[EAI 21st Presidential Election Poll Analysis] ③ Reconstruction of Division: The Ideological Landscape of Conservatism-Progressivism Drawn by the Voting Base of Four Candidates

YouTube Link: https://www.youtube.com/watch?v=UGH_VyCIj8g

Video Script

Ideological Landscape of Korean Voters and Analysis of Voting Blocs by Candidate

According to a large-scale survey conducted in 2024, the most significant conflict felt by citizens in Korea is the ideological confrontation between progressives and conservatives. We will examine the extent to which voters for each candidate in the 2025 presidential election are ideologically divided. The most common way to measure ideology is to ask individuals to subjectively place themselves on a scale from 0 to 10, with 0 representing the most conservative and 10 representing the most progressive. The following visualizes the distribution of responses to this question, segmented by the candidate each individual voted for. The bloc voting for Lee Jae-myung is in the upper left, the red bloc represents voters for Kim Moon-soo, the orange bloc represents voters for Lee Jun-seok, and the last bloc represents voters for Kwon Young-se. We can see that the voters for Lee Jae-myung are concentrated to the left of 5. Conversely, voters for Kim Moon-soo are concentrated to the right of 5.

While voters for Lee Jun-seok are considerably concentrated to the right of 5, there is also a notable presence to the left. Voters for Kwon Young-se are significantly concentrated on the left. Based on this, we can conclude that there is considerable ideological division among progressive and conservative voters depending on their candidate choice. Let us examine this in more detail. The following presents a box plot comparison of Lee Jun-seok's voters by bloc. A striking observation is the bloc of Lee Jun-seok's voters. They are positioned between the blocs of Lee Jae-myung and Kim Moon-soo, or closer to the center than Kim Moon-soo's voters. Looking at the median, the median for Lee Jun-seok's voters across all age groups is 5. This indicates that 50% of these voters are located around this point.

We can also differentiate this by age group. When segmented by age group, the box plot on the left shows that the 39 and under demographic, i.e., young voters, exhibit a slightly more conservative stance internally, with a median value around 6, compared to the overall age group. Comparing the overall age group and the 39 and under demographic for Lee Jun-seok's voters using histograms, the overall age group is represented by light gray and the 39 and under group by orange. We can ascertain that the 39 and under demographic is internally more conservative. This time, instead of a simple progressive-conservative scale from 0 to 10, we will measure their ideology using policy attitude questions. These policy attitude questions address short-term issues, such as

Measurement of Socio-Cultural and Economic Security Policy Attitudes

Unlike short-term and specific policy issues that are important to our society, such as martial law, pension reform, or soaring real estate prices, these questions were designed to reflect longer-term and more enduring value orientations. First, the socio-cultural policy attitude questions included items on the rights of social minorities, such as homosexuals, foreign workers, and the disabled. Second, the questions addressed attitudes towards responding to the climate crisis. Third, questions on attitudes towards gender discrimination were included as socio-cultural policy attitude items.

The economic security policy questions included items related to security against North Korea, items constituting market principles concerning market and corporate competitiveness, and finally, items constituting meritocracy. For the first, third, and last items—opposition to discrimination, climate, and gender equality—we reverse-coded them and calculated the proportion of respondents who answered 4 or 5, corresponding to conservative positions. The numbers for the third, fourth, and fifth items are mixed up; please refer to the main text for the exact numbers.

Comparative Analysis of Ideological Characteristics of Candidate Support Blocs

Examining the analysis results, the supporters of Kim Moon-soo and Lee Jun-seok exhibit common strong conservative tendencies in the economic security domain, including security, economy, and meritocracy. Meanwhile, these two supporter blocs reveal distinct differences in socio-cultural issues. Lee Jun-seok's supporters tend to be more conservative in socio-cultural aspects than Kim Moon-soo's supporters. In other words, while they are closer to the center in the economic security domain compared to Kim Moon-soo, they are closer to the conservative end in the socio-cultural domain.

Conversely, supporters of Lee Jae-myung and Kwon Young-se showed clear support for progressive values in socio-cultural issues and also supported progressive values in the economic security domain. Using these data, we categorized the six items into socio-cultural and economic security dimensions using a four-dimensional item response theory and measured the positions of respondents in each dimension. The respondents' positions were standardized to follow a normal distribution centered at 0.

Therefore, a position close to 0 does not necessarily mean absolutely centrist; rather, it signifies being relatively centrist when compared to the entire population. Based on these analytical results, voters were placed in a two-dimensional policy attitude ideological space. In the socio-cultural dimension, positions higher on the x-axis are more conservative, and positions further to the right on the y-axis are more conservative in the economic security domain. The black dots on the left represent the distribution of all respondents.

Ideological Differentiation and Characteristics of Lee Jun-seok and Kwon Young-se Support Blocs

We then segmented the voters of all age groups by gender. Blue represents males, and red represents females. There is considerable overlap. To reduce this overlap, we compare the age group to 39 and under. The overlap is significantly reduced, and the distribution widens vertically. For the 29 and under demographic, the top 25% density areas are considerably spread out vertically with almost no overlap. This time, we indicate the top 25% density by candidate support bloc. Looking at the left graph for individuals who voted for Lee Jae-myung and Kim Moon-soo, there is substantial overlap. Simultaneously, they are spread out horizontally. This means that the economic security domain, which has traditionally been considered a key component of ideology in Korea, is an important ideological component for them. In contrast, comparing Lee Jun-seok's and Kwon Young-se's voter blocs, Lee Jun-seok's voters are widely distributed vertically. This indicates that they are sensitive to socio-cultural issues. Let us narrow this down again to the 39 and under age group. The overlap decreases even when supporting Lee Jae-myung or Kim Moon-soo. We can see the spread at a 45-degree angle. Ah, this is for the 29 and under demographic. This is for the 39 and under demographic, comparing Lee Jun-seok and Kwon Young-se. For these groups, the top 25% density areas hardly overlap. Considering why people voted for Lee Jun-seok instead of Kim Moon-soo, both Kim Moon-soo and Lee Jun-seok can be classified as conservative, but what was the criterion that distinguished them? As discussed earlier, socio-cultural ideology played a certain role for them. We can draw this conclusion for Lee Jun-seok's voters.

Analysis of Key Determinants for Lee Jun-seok's Supporters

The analysis results show that perceptions of emergency martial law are significant. The more negatively one perceives emergency martial law, the higher the probability of voting for Lee Jun-seok. This probability increases for males aged 39 and under. In the economic security dimension, the closer one is to the center, the higher the probability. Conversely, in the socio-cultural dimension, the more conservative one is, the higher the probability. Finally, regarding the political efficacy of men in their 20s (Idaenam), which is frequently mentioned in the media, men across all age groups have higher efficacy than women. We also found that political efficacy is unlikely to be a differentiating factor between Kim Moon-soo and Lee Jun-seok. Other factors, such as education level and economic class, showed little influence.

The analysis results show that perceptions of emergency martial law are significant. The more negatively one perceives emergency martial law, the higher the probability of voting for Lee Jun-seok. This probability increases for males aged 39 and under. In the economic security dimension, the closer one is to the center, the higher the probability. Conversely, in the socio-cultural dimension, the more conservative one is, the higher the probability. Finally, regarding the political efficacy of men in their 20s (Idaenam), which is frequently mentioned in the media, men across all age groups have higher efficacy than women. We also found that political efficacy is unlikely to be a differentiating factor between Kim Moon-soo and Lee Jun-seok. Other factors, such as education level and economic class, showed little influence. This concludes the presentation.

Conversely, in the socio-cultural dimension, the more conservative one is, the higher the probability. Finally, regarding the political efficacy of men in their 20s (Idaenam), which is frequently mentioned in the media, men across all age groups have higher efficacy than women. We also found that political efficacy is unlikely to be a differentiating factor between Kim Moon-soo and Lee Jun-seok. Other factors, such as education level and economic class, showed little influence. This concludes the presentation.

Goo Se-jin, Professor of Political Science and International Relations, Inha University.

Managed and Edited by: Song Chae-rin, EAI Research Fellow

Inquiries: 02 2277 1683 (ext. 211) | crsong@eai.or.kr