Why Did the 180-Seat Giant Ruling Party Face Judgment in Just Two Years: Focusing on the Choices of Defecting Democrats

1. The 180-Seat Giant Ruling Party Faced Judgment in Two Years

In the 21st general election in 2020, the Democratic Party of Korea achieved a landslide victory for the third consecutive time, following the 2017 presidential election and the 2018 local elections, demonstrating that the '20-year ruling theory' was not an unrealistic goal. Until the mid-term of the Moon Jae-in administration, there was a pattern of alternating ruling and opposition parties every ten years after democratization, and the next presidential election was scheduled five years after the Democratic Party of Korea's term. Given that the incumbent administration and ruling party maintained high approval ratings for state affairs and an overwhelming advantage in party support, extending their term in office was perceived as a manageable task.

However, the Democratic Party of Korea, which had secured a landslide victory with 180 seats, far exceeding the threshold to block constitutional amendments, suffered a crushing defeat in the April 7, 2021, by-elections in Seoul and Busan within less than a year. Less than two years later, on March 9, 2022, they had to hand over power in the presidential election. Conversely, the People Power Party, whose affiliated president was impeached due to public resistance and which had to face criticism for judging the opposition party instead of the usual 'judgment of the administration' during the Moon Jae-in administration's term, unexpectedly regained power after only five years.

The 180-seat giant ruling party was reduced to an object of public judgment in the short span of two years. What enabled the People Power Party, which had been criticized for 'obstruction,' 'ideological excess,' and 'politically abusive language' without crossing the 'river of impeachment,' to re-emerge as the ruling party? This article argues that the emergence of "swing Democrats," who supported the Democratic Party of Korea in the last general election but withdrew their support after the April 7 by-elections, dismantled the voter coalition favoring the Democratic Party of Korea and led to the victory of candidate Yoon Suk-yeol, based on an analysis of data from the EAI Election Panel Survey (KEPS 2022) conducted twice before and after the election.

First, a comparison between swing Democrats and remaining Democrats confirms that defections were particularly high among younger generations in their 20s and 30s, in the Gyeongin region, and among supporters in the Daegu/Gyeongbuk region. Generally, Democratic Party supporters with moderate/conservative tendencies defected after the general election. This means that approximately three out of ten supporters who voted for the Democratic Party of Korea two years ago withdrew their support. The analysis empirically examined the key factors that led them to withdraw their support for the Democratic Party and switch to supporting candidate Yoon in the election. The real estate/Daejang-dong issue and allegations of corruption involving the candidate's spouse appear to have influenced the choices of swing Democrats. For a rigorous analysis, the results of a logistic regression analysis verifying the influence of factors differentiating remaining Democrats from swing Democrats are presented.

The logistic regression analysis results indicate that demographic factors influenced the fissures among Democratic Party supporters. Female Democratic Party supporters defected more than male supporters, but the gender gap varied significantly by age group. Regionally, defecting tendencies were stronger among Democratic Party supporters residing in the Gyeonggi/Incheon and Daegu/Gyeongbuk regions compared to the Honam region. Most importantly, the primary reasons for the defection of swing Democrats were the "strengthening of the administration judgment narrative" and "antipathy towards candidate Lee Jae-myung," with the influence of an ideological stance prioritizing growth over welfare also confirmed.

2. The Emergence of Swing Democrats, the Main Force in Dismantling the Impeachment Political Coalition

One of the most effective methods for analyzing rapid shifts in voters' political support is to track the factors behind these shifts by focusing on the groups experiencing changes in party affiliation, which are key determinants of political attitudes. By comparing the "remaining voters" who maintained their support for a party before and after a significant shift in voter party support with the "swing voters" who switched their support, one can easily understand the reasons for the dissolution and realignment of existing voter-party coalitions. Vote defection can be further classified into "demobilization" (defecting to become unaffiliated or undecided) and "conversion" (defecting to support a competing party) (Hawley and Sagarzazu 2012, Norpoth and Rusk. 2007).

From the 2017 presidential election to the 2020 general election, the overwhelming political landscape favoring the Democratic Party of Korea, which emerged from the candlelight protests and the impeachment of former President Park Geun-hye, was based on a new voter political coalition combining the Honam + non-Daegu/Gyeongbuk regions, the 20s-50s age groups excluding those over 60, and progressives + centrists excluding conservatives (for convenience, the author refers to this as the 'Impeachment Voter Political Coalition'). The key factor in the dismantling of the previous conservative party-dominated structure, which relied on a large bloc of conservative voters, and the emergence of the Democratic Party of Korea-led 'Impeachment Political Coalition' was the division within the (former) Saenuri Party's support base, which accounted for 40-50%. This period was characterized by a comparison between the "remaining conservatives" who maintained their support for the Saenuri Party and its successors (Liberty Korea Party, United Future Party) even during the impeachment process, and the "defected conservatives" who withdrew their support from conservative parties, explaining the birth and consolidation of the Impeachment Political Coalition.[1]

Figure 1-(1) shows that the Saenuri Party's (conservative party) approval rating, which held an overwhelming advantage until the 20th general election in 2016 when the Impeachment Political Coalition was formed, lagged behind the Democratic Party of Korea by 20% to 40%p during the period of impeachment and the Moon Jae-in administration's local and general elections from 2017 to 2020. However, Figure 1-(2), which shows the changes in party support after the 2020 general election, indicates that the Democratic Party of Korea's approval rating, which fluctuated between 40% and 50%, dropped to the 30% range. Conversely, the approval rating for the Liberty Korea Party and the United Future Party, predecessors of the People Power Party, which had remained at 10-25%, recovered to over 30% and approached 40% around the April 7, 2021, by-elections. The realignment of the voter landscape, which had favored the Democratic Party of Korea, was completed during the 2021-2022 20th presidential election competition, resulting in a balanced two-party system.

Ultimately, the dismantling of the Impeachment Political Coalition can be said to have begun with the division of the Democratic Party's supporters, who had exceeded 40% and gave the Democratic Party 180 seats, into "Remaining Democrats" who continued to support the Democratic Party during the by-elections and presidential election, and "Swing Democrats." As the Impeachment Political Coalition moved from formation to dissolution, the focus of analysis shifted from "swing conservatives" to "swing Democrats." Understanding these swing Democrats is expected to provide clues to understanding the factors behind the victory of candidate Yoon Suk-yeol, who won the presidential election by a narrow margin.

[Figure 1] Changes in Party Support Rates (%) During the Formation and Dissolution Periods of the Impeachment Political Coalition

(1) Impeachment Political Coalition Formation Period (2016.2-2020.4) (2) Dissolution Period After General Election (2020.7-2022.3)

3. Swing Democrats' Characteristics and Voting Choices as Seen Through the EAI Presidential Election Panel Survey

➀ Three out of ten proportional vote supporters for the Democratic Party in the general election withdrew their support.

Table 1 presents the cross-tabulation of proportional vote results in the 21st general election and the current party support from the first wave of the presidential election panel survey (January 12-15) to classify remaining Democrats and swing Democrats based on the results of this panel survey. Among respondents who supported the Democratic Party of Korea's proportional representation party in the general election, 39% of the analysis subjects (901 people) supported it. Respondents who voted for the United Future Party at the time accounted for 26% of all analysis subjects, totaling 234 people. Among those who voted for the Democratic Party of Korea's proportional representation party (350 people), 71% (249 people) were classified as "remaining Democrats" who still supported the Democratic Party as of January. The remaining 29% (101 people) were classified as "swing Democrats" who withdrew their support from the Democratic Party (15% defected to become unaffiliated, and the remaining 14% converted their support - 8% to the People Power Party, 4% to the Justice Party, 2% to the People's Party, and 1% to other parties).[2]

In contrast, 82% (193 people) of those who voted for the United Future Party's proportional representation party (234 people) maintained their support for its successor, the People Power Party, showing a higher vote retention rate than the ruling party. Furthermore, among the 317 People Power Party supporters at the time of the first survey (35% of all analysis subjects), 124 people (39%) were classified as "new People Power Party supporters" who newly joined after the general election, excluding the 193 people (61%) who maintained their support. This means that approximately four out of ten current People Power Party supporters are new conservatives who joined after the general election. On the other hand, for the Democratic Party, 75% (249 people) of the 333 supporters at the time of the first survey were consistent supporters who had maintained their support since the general election's proportional vote, with only 25% (84 people) being new Democratic Party supporters who joined after the general election. As swing Democrats outnumbered defected conservatives, and the People Power Party had an advantage in attracting new supporters, the ruling party's advantage in party support ratings disappeared, and a two-party competitive structure was formed.

[Table 1] Cross-tabulation of Proportional Vote Party in the 21st General Election and Party Support at the Time of the First Survey

f2e34dd1a320d063

f2e34dd1a320d063

f2e34dd1a320d063

➁ Who are the Defected Democrats and Remaining Democrats? - The 20s/Gyeongin/Centrist-Conservative Bloc Led the Defection

If so, from which group of supporters of the ruling party, which pushed for 180 seats during the general election period, did the defection phenomenon concentrate? We classified 350 supporters of the Citizen Party of the Blue Sky (Minjoo Party of Korea's predecessor) during the 2020 general election by generation, region, and ideological tendency. Table 2 allows us to see which generation experienced concentrated defection and retention by comparing the generational, regional, and ideological composition of the 'Defected Democrats' and 'Remaining Democrats' in the first survey period with the generational, regional, and ideological composition of the Citizen Party of the Blue Sky supporters in 2020.

[Table 2] Composition of Defected Democrats and Remaining Democrats by Generation, Region, and Ideological Tendency

f2e34dd1a320d063

f2e34dd1a320d063

f2e34dd1a320d063

Looking at generational breakdown, among those who voted for the Democratic Party in the general election, only 13% in their 20s and 17% in their 30s voted, while those in their 40s and above were around 20-25% each. This indicates that the 20s and 30s demographic has consistently constituted a smaller proportion compared to other age groups since the 20th general election.[3]However, when examining the generational composition of those who defected from the Democratic Party and those who remained loyal, the proportion of those in their 20s and 30s significantly increased in the defecting group compared to 2020 (24% each). Conversely, in the loyal group, the proportion of those in their 20s and 30s did not reach the 2020 level, while the proportion of those in their 40s and above, and 60s and above, surpassed the 2020 supporter composition. This highlights that the Democratic Party's support base among those in their 40s and above showed strong loyalty, while there was a significant outflow from the 20s and 30s demographic.

Regionally, compared to the supporter composition in 2020, the defecting Democratic group showed a higher proportion than average in the Gyeongin and Daegu-Gyeongbuk regions. For the loyal Democratic group, the proportion of residents in the Honam region exceeded the 2020 proportion. The composition in Seoul/Chungcheong/Busan-Ulsan-Gyeongnam showed no significant difference compared to 2020. In terms of ideological orientation, the defecting Democratic group saw a significant increase in the proportion of centrist and conservative voters compared to 2020. Conversely, the loyal Democratic group was predominantly progressive, with 52% being progressive, an increase in proportion compared to 2020. This implies that while the Democratic Party garnered a majority of centrist voters and some conservative voters in the 2020 general election, a significant portion of centrist and conservative voters defected in the presidential election, leading to a realignment of the loyal Democratic base centered around progressive voters. In summary, the results indicate a serious defection from supporters in their 20s and 30s, the Gyeongin region, and among centrist voters.

➂ The architects of Yoon Suk-yeol's victory: Half of the defecting Democrats switched to Yoon Suk-yeol.

Before examining the shifts in voter intention between the loyal and defecting Democratic groups, let us first analyze the changes in voter intention among all respondents. Leveraging the panel survey design, we will examine the changes in presidential candidate support among the 1,103 individuals who participated in both the first survey conducted in January before the election and the second survey conducted immediately after the election. In the first survey in mid-January, 36% (401 people) indicated they would vote for candidate Lee Jae-myung, while 35% (388 people) stated they would vote for candidate Yoon Suk-yeol, who had successfully regained support with campaigns such as 'abolishing the Ministry of Gender Equality and Family' and 'additional deployment of THAAD.' Support for candidate Ahn Cheol-soo was 12% (136 people) ([Table 3], last column, 'All respondents at the time of the 1st survey'). In the second survey conducted immediately after the election, with candidate Ahn Cheol-soo removed due to the merger, the final vote showed Lee Jae-myung at 45%, an increase of only 9 percentage points from the first survey, while Yoon Suk-yeol rose by 13 percentage points to 48%, achieving a reversal ([Table 3], last row, 'All respondents at the time of the 2nd survey').

Of the 401 supporters of Lee Jae-myung in the first survey, 92% (368 people) voted for Lee Jae-myung in the actual election, with only 5% (19 people) defecting to Yoon Suk-yeol. Meanwhile, among the 38 supporters of Sim Sang-jung in the first survey, 63% (24 people) were absorbed by Lee Jae-myung's support, along with 38% of the 21 supporters of other candidates. Lee Jae-myung also newly absorbed 35% (34 people) of the 98 undecided individuals and 48% (10 people) of the 21 individuals who were undecided or refused to answer. On the other hand, Yoon Suk-yeol maintained 95% (369 people) of his 388 supporters from the first survey, with only 2% (6 people) defecting to Lee Jae-myung. While Yoon Suk-yeol was less successful than Lee Jae-myung in absorbing votes from supporters of Sim Sang-jung and other candidates in the first survey, he gained an advantage by securing more votes than Lee Jae-myung from the larger groups of Ahn Cheol-soo supporters (76 people, 56%), undecided individuals (43 people, 44%), and undecided/refused respondents (11 people, 52%) from the first survey, which laid the foundation for his victory.

The reversal in support rates between the first and second surveys is not statistically significant. The truly noteworthy point is that both the loyal and defecting Democratic groups were the same groups that overwhelmingly supported the ruling party and candidate Moon Jae-in just two years prior in the general election and five years prior in the presidential election. In the general election, they formed the support base for the Democratic Party's acquisition of 180 seats, having overwhelmingly supported the Democratic Party together. Examining the 2017 presidential election voting trends in [Figure 2], both the loyal and defecting Democratic groups showed overwhelming support for candidate Moon Jae-in at the time, with 90% of the loyal Democratic group and 83% of the defecting Democratic group supporting him.

[Table 3] Changes in Presidential Candidate Support Among All Respondents (1,104 people): 1st-2nd Panel Survey

f2e34dd1a320d063

f2e34dd1a320d063

f2e34dd1a320d063

[Figure 2] Support Rate (%) for Candidate Moon Jae-in by Residual Democratic Camp and Defected Democratic Camp in the 2017 Presidential Election

However, the supporters who overwhelmingly backed candidate Moon Jae-in in the 19th presidential election and vocally supported the ruling party in the 20th general election have been divided into the residual democratic camp and the defected democratic camp ahead of the presidential election, and the voting choices of these two groups are starkly contrasted. In [Figure 3], for the residual democratic camp, 86% expressed support for candidate Lee Jae-myung in the January survey, 3% for candidate Yoon Suk-yeol, and 4% for candidate Ahn Cheol-soo. In the actual vote, 89% supported candidate Lee Jae-myung, a 3%p increase from the first survey. For candidate Yoon Suk-yeol, support increased by 5%p to 8%, absorbing support for candidate Ahn Cheol-soo. This confirms that support for candidate Lee Jae-myung, the presidential candidate of the Democratic Party of Korea, has been maintained.

The change in support rates between the first and second surveys for the defected democratic camp in [Figure 4] is noteworthy. In the first survey, 26% of the defected democratic camp indicated support for candidate Ahn Cheol-soo, 25% for candidate Lee Jae-myung, and 22% for candidate Yoon Suk-yeol, revealing that the voting intentions of the defected democratic camp were extremely dispersed. After candidate Ahn Cheol-soo withdrew, in the actual vote, 44% of the defected democratic camp responded that they voted for candidate Lee Jae-myung, and 45% voted for candidate Yoon Suk-yeol. Compared to the first survey, support for candidate Lee Jae-myung increased by 19%p from 25% to 44%, while support for candidate Yoon Suk-yeol increased by 23%p from 22% to 45%. Although within the margin of error, the support rates for candidate Lee Jae-myung and candidate Yoon Suk-yeol were reversed within the defected democratic camp. This is a clear indication that the shift in the voting intentions of the defected democratic camp, who withdrew their support for the Democratic Party, played a crucial role in candidate Yoon Suk-yeol's victory in a closely contested race, with half of this group converting their vote to candidate Yoon Suk-yeol.

[Figure 3] Change in Presidential Support for Residual Democrats (249 people) (%) [Figure 4] Change in Presidential Support for Defected Democrats (101 people) (%)

4. What Created the "Leavers" in the Democratic Party?

➀ Divergent Perceptions of the Moon Jae-in Administration: "Retained Democrats" view it as successful, while "Leavers" hold ambivalent assessments.

The most significant difference between the "Retained Democrats," who showed overwhelming support for the Democratic Party candidate Lee Jae-myung in this presidential election, and the "Leavers," who showed a defection of more than half of their support, lies in their evaluation of the Moon Jae-in administration. When asked to rate the Moon Jae-in administration's performance on a scale of 0-100, the "Retained Democrats" gave an average score of 77, while the "Leavers" gave a lukewarm score of 52 ([Figure 5]). A breakdown of agreement with statements such as "COVID-19 containment was successful," "The comprehensive real estate holding tax is excessive," and "This presidential election is a referendum on the Moon Jae-in administration," on a scale of 0 (strongly disagree) to 10 (strongly agree), reveals clearer distinctions. The "Retained Democrats" showed strong agreement with the statement 'COVID-19 containment was successful' with a score of 7.8, while they tended to strongly disagree with the statements 'The comprehensive real estate holding tax is excessive' and 'This presidential election is a referendum on the Moon Jae-in administration,' with scores of 3.6 and 2.5, respectively. In contrast, the "Leavers" generally agreed with the statement 'COVID-19 containment was successful' (5.8 points), but at the same time, they tended to largely accept the statements that 'The comprehensive real estate holding tax is excessive' (5.7 points) and that it is a 'Referendum on the Moon Jae-in administration' (5.4 points) ([Figure 6]). The "Leavers" are contrasted with the "Retained Democrats" in that they exhibit ambivalent attitudes, acknowledging both the achievements and failures of the Moon Jae-in administration.

[Figure 5] Presidential Approval Ratings of Moon Jae-in by Retained and Leaver Democrats (0 = Very Poor, 100 = Very Good)

[Figure 6] Scores on Election Perceptions and Issue Evaluations by Retained and Leaver Democrats (0 = Strongly Disagree, 10 = Strongly Agree)

➁ Dislike of Both Candidates: "Leavers" express a "both sides are bad" sentiment, dividing between "Retained" and "Converted" Democrats.

In terms of emotional attitudes toward the candidates of both camps, "Retained Democrats" exhibit strong partisan sentiment. As shown in [Figure 7], which measures candidate/party favorability on a scale of 0-10, "Retained Democrats" show high favorability towards President Moon Jae-in (7.9 points), candidate Lee Jae-myung (7.5 points), and the Democratic Party (6.8 points), while displaying strong dislike towards candidate Yoon Suk-yeol (1.8 points), the People Power Party (1.7 points), and its leader Lee Jun-seok (1.5 points). Conversely, among "Leavers," favorability towards President Moon Jae-in is relatively high (5.3 points), but favorability towards candidate Lee Jae-myung (4.2 points) and the Democratic Party (4.0 points) falls below the midpoint. While dislike of President Moon Jae-in was present compared to "Retained Democrats," the dislike sentiment towards candidate Lee Jae-myung also played a role in the support defection. However, dislike towards the opposition candidates/party/leader is also considerable. Candidate Yoon Suk-yeol scored 3.7 points, the People Power Party 3.4 points, and leader Lee Jun-seok 3.2 points, indicating even lower favorability. Among "Leavers," excluding favorability towards President Moon Jae-in, the favorability scores for other candidates/parties are within the margin of error, suggesting a coexistence of dislike towards both sides. This explains why the "Leavers" developed a lukewarm sentiment towards President Moon Jae-in, candidate Lee Jae-myung, and the Democratic Party after the general election, leading to significantly lower approval ratings for candidate Lee Jae-myung compared to "Retained Democrats." Simultaneously, while a Hankook Ilbo poll in February showed "Leavers" consolidating behind Yoon Suk-yeol, creating a lead for him, the final result was a close contest. This can be interpreted as a balancing effect caused by a significant portion of these voters returning to support Lee Jae-myung (Hankook Ilbo 2022/02/22).

[Figure 7] Emotional Favorability Towards Leaders and Parties of Both Camps by Retained and Leaver Democrats (0-10 points)

➂ Impact of Negative Issues: "Leavers" focus on real estate/Daejang-dong vs. "Retained" focus on Kim Keon-hee's suspicions/shamanism controversy.

What factors then reinforced the ambivalent stance of the "Leavers"? Let us examine the impact of negative issues that swayed the campaign competition between the two candidates through [Figure 8]. In this panel survey, respondents were asked to identify factors influencing their choice of candidate, with up to two overlapping responses allowed. Among "Retained Democrats," 55% of the total responses cited 'Kim Keon-hee's false resume and stock manipulation controversy,' and 34% cited 'Yoon Suk-yeol's shamanism controversy.' In contrast, only 12-15% cited 'real estate policy' and 'Daejang-dong development suspicions,' which were damaging to the ruling party and candidate Lee Jae-myung. On the other hand, 37% of "Leavers" cited 'failure of the Moon Jae-in administration's real estate policy,' 30% cited 'Daejang-dong development suspicions,' and 25% cited 'Yoon Suk-yeol-Ahn Cheol-soo candidate consolidation.' The response 'Kim Keon-hee's false resume and stock manipulation controversy,' which was of greatest concern to "Retained Democrats," accounted for only 25%, half the proportion, and the 'shamanism controversy' remained at around 10%.

Ultimately, from the perspective of the ruling party and government, to prevent the defection of "Leavers" and mitigate the narrative of a referendum on the administration, they should have focused on sincere apologies and diagnosis-and-solution presentations for the 'real estate issues' and 'Daejang-dong suspicions,' which were the primary reasons for the "Leavers"' withdrawal of support. However, the Democratic Party's negative campaigning focused on the concerns of "Retained Democrats." Furthermore, instead of concentrating on the 'Kim Keon-hee's false resume and stock manipulation controversy,' which was the biggest source of dissatisfaction, they became bogged down in the 'shamanism controversy' or the so-called 'Juliet controversy,' resulting in a campaign strategy that failed to appeal to either "Retained" or "Leaver" Democrats.

According to [Table 4], when asked about the criteria for selecting their candidate among the 1,050 respondents who reported voting, 64% of the 494 respondents who supported Lee Jae-myung cited ability and experience, indicating that national governance capacity is Lee Jae-myung's strength. Conversely, only 1% cited moral issues, the root of negative campaigning, clearly highlighting this as a weakness. Nevertheless, the intensification of negative campaigning close to the election can be seen as a factor that made it difficult to regain the support of "Leaver" Democrats.[4]

[Figure 8] Issues Influencing Candidate Choice by Retained and Leaver Democrats (1st and 2nd choice, multiple responses allowed, based on number of respondents, %)

[Table 4] Reasons for Voting for the Chosen Candidate by Voted Candidate (%)

f2e34dd1a320d063

f2e34dd1a320d063

f2e34dd1a320d063

5. Empirical Analysis

To verify the above discussion through a rigorous empirical verification model, a logistic regression analysis was conducted to distinguish between those who maintained their support for the Democratic Party in the 20th presidential election (coded as 0) and those who withdrew their support (coded as 1), among the 333 respondents who did not refuse or reserve their opinion out of the 350 who supported the Democratic Party in the last general election.

The analysis examined the impact on the withdrawal of current support among past ruling party supporters in the last general election (1 for withdrawal, 0 for remaining Democratic supporters) by including the following variables in the equation: (1) Generation/Gender factors (gender, age group, and interaction term between gender and age group), (2) Region (based on Gwangju-Jeolla), (3) Election Framework factor (attitude towards government judgment, 0-10), (4) Ideology factors (subjective ideology factor, objective ideology indicators - North Korea policy/welfare vs. growth), (5) Candidate Favorability (favorability scores for Lee Jae-myung, Yoon Suk-yeol, Ahn Cheol-soo, Sim Sang-jung, and Lee Jun-seok, each 0-10 points), and (6) Election Issues (1 if selected as the primary voting criterion, 0 otherwise).[5]

[Table 5] Analysis of Factors for Withdrawal of Democratic Party Support Since the 21st General Election (Withdrawal=1, Remaining=0): Logistic Regression Model

f2e34dd1a320d063

f2e34dd1a320d063

f2e34dd1a320d063

Note 1) N=333, Model Fit Chi-Square Test p=.000, Nagelkerke’s R2=.484, Classification Accuracy 81.6%

Note 2) + (p<0.1), * (p<0.05), ** (p<0.01), *** (p<0.001)

Although a more detailed interpretation is needed, the analysis results indicate that demographic factors influenced the fragmentation of Democratic Party supporters. Female Democratic Party supporters defected more than male supporters, but the gender difference varied by age group (the interaction effect of gender*age was significant at the 90% confidence level). Logistic regression analysis did not show a clear linear relationship by age group. To more directly examine the moderating effect of age group on gender differences, the predicted probability values estimated by the logistic regression equation were plotted by age*generation group. As shown in [Figure 9], women showed a higher defection rate compared to men in the traditional core support base of the Democratic Party (ages 30-50s). However, in the 20s and 60s and above, the defection probability was higher among male supporters than female supporters. Among the 30s-50s, the defection probability for male and female supporters in the 30s-40s differed by only 2 percentage points. However, in the 50s, which can be considered the core base of the Democratic Party's support, the defection probability for female supporters was 9 percentage points higher than that for male supporters (48% for 50s women, 39% for 50s men). In contrast, in the 20s and 60s and above, the defection rate was higher among male supporters, confirming that the gender difference in Democratic Party support defection varies by generation.

Regionally, supporters residing in the Gyeonggi/Incheon and Daegu/Gyeongbuk regions showed a stronger tendency to withdraw their support for the Democratic Party compared to supporters in the Gwangju/Jeolla region. Comparing the probability of defection among Democratic Party supporters during the general election period by region, calculated using the logistic regression model, reveals that the defection probability for supporters in the Gwangju region was only about 25%. In contrast, it was 41% for supporters in the Busan/Ulsan/Gyeongnam region, 47% for supporters in the Chungcheong region, and 49% for supporters in the Gangwon/Jeju region. For supporters in the Gyeonggi/Incheon region, it was 54%, and for supporters in the Seoul region, it was 56%, indicating that the defection probability for supporters in the Seoul metropolitan area exceeded 50%. Although the numbers are small, the probability of supporters in the Daegu/Gyeongbuk region, which has the strongest conservative tendency, being classified as defectors after the general election was a very high 69%.

[Figure 9] Predicted Probability of Defection among Democratic Party Supporters during the General Election Period by Gender and Generation (0: Retention, 1: Defection)

[Figure 10] Predicted Probability of Defection among Democratic Party Supporters during the General Election Period by Region of Residence (0: Retention, 1: Defection)

Meanwhile, within the electoral district context, attitudes toward the "regime judgment" narrative also emerged as a key factor strengthening defection. This showed a clear influence on the defection of Democratic Party supporters during the general election period. Regarding ideological orientation, objective ideological orientation, rather than subjective ideological identity, was found to have a statistically significant influence. That is, among supporters who backed the Democratic Party during the general election, those who sympathized more with the "regime judgment" narrative due to disappointment with the Moon Jae-in administration were more likely to defect (B=+.135, p<0.01). In particular, those with a stronger preference for growth over welfare were more likely to be classified as defectors (B=+.524, p<0.01). As shown in [Figure 11], the probability of being classified as a defector was only 33% among the group prioritizing welfare, while it was nearly double, at 65%, among Democratic Party supporters who prioritized growth.

Among candidate-related factors, attitudes toward candidate Lee Jae-myung were confirmed to be the criterion distinguishing between defection and retention. [Figure 12] clearly illustrates the distribution of defection probabilities classified by Lee Jae-myung's favorability score. Just as the defection of party support influenced the defection of support for candidate Lee Jae-myung, this result suggests that, conversely, attitudes toward candidate Lee Jae-myung influenced the defection of party support. On the other hand, the influence of the real estate issue, which was cited as a major factor in the decline of President Moon Jae-in's approval ratings and the Democratic Party's support, as well as negative issues concerning both candidates that became points of contention during the election campaign, were difficult to consider as significant variables. Although the influence of these issues on the division of the Democratic Party's support base was not statistically significant, it is difficult to definitively conclude that these variables had no impact. In fact, attitudes toward the failure of real estate policy were not only a key factor reinforcing the "regime judgment" narrative but also highly correlated with candidate Lee Jae-myung's favorability regarding negative issues related to his integrity. This means that although the direct effect was not statistically significant, the possibility cannot be ruled out that the impact of these issue-related factors was suppressed through indirect effects via the "regime judgment" narrative or attitudes toward candidate Lee, or conversely, by the influence of the "regime judgment" or candidate Lee's favorability variables.[6]

[Figure 11] Average Defection Probability by Welfare-Growth Attitude

[Figure 12] Distribution of Predicted Defection Probability by Lee Jae-myung Favorability

6. Conclusion: After the Dissolution of the Impeachment Political Coalition, What Order Will Emerge?

The preceding sections have explored the reasons why the ruling party, holding a supermajority of 180 seats, became the target of a regime change within two years, focusing on the voting behavior of the "defector" group who supported the ruling party in the last general election but subsequently withdrew their support. The "defector" group, primarily composed of individuals in their 20s and 30s, residing in the Gyeonggi/Incheon region, and holding moderate political views, maintained relative favorability towards President Moon Jae-in but simultaneously harbored sentiments for "regime judgment" due to the failure of real estate policies. While candidate Lee Jae-myung demonstrated strengths in governance capabilities, and the controversies surrounding Yoon Suk-yeol's shamanism and "Juliet" allegations during the election campaign helped rally the "retained" Democratic supporters, they did not seem to help in regaining the support of the "defector" group. When the Daejang-dong issue and integrity problems arose, the governance capability issue was suppressed, ultimately appearing to have backfired on candidate Lee Jae-myung. Ultimately, the initial survey in early January confirmed a close contest between candidates Lee and Yoon among the "defector" group, and in the final outcome, this close contest could not be overcome, failing to create a lead in support.

This paper focuses on summarizing the key characteristics of the voting behavior of the defecting democratic base and empirically examining, at a hypothetical level, the factors that may have led to the defeat of the Democratic Party of Korea. However, the voter landscape in which the new government's state administration and the next local and general elections will be held remains a matter of utmost concern. The trajectory of the defecting democratic base is likely to emerge as the biggest variable in answering this question, as it remains fluid whether the defecting democratic base will be restored to support the Democratic Party, be absorbed into the support base of the new ruling party, or maintain a state of unstable equilibrium as in the recent election.

The conservative party, which experienced division and defection among its supporters during the impeachment process, only succeeded in restoring its former support base size after suffering unprecedented defeats in three national elections, and was able to regain power after five years. However, the presidential election, which ended with a narrow margin of 0.7%, and the Yoon Suk-yeol administration, which is launching amidst a check-and-balance mentality as strong as the expectations, despite being in the early part of its term, faces a dilemma of having to manage state affairs amidst a balance of power between the ruling and opposition parties. Conversely, can the Democratic Party of Korea, which dreamed of a '20-year rule' with an unprecedented landslide victory, succeed in restoring the defecting democratic base that left after the general election? If so, how quickly? To predict the success of the new government and the future of the Post-Impeachment Political Alliance, which will emerge after the Impeachment Political Alliance, in-depth research on the changes in perception and policy preferences of the defecting democratic base will be necessary. This is why the analysis of the defecting democratic base in this presidential election should not be concluded as a one-off study. ■

References

Kim, Eun-ji. 2022. “You Who Voted Diligently, Why Did You Vote and Why Didn't You Vote?” <SisaIN> Issue No. 758 (2022/03/24).

Hankook Ilbo. 2022. “Lee Jae-myung's Crisis Lies with the 'Defecting Democrats'... Disappointed with the Moon Administration, They Rally Behind Yoon Suk-yeol” (2022/03/22).

Hankook Ilbo. 2022. “Lee Jae-myung's Lack of Ethics 70.4%, Yoon Suk-yeol's Lack of State Administration Capacity 69.8%” (2022/01/01).

Jeong, Han-ul. 2021. “Analysis of the April 7th By-elections: The Unification Variable and the Weaknesses of the Democratic Party of Korea: In Seoul, the opposition has an advantage in actual voting if the polls are tied.” <Fair Election Research ➀> (2021.3.16.).

_____. 2020. “The 21st General Election from the Perspective of Judging the Ruling and Opposition Parties and the Dilemma of Conservative Reform.” 1-20 (2020.5.8.).

Jeong, Han-ul and Kang, Woo-chang. 2017. “Cracks in the Conservative Base: Causes and Consequences of the Emergence of Swing Conservatives.” In Kang, Won-taek (ed.), <Changing Korean Voters 6: Candlelight Vigils, Impeachment Crisis, and the 19th Presidential Election>. Seoul: East Asia Institute.

Hawley, George and Inaki Sagarzazu. 2012. “Where Did the Votes go? Reassessing American Party Realignment via Vote Transfers between Major Parties from 1860 to 2008. Electoral Studies 31, 726-739.

Norpoth, Helmut, and Jerrold G. Rusk. 2007. "Electoral Myth and Reality: Realignments in American Politics." Electoral Studies 26, 2: 392-403.

[1] For an explanation of voter behavior changes in the last three national elections from the perspective of 'defector conservatives' and 'retained conservatives,' see (Jeong Han-ul 2020; Jeong Han-ul & Kang Woo-chang 2017).

[2] This panel survey did not include a question about party support during the general election period, so the proportional representation party at the time is used as a proxy for the party supported at that time. Among the 1,104 respondents who participated in both the first (January 11-14) and second surveys (March 10-15), 901 responses were classified after excluding 203 individuals who stated they were not eligible to vote, refused to answer, or could not recall. Cross-tabulated responses from the Hankook Ilbo/ 한국리서치 New Year's Survey (December 2021) and the Presidential Election Poll (February 2022), comparing party support at the time of the general election with current party support, indicated that 35-40% of Democratic Party supporters at the time of the general election were classified as defectors. While this discrepancy may be due to differences in wording for proportional representation party and supported party questions, or differences in timing, considering the effect of Lee Jae-myung's support consolidation towards the end of the election, it is highly probable that some of the defecting supporters consolidated behind candidate Lee Jae-myung. However, this report focuses on comparing with the retained Democratic supporters rather than precisely estimating the scale of the defecting Democrats.

[3] In fact, for the 20s and 30s age groups, exit polls from the 2020 general election showed high support rates around 60% for the Democratic Party, which might lead one to believe there was no significant defection. However, according to the age-specific voter turnout rates announced later, the turnout for the 20s and 30s age groups was significantly lower compared to those in their 60s and above. This suggests that while there may not have been a large number of voters who switched parties, there was a considerable number who abstained from voting. The phenomenon of younger generations switching parties was particularly evident in the April 7 by-elections (Jeong Han-ul 2021).

[4] While candidate Lee Jae-myung showed an advantage in governance capabilities, public opinion consistently indicated that the integrity assessment was neck-and-neck, or that negative sentiment towards candidate Lee Jae-myung was stronger, as confirmed through various opinion polls during the presidential election campaign (Hankook Ilbo 2022/01/01; Kim Eun-ji 2022).

[5] Objective ideological factors were coded as follows: North Korea policy: -1 for maintaining/strengthening dialogue and cooperation, 0 for don't know, +1 for maintaining/strengthening hardline policy. For welfare versus growth: -1 for prioritizing welfare, 0 for don't know, +1 for prioritizing growth. For election issue factors, among the first and second ranked responses to the question about issues influencing voting choice, the 'failure of real estate policy' or 'Daejang-dong issue' were selected as the first choice by the 'Real Estate Issue' response group. The 'Lee Jae-myung's sister-in-law's abusive language' or 'Kim Hye-kyung's corporate card controversy' were selected by the 'Lee_Integrity' issue response group. The 'Kim Keon-hee's career and speculation allegations' or 'shamanism controversy' were selected by the 'Yoon_Integrity' issue response group.

[6] For a more detailed analysis of these issue-related factors, please refer to the chapter on issue factors (e.g., real estate policy) in this monograph. A precise analysis of these factors goes beyond the scope of this chapter and will be addressed in subsequent work.

■ Author: Jeong Han-ul_A professional poll analyst and research designer at Hankook Research. He holds a Ph.D. in Political Science and International Relations from Korea University. He has served as Deputy Director of the EAI Public Opinion Analysis Center, Deputy Director of the Center for Foreign Affairs and National Security, and Secretary-General. His main research areas include elections and generational politics, national identity and security perceptions, and CSR research. His major publications (co-authored) include "20s Men," "20s Women," "Market and Democracy in the Age of Inequality," and "The Park Geun-hye Phenomenon."

■ Editor: Jeon Ju-hyun_EAI Researcher

Contact: 02-2277-1683 (ext. 204) jhjun@eai.or.kr