[22nd General Election Research Series] The Choice of Conflicting Voters in an Era of Polarization: Government Scrutiny vs. Government Stability

1. Introduction

Parliamentary elections held in the middle of a presidential term typically use the evaluation of the incumbent party's governance as a key criterion for judgment. The results of the 22nd general election clearly demonstrated this pattern. President Yoon Suk-yeol's approval rating started at 54% upon taking office, but after hitting a low of 28% three months into his term on August 28, 2022, it remained in the mid-to-high 30% range until the 22nd general election.[1]This situation, where only one-third of respondents supported the president's governance, meant that the general election was inherently unfavorable to the ruling party from the outset.

However, another interesting fact is that the public opinion distribution regarding government scrutiny and government stability, which were the main arguments put forth by the ruling and opposition parties before the general election, showed no significant difference. Depending on the polling organization, the evaluation of these two perspectives was slightly in favor of government scrutiny or showed a balanced equilibrium just before the general election.[2]This situation indicates that voters' perceptions are divided as to whether the responsibility for the current state of governance lies with the president and the executive branch or with the opposition party, which holds a majority in the National Assembly.

This seemingly contradictory situation can be explained by the fact that negative evaluations of governance, influenced by partisan polarization, affected voters' perceptions differently. Although the evaluation of the ruling party's governance was negative, in the reality of a divided government, voters' perceptions of responsibility varied depending on their party affiliation. This is a classic explanation in political psychology, suggesting that voters perceive reality not objectively but from the perspective of their preferred party (Campbellet al. 1060; Converse 1976; Lewis-Beck et al. 2008).

This paper focuses on classifying voters based on their attitudes toward government scrutiny and government stability in the current election and examining how these attitudes differ and how they influence voting behavior. In a situation where the impact of partisan polarization has increased, voters are likely to agree differently with government scrutiny and government stability depending on their party preference. However, the reality of a divided government, with power split between the executive and legislative branches, creates the possibility of voters who oppose or approve of both perspectives, i.e., voters with ambivalent attitudes. If such voters are substantial in number, the election outcome may have been determined by their choices. Above all, in the context of fierce electoral competition caused by extreme partisan polarization, the choices of partisan voters are unlikely to show significant changes. In this context, it is necessary to closely examine the characteristics and political behavior of ambivalent voters. Who are they, and what voting behavior did they exhibit in this election?

2. Voters' Attitudes Toward Political Objects: Unidimensional vs. Bidimensional Perception

Traditional election research has identified preferences for political objects as a significant factor determining individual political behavior. This assumption, that preferences for political objects determine political behavior, stems from the 'traditional discussion of attitudes (e.g., cognitive consistency theory, Festinger 1957)' which views human attitudes unidimensionally/dichotomously. From this perspective, an individual's attitude toward an object is composed of favorable or unfavorable elements, and these two elements tend to be 'reciprocal.' That is, like or dislike for a political object is unidimensionally linked, and an increase in favorable attitudes implies a decrease in unfavorable attitudes, and vice versa. Ultimately, voters' attitudes are determined by the relative difference in like or dislike for political objects measured unidimensionally/dichotomously, and the resulting relative preference or relative dislike for political objects has been understood as a key determinant of political behavior.

New findings in cognitive science and political psychology directly question the reciprocal interpretation of voter attitudes. For example, Cacioppo et al. (1997) point out that 'reciprocity' is merely one of the possible combinations of the two elements of attitude, favorable and unfavorable. They propose a bidimensional/bivariate perspective on individual attitudes, arguing that the two elements of attitude can sometimes be separated as distinct entities, increasing or decreasing simultaneously, and coexisting.et al. 1997) posits that 'offsetting' is merely one combination through which the two components of an attitude, favorable/unfavorable, can be expressed. They argue that the two components of an attitude can sometimes increase or decrease independently, as if separate entities, and can coexist, thereby presenting a two-dimensional/ambivalent perspective on individual attitudes.

These studies suggest that actual voter attitudes may be far more complex than traditional unidimensional interpretations. For instance, unlike the traditional understanding that a voter's preference for a party or candidate is directly linked to their dislike of competing parties or candidates, voters' like or dislike for political objects can sometimes exist independently and coexist. This state where opposing elements coexist without canceling each other out has been conceptualized as 'ambivalent attitude,' and numerous studies have confirmed its diverse effects on individual behavior (Jonas et al. 1997; Maio et al. 1996). Ambivalent attitudes are frequently observed even among voters with strong emotional identification with a party, and similar to 'cross-pressures' and 'attitude conflict' suggested in classical election research theories, they cause psychological distress in decision-making, directly impacting individual political behavior (Basinger and Lavine 2005; Cacioppo et al. 1997; Hochschild 1981, 1993; Lavine 2001; McGraw et al. 2003; Yoo 2010).

This perspective on individual political attitudes can be directly applied to the contrast between government scrutiny and government stability. That is, a preference for government scrutiny does not immediately imply opposition to government stability, and voters can sometimes hold both positive and negative feelings toward both. Particularly, an electoral environment like the current general election, where presidential and legislative powers were divided between different parties, increases the likelihood that voters will hold negative attitudes toward both simultaneously. Applying the bidimensional/bivariate discussion of attitudes, voters can be classified as shown in <Figure 1> based on their attitudes toward government scrutiny and stability.

<Figure 1> Classification of Voters by Attitudes Toward Government Scrutiny/Stability

In <Figure 1>, voters who agree with government scrutiny and oppose government stability, or vice versa, have one-sided attitudes. In contrast, voters who agree with or oppose both government scrutiny and government stability have ambivalent attitudes. In traditional attitude research, favorable and unfavorable elements toward a political object are reciprocal, so most voters have one-sided attitudes, with only a very small minority exhibiting neutral attitudes where the two elements completely cancel each other out.[3]However, in the bidimensional attitude discussion, there are voters with ambivalent attitudes where favorable and unfavorable elements coexist without canceling each other out, and these ambivalent attitudes directly influence their political behavior.

So, what was the situation in this election? Voter perception surveys conducted immediately after the election show that the bidimensional discussion of attitudes is more consistent with the actual situation in this election.[4]<Table 1> presents the classification of respondents based on their attitudes toward both government scrutiny and stability perspectives.[5]The most significant characteristic of the distribution is that the majority of all respondents hold one-sided attitudes. Respondents with one-sided attitudes, who agree with one perspective and oppose the other, numbered 979, accounting for 64% of all respondents. Those who agree or disagree with one perspective but are neutral toward the other (quasi-unilateral) numbered 227 in total, accounting for approximately 15% of all respondents. According to traditional attitude research, favorable or unfavorable elements are reciprocal, so the respondents classified as quasi-unilateral in <Table 1> would ultimately be considered unilateral respondents. If these two groups are combined, nearly 80% of all respondents can be classified as voters with a unilateral attitude, preferring either government scrutiny or government stability. This result is understandable in the current electoral environment of extreme partisan polarization in South Korea.

However, another important finding from <Table 1> is that about 15% of all respondents hold ambivalent attitudes. Respondents who agree with or oppose both government scrutiny and stability numbered 235 in total, accounting for 15.4% of all respondents, with the majority of these being respondents who oppose both perspectives (154 people). This pattern demonstrates that a significant proportion of our voters, consistent with the bidimensional discussion of attitudes, hold ambivalent attitudes where conflicting emotions coexist. Given the intense electoral competition between the Democratic Party and the People Power Party amidst extreme partisan polarization, the existence of these voters and their proportion of the total electorate is a figure that cannot be ignored when considering the election results. <Figure 2> visually presents the table results as percentages for easier comparison.[6]

<Table 1> Classification of Respondents by Attitudes Toward Government Scrutiny/Stability

<Figure 2> Distribution of Respondents by Attitudes Toward Government Scrutiny/Stability

3. Who are the Ambivalent Voters?

As described above, despite the overwhelmingly negative evaluation of the president's governance immediately before the election, voters' attitudes toward government scrutiny and government stability were in a close balance. This was interpreted as partisan, differentiated perceptions of reality operating within partisan polarization. If this is the case, then partisan voters should generally hold unilateral attitudes toward government scrutiny and stability. To confirm this, we examined how respondents' attitudes toward government scrutiny/stability varied by partisanship, and the results are presented in <Table 2>.

<Table 2> Attitudes Toward Government Scrutiny/Stability by Partisanship

The results in the table show that partisanship and unilateral attitudes are indeed linked. Approximately 70% of respondents supporting the Democratic Party and the People Power Party showed attitudes leaning toward either government scrutiny or stability. However, even here, the proportion of respondents with ambivalent attitudes reached 10%, indicating that partisanship does not explain the entirety of voter attitudes, even within partisan polarization. Supporters of the Cho Kuk Innovation Party showed a similar pattern to supporters of the two major parties. In contrast, about 30% of those supporting other parties and those with no party affiliation held ambivalent attitudes. Ultimately, while ambivalent voters were found in higher proportions among third-party supporters and independents, they were also present among supporters of the two major parties. The fact that ambivalent attitudes appear regardless of partisanship, even amidst extreme partisan polarization, demonstrates that ambivalent attitudes are a characteristic that can be found in all voters, independent of partisanship. However, the higher proportion of ambivalent attitudes among third-party supporters and independents (excluding the Cho Kuk Innovation Party) suggests that they hold more negative attitudes toward the two major parties.

So, who holds ambivalent attitudes toward government scrutiny and stability? <Table 3> summarizes the socio-demographic characteristics of the four groups classified by their attitudes toward government scrutiny and stability. Comparing the ambivalent voter group and the unilateral voter group, the former has a higher proportion of males, is relatively younger, and has a higher proportion of residents outside the Seoul metropolitan area. However, excluding age, the differences between the two groups are not significant enough to be considered striking.[7]

<Table 3> Socio-demographic Characteristics of Ambivalent Voters

What about their political characteristics? <Table 4> compares respondents with ambivalent attitudes and those with unilateral attitudes in terms of party support, ideological orientation, and political knowledge.[8] In terms of party support, the two groups show distinct differences. While the proportion of supporters of the Democratic Party and the People Power Party among unilateral voters was nearly 80%, among ambivalent voters, support for the two major parties was less than 50%, with over half comprising third-party supporters (23.7%) and independents (28.6%). These characteristics of ambivalent voters also showed significant differences compared to the overall respondents.

<Table 4> Political Characteristics of Ambivalent Voters

The two attitude groups also showed statistically significant differences in ideological orientation and political knowledge. On average, respondents with ambivalent attitudes had a slightly more conservative ideological orientation than those with unilateral attitudes, and this was also true when compared to the overall respondents (F=3.941, p<.01). Considering that this election was held in the middle of the presidential term and the framework of evaluating the administration was at play, this result suggests that respondents with conservative leanings had difficulty forming clear positions on government scrutiny and government stability. Meanwhile, in terms of political knowledge, the group with the highest level of political knowledge was unilateral respondents. Ambivalent voters showed lower scores than the average for all respondents, indicating that they are a group with relatively low political knowledge (F=38.394, p < .01).[9]

How do voters with ambivalent attitudes toward the theories of regime alternation and regime stability differ in their evaluations of political actors? Table 5 compares the two groups in terms of their evaluations and favorability toward political targets.[10]First, as can be seen from the average scores of all respondents, the evaluations of our voters regarding the President's state administration and their local district National Assembly members' activities in the Assembly were generally at a negative level, and their favorability toward political parties and politicians was also not very positive. The average score for the President's state administration among all respondents was slightly above 3 out of 10 points, and the evaluation of the National Assembly members' legislative activities was 4.67 out of 10 points, which was higher than that for the President but still within the range of negative evaluations. Regarding favorability toward political parties, the Democratic Party of Korea scored an average of 43.24 points and the People Power Party scored 35.37 points among all respondents, indicating generally negative sentiment. Favorability toward President Yoon Suk-yeol and Representative Lee Jae-myung averaged 27.50 and 33.27 points, respectively, remaining even more negative.

The respondents with ambivalent attitudes, who are the focus of this paper, exhibited the most negative attitudes in their evaluations and perceptions of political targets. This clearly shows that ambivalent voters hold the most negative stance toward political actors and parties. In contrast, respondents with unilateral attitudes were more negative than the overall average in their evaluation of the President's state administration, but showed more positive perceptions than the overall average in their evaluation of National Assembly members and their favorability toward parties and politicians. This can be understood as a result of the ambivalent voters viewing the theories of regime alternation and stability from the perspective of their supporting party, leading to a cancellation of favorability and unfavorability. However, the fact that evaluations of political actors remained at a negative level even among respondents with unilateral attitudes is significant.[11]

<Table 5> Political Attitudes of Ambivalent Voters

4. Voting Choices of Ambivalent Voters

What, then, are the voting behaviors of voters with ambivalent attitudes? Based on the previous analysis, these are the voters who hold the most negative evaluations and favorability toward political targets. If their negative perceptions lead these ambivalent voters to abstain from voting, their role in the upcoming general election will inevitably be limited.

The results summarized in Table 6 show that, despite their negative perceptions of political targets, the vast majority of them participated in voting and made political choices. According to the perception survey, their abstention rate exceeded 20%. Of course, this is three times the abstention rate of respondents with unilateral attitudes and is also higher than the overall average for all respondents.[12] However, the voter turnout among voters with ambivalent attitudes was close to 80%, meaning that the vast majority of them voted despite the ambivalence in their perceptions.

As the results regarding the timing of voting decisions show, it is true that ambivalence in perceptions caused psychological distress in their voting decisions. While about half of the respondents with unilateral attitudes had already decided on a candidate or party one month before the election, less than 30% of respondents with ambivalent attitudes had done so, and nearly one in four decided on their choice on election day. This difference is also pronounced when compared to all respondents.[13] The fact that respondents with ambivalent attitudes, compared to those with unilateral attitudes and all respondents, chose their candidates and parties closer to the election indicates that they deliberated on whether to vote until the last moment. Nevertheless, the fact that nearly 80% of ambivalent voters participated in the election suggests that their decisions had a significant impact on the election outcome.

<Table 6> Voting Behavior of Ambivalent Voters

What, then, were the choices of ambivalent voters who participated in the election? The results in Table 7 show how ambivalent voters and unilateral voters made their choices in district and party-list elections. In district elections, ambivalent voters relatively more frequently chose candidates from third parties than from the two major parties. Their rate of choosing third-party candidates reached 30%, which is twice the rate of all respondents and more than three times the rate of voters with unilateral attitudes. In party-list voting for proportional representation, ambivalent voters also showed a different pattern compared to other groups. Unlike unilateral voters, whose choices were divided between the Democratic Party of Korea and the People Power Party, ambivalent voters concentrated their choices on third parties, and among third parties, they chose parties other than the Justice Party, which gained significant momentum in this general election. The results in Table 4, which summarized party support, indicated that more than half of ambivalent voters consisted of supporters of third parties and unaffiliated voters, which can be seen as leading to the high proportion of third-party choices. It is also noteworthy that while nearly 30% of unilateral voters chose the Justice Party, a high proportion of ambivalent voters chose third parties other than the Justice Party.

<Table 7> Voting Choices of Ambivalent Voters

The results in Figure 3 show that their choices were not spontaneous decisions made at the polling station. The figure summarizes the results of whether voter groups, divided by their attitudes toward regime alternation and regime stability, made different choices in district and party-list elections.[14] If the choices of ambivalent voters differed between district and party-list elections, meaning they engaged in split voting, it would imply that they took a stance on one of these despite their ambivalent attitudes and made a strategic choice. However, the majority of voters with ambivalent attitudes showed consistent choices in both district and party-list elections, and the proportion was not significantly different from that of unilateral voters. More specifically, 68.6% of respondents with ambivalent attitudes had consistent choices in district and party-list voting, a figure that exceeds that of voters with unilateral attitudes (64.7%). This fact indicates that their negative perceptions of the two major parties and political actors led ambivalent voters to make consistent choices in both district and party-list elections.

The results above indicate that ambivalent voters, who did not choose either the theory of regime stability or the theory of regime alternation, which were the main frameworks of this general election, held negative perceptions of the established political circles and, despite the difficulties in making choices due to these perceptions, participated in the vote and actively expressed their dissatisfaction with the two major parties.

<Figure 3. Attitudes toward Regime Alternation/Stability and Split Voting>

Then, can all ambivalent voters be considered as those who voted based on their dissatisfaction with the established political circles? In this regard, it is also necessary to consider that ambivalent voters consist of two different groups: those who agree with both perspectives on the general election and those who oppose both.[15] Logically, negative perceptions of the established political circles are likely to lead to opposition to both perspectives on the general election, and if so, it is highly probable that this group would have shown different behaviors depending on the type of ambivalence.

<Table 8> Two Types of Ambivalence: Agreement with Both vs. Opposition to Both

The results in Table 8 support this logical inference. While the voting behavior of ambivalent voters who agreed with both frameworks of the general election showed no significant difference from that of unilateral voters and all voters, ambivalent voters who opposed both frameworks showed a higher tendency to choose candidates from third parties in district elections and reached nearly half in choosing other parties in party-list elections. This fact indicates that the perceptions and behaviors of ambivalent voters examined in this paper were driven by those who opposed both perspectives of the election.

5. Statistical Analysis of Voting Choices

The analysis results thus far show that attitudes toward regime alternation/stability are a characteristic found across all voter groups, regardless of their party affiliation, within the context of partisan polarization, and that these voters hold negative perceptions of political actors such as the President and the two major parties. Furthermore, compared to voters with unilateral attitudes, voters with ambivalent attitudes experience difficulties in making voting decisions and some may choose to abstain, but they still participate in voting at a high rate and actively express their negative perceptions of existing political actors by choosing third-party candidates and parties. While these analyses are meaningful in themselves, they have limitations as they are bivariate analyses conducted without controlling for other variables. Therefore, we will set up a statistical model for voting choices and confirm whether the results of the bivariate analysis appear in the multivariate analysis as well.

The statistical model analysis was conducted separately for district and proportional representation elections in this general election. The dependent variable of the model is the voter's choice in the election. For district voting, the choices are divided into four categories: candidates from the Democratic Party, candidates from the People Power Party, candidates from third parties, and abstention. For proportional representation elections, the voter's choice is divided into five categories: the Democratic Alliance, the Future Party, the Justice Party, other parties, and abstention. Given the nature of the dependent variable, a multinomial logit analysis was performed.[16] The independent variables of the model included gender (female), age group, education level, political knowledge level, political interest, presidential approval rating, party affiliation, and attitudes toward regime alternation/stability.[17] As mentioned, the statistical analysis was conducted separately with district voting and party-list voting as dependent variables, and the results are summarized in Tables 9 and 10.

First, let's look at the results of Table 9 regarding district voting. Before interpreting the results, it is important to keep in mind that, due to the nature of multinomial logit analysis, each coefficient represents the statistical probability of choosing a particular category compared to the reference category. The results in the first column indicate the effect of each independent variable on choosing the People Power Party versus the Democratic Party. According to these results, a more positive evaluation of the President's state administration and being a supporter of the People Power Party increased the likelihood of choosing a People Power Party candidate in the district election rather than a Democratic Party candidate. Conversely, supporters of the Democratic Party and supporters of the Justice Party had a statistically significant decrease in the likelihood of choosing a People Power Party candidate.

<Table 9> Attitudes toward Regime Alternation/Stability and Voting Choices (District Voting)

Next, let's examine the results of the attitude variables toward regime alternation and stability, which are the focus of this paper. The analysis results show that all three attitude variables are statistically significant. The fact that these variables showed statistically significant results after controlling for partisanship means that they exert an independent influence. Looking at the content, respondents with neutral or ambivalent attitudes, and those with unilateral attitudes toward regime stability, were more likely to choose People Power Party candidates than Democratic Party candidates, compared to those with unilateral attitudes toward regime alternation.

A more interesting result is observed in the second column, which shows the choice of third-party candidates versus Democratic Party candidates. The statistically significant variables were party support and the three attitude variables. The influence of party support variables can be easily understood as a difference between the reference category and the chosen category. A tendency for respondents with neutral, ambivalent, and unilateral stability attitudes to choose third-party candidates over Democratic Party candidates was confirmed. The influence of attitude variables arises for different reasons: unilateral stability attitudes can be interpreted as reflecting dislike for the Democratic Party, while ambivalent attitudes can be interpreted as expressing dissatisfaction with established political actors. The influence of attitude variables was also statistically confirmed in the last column, which analyzes the choice of abstention versus Democratic Party candidates.

These results, as shown in the previous bivariate analysis, confirm once again in the statistical analysis the tendency of ambivalent voters to choose third-party candidates over major party candidates and to abstain in the district elections.

<Table 10> Attitudes toward Regime Alternation/Stability and Voting Choices (Party-List Voting)

Table 10, which summarizes the results of party-list voting, shows that the results for proportional representation elections are similar to those for district elections, with a few exceptions. First, regarding differences from district election results, the statistical significance of gender was confirmed among sociodemographic variables. Women were more likely to choose the Democratic Alliance, the satellite party of the Democratic Party, than men in all choice categories. In addition, the statistical significance of age groups was confirmed, with older age groups being statistically significantly more likely to choose the Future Party, the Justice Party, and other parties than the Democratic Alliance, indicating antipathy toward the Democratic Party among older voters in party-list voting.

The patterns of variables related to presidential approval ratings, party support, and attitudes toward regime alternation/stability were similar to the results of district elections. A positive evaluation of the President's state administration increased the probability of choosing the Future Party, other parties, and abstention over the Democratic Alliance. In terms of party support, supporters of the Democratic Party had a statistically significant higher probability of supporting the Democratic Alliance in all choice categories, while supporters of the People Power Party were more likely to choose the Future Party than the Democratic Alliance, and supporters of the Justice Party had a higher probability of choosing the Justice Party. This is understood as a natural consequence of partisanship. Attitude variables also showed similar results to the previous district election analysis. The results in the first column show that respondents with neutral, ambivalent, and unilateral stability attitudes were more likely to choose the Future Party than the Democratic Alliance, compared to respondents with unilateral attitudes toward regime alternation. However, this tendency was not observed in the choice of the Justice Party category, but was clearly evident in the choice of the other parties category. In particular, the tendency for ambivalent voters to choose other parties or abstain rather than the Democratic Alliance precisely matches the results of the bivariate analysis.

6. Conclusion

This paper examined the results of the general election by focusing on voters' attitudes toward the two frameworks of regime alternation and regime stability. The influence of partisan polarization created a situation where the two perspectives of the general election clashed intensely, despite negative evaluations of the ruling party and the administration, in this general election, which was held in the middle of the presidential term and thus had a strong character of a mid-term evaluation. Nevertheless, those who could not choose either of the two electoral frameworks accounted for about 15% of all voters, and even among respondents with partisan affiliation, this proportion reached 10%, creating conditions where ambivalent attitudes could exert a meaningful influence amidst partisan polarization. Ambivalent voters showed distinct differences from unilateral voters in their evaluations of political actors and voting behavior, actively expressing their dissatisfaction with mainstream political actors by choosing third-party candidates in district elections and other parties (not the two major parties or the Justice Party) in party-list voting.

As examined in this paper, their ambivalence stemmed from negative perceptions of the established political circles, and the fact that many of them, despite difficulties in making political choices, actually participated in the vote and actively expressed their dissatisfaction by choosing third-party candidates and parties rather than the two major parties is highly significant. Furthermore, the fact that ambivalent voters showed distinct perceptions and behaviors even amidst deepening partisan polarization and intense electoral competition centered around the two major parties has the following important implications.

First, the analysis in this paper shows that ambivalent attitudes toward political actors and electoral frameworks must be considered as variables explaining voters' voting choices. Above all, ambivalent attitudes are significant because they are a factor that can be found in all voters, regardless of their party affiliation, and create behaviorally meaningful differences.

Second, voters with ambivalent attitudes are based on negative perceptions of mainstream parties and political actors and actively seek alternatives in elections. In the case of this general election, they clearly showed a tendency to choose third-party candidates in district elections and other parties in party-list voting. This means that they are voters who actively seek alternatives to resolve their dissatisfaction, and they showed a tendency to choose available alternatives even within South Korea's electoral environment, which is overwhelmingly favorable to major parties. The fact that the Justice Party was not their choice in party-list voting implies that the most important condition for the political choices sought by ambivalent voters is differentiation from established parties. If the electoral system reform in our electoral environment allows for the emergence of diverse parties, it is not difficult to imagine where their choices will lead.

Finally, the political behavior of voters with ambivalent attitudes, as shown in this paper, has normative implications for the development of democracy. Normatively, democracy includes citizens who actively respond to environmental changes and express themselves as an important condition. In this respect, ambivalent voters provide an important driving force for the smooth functioning of democracy by actively responding to the characteristics of political actors and participating actively.

Based on the analysis results of this paper, it is undeniable that the choices of ambivalent voters played a crucial role in the election results in a situation where electoral competition has intensified due to the solidification of the two-party system. However, if the situation where their political dissatisfaction is not resolved continues in the future, it is necessary to observe whether they will continue to participate in elections and express their opinions actively. It is self-evident that if the expressed political dissatisfaction is not resolved and they leave the electoral arena, the dynamism of South Korean elections will be significantly diminished.■

References

Alvarez, Michael and John Brehm. 1997. “Are Americans Ambivalent towards Racial Policies?” American Journal of Political Science 41: 345-74.

Basinger, Scott and Howard Lavine. 2005. “Ambivalence, Information, and Electoral Choice.” American Poltiical Science Review 49: 169-84.

Berelson, Bernard, Paul Lazarsfeld, and William McPhee. 1954. Voting. Chicago: University of Chicago Press.

Cacioppo, John, Wendy Gardener, and Gary Berntson. 1997. “Beyond Bipolar Conceptualizations and Measures: The Case of Attitudes and Evaluative Space.” Personality and Social Psychology Review 1: 3-25.

Campbell, Angus, Philip Converse, Warren Miller, and Donald Stokes. 1960. The American Voters. New York: Wiley.

Converse, Philip. 1976. The Dynamics of Party Support. London: Sage.

Festinger, Leon D. 1957. A Theory of Cognitive Dissonance. Stanford: Stanford University Press.

Hochschild, Jennifer. 1981. What’s Fair? American Beliefs about Distributive Justice. Cambridge: Harvard University Press.

Hochschild, Jennifer. 1993. “Disjunction and Ambivalence in Citizens’ Political Outlooks.” George E. Markus and Russell L. Hanson (eds.) Reconsidering the Democratic Public. PA: Pennsylvania State University Press.

Jonas, Klaus, Michael Diehl, and Philip Bromer. 1997. “Effects of Attitudinal Ambivalence on Information Processing and Attitude-Intention Consistency.” Journal of Experimental Social Psychology 33: 190-210.

Lavine, Howard. 2001. “The Electoral Consequences of Ambivalence toward Presidential Candidates.” American Journal of Political Science. 45: 915-29.

Lewis-Beck, Michael S., William Jacoby, Helmut Norpoth, and Herbert F. Weisberg. 2008. The American Voter Revisited. Ann Arbor: University of Michigan Press.

Maio, Gregory, David Bell, and Victoria Esses. 1996. “Ambivalence and Persuasion: The Processing of Messages about Immigrant Groups.” Journal of Experimental Social Psychology.  32: 513-36.

Yoo, Sung-jin. 2010. “Two Types of Neutrality: Ambivalence, Indifference, and Political Particiation.” Journal of Politics. 72(1).

■ Author: Yoo Sung-jin, Professor, Ewha Womans University, School of Scrantons.

■ Contact and Editing: Kim Sun-hee, Senior Researcher, EAI.

Inquiries: 02-2277-1683 (ext. 209), shkim@eai.or.kr

[1] These figures are based on the National Barometer Survey (NBS), jointly conducted by Embrain Public, CaseStat Research, Korea Research, and Korea Research (http://nbsurvey.kr/). An evaluation of the president's job performance surveyed by Gallup Korea shows similar figures.

[2] The National Barometer Survey in the first week of April, just before the general election, showed almost no difference between the 'regime control' and 'regime stability' arguments, at 47% and 46% respectively. In contrast, a survey by Gallup Korea in the last week of March showed the 'regime control' argument prevailing with 49% compared to 40% for 'regime stability'.

[3] This is even more likely given the current severe partisan polarization.

[4] The analysis in this article utilized a post-election voter perception survey conducted by Korea Research at the request of the East Asia Institute (EAI). The survey was conducted from April 12 to 16 via web survey among the Korea Research Political and Social Panel (49,889 members), with a total of 1,528 respondents.

[5] In the survey conducted immediately after the election, respondents were asked to indicate their agreement with two perspectives on the general election: "The Yoon Suk-yeol administration should be empowered (regime stability argument)" and "The opposition party should be empowered (regime control argument)." Each attitude was measured on a 7-point scale (1 point (strongly disagree) to 7 points (strongly agree)), and respondents were divided into four groups based on their answers. "One-sided" voters were those who agreed (5-7 points) with the regime stability argument while disagreeing (1-3 points) with the regime control argument, or vice versa. "Conflicting" voters were those who agreed (5-7 points) or disagreed (1-3 points) with both positions. Those who were neutral (4 points) on one position and agreed or disagreed with the other were classified as "quasi-one-sided" voters, and those who were neutral (4 points) on both positions were classified as "neutral" voters.

[6] Respondents with a neutral attitude, who held a neutral position (4 points) on both perspectives, numbered 87 in total, accounting for 5.4% of the total.

[7] Differences between groups were not statistically significant, except for age (Age: F=38.659 (p < .01), Gender: χ2= 3.469 (p = .325), Region of Residence: χ2= 15.790 (p = .201)).

[8] Ideological orientation was measured on an 11-point scale from "0 (very progressive) to 10 (very conservative)," and political knowledge was measured by the number of correct answers to four items: the size of the national budget in 2024, the number of local constituency members of parliament, the voting age, and the name of the Prime Minister.

[9] This pattern is similarly observed in interest in politics. The proportion of respondents who answered that they were not very or not at all interested in politics was about 25% among those with conflicting attitudes, compared to about 12% among those with one-sided attitudes.

[10] The president's job performance was measured on an 11-point scale from "0 (very poorly) to 10 (very well)," and the evaluation of members of the National Assembly was measured on an 11-point scale regarding the past four years of parliamentary activities of their local constituency members, from "0 (very poorly) to 10 (very well)." Favorability towards political parties and politicians was measured using a thermometer scale from "0 (very unfavorable) to 100 (very favorable)."

[11] Differences between groups in political attitudes were statistically significant for all measures except for the evaluation of National Assembly members and favorability towards the People Power Party. The statistical test results for the differences between groups for each attitude are as follows: Presidential approval: F=2.916 (p < .05), National Assembly member evaluation: F=.684 (p = .562), Democratic Party favorability: F=8.538 (p < .01), People Power Party favorability: F=2.248 (p = .081), Yoon Suk-yeol favorability: F=4.906 (p < .01), Lee Jae-myung favorability: F=20.078 (p < .01).

[12] Differences between groups in terms of voting behavior are statistically significant (χ2= 63.299 (p < .01)).

[13]Statistical verification of the differences between groups regarding the timing of decisions for district candidates and proportional representation parties showed χ2= 144.015 (p < .01) and χ2= 98.851 (p < .01), respectively.

[14]Consistent voting was defined as choosing the same party for the district candidate and the proportional representation party. Split voting was defined otherwise. In this classification, the satellite parties of the Democratic Party of Korea, the Democratic United Party, and the People Power Party, People's Future, were considered the same choice. For proportional representation, the analysis was conducted such that voters who chose other parties in the district election and then chose the Cho Kuk Innovation Party in the proportional representation election were classified as split voters, distinguishing the Cho Kuk Innovation Party from other parties.

[15]The 235 conflicted voters were composed of 154 who opposed both electoral structures and 81 who supported both. In other words, voters with a negative conflicted attitude accounted for 66% of conflicted voters.

[16]The results of the multinomial logistic analysis statistically indicate the likelihood of choosing each option in comparison to the reference category. The reference category here is the candidate of the Democratic Party of Korea for district voting and the Democratic United Party for party voting.

[17]The measurement methods for each independent variable are as follows. Gender was measured as a dummy variable with males as the reference category. Age was measured on a 6-point scale (18-29, 30-39, 40-49, 50-59, 60-69, 70 and older). Political knowledge level was measured by the number of correct answers (0-4) to four political knowledge questions. Political interest was measured on a 5-point scale (1=not interested at all, 2=not very interested, 3=somewhat interested, 4=fairly interested, 5=very interested). Presidential approval rating was measured on an 11-point scale (0=very poorly, 10=very well). Party affiliation was measured using four dummy variables with independents as the reference category. Finally, attitudes toward government check/stability were categorized according to a dichotomy of neutral attitude, conflicted attitude, and unilateral attitude. Considering the nature of the dependent variable, attitudes of quasi-unilateral support for checks and balances were included in the unilateral attitude category. Consequently, this variable is divided into three dummy variables with a unilateral attitude toward government checks and balances as the reference category.