. These findings, when considered together, do not accord with what is known from the studies conducted in other regions. According to these studies, partisanship is uniformly more influential in orienting the masses toward democratic institutions and values in emerging democracies than established democracies. (33) In striking contrast, it contributes more only to cultural democratization in the former than the latter.

Conclusion

This article has explored whether mass partisanship contributes to the process of democratization in East Asia. Our analyses of the EAB surveys indicate that attachments to political parties among East Asian publics do support the process of democratization in the region. As the socialization and cognitive dissonance theories hold, partisan attachments tend to contribute to institutional democratization by motivating East Asians to recognize a newly installed political system as a democracy and appreciate its virtues. Those attachments also tend to contribute to institutional democratization by encouraging citizens to tolerate the shortcomings of key democratic institutions and thereby ensure the continuity of those institutions. As the rational choice theory of partisanship holds, identification with the governmental party facilitates this process of institutional democratization by endorsing limited democratic rule and supporting its expansion to the greatest extent. (34)

By and large, partisanship also contributes to the process of cultural democratization. It motivates East Asian citizens to endorse the practices of democratic politics and embrace it as the preferred system of government. In orienting them away from the virtues of authoritarian rule, however, identification with either a winning or losing party matters much less. Yet, in every East Asian country, partisanship contributes independently to positive citizen reactions to at least one of the three surveyed domains of cultural democratization. Only in the domains of cultural democratization, partisanship contributes consistently less in Japan, an established democracy, than the other five new democracies.

On the basis of these findings, we conclude that mass partisanship fosters the legitimization of nascent democratic rule, and promotes its expansion and consolidation in East Asia. For further research, we need to ask why partisanship plays a divergent role across the different domains and dimensions of democratization. We need to also ask why it plays a divergent role across countries in varying levels of democratic change.

Notes

(1.) "Democracy Without Parties? Political Parties and Regime Collapse in Fujimori's Peru," paper presented at the Congress of the Latin American Studies Association, Washington, DC, September 6-8, 2001, p. 5.

(2.) Scott Mainwaring, Rethinking Party Systems in Third Wave of Democratization (Stanford: Stanford University Press, 1999); Geoffrey Pridham and Paul G. Lewis, eds., Stabilising Fragile Democracies: Comparing the New Party Systems in Southern and Eastern Europe (London and New York: Routledge, 1966).

(3.) Russell J. Dalton and Steven Weldon, "Partisanship and Party System Institutionalization," Party Politics 13, no. 2 (March 2007); Vicky Randall and Lars Svasand, "Party Institutionalization in New Democracies," Party Politics 8, no. 1 (January 2002): 5-29.

(4.) Ian McAllister and Stephen White, "Democracy, Political Parties and Party Formation in Postcommunist Russia," Party Politics 1 (1995): 49-72.

(5.) Timothy J. Colton, "Babes in Partyland: The Riddle of Partisanship in Post-Soviet Russia" (unpublished manuscript); Richard Gunther, Jose Ramon Montero, and Juan Linz, eds., Political Parties: Old Concepts and New Challenges (Oxford: Oxford University Press, 2002).

(6.) Russell J. Dalton, Citizen Politics: Public Opinion and Political Parties in Advanced Industrial Democracies, 4th ed. (Washington, DC: CQ Press, 2005), p. 174. See also Michael Bratton, Robert Mattes, and E. Gyimah-Boadi, Public Opinion, Democracy, and Market Reform in Africa (Cambridge: Cambridge University Press, 2005), p. 39.

(7.) Christopher Anderson and Christine A. Guillory, "Political Institutions and Satisfaction with Democracy," American Political Science Review 91, no. 1 (1997): 66-81; Larry Bartels, "Beyond the Running Tally: Partisan Bias in Political Perceptions," Political Behavior 24, no. 2 (2002): 117-150.

(8.) Christian Anderson, "Parties, Party Systems, and Satisfaction with Democratic Performance in the New Europe," Political Studies 46, no. 4 (1998): 572.

(9.) Christopher Anderson et al., Loser's Consent: Elections and Democracy Legitimacy (Oxford: Oxford University Press, 2005); Jack Dennis and Diana Owen, "Popular Satisfaction with the Party System and Representative Democracy in the United States," International Political Science Review 22, no. 4 (2001): 399-415; Soren Holmberg, "Are Political Parties Necessary?" Electoral Studies 22, no. 2 (2003): 287-299.

(10.) Richard Rose, William Mishler, and Christian Haerpfer, Democracy and Its Alternatives (Baltimore: Johns Hopkins University Press, 1998), p. 157.

(11.) Bratton, Mattes, and Gyimah-Boadi, Public Opinion, Democracy, and Market Reform, pp. 259-261.

(12.) The study reported in this paper is based on the first wave of the EAB surveys conducted in 2002 and 2003. These multinational public opinion data consist of responses collected through face-to-face interviews with randomly selected voters in Japan (N = 1,360), South Korea (N = 1,500), Mongolia (N = 1,144), the Philippines (N = 1,199), Taiwan (N = 1,350), and Thailand (N = 1,546) (though surveys were conducted in China and Hong Kong, these are not included in this study). Further information about questionnaire design, sampling methodology, and fieldwork is available at www.eastasiabarometer.org.

(13.) Holmberg, "Are Political Parties Necessary?"

(14.) Anderson et al., Loser's Consent, p. 76; Aida Paskeviciute and Christopher J. Anderson, "Political Parties, Partisanship, and Attitudes Toward Government in Democracies," presented at the annual meeting of the Midwest Political Science Association, Chicago, Illinois, April 2004.

(15.) Anderson et al., Loser's Consent, p. 75; Christopher Anderson and Yuliya Tverdova, "Winners, Losers, and Attitudes About Government in Contemporary Democracies," International Political Science Review 22, no. 4 (2001): 323.

(16.) Christopher Anderson and Yuliya V. Tverdova, "Corruption, Political Allegiances, and Attitudes Toward Government in Contemporary Democracies," American Journal of Political Science 47, no. 1 (2003): 91-109.

(17.) Angus Campbell et al., The American Voter (New York: Wiley, 1960); Morris Fiorina, "Parties and Partisanship: A Forty Year Retrospective," Political Behavior 24, no. 2 (2002): 93-115.

(18.) Russell Dalton, Democratic Challenges, Democratic Choice: The Erosion of Political Support in Advanced Industrial Democracies (Oxford: Oxford University Press), p. 33; see also Steven Greene, "The Social-Psychological Measurement of Partisanship," Political Behavior 24, no. 3 (2002): 171-197.

(19.) Ruling parties are Japan's Liberal Democratic Party; Korea's Uri Party; Mongolia's People's Revolutionary Party; the Philippines' governing coalition of Gloria Arroyo Party, Lakas Ng Tao, Liberal Party, and People Power Coalition (48 percent); Taiwan's Democratic Progressive Party (56 percent); and Thailand's Thai Rak Thai Party (90 percent).

(20.) Valerie Bunce, "Rethinking Recent Democratization: Lessons from the Postcommunist Experience," World Politics 55, no. 2 (2003): 167-192; Laurence Whitehead, Democratization: Theory and Experience (Oxford: Oxford University Press, 2002).

(21.) Russell Dalton, "Political Support for Advanced Industrial Democracies." In Pippa Norris, ed., Critical Citizens (Oxford: Oxford University Press, 1999), pp. 57-77.

(22.) Hans-Dieter Klingemann, "Mapping Political Support in the 1990s: A Global Analysis," in Norris, ed., Critical Citizens, pp. 31-56; Doh Chull Shin, "Democratization: Perspectives from Global Citizenries." In Russell Dalton and Hans-Dieter Klingemann, eds., The Oxford Handbook of Political Behavior (Oxford: Oxford University Press, 2007), pp. 259-282.

(23.) William Mishler and Richard Rose, "Political Support for Incomplete Democracies," International Political Science Review 22, no. 4 (2001): 303-320.

(24.) Rose, Mishler, and Haerpfer, Democracy and Its Alternatives; Doh Chull Shin, Mass Politics and Culture in Democratizing Korea (Cambridge: Cambridge University Press, 1999).

(25.) Seymour Lipset and Stein Rokkan, eds., Party Systems and Voter Alignments: Cross National Perspectives (New York: Free Press, 1967). See also Ian McAllister's article in this issue.

(26.) Wolfgang Sachsenroder and Ulrike Frings, eds., Political Party Systems and Democratic Developments in East and Southeast Asia (London: Ashgate, 1998).

(27.) Aurel Croissant, "Parliamentary Elections in Thailand, March 2000 and January 2001," Electoral Studies 22, no. 1 (March 2003): 153-160.

(28.) Of the six East Asian democracies, the Philippines is the only country where a majority of winning partisans refused to endorse the current regime. This is probably because the congress impeached President Estrada and the Supreme Courts swore in Gloria Arroyo as acting president on January 20, 2001. For further details, see Steven Rogers, "Philippine Politics and the Rule of Law," Journal of Democracy 15, no. 4 (October 2004): 111-125.

(29.) Christopher Anderson and Silvia Mendes, "Learning to Lose: Election Outcomes, Democratic Experience and Political Protest Potential," British Journal of Political Science 36, no. 4 (2005): 91-111; Anderson and Tverdova, "Winners, Losers, and Attitudes"; Paskeviciute and Anderson, "Political Parties, Partisanship, and Attitudes."

(30.) Bratton, Mattes, and Gyimah-Boadi, Public Opinion, Democracy, and Market Reform; Rose, Mishler, and Haerpfer, Democracy and Its Alternatives; Shin, Mass Politics and Culture in Democratizing Korea.

(31.) Age, school, and class standing were measured in terms of five categories. Community type was measured in terms of two categories, rural and urban.

(32.) Frank Andrew, James Morgan, and James Gouquist, Multiple Classification Analysis (Ann Arbor: Survey Research Center, the University of Michigan, 1973).

(33.) Anderson, "Parties, Party Systems, and Satisfaction"; Anderson et al., Losers' Consent (2005); Holmberg, "Are Political Parties Necessary?"

(34.) Anderson et al., Losers' Consent, p. 527.

Doh Chull Shin is professor of political science at the University of Missouri at Columbia and holds the Korea Foundation and Middlebush chairs. Since 1988 he has directed the Korean Democracy Barometer program and is one of the founding members of the East Asia Barometer program. His scholarly interests include cultural values, democratic regime change, market reform, and the quality of life. His recent publications include Mass Politics and Culture in Democratizing Korea (1999), Institutional Reform and Democratic Consolidation in Korea (2000), The Quality of Life in Korea (2003), Economic Crisis and Dual Transition in Korea (2004), and Citizens, Democracy and Markets Around the Pacific Rim (2006).

Rollin F. Tusalem is a doctoral student in political science at the University of Missouri at Columbia. His interests include the comparative study of civil society and governance, democratization, and the politics of East Asia.

Table 1 Variants of Partisanship Identification Among East Asians

 Nonpartisans Losing Winning Total N
 (%) Partisans Partisans (%)
 (%) (%)

Japan 46.5 21.6 31.9 100 1,360
Korea 26.7 34.9 38.4 100 1,500
Mongolia 9.8 46.4 43.8 100 1,144
Philippines 46.0 31.8 22.1 100 1,199
Taiwan 43.2 33.8 23.0 100 1,350
Thailand 58.6 19.1 22.3 100 1,546

Source: East Asia Barometer Surveys, 2001-2003.

Table 2 Perceptions of a Well-Functioning Democracy by
Types of Partisanship

 Nonpartisans Losing Winning Eta
 (%) Partisans Partisans
 (%) (%)

Japan 42.6 44.7 56.8 .13 *
Korea 44.4 48.5 65.7 .19 *
Mongolia 48.1 53.9 58.7 .06
Philippines 37.9 40.4 47.9 .08 *
Taiwan 41.4 44.2 56.3 .12 *
Thailand 83.9 84.5 90.0 .07 *

Source: East Asia Barometer Surveys, 2001-2003.

Note: * indicates statistically significant at the 0.05 level.

Table 3 Types of Partisanship and Levels of Practical Support
for Democracy

 Nonpartisans Losing Winning Eta
 (%) Partisans Partisans
 (%) (%)

Japan 64.5 65.4 69.1 .04
Korea 36.9 31.5 44.7 .12 *
Mongolia 40.0 57.8 37.1 .20 *
Philippines 36.8 38.8 41.0 .03
Taiwan 26.5 29.1 36.5 .09 *
Thailand 77.4 71.4 78.5 .06

Source: East Asia Barometer Surveys, 2001-2003.

Note: * indicates statistically significant at the 0.05 level.

Table 4 Types of Partisanship and Levels of Opposition
to Authoritarian Politics

 Nonpartisans Losing Winning Eta
 (%) Partisans Partisans
 (%) (%)

Japan 57.3 64.2 57.1 .06
Korea 77.3 70.6 67.7 .08 *
Mongolia 45.5 52.2 33.3 .18 *
Philippines 43.7 34.4 41.2 .08 *
Taiwan 53.0 69.4 59.7 .15 *
Thailand 46.1 49.3 45.1 .03

Source: East Asia Barometer Surveys, 2001-2003.

Note: * indicates statistically significant at the 0.05 level.

Table 5 Multiple Classification Analysis of the Effects
of Partisanship on Mass Reactions to Democratization

Democratization
Domains Japan Korea Mongolia

Regime .12 * .18 * .07 *
 performance
Institutional .16 * .10 * .24 *
 trust
Normative .02 .07 * .13 *
 support
Practical .08 * .11 * .20 *
 support
Opposition to .04 .07 * .17 *
 authoritarianism

Democratization
Domains Philippines Taiwan Thailand

Regime .08 * .13 * .07 *
 performance
Institutional .06 .13 * .11 *
 trust
Normative .04 .08 * .04
 support
Practical .04 .10 * .04
 support
Opposition to .09 * .06 .09 *
 authoritarianism

Source: East Asia Barometer Surveys, 2001-2003.

Notes: Entries are beta coefficients for the relationships
between partisanship and five domains of democratic change
after holding five other factors constant.

* indicates statistically significant at the 0.05 level.