[The 21st Presidential Election and Korean Democracy: Crisis, Division, and Realignment] ⑦ Dissolution of the 6070 Conservative Alliance and the Roots of the 4050 Pro-Democratic Party Voting Tendency

I. Introduction

This chapter deals with martial law and impeachment among generations other than the 2030s in the 21st presidential election voting, and voting behavior in the 21st presidential election. Using the analysis of the 21st presidential election vote intention commissioned by the East Asia Institute to Korea Research, conducted from June 4-5, 2025, and other survey data available to the author, this chapter compares generational voting patterns in the 21st and 20th presidential elections and raises the need for a detailed analysis of the voting behavior of generations other than the 2030s, namely the 4050s and 6070s. While recent years have focused on the new phenomenon of gender conflict among the 2030s in Korean society, leading to concentrated attention on generational voting in this demographic, the results of the last two presidential elections show several noteworthy phenomena in generational voting.

First, comparing the generational voting patterns confirmed in the 2022 and 2025 exit polls, the U-shaped generational voting pattern, which has strengthened since the mid-2010s, is being maintained. In the 20th presidential election [Figure 1], the 4050 generation, which is the focus of this chapter, showed a preference for the Democratic Party candidate Lee Jae-myung, the 2030 generation showed a tight contest between Lee Jae-myung and Yoon Suk-yeol, and the 6070 generation showed a clear preference for Yoon Suk-yeol. Among the 40s, Lee Jae-myung had 61% vs. Yoon Suk-yeol 35%; among the 50s, Lee Jae-myung had 52% vs. Yoon Suk-yeol 44%, with a preference for Lee Jae-myung. Among the 20s, it was a very close race with Lee Jae-myung at 48% vs. Yoon Suk-yeol at 46%, and among the 30s, Lee Jae-myung at 46% vs. Yoon Suk-yeol at 48%. In contrast, among the 60s, Lee Jae-myung had 33% vs. Yoon Suk-yeol 65%, and among those over 70, Lee Jae-myung had 29% vs. Yoon Suk-yeol 70%, showing a clear preference for the conservative candidate.

According to the exit polls of the three major broadcasters for the 21st presidential election [Figure 2], support for the Democratic Party candidate Lee Jae-myung was overwhelming at 72.7% and 69.8% among the 40s and 50s, respectively, while support for the People Power Party candidate Kim Moon-soo was only 22.2% and 25.9%. The intensity of pro-Lee Jae-myung support has strengthened compared to the 20th presidential election. On the other hand, the most noticeable change is among the 60s. Among those over 70, candidate Kim Moon-soo's support rate was 64%, overwhelming Lee Jae-myung's 34%, showing a continued strong pro-conservative tendency, although its intensity has weakened compared to the previous election. However, among the 60s, who overwhelmingly supported the People Power Party's Yoon Suk-yeol in the previous election, the support rates for the Democratic Party's Lee Jae-myung and the People Power Party's Kim Moon-soo were neck and neck at 48% vs. 49%. Meanwhile, the 2030 generation, which saw a very close race between Lee Jae-myung and Yoon Suk-yeol in the previous election, continues to show a competitive landscape in terms of support for Lee Jae-myung versus non-Lee Jae-myung candidates. However, a characteristic feature is that the votes for non-Lee Jae-myung candidates are dispersed among Kim Moon-soo and the Rebuilding Korea Party's Lee Jun-seok. While Lee Jae-myung's support rate was 41% and 48% among the 20s and 30s, respectively, surpassing Kim Moon-soo's 31% and 33%, the competitive landscape can be considered maintained when adding the Rebuilding Korea Party's Lee Jun-seok's support rate of 24% and 18%.

[Figure 1] Generational Voting in the 20th Presidential Election (%), [Figure 2] Generational Voting Patterns in the 21st Presidential Election (%)

Source: 2022 Presidential Election Three-Broadcaster Exit Poll, Source: 2025 Presidential Election Three-Broadcaster Exit Poll

From the perspective of generational gender voting in [Figures 3] and [Figures 4], there are clear differences in voting tendencies between men and women in the 2030 generation, while there are no differences in support for candidates between genders in the 4050 and 6070 generations. However, the 30s are noteworthy. Comparing generational gender voting trends during the 20th presidential election, the gender gap was mainly concentrated in the 20s in the 20th presidential election, but in the 21st presidential election, the difference in voting behavior between men and women became clear not only in the 20s but also in the 30s.[1]The gap between men and women in their 20s was -22 percentage points for Lee Jae-myung's support and +25 percentage points for Yoon Suk-yeol's support in the 20th presidential election. In the 21st presidential election, the gap for Lee Jae-myung's support was -34 percentage points, for Kim Moon-soo's support +12 percentage points, and for Lee Jun-seok's support +27 percentage points, widening the gap. In contrast, the gap between men and women in their 30s was only -7 percentage points for Lee Jae-myung's support and +9 percentage points for Yoon Suk-yeol's support in the 20th presidential election. However, in the 21st presidential election, the gap widened to -19 percentage points for Lee Jae-myung's support, +4 percentage points for Kim Moon-soo, and +17 percentage points for Lee Jun-seok's support.[2]

[Figure 3] Generational x Gender Voting in the 20th Presidential Election (%), [Figure 4] Generational x Gender Voting in the 21st Presidential Election (%)[3]

Source: 2022 Presidential Election Three-Broadcaster Exit Poll, Source: 2025 Presidential Election Three-Broadcaster Exit Poll

These generational voting changes raise several research questions. First, the consistent pro-Democratic Party political tendency and voting behavior observed in the 4050 generation are themselves of research interest. Why do the 4050s consistently show a pro-Democratic Party tendency? Second, in the 6070 generation, which has been the base of conservatism in Korean society, the tendency to support conservative candidates remained strong among those over 70, similar to the 4050s. However, among the 60s, support for candidate Lee Jae-myung increased significantly compared to the previous election. This presidential election shows clear signs of the dissolution of the 6070 conservative alliance. Why is there an erosion of the conservative base among the 60s? Is it a temporary phenomenon or a structural one? Third, what is the reason for the spread of the gender voting divide among the youth generation, symbolized by the conflict between 'Lee Dae-nam' and 'Lee Dae-nyeo'?

This chapter will comprehensively examine three significant changes in generational political behavior (① strengthening of the pro-Democratic tendency among the 4050s, ② dissolution of the 6070 conservative alliance, and ③ expansion of the gender conflict among the 20s) and then focus on the 'ACP effect (Aging, Cohort, Period)' related to age effects to empirically analyze the factors behind these changes. In generational voting research, one perspective argues for the 'aging effect,' suggesting that political tendencies change with age, while another perspective, the 'cohort effect' theory, posits that political tendencies imprinted during the formative period are maintained as individuals age. Meanwhile, the period effect refers to simultaneous attitude changes that affect all generations, such as the 'end of the Cold War' or the '9/11 terrorist attacks,' rather than phenomena concentrated in specific age groups or generational positions (Kang Won-taek 2010; Park Won-ho 2013; Bhatti and Hansen 2012; Mannheim 1997; Tilley and Evance 2013).

In Korea, it is common to define the 6070 generation as the 'war, Yushin, and industrialization generation' and explain their conservative tendencies, or to define those born in the 1960s and 1970s as the '386 generation' or 'Roh Moo-hyun generation' and explain their progressive/pro-Democratic Party tendencies. These are typical explanations based on the cohort effect. On the other hand, the influence of issues such as 'martial law and impeachment,' including the martial law of December 3, 2024, and the Constitutional Court's impeachment ruling in 2025, can be seen as a kind of 'period effect' that has influenced recent generational political behavior (Noh Hwan-hee et al. 2013; Park Jae-hong 2009; Bae Jin-seok 2022; Lee Sang-shin et al. 2020; Jeong Han-wool 2020; Heo Seok-jae 2014).

Finally, we will focus on the possibility that the expansion of the gender voting gap among the 20s to other generations is a product of social changes where feminist attitudes, which influence gender voting, are spreading beyond the 20s to other generations. Although separate research is needed, the process of conservatism among men in their 20s can be seen as originating from antipathy towards feminism, leading to opposition to President Moon Jae-in and the Democratic Party, who declared themselves 'feminist presidents,' and subsequently adopting an ideological conservative identity in policy preferences (Kook Seung-min et al. 2022; Chun Gwan-yul and Jeong Han-wool 2019).[4]

II. Changes in Generational Alliances During the Martial Law and Impeachment Periods

1. The 22nd Presidential Election and Generational Voting

First, using the presidential election vote intention analysis survey by the East Asia Institute and other survey data, we confirm the strengthening of the pro-Democratic tendency among the 4050s and the decoupling phenomenon between the 60s and 70s observed in the exit polls. In [Table 1], looking at the preferred candidates of the 1,238 respondents who answered they voted in the 2022 presidential election by age group, similar to the exit poll results, the gap between Lee Jae-myung's support rate (28-31%) and Yoon Suk-yeol's support rate (28-31%) was less than 10 percentage points in the 2030 generation, indicating a competitive situation, and the undecided voters, such as 'abstention/reserved response,' were the most fluid generation with 24-33%. In contrast, among the 4050 generation, Lee Jae-myung's support rate (52%) was nearly double that of Yoon Suk-yeol's support (24-30%), showing a clear advantage. In the case of the 6070 generation, Yoon Suk-yeol's support rate (45-68%) was close to or exceeded half, and Lee Jae-myung's support rate (22-37%) was at a disadvantage, indicating that the conservative candidate support alliance of the 6070 generation was in operation.[5]

However, in the 2025 21st presidential election [Table 2], several changes are confirmed, as observed earlier. First, among the 4050 generation, Lee Jae-myung's support rate (60% each) significantly surpassed the People Power Party's Kim Moon-soo support (20-26%). While the 20s among the 2030 generation still showed a competitive pattern (Lee Jae-myung 25%, Kim Moon-soo 27%, Lee Jun-seok 18%, Kwon Young-guk 3%), the pro-Democratic tendency strengthened in the 30s as well, with Lee Jae-myung's support rate rising to 46%. The departure of support from the 60s is clearly confirmed in this survey as well. Among the 70s, 60% supported Kim Moon-soo and 30% supported Lee Jae-myung, a two-fold difference. However, among the 60s, 45% supported Lee Jae-myung, 42% supported Kim Moon-soo, and 2% supported Lee Jun-seok, making Lee Jae-myung's support rate comparable to the conservative candidate's support rate.

[Table 1] Preferred Candidate in the 2022 20th Presidential Election by Age Group (1,238 respondents who voted)

EAI·한국리서치 <Analyse des intentions de vote pour la 20e élection présidentielle>

[Tableau 2] Candidat préféré pour la 21e élection présidentielle de 2025 par groupe d'âge (1 310 répondants ayant voté)

EAI·한국리서치 <Analyse des votes pour la 21e élection présidentielle>

Les orientations politiques générationnelles sont également clairement 확인ées dans la perception des candidats. Le [Figure 5] présente l'évaluation de l'attrait pour les candidats Lee Jae-myung, Kim Moon-soo et Lee Jun-seok sur une échelle de 0 à 10 (0 point : très peu attrayant ~ 5 points : neutre ~ 10 points : très attrayant).

Dans la tranche d'âge 40-50 ans, l'attrait pour le candidat Lee Jae-myung était de 6,0 et 5,5 points, tandis que l'attrait pour le candidat Kim Moon-soo était faible, à 3,2 et 3,5 points respectivement, et l'attrait pour le candidat Lee Jun-seok était très froid, à 2,8 et 2,5 points. Pour le vote, l'attrait pour le candidat Lee Jae-myung était de 5,0 points chez les 30 ans, qui ont soutenu les deux camps de candidats en concurrence avec les 20 ans, dépassant la marge d'erreur, tandis que Kim Moon-soo était à 3,0 et Lee Jun-seok à 3,1 points. En revanche, chez les plus de 70 ans, en tant que base de la politique conservatrice, l'attrait pour le candidat Kim Moon-soo était de 5,9 points, Lee Jae-myung de 3,3 points et Lee Jun-seok de 2,9 points, montrant un sentiment pro-Kim Moon-soo clair.

Les 20 et 60 ans ont également été en concurrence dans leur attitude envers les candidats. Chez les 20 ans, Lee Jae-myung était à 4,2 points, Lee Jun-seok à 4,1 points et Kim Moon-soo à 3,9 points, une compétition extrêmement serrée dans la marge d'erreur. En revanche, chez les 60 ans, l'attrait pour le candidat Lee Jae-myung et l'attrait pour le candidat Kim Moon-soo étaient respectivement de 4,4 et 4,5 points, une situation extrêmement serrée, et Lee Jun-seok n'a atteint que 2,9 points. Le fait que le candidat Lee Jun-seok soit un objet de désapprobation autant que le candidat Lee Jae-myung dans la tranche des 60-70 ans est significatif.

[Figure 5] Attrait des candidats à la présidentielle par groupe d'âge (0 point : très peu attrayant ~ 5 points : neutre ~ 10 points : très attrayant)

Source : EAI·한국리서치 <Analyse des votes pour la 21e élection présidentielle>

Le [Figure 6] montre le pourcentage de réponses multiples pour les questions qui ont influencé la décision du candidat soutenu (1er + 2ème choix). Globalement, la question qui a reçu le plus de réponses est « La phase de loi d'exception et de destitution du gouvernement Yoon Suk-yeol », représentant 58 % par rapport à un total de 200 % de réponses multiples. Les réponses indiquant « le risque moral et judiciaire du candidat Lee Jae-myung » représentaient 47 % du total des cas de réponse, « l'orientation de coopération avec la Chine du candidat Lee Jae-myung » 35 %, « la unification de Kim Moon-soo et Han Duck-soo et le processus de sélection des candidats présidentiels » 32 %, le pourcentage de ceux qui ont réagi à « la ligne indépendante du candidat Lee Jun-seok dans la troisième voie et la critique du système bipartite » était de 12 %, le pourcentage de ceux qui ont choisi « la politique de la jeunesse du candidat Lee Jun-seok » était de 9 %, et « autres » 7 % (total 200 %).

En examinant les candidats effectivement votés pour chaque question choisie, on peut confirmer quelle question était la plus favorable à quel candidat. Pour la question « Loi d'exception et destitution », 71 % des répondants ont choisi cette option, et ceux qui ont choisi « l'orientation de coopération avec la Chine du candidat Lee Jae-myung » et « l'unification de Kim Moon-soo et Han Duck-soo et le processus de sélection des candidats présidentiels » ont soutenu le candidat Lee Jae-myung à 57 % et 53 % respectivement, tandis que le soutien à Kim Moon-soo n'a atteint que 33 %. En revanche, parmi ceux qui ont choisi « la controverse sur la moralité et le risque judiciaire du candidat Lee Jae-myung », 62 % ont soutenu Kim Moon-soo, et pour « la ligne indépendante du candidat Lee Jun-seok dans la troisième voie », 38 % ont soutenu Kim Moon-soo, 32 % Lee Jae-myung et 20 % Lee Jun-seok. Parmi ceux qui ont choisi « la politique de la jeunesse du candidat Lee Jun-seok », 36 % ont soutenu Lee Jun-seok et 17 % Lee Jae-myung. Le candidat Lee Jae-myung semble avoir bénéficié d'une augmentation de son soutien grâce à l'influence écrasante de la question de la loi d'exception et de la destitution, tandis que le candidat Kim Moon-soo n'a pas réussi à attirer de soutien en dehors des questions négatives concernant les candidats adverses. Au contraire, le candidat Lee Jun-seok a montré sa présence en générant des votes grâce à sa ligne indépendante et à sa politique de la jeunesse.

[Figure 6] Questions ayant influencé la décision du candidat soutenu : ratio combiné des 1er et 2ème choix, %)

Source : EAI·한국리서치 <Analyse des votes pour la 21e élection présidentielle>

[Tableau 3] Candidats présidentiels soutenus par question ayant influencé la décision de soutien

Source : EAI·한국리서치 <Analyse de l'électorat de la 21e élection présidentielle>

Il est intéressant de noter que les 20-30 ans ont eu un taux égal de choix entre les questions favorables au Parti Démocrate comme « l'état de siège et la destitution » et les questions favorables à la droite comme « la moralité et les risques judiciaires du candidat Lee Jae-myung ». Le taux de choix de « la ligne indépendante du candidat Lee Jun-seok dans la troisième voie » ou « les politiques de jeunesse du candidat Lee Jun-seok » était plus élevé que dans les autres groupes d'âge. En revanche, les 40-50 ans ont majoritairement choisi « la situation de l'état de siège et de la destitution du gouvernement Yoon Suk-yeol » (63% pour les 40 ans, 70% pour les 50 ans), suggérant que cela est devenu un facteur de soutien écrasant pour le Parti Démocrate. Chez les 70 ans et plus, le taux de choix de « la controverse sur la moralité et les risques judiciaires du candidat Lee Jae-myung », qui sert de justification au soutien au Parti du Pouvoir du Peuple, était le plus élevé à 69%. Cependant, chez les 60 ans, la question la plus influente était « l'état de siège et la destitution (60%) », et le taux de choix de « la moralité et les risques judiciaires du candidat Lee Jae-myung », qui a renforcé le vote anti-Démocrate, était également de 54%, ce qui est comparable. Le fait que les voix des 60 ans se soient détournées pour soutenir le candidat Lee Jae-myung, au même titre que le soutien au candidat Kim Moon-soo, semble être dû au fait qu'un nombre important de personnes de 60 ans ont réagi à la question de « l'état de siège et de la destitution ».

[Tableau 4] Analyse des facteurs de décision du candidat soutenu par groupe d'âge, analyse des réponses multiples (1+2e choix)

Source : EAI·한국리서치 <Analyse des votes pour la 21e élection présidentielle>

2. Évolution des alliances générationnelles dans le soutien aux partis politiques

Lors de l'élection présidentielle, la division générationnelle du soutien aux partis politiques se reflète précisément dans la configuration où les 40-50 ans soutiennent le Parti Démocrate, les 70 ans et plus soutiennent le Parti du Pouvoir du Peuple, et les 20-30 ans et les 60 ans sont en concurrence entre le soutien au Parti Démocrate et le rejet de ce parti. Comme le montre le graphique 7 sur le soutien aux partis par génération, les taux de soutien au Parti Démocrate étaient respectivement de 53 % et 49 % chez les 40 et 50 ans, et en ajoutant le soutien au Parti de l'Innovation de Cho Kuk (8 % et 11 %) et au Parti Progressiste, le pourcentage de soutien au camp démocrate/progressiste était globalement écrasant. Le soutien au Parti du Pouvoir du Peuple était respectivement de 17 % et 23 %, et le soutien au Parti de la Réforme était d'environ 3 % pour chaque groupe d'âge.

Chez les 20-30 ans, le soutien au Parti Démocrate était de 33 % et 43 %, inférieur à celui des 40-50 ans, et le taux de soutien au Parti du Pouvoir du Peuple n'était que de 17 % et 21 %. Cependant, il est caractéristique que le taux de soutien au Parti de la Réforme (20 % chez les 20 ans, 10 % chez les 30 ans) et le pourcentage d'indécis (21 % chez les 20 ans, 20 % chez les 30 ans) soient plus élevés dans le groupe des 20-30 ans que dans les autres générations. Il est à noter que des différences apparaissent également dans le soutien aux partis entre les 60 et 70 ans. Chez les 70 ans et plus, le soutien au Parti du Pouvoir du Peuple était de 57 %, celui au Parti de la Réforme de 3 %, tandis que le soutien au Parti Démocrate n'était que de 26 %, celui au Parti de l'Innovation de Cho Kuk de 2 % et au Parti Progressiste de 2 %, avec environ 10 % d'indécis. Cela confirme clairement qu'il s'agit toujours d'une base de soutien pour le Parti du Pouvoir du Peuple. En revanche, chez les 60 ans, le soutien au Parti Démocrate était de 34 %, celui au Parti de l'Innovation de Cho Kuk de 10 % et au Parti Progressiste de 1 %, soit un total de 45 % pour le camp démocrate. Le soutien aux partis conservateurs, incluant le Parti du Pouvoir du Peuple (37 %), le Parti de la Réforme (5 %) et autres (1 %), n'atteignait que 43 %, ce qui montre une nette différence par rapport à l'époque où il s'agissait d'une base de soutien pour les partis conservateurs.

[Graphique 7] Répartition du soutien aux partis par génération (%)

Source : EAI·Korea Research <Enquête sur l'analyse des votes de la 21e élection présidentielle>

Ces changements dans les modèles de clivage générationnel semblent être le résultat de l'accumulation continue de changements dans les alliances de soutien aux partis au cours des cinq dernières années. Le graphique 8 présente les taux de soutien aux partis de l'enquête nationale (NBS), agrégés annuellement de juillet 2020 à août 2025. Jusqu'aux élections législatives de 2020, le soutien au Parti Démocrate était supérieur à celui du Parti du Pouvoir du Peuple chez les 20-30 ans, mais ils sont devenus une génération en concurrence après l'élection présidentielle de 2022. Durant le mandat de Yoon Suk-yeol et après les élections législatives de 2024, en particulier après le décret de loi martiale du 12.3, la tendance générale est à la concurrence, mais le soutien au Parti Démocrate montre une légère supériorité.

En revanche, le soutien au Parti Démocrate s'est maintenu chez les 40-50 ans et s'est même renforcé après les élections législatives de 2024 et le décret de loi martiale du 12.3. Chez les 40 ans, le soutien au Parti Démocrate est resté stable, indépendamment de la période. Chez les 50 ans, le soutien au Parti du Pouvoir du Peuple s'est rapproché du soutien au Parti Démocrate dans la marge d'erreur après l'élection présidentielle de 2022, montrant une tendance de concurrence. Durant le mandat de Yoon Suk-yeol, après les élections législatives et le décret de loi martiale, le soutien au Parti du Pouvoir du Peuple s'est stabilisé/a diminué, tandis que le soutien au Parti Démocrate a fortement augmenté, consolidant ainsi l'alliance pro-démocrate chez les 40-50 ans.

Les changements dans le soutien aux partis ont été les plus importants chez les 60-70 ans. Après les élections législatives de 2020 et jusqu'à l'élection présidentielle de 2022, le soutien au Parti Démocrate chez les 60-70 ans a chuté de la fin de la vingtaine à la début de la vingtaine, tandis que le soutien au Parti du Pouvoir du Peuple a augmenté pour atteindre 50-60 %. Durant cette période, l'alliance conservatrice des 60-70 ans s'est consolidée. Cependant, durant le mandat de Yoon Suk-yeol, après les élections législatives et les phases de décret de loi martiale et de destitution, le soutien au Parti Démocrate chez les 60 ans et les 70 ans et plus a augmenté pour atteindre 30-37 %, tandis que le soutien au Parti du Pouvoir du Peuple a continuellement diminué. En particulier, chez les 60 ans, le soutien au Parti du Pouvoir du Peuple, qui dépassait 50 % en 2022, est tombé à un peu plus de 40 % en 2025, et le soutien combiné au Parti de l'Innovation de Cho Kuk et au Parti Démocrate rivalise avec le soutien combiné au Parti du Pouvoir du Peuple et au Parti de la Réforme, devenant une génération en concurrence. Chez les 70 ans, le soutien au Parti du Pouvoir du Peuple, qui approchait 60 % en 2025, est tombé à un peu plus de 50 %, et le soutien au Parti Démocrate a augmenté de 20 % à près de 30 %, mais l'écart de soutien entre les deux partis reste supérieur à 20 points de pourcentage, maintenant une prédominance conservatrice.

En examinant les taux de soutien aux partis durant le processus des élections législatives de 2024, du décret de loi martiale et de la destitution, il existe des différences dans l'ampleur des fluctuations selon les générations. Cependant, dans toutes les générations, le soutien au Parti du Pouvoir du Peuple a stagné ou diminué, tandis que le soutien au Parti Démocrate a constamment augmenté, démontrant que les questions de décret de loi martiale et de destitution ont agi comme un « effet de période » commun à toutes les générations.

[Graphique 8] Évolution du soutien aux partis par génération au cours des cinq dernières années

Source : Enquête nationale de référence (NBS) (juillet 2020 - août 2025)

III. Facteurs de fluctuation du clivage générationnel : Attitudes face au décret de loi martiale et à la destitution et orientation politique

1. Facteur de fluctuation à court terme : Effet de période des questions de décret de loi martiale et de destitution

Examinons maintenant les différences générationnelles dans la perception du décret de loi martiale et de la destitution. Selon l'enquête <2e Enquête sur l'agenda national> menée par le Progressive Policy Institute, le Korea People's Research Institute et Korea Research, comme le montre le graphique 9, 84 % des 40-50 ans ont considéré la décision de la Cour constitutionnelle de destituer le président Yoon Suk-yeol comme une bonne décision. Chez les 20-30 ans, 73 % à 76 % étaient également d'accord avec la destitution. Il est à noter que même chez les 70 ans et plus, l'opinion publique considérant la décision comme bonne était de 52 %, tandis que 43 % la considéraient comme mauvaise, montrant une légère prédominance de l'opinion favorable à la destitution. Chez les 60 ans en particulier, 62 % étaient favorables à la destitution.

En ce qui concerne la gestion de l'incident du tribunal de district ouest, qui a suscité une controverse d'extrême droite en Corée, et le traitement des personnes arrêtées, l'opinion selon laquelle « ils doivent être sévèrement punis car il s'agit de violence illégale » a largement dépassé l'opinion selon laquelle « il faut faire preuve de clémence car il s'agit d'une expression légitime de mécontentement face à l'arrestation abusive du président Yoon ». Chez les 40-50 ans en particulier, la position selon laquelle une punition sévère est nécessaire était forte, à 85-86 %, et même chez les 20-30 ans, contrairement aux inquiétudes concernant l'extrémisation, 79 % à 82 % ont exprimé une opinion de non-tolérance. Même chez les 70 ans, qui montrent la plus forte tendance d'extrême droite dans les recherches récentes, l'opinion favorable à la clémence au motif d'une « expression légitime de mécontentement » n'était que de 43 %, tandis que la majorité, soit 54 %, adoptait la position selon laquelle « il faut punir sévèrement car il s'agit de violence illégale ». Chez les 60 ans, qui montrent des signes de rupture avec l'alliance conservatrice des 60-70 ans, plus de la moitié, soit 68 %, ont adopté une attitude principielle de non-tolérance face aux violences du tribunal de district ouest. Si les voix fortes en faveur du décret de loi martiale et de la destitution se renforcent, comme en témoigne l'élection du président du Parti du Pouvoir du Peuple après l'élection présidentielle, la tendance à la rupture des 60 ans avec le camp conservateur devrait s'accélérer.

En examinant les attitudes des hommes et des femmes par génération face au décret de loi martiale et à la destitution dans les tableaux 5 et 6, même chez les hommes de 20 ans, qui sont critiqués pour leur extrême droite, 70 % ont répondu que la « décision de la Cour constitutionnelle d'accepter la destitution » était une « bonne décision », et 81 % ont répondu qu'il fallait « punir sévèrement car il s'agit de violence illégale contre un organe de l'État, le tribunal ». Par conséquent, en termes d'attitudes face au décret de loi martiale et à la destitution, il est problématique de classer ces groupes comme étant d'extrême droite, du moins par rapport aux femmes de 20 ans. Pour les analyses récentes sur l'extrémisation des hommes de 20 ans, veuillez vous référer à Kuk Seung-min (2025), Kim Chang-hwan (2025), Jeon Hye-won (2025), Choi Young-jun et al. (2025), Jeong Han-wool (2025). Alors que Kim Chang-hwan, Jeon Hye-won, Choi Young-jun et al. reconnaissent généralement la possibilité d'une orientation conservatrice ou d'extrême droite chez les hommes de 20 ans, Kuk Seung-min (2025) et Jeong Han-wool (2025) soulignent qu'il est déraisonnable ou exagéré de diagnostiquer ces groupes comme des groupes d'extrême droite sur la base de leurs attitudes envers la démocratie, le décret de loi martiale et la destitution.

[Graphique 9] Attitudes face à l'acceptation de la destitution par la Cour constitutionnelle par génération (%), [Graphique 10] Traitement des incidents du tribunal de district ouest et des personnes arrêtées par génération (%)

Source : Progressive Policy Institute·Korea People's Research Institute·Korea Research <2e Enquête sur l'agenda national> (2-4 mai 2025)

[Tableau 5] Attitudes face à la destitution par la Cour constitutionnelle par sexe et par génération (%), [Tableau 6] Traitement des incidents du tribunal de district ouest par sexe et par génération (%)

Source : Institut d'études progressistes·Institut d'études coréennes·Korean Research <Enquête sur les tendances électorales de la 21e élection présidentielle> (2 mai - 4 mai 2025)

2. Facteurs de changement à long terme : Réalignement idéologique ou effet de cohorte ?

Dans l'enquête <Analyse des tendances électorales de la 21e élection présidentielle> de l'Institut d'études sur l'Asie de l'Est, en ventilant les tranches d'âge par intervalles de 5 ans et en calculant la moyenne de leur score idéologique subjectif, nous constatons que les générations 20-30 se rapprochent du centre avec 5,0 à 5,1, tandis que les générations 40-50 sont légèrement plus progressistes avec 4,5 à 4,7. Pour les 70 ans et plus, la première moitié (jusqu'à 74 ans) est de 6,1, et les personnes âgées de 75 ans et plus sont de 6,3, montrant une inclinaison nette vers le conservatisme. Ce qui est intéressant, ce sont les 60 ans et plus qui retiennent notre attention. Jusqu'à la première moitié des 60 ans (60-64 ans), qui constituent le groupe principal de la génération 386, le score est de 0,49, se rapprochant du centre progressiste. En revanche, la seconde moitié des 60 ans (65-69 ans) a un score de 5,6, montrant une orientation conservatrice relativement plus forte que la première moitié. Ces résultats suggèrent que le changement de la génération des 60 ans et plus, qui était auparavant un pilier de l'alliance conservatrice, a été largement influencé par la première moitié des 60 ans et plus, dont la tendance vers le centre progressiste s'est renforcée.

Quelle est la raison de la tendance relativement centriste et progressiste de la première moitié des 60 ans, par rapport à la tendance relativement conservatrice de la seconde moitié des 60 ans et des 50 ans ? Une hypothèse est qu'il pourrait s'agir d'un phénomène survenu avec l'entrée de la génération relativement progressiste « 386 » dans la soixantaine. Compte tenu que l'aîné de la génération 386 (né en 1960) a 65 ans, cela peut être considéré comme le reflet de la différence idéologique entre la génération née dans les années 1950, juste avant la génération 386, et la génération 386. Si nous définissons par convention les personnes nées dans les années 1960 comme la génération 386, elles s'étendent de 56 ans (né en 1969) à 65 ans (né en 1960), ce qui signifie qu'environ 60 % des personnes dans la soixantaine (60-65 ans sur 60-69 ans) sont de la génération 386. En fait, le [Tableau 7] montre que parmi les répondants dans la soixantaine, 37 % sont nés dans les années 1950 et 63 % sont nés dans les années 1960, c'est-à-dire classés comme « génération 386 ». Dans seulement quatre ans, les personnes dans la soixantaine en Corée seront entièrement remplacées par la génération 386.

[Figure 11] Score moyen de la tendance idéologique subjective par tranche d'âge (0 point très progressiste - 5 points centre - 10 points très conservateur)

Source : EAI·Korean Research <Analyse des tendances électorales de la 21e élection présidentielle>

[Tableau 7] Taille des cohortes de naissance selon notre enquête

Source : EAI·Korea Research <Analyse du vote pour la 21e élection présidentielle>

Bien sûr, certains soutiennent que si l'on analyse le groupe des 386 en détail, il n'est pas aussi progressiste qu'on le croit communément, et qu'il n'a pas de tendances aussi progressistes (vote présidentiel, auto-évaluation idéologique, etc.) (Bae Jin-seok 2017, 2022). Parallèlement, il est également vrai que des recherches ont été constamment soulevées, démontrant que, du moins en ce qui concerne la politique électorale (vote présidentiel, auto-évaluation idéologique, etc.), le soi-disant « effet de la génération 386 » est confirmé de manière globale ou conditionnelle, car ils préfèrent les candidats ou les partis anti-conservateurs et montrent une progressivité relative sur des questions telles que le bien-être ou les relations avec la Corée du Nord, par rapport aux personnes nées dans les années 1940 ou 1950 qui ont grandi à l'ère de l'autoritarisme (Oh Se-jae·Lee Hyun-woo 2014 ; Noh Hwan-hee et al. 2013).[6]

Examinons d'abord brièvement les résultats des deux dernières élections présidentielles du groupe des 386 (nés dans les années 1960). En général, à l'exception de 2007 où le soutien aux candidats conservateurs était prédominant, il y avait une relation linéaire où plus on était jeune, plus la tendance démocrate/progressiste était forte, et plus on était âgé, plus on soutenait les candidats conservateurs. Cependant, depuis l'élection de 2017, il est connu qu'un schéma en U s'est progressivement formé, où les personnes âgées de 40 et 50 ans montrent le comportement de vote le plus progressiste. En termes de taux de soutien aux candidats du Parti Démocrate ou des partis progressistes lors des élections présidentielles ou d'autres élections, les 386 ne sont pas les plus progressistes, et les personnes nées dans les années 1970 représentent le pic du vote pro-démocrate/pro-progressiste (Bae Jin-seok 2017 ; Jeong Han-wool 2020).

En fait, les résultats de l'enquête de l'EAI menée cette fois-ci montrent que, lors de la 20e élection présidentielle en 2022 et de la 21e élection présidentielle en 2025, les 386, contrairement aux générations nées dans les années 1940 et 1950 qui soutenaient massivement Yoon Suk-yeol (61-76%) et Kim Moon-soo (55-71%), montrent une progressivité relative, le soutien au candidat Lee Jae-myung du Parti Démocrate étant prédominant (45% pour les 20 ans, 54% pour la 21e élection présidentielle) par rapport au soutien à Yoon Suk-yeol (35%) ou Kim Moon-soo (32%), Lee Jun-seok (3%). Cependant, par rapport aux personnes nées dans les années 1970 et 1980, leur comportement de vote pro-démocrate/pro-progressiste est relativement plus faible. Les personnes nées dans les années 1970 ont vu leur soutien à Lee Jae-myung se renforcer lors de la 20e élection présidentielle (Lee Jae-myung 45%, Yoon Suk-yeol 35%) et de la 21e élection présidentielle (Lee Jae-myung 60%, Kim Moon-soo 22%, Lee Jun-seok 5%). Les personnes nées dans les années 1980 ont également montré un soutien écrasant à Lee Jae-myung lors de la 20e élection présidentielle (Lee Jae-myung 46% contre Yoon Suk-yeol 20%) et de la 21e élection présidentielle (Lee Jae-myung 56%, Kim Moon-soo 21%, Lee Jun-seok 4%). Elles peuvent être considérées comme la génération la plus opposée au Parti du Pouvoir du Peuple. Chez les personnes nées dans les années 1990 et 2000, les taux de soutien à Lee Jae-myung et aux candidats conservateurs étaient serrés, et en particulier, le taux de soutien à Lee Jun-seok lors de la 21e élection présidentielle était de 13-20%, montrant une différence nette par rapport aux générations nées dans les années 1970-1980.

Ces résultats confirment ce que j'ai souligné dans mon article de 2013 : alors que les générations jeunes et d'âge moyen (nés entre 1963-1967 et après 1968-1972, le groupe principal des 386) ont montré un effet de cohorte clair, se concentrant brièvement sur le candidat conservateur Lee Myung-bak lors des 16e, 17e et 18e élections présidentielles avant de revenir à une préférence pour le candidat du Parti Démocrate Moon Jae-in, les personnes nées avant 1958-1962, qui avaient alors plus de 50 ans, ont montré un effet d'âge renforcé en faveur des candidats conservateurs (Lee Nae-young·Jeong Han-wool 2013). En examinant le comportement de vote depuis lors, le groupe des 386 a montré un comportement de vote en tant que cohorte relativement pro-progressiste/pro-démocrate, en soutenant massivement Moon Jae-in lors de l'élection présidentielle de 2017 avec le jugement de destitution, et en maintenant une préférence pour le candidat du Parti Démocrate Lee Jae-myung lors des élections présidentielles de 2022 et 2025.

[Tableau 8] Résultats des 20e et 21e élections présidentielles par cohorte de naissance (%)

Indeed, in this survey, when we look at the average subjective ideological score not only by age group but also by year of birth, those born before 1940 (76 years and older) show strong conservatism with an average score of 6.7, and those born in the 1950s (66-75 years old), who can be called the 'post-war industrialization generation,' are clearly conservative with a score of 5.9 [Figure 12]. Meanwhile, the so-called '386 generation' (born in the 1960s), corresponding to those aged 56-65, show a progressive tendency close to the center with a score of 4.9. Those born in the 1970s (46-55 years old) are the most progressive with a score of 4.6, and those born in the 1980s (36-45 years old) also confirm a progressive tendency with a score of 4.8. Those born in the 1990s show a centrist-conservative tendency with a score of 5.2, and those born in the 2000s up to 2007 show progressiveness with a score of 4.8 [Figure 13].

[Figure 12] Subjective Ideological Tendency Score by Age Group (Points), [Figure 13] Subjective Ideological Tendency Score by Birth Year (Points)

Source: EAI·Korea Research <Analysis of Voting Behavior in the 21st Presidential Election>, Source: EAI·Korea Research <Analysis of Voting Behavior in the 21st Presidential Election>

[Figure 14] shows how the subjective ideology distribution of birth cohorts, divided into 5-year birth year groups for each age group based on the 2002 election, has changed over the course of six presidential elections from 2002 to 2025 as people age.[7]This shows the trend of changes in subjective ideological evaluation scores for each birth cohort over the past six presidential elections, spanning 23 years (excluding two years due to early elections). An upward trend in the graph suggests a conservative shift with age, a downward trend suggests a progressive shift with age, and a horizontal movement suggests a cohort effect where political attitudes from youth are sustained.

The survey results show, first, that generations prior to the '386 generation' we focus on, namely those born in the 1940s and 1950s, tend to become more conservative with age (upward trend). While the trend of change, which cannot strictly distinguish between age, cohort, and period effects, is a combined result of age and period effects, we can infer that the age effect is relatively at play compared to other generations (Lee Nae-young·Jeong Han-wool 2013).

Second, looking at the 'born 1958-1962', 'born 1963-1967', and 'born 1968-1972' birth cohorts, which are close to the 386 generation we are interested in, unlike the previous generations, they also show some conservatism until 2007 and 2012 (upward movement), but after 2017, they decline again, returning to their ideological positions from 2002. While there are fluctuations due to life cycle and period effects, the characteristic of a 'cohort effect' is evident in that their political attitudes from youth are generally maintained and they return to those positions. However, it is noteworthy that in the 2002 survey results, the subjective ideological position of the 386 generation was relatively progressive (between -0.5 and 0 points) compared to the generations above them, born in the 1940s and 1950s, who were at +0.5 to +1.3, but they were generally located near the center, between -0.2 and +0.5.[8]

Meanwhile, generations after the 386 generation, born in the 1970s to 1990s, started from a progressive position (-) compared to the 386 generation during their 20s and have generally maintained their progressive tendencies to this day. However, it is observed that recent cohorts born in the 2000s tend to start from a relatively centrist-conservative position from their early twenties, influenced by the conservatism of the male population.

The history of Korean elections is still short, and there is insufficient data to explain the changes in voting behavior of each generational cohort. Therefore, it is difficult to definitively predict whether the tendency for those born in the 1940s-1950s to become more conservative with age will continue, whether the 386 generation, which has shown cohort effect characteristics so far, will shift to a conservative path like the older generations, whether the progressiveness of those born in the 1970s and 1980s, who showed more progressive attitudes (in subjective ideology and voting choices) than the 386 generation, will be maintained, and how the conservatism of those born after the 2000s will change.

[Figure 14] Changes in Subjective Ideological Tendencies of Birth Cohorts over 23 Years of Presidential Elections (Points)

Source: Jeong Han-wool (2020, p89 Figure 9), EAI Election Polls (2002; 2025), EAI Panel Surveys (2007; 2012; 2017; 2022) combined

3. Factors Expanding Gender Voting: Period Effect of Martial Law and Impeachment Issues

As examined earlier, it has been confirmed that the gender voting gap is primarily spreading from the 20s to the 30s. Why did this change occur? Above all, feminist conflicts are spreading across all generations, and it is presumed to be related to the phenomenon of feminist conflicts overlapping with ideological conflicts. First, [Figure 15] shows the trend of changes in the average scores of men and women by generation using a feminism index (from -12 points for very anti-feminist to +12 points for very pro-feminist) based on six questions developed by the author through the '20s Male Project' in 2019.

While the perception gap between men and women in their 20s, where feminist conflicts were concentrated, is becoming entrenched without change, the tracking surveys conducted annually show that the gender perception gap in feminism is spreading beyond the 20s to the 30s, 40s, and 50s. In particular, the perception gap in the 30s was comparable to that of the 20s, and this perception gap is presumed to have influenced political attitudes. Cheon Gwan-yul described this as the gender conflict between men and women in their early 20s ('20s men') and women in their early 20s ('20s women') spreading 'at the intersection of power and gender,' and Guk Seung-min pointed out the phenomenon of gender conflict overlapping with ideological conflict (Cheon Gwan-yul·Jeong Han-wool 2019; Guk Seung-min 2022). [Figure 16] is a graph drawn by Guk Seung-min, synthesizing objective indicators from the Sisain·Korea Research survey '<20s Women>' to show the correlation between the progressive ideology index and the feminism index by generation. In all generations, the feminism index strengthens progressive attitudes. [Figure 17] visualizes the relationship between the subjective ideology score and the feminism index using recent data from the Progressive Policy Institute's National Agenda Survey. In Guk Seung-min's graph, the overlap between feminism and progressive ideology is a phenomenon across all generations, but in terms of slope (correlation strength), the 20s and 30s show a greater influence than those aged 40 and above. Based on subjective ideology, the relationship between feminism and ideological identity is prominent mainly in the 20s and 30s, and has not spread to the 40s and 50s or 60s and 70s.

In conclusion, this implies that gender conflict (feminist conflict) is connecting to political and ideological divides. As gender conflict (feminist conflict) spreads across all generations and the phenomenon of gender conflict overlapping with ideological conflict intensifies, it is a natural prediction that this will expand into phenomena where political behavior clashes based on gender. What is important here is that mobilization by the political sphere and triggers from the media, SNS, and online communities appear to be necessary conditions. If this hypothesis is correct, while the gender voting phenomenon was concentrated in the 20s and 30s in this presidential election, the possibility of a gender voting gap spreading to the 40s and 50s cannot be ruled out.

[Figure 15] Attitudes toward Feminism by Gender and Generation

Source: Korea Institute for Social Research·Korea Research Political and Social Survey DB (2019-2025)

[Figure 16] Feminism Index and Progressive Ideology Index, [Figure 17] Feminism Index and Subjective Ideology Index

Source: Guk Seung-min (2022), Source: Progressive Policy Institute (2024)

VI. Exploratory Discussion: The Roots of Pro-Democratic Tendencies in the 40s-50s and Conservatism in the 70s

1. Cohort Classification Based on Political Symbols (Presidents with whom people feel a sense of unity)

Due to the short history of democratization and political parties, attempts to capture the characteristics of political cohorts have been insufficient. While the characteristics of each generation's political tendencies and behaviors are analyzed, research on why and how these generational characteristics arise has been lacking. The author estimates that the historical experiences during the formative period of political attitudes, roughly between the ages of 18 and 25, and the emotional attachments formed at that time, solidify into a unique 'generational identity' and influence long-term political preferences and behaviors. Existing research suggests that shared historical events and the first voting experiences during early adulthood (first elections effect) during the impressionable years lead to the formation of values and orientations that differentiate them from other generational groups (Alwin and Krosnick 1991; Mannheim 1997; Pilcher 1994). For example, generational cohorts are sometimes named based on the historical events and zeitgeist of their formative period, such as the 'Korean War generation,' 'industrialization generation,' and '386 generation.'

However, there has been insufficient exploration into whether these generational cohorts possess an identity as homogeneous political cohorts or what factors create such an identity (Lee Nae-young·Jeong Han-wool 2013; Jeong Han-wool 2020). The author proposes that in Korea, as an analytical indicator for understanding the characteristics of political cohorts, the president voted for during the early period of political socialization should be considered as the object of political identification, and the orientation of political cohorts should be captured by focusing on attitudes toward past presidents. In the US, the generation that voted for Roosevelt during the New Deal and Great Depression is understood as the 'New Deal Democratic' cohort. In Korea, the tendency is for political preferences and behaviors to be expressed through identification with the political symbol that dominated one's formative period of political attitudes, primarily the president voted for or experienced during youth. Therefore, the author suggests using the president (symbol) who governed during that ideological period.

Pour les personnes nées dans les années 1940 et 1950, il n'est pas exagéré de dire que le « Président Park Chung-hee » était le symbole politique qui définissait leur expérience historique et leur attitude politique lors des processus de reconstruction d'après-guerre et d'industrialisation. Bien qu'il ne soit pas utilisé dans un contexte académique, le terme « génération Park Chung-hee » serait plus approprié pour décrire leur attitude politique que des désignations socio-économiques telles que génération d'après-guerre ou génération de l'industrialisation. De plus, je pense que le Président Roh Moo-hyun est le symbole de la cohorte politique des 40-50 ans actuelle. Non seulement la majorité de cette génération a soutenu le Président Roh lors des premières élections présidentielles de leur socialisation politique, mais en tant que génération qui a réalisé la réforme politique représentée par l'anti-dictature et l'anti-régionalisme à travers le symbole de Roh Moo-hyun, elle a créé un élan de « politique générationnelle » qui a remplacé le régionalisme national destructeur. Le Président Roh Moo-hyun était le symbole et le point central de cette politique. Bien qu'il n'ait pas non plus été largement utilisé dans le milieu universitaire, le terme « génération Roh Moo-hyun » a été fréquemment utilisé dans la réalité pour désigner ceux qui correspondent à la génération Park Chung-hee.

Si le Président Park Chung-hee symbolise le régime anticommuniste de la Guerre Froide et l'ère de l'industrialisation partagés par les générations nées dans les années 1940 et 1950, le Président Roh Moo-hyun, après le changement de régime, est une génération qui a (1) fait l'expérience d'une génération et d'une politique basées sur l'idéologie qui ont remplacé le régionalisme destructeur (Kang Won-taek 2002 ; Lee Nae-young 2002), (2) s'est familiarisée avec de nouvelles formes de participation telles que la politique sur Internet et la culture de la Coupe du Monde, qualifiée de gouvernement participatif (Yoon Sung-yi 2003), et (3) a partagé une expérience historique organisationnelle par le biais de campagnes politiques menées par des groupes comme Nohsamo et des forums publics sur Internet (Cho Hwa-soon 2008 ; Yoon Yong-hee 2003). Compte tenu de ces caractéristiques, je propose de classer la génération actuelle des 40-50 ans (qui était la génération des 20-30 ans avant l'élection présidentielle de 2002) comme la « génération Roh Moo-hyun / cohorte Roh Moo-hyun ».

2. Political Symbols Representing Each Generation: "Park Chung-hee Generation" vs. "Roh Moo-hyun Generation"

First, we will examine whether former Presidents Park Chung-hee and Roh Moo-hyun actually function as political symbols that foster generational unity by looking at the favorability ratings (psychological attachment) each generation has toward the presidents of each camp. Let's look at the results of the favorability survey toward past presidents included in the '2nd National Agenda Survey' conducted by the Progressive Policy Institute·Korea Institute for Social Research·Korea Research in May 2025. Based on the proportion who answered 'very + somewhat favorable' in [Figure 18], President Roh Moo-hyun of the progressive/Democratic Party received a favorability rating of 72%, followed by President Kim Dae-jung at 61%, and President Moon Jae-in at 46%. On the conservative side, President Park Chung-hee received 50%, President Kim Young-sam 44%, President Syngman Rhee, revered as the father of the nation by conservatives, 32%, and President Lee Myung-bak 30%. President Park Geun-hye, who was impeached, received 27%, and President Yoon Suk-yeol received 21%. Presidents Chun Doo-hwan and Roh Tae-woo, who were punished for coups, received favorability ratings of only 16-18%.

[Figure 18] Favorability Ratings of Past Presidents (%)

Source: Progressive Policy Institute·Korea Institute for Social Research·Korea Research <2nd National Agenda Survey> (May 2-4, 2025)

Let's examine the favorability ratings of past presidents by age group. The important factor is the extent to which each generation has favorability and attachment to the president during their period of political socialization. According to cohort theory, political experiences during the formative period of political attitudes and the first elections connect to identification with the respective political forces, shaping one's values and attitudes over the long term. Therefore, it is unlikely that a generation would accept the orientations shared by that group without positive evaluations and attachment to the political objects (in this case, the president and historical events of the time) that symbolize them.

Seen in this light, the attachment of the 40s-50s generation to President Roh Moo-hyun may suggest it as the source of the cohort effect influencing their pro-Democratic Party orientation and political behavior. For the 40s-50s generation, who underwent their early political socialization during President Roh Moo-hyun's term, the favorability rating for President Roh Moo-hyun was overwhelmingly high at 80-82%, and for the 20s-30s generation, it was 69-75%. Notably, even among those aged 70 and above, who are the most conservative, the favorability rating was 54%. It is noteworthy that among the 60s generation, who are deviating from the conservative coalition, the favorability rating for President Roh Moo-hyun was 69%, comparable to that for President Park Chung-hee. Following him was President Kim Dae-jung, with favorability ratings also high among the 40s-50s generation at 70-72%, and among the 20s-30s at 49-63%, among the 60s at 62%, and lower among those aged 70 and above at 46%.

Conversely, among conservative presidents, President Park Chung-hee is the political symbol and focal point. He received high favorability ratings from those aged 70 and above (75%), the 60s (67%), and the 50s (52%). However, among those in their 20s to 40s, the rating remained around 32-36%. Among the 20s-40s generation, President Kim Young-sam received relatively high favorability. For other conservative presidents, generally only those aged 70 and above recorded relatively high favorability ratings, and except for President Park Chung-hee, there were no conservative presidents with high favorability among the 20s-30s, 40s-50s, or even the 60s. The favorability ratings for Presidents Chun Doo-hwan and Roh Tae-woo, who were leaders of the coup, were only around 34-36% even among those aged 70 and above, and between 10-22% for other age groups.

[Table 9] Favorability Ratings of Past Presidents by Age Group "Very + Somewhat Favorable" Rate (%)

Source: East Asia Institute, Korea Institute for National Studies, and Korea Research <2nd National Agenda Survey> (May 2-4, 2025).

What is noteworthy is that, as shown in [Figure 19], before the impeachment of President Park Geun-hye in 2016, President Park Chung-hee showed similar levels of emotional connection and favorability among citizens as President Roh Moo-hyun. This is a result of the decreased respect and attachment towards the conservative forces due to the aftermath of President Park Geun-hye's impeachment in 2017. Although it has been gradually improving since then, it appears to be a result of the low favorability among the 2040 young generation. If the recent weakening of the conservative support base (a representative example being the departure of the 60s from the political conservative coalition) relies solely on the impact of short-term issues like martial law and impeachment (after crossing the so-called river of impeachment), it would be reasonable to view the existing conservative political orientation as a short-term realignment that can be restored. However, if, as I hypothesize, the favorability of past presidents reflects the influence of political cohorts in determining the political attitudes and behaviors of each political cohort, then the weakening of the conservative base and the departure of supporters should be considered a phenomenon that will continue for a considerable period.

In other words, the 40s and 50s generation, represented by Roh Moo-hyun and Kim Dae-jung, still have emotional attachment to these two political symbols, and conversely, do not feel a sense of unity with conservative presidents. On the other hand, among the 60s and 70s, a sense of unity with conservative presidents only exists among those aged 70 and above. For the 60s generation, the favorability towards Presidents Roh Moo-hyun and Kim Dae-jung is comparable to that towards Presidents Park Chung-hee and Kim Young-sam. This implies that the 60s generation can no longer serve as a strong base for the conservative camp as it did in the past.

[Figure 19] Changes in Favorability (%) towards Political Symbols of Progressivism and Conservatism

Source: East Asia Institute DB (2012-2016), Korea Institute for National Studies · Korea Research DB (2021-2025).

V. Conclusion

Attention is focused on the '20-something male' and '20-something female' phenomenon in generational voting, and the country is quickly normalizing after putting aside concerns about political civil war that were feared after martial law and impeachment. However, although the conservative party performed relatively well in the presidential election, its support base is significantly shaken in terms of party support rates and various political indicators after the election. In fact, the crisis of conservative politics is evident in the generational voting patterns of the 21st presidential election.

In the presidential election held after martial law and impeachment, candidate Kim Moon-soo performed well with 41%, and if we add candidate Lee Jun-seok's 8%, it seems to have been a close contest with candidate Lee Jae-myung. This illusion has accelerated the tendency of the People Power Party to rely more on 'Yoon Again' and asphalt conservatism that opposes impeachment, rather than on self-reflection, introspection, and reform for conservative innovation. However, as examined in this report, the relative success of the conservative candidate in the last presidential election was not a foregone conclusion until before martial law. It was merely the result of a conservative mobilization based on antipathy towards candidate Lee Jae-myung's judicial risks and anti-Democratic Party sentiment; the winner would not have changed. Rather, the conservative support base has been rapidly collapsing since the presidential election results. Despite the setbacks in the appointment process of the Lee Jae-myung administration and the Democratic Party, the president's approval rating remains high, and there has been no significant blow to the support for the Democratic Party, its political base. Above all, the People Power Party's approval rating has fallen to the 10% range. Just as the Democratic Party achieved overwhelming victories in the 2017 presidential election, 2018 local elections, and 2020 general elections with the power of the impeachment voter political coalition formed after the first impeachment in 2016-2017, a political structure dominated by the Democratic Party has once again been formed through martial law and impeachment.

A more fatal crisis factor for conservative parties is, first, that the 40s and 50s generation, which served as the base for the second impeachment voter political coalition, has become much more solidified in their pro-Democratic Party and anti-conservative party sentiment through martial law and impeachment. Support for the Democratic Party candidate in the presidential vote has strengthened, and their attitude against the People Power Party has also solidified. The fact that the core support base of the new government is consolidated suggests that the current Democratic Party-dominated structure can be maintained quite stably. Although raised as a theoretical argument later in this paper, the 40s and 50s generation, centered around the political symbol of "Roh Moo-hyun," shares generational identity as a group that is more homogeneous and action-oriented than any other political cohort, based on their shared experiences since the change of government: (1) the era's task of political reform and the end of authoritarian mobilization and the old regional mobilization politics of the three Kims; (2) the emergence of a new generation represented by internet/World Cup culture; and (3) new political participation experiences through Roh-mo-sa/internet public forums. This generation has supported the current regime change by the Democratic Party and is strongly supporting the state affairs of the Lee Jae-myung government.

Second, it is a painful blow to the People Power Party that the 60s generation, which was unwavering during the first impeachment in 2016-2017, is now wavering. The conservative coalition of the 60s and 70s generation, which was as strong as the 40s and 50s, served as the base that enabled the change of government five years after the first impeachment. A significant finding of this report is the confirmation of a crack in the strong 60s-70s conservative coalition through martial law and impeachment. The 60s generation, which overwhelmingly supported the People Power Party in the election three years ago, has now become a competitive generation similar to the 20s and 30s generation in this election. The fact that this change is based on a weakening of party support and subjective conservative identity suggests that the departure of support from the 60s generation may not be a temporary phenomenon. The current 60s generation is distinctly different from the previous 60s generation. The change in the 60s generation is also difficult to view as a temporary phenomenon, as it is the result of the so-called 386 generation, which has a relatively strong pro-Democratic Party/progressive tendency, aging and replacing the main force of the 60s generation. The 386 cohort, born in the 1960s, experienced a concentrated age effect where conservative tendencies strengthened as those born in the 1940s and 50s entered their 60s. In contrast, the 386 generation, which is currently replacing the 60s generation, unlike the previous generation, has been identified as a political cohort that has maintained a progressive tendency through six presidential elections from 2002 to 2025.

Third, this chapter pointed out that the gender voting cleavage, which was prominent in the last presidential election, has spread to the 30s generation and has the potential to expand to the 40s and 50s generation. It has been confirmed in various studies that the process of expanding gender politics in Korea has increased political interest and influence as gender conflict surrounding feminism overlaps with political and ideological cleavages. The author intended to point out that the conflict surrounding feminism is intensifying not only among the 20s but also among the 30s and 40s, and even among the 50s, and that gender conflict politics can spread to all generations if certain conditions are met. Observing the process of gender politics spreading throughout society in 2022, it required enormous social costs due to the combination of mobilization by political circles and the media, and an environment that triggers gender conflict in online spaces represented by SNS and various communities. Although gender conflict mobilization seemed to subside and lie dormant during the Yoon Suk-yeol administration, it was confirmed that the storm of gender politics has once again spread to the 30s generation in this presidential election. Regardless of political gains and losses, the intense heat of social conflict and hate politics caused by gender conflict over the past 5-6 years has indeed had a shock that has overshadowed other important social conflicts or discrimination issues in Korean society.

The changes in generational political behavior in Korea discussed above are important research questions that need empirical verification. This chapter focused on sharing my own hypothetical interpretations and opinions on these questions using survey data. I do not believe that the discussions in this chapter have sufficiently verified my concerns and arguments. Rather, I focused on raising an agenda for discussion, and further data collection and more in-depth analysis and interpretation will be necessary for the discussions in this chapter to gain credibility. Although the discussion is currently limited to a theoretical level, given the importance of these questions, follow-up research and verification cannot be postponed for long. While future research on the questions and hypotheses raised in this chapter will be deferred, I hope that the political circles will not postpone their interest and consideration of the implications of generational politics raised in this chapter.

The strong cohesion of the 40s and 50s generation and the collapse of one pillar of conservative support will primarily pose a significant crisis and challenge for conservative parties. The crisis of conservatism, brought about by martial law and impeachment that led to the departure of conservatives, is severe, yet the People Power Party's leadership election raises questions about whether it recognizes these crisis factors. Consequently, the normalization of conservative politics is bound to be delayed, and if conservative politics is not normalized, the normalization of Korean politics will also be difficult. Meanwhile, for the Lee Jae-myung government and the Democratic Party, which must lead the stability and normalization of state affairs beyond martial law and impeachment, the current situation is not entirely optimistic. It is hoped that the favorable political environment will not lead to arrogance and unilateralism by the Lee Jae-myung government and the ruling party. National challenges such as finding a way forward for South Korea amidst US-China competition, seeking national competitiveness in the AI era, low birth rates-aging population, and regional extinction, which are of concern to the majority of the public, are also serious. If the focus shifts to partisan tasks, neglecting the national tasks that concern the majority of the public, even high approval ratings that seemed likely to last for over 20 years can disappear in an instant. A time for learning from the mistakes of the Moon Jae-in administration seems necessary.■

VII. References

Kim Ki-dong and Lee Jae-mok. 2021. “Partisan Identity and Affective Polarization of Korean Voters.” Korean Political Science Review 55(2): 57–87.

Kook Seung-min. 2025. “Extreme Right-Wing Radicalization of Men in Their 20s? Insufficient Evidence.” Sisa IN 929 (June 26, 2025).

______. 2022. “Gender Conflict That Will Change the Political Landscape of Korea.” In *Women in Their 20s*. Sisa IN Books.

______. 2025a. “‘China Interferes in Korean Politics.’ The More You Believe It, the More Anti-Democratic It Is.” Sisa IN 911 (March 5, 2025).

Kook Seung-min, Kim Da-eun, Kim Eun-ji, and Jeong Han-ul. 2022. *Women in Their 20s*. Seoul: Sisa IN Books.

Kim Chang-hwan. 2025. “Why Have Young Men Become Conservative?” Sisa IN Issue 930 (July 4, 2025).

Noh Hwan-hee, Song Jeong-min, and Kang Won-taek. 2013. “Generational Effects in Korean Elections.” *Korean Party Studies Review* 12(1): 113–39.

Park Sun-kyung. 2022. “Populism Driven by Economic Inequality? An Analysis of Perceptions of Economic Inequality and Populist Tendencies among the Economically Vulnerable.” *Journal of 21st Century Political Science* 32(1): 1–23.

Park Won-ho. 2013. “The Transformation of Generational Theory: The 18th Presidential Election and Generations.” In *The Choice of Korean Voters 2: The 18th Presidential Election*, edited by Park Chan-wook, Kim Ji-yoon, and Woo Jeong-yeop. Seoul: Asan Institute for Policy Studies.

Park Jae-heung. 2009. “A Critical Review of Generational Labels and Generational Conflict Discourse.” *Economy and Society*, Spring Issue 81.

Park Jeong-ui, Yoon Seong-tae, and Seong Ji-yeon. 2020. “Social Dominance Orientation and Multicultural Acceptance: A Focus on University Students.” *Multiculturalism and Peace* 14(2): 160–183.

_____. 2001. “Theoretical and Methodological Issues in Generational Research.” *Korean Journal of Population Studies* 24(2).

Bae Jin-seok. 2022. 2020. “The 86 Generation and the End of Generational Effects: Analysis of Presidential Elections from 1992-2022.” In *Changing Korean Voters 7: The 2022 Presidential Election and Korean Politics*, edited by Kang Won-taek. Seoul: East Asia Institute.

_____. 2017. 2020. “Generational Factors in Voting Choice and Ideological Orientation: Analysis of Presidential Elections from 1992 to 2017.” In *Changing Korean Voters 6: The Candlelight Protests, Impeachment Crisis, and the 19th Presidential Election*, edited by Kang Won-taek. Seoul: East Asia Institute.

Byun Jin-kyung. 2025. “Japchae to Share with Someone [Editor’s Letter].” Sisa IN 910 (February 17, 2025).

Seo Bu-won. 2025. “To Those Who Say the Extreme Right-Wing Radicalization of Boys' High Schools is Exaggerated: [Children Are My Teachers] School Hierarchy, Racial Discrimination, Anti-Feminism... The Dangerous Values of Male Students.” Ohmynews (July 5, 2025).

Shin Jin-wook. 2025. “The ‘Extreme Right’ Issue Must Be Viewed Correctly [Shin Jin-wook’s Perspective].” Hankyoreh (July 2, 2025).

Oh Se-jae and Lee Hyun-woo. 2014. “Conditional Generational Effects of the 386 Generation: Focusing on Ideological Orientation and Presidential Voting.” *Journal of Parliamentary Research* 20(1): 199–234.

Yoon Sung-ee and Lee Min-kyu. 2014. “Comparison of Generational Characteristics of Ideological Conflict in Korean Society: Focusing on Determinants of Subjective Ideology and Ideological Self-Placement.” *Journal of 21st Century Political Science* 24(3): 272–292.

Lee Nae-young and Jeong Han-ul. 2013. “Components of Generational Cleavage: Cohort Effects and Age Effects.” *Journal of Parliamentary Research* 19(3): 39–83.

Lee Sang-shin, Min Tae-eun, Yoon Kwang-il, Gu Bon-sang, and Peter Gries. 2020. *KINU Unification Consciousness Survey 2020: Comparative Study of Perceptions of Neighboring Countries*. Seoul: Seoul National University Institute for Peace and Unification Studies.

Jeon Sang-jin. 2004. “The Overuse of the Concept of Generation, the Poverty of Generational Research: A Review of Generational Research Methods.” *Korean Sociological Association* 36(5): 31–52.

Jeon Hye-won. 2025. “The Higher the Economic Status in Seoul, the Higher the Probability of Being an Extreme Right-Wing Youth.” Sisa IN 928 (July 2, 2025).

Cho Hwa-soon. 2008. *Internet Networks and Political Participation of the Online Generation*

Jeong Han-ul. 2020. “Trends and Prospects in Generational Voting Research and the 21st General Election.” *Trends and Prospects* 109: 74–96.

Cheon Gwan-yul and Jeong Han-ul. 2019. *Men in Their 20s*. Seoul: Sisa IN Books.

Choi Young-jun, Kim Min-jeong, Oh Seo-eun, and Kwon Tae-jin. 2025b. “Reading the Social and Political Extreme Right-Wing Tendencies in Korea: Generational Tendencies and Social Factors.” Keynote presentation at the Seminar on Diagnosing and Prospecting Extreme Polarization in Korean Society and Politics.

Heo Seok-jae. 2014. “Changes in Party Identification in Korea: Generational Replacement or Life Cycle?” Korean Political Science Review 13(1): 65–93.

Alwin, Duane F. and Krosnick, Jon A. 1991. “Aging, Cohorts, and the Stability of Sociopolitical Orientations Over the Life Span.” American Journal of Sociology 97(1). 169–195.

Bhatti, Yosef and Kasper M. Hansen. 2012. “The Effect of Generation and Age on Turnout to the European Parliament-How Turnout Will Continue to Decline in the Future.” Electoral Studies 31: 262–272.

Koo, Bon-sang, Kwang-Il Yoon, and Sang Sin Lee. “Beyond generational discourse: An age-period-cohort analysis of South Koreans’ attitudes toward reunification with North Korea..” The Social Science Journal https://doi.org/10.1080/03623319.2024.2334888

Mannheim, Karl. 1997. “The Problem of Generations.” in Collected Works of Karl Mannheim 5. London: Routledge: 276–322.

Pilcher, Jane. 1994. “Mannheim’s Sociology of Generations: An Undervalued Legacy.” The British Journal of Sociology. 45(3): 481–495.

Tilley, James and Geoffery Evance. 2013. “Ageing and Generational Effects on Vote Choice: Combining Cross-sectional and Panel Data to Estimate APC Effects.” Electoral Studies 33: 19–27.

[1]The difference between Candidate A’s support rate among men in their 20s and Candidate A’s support rate among women in their 20s represents the gender gap in Candidate A’s support rate. (+) indicates that men’s support rate was higher than women’s, and (-) indicates that women’s support rate was higher.

[2]As the difference in voting propensity between men and women among those in their 20s is discussed in a separate chapter, this chapter focuses on the fact that the difference in voting propensity between genders has expanded beyond the 20s to the 30s, compared to the last presidential election, but there is still no gender difference among those in their 40s and 50s, which is also noteworthy in the voting behavior of the 40s and 50s [Figure 2].

[3]The classification criteria for support rates between genders by age group in the exit polls of the three major broadcasters are inconsistent. Sometimes, the classification includes those over 60 years old (2020 Presidential Election), and sometimes the results for those over 70 years old are announced (2020 Presidential Election). As there is no raw data to reconcile this, the results as reported in the media are presented here.

[4]Cheon Gwan-yul described the conflict between men in their 20s and women in their 20s as occurring at the intersection of 'power (attitude toward political forces)' and 'gender (feminism)' (Cheon Gwan-yul and Jeong Han-ul 2019).

[5]Based solely on voters, excluding those who abstained or did not respond, Yoon Suk-yeol received 53% support and Lee Jae-myung received 44% support among those in their 60s.

[6]For a discussion on this, refer to Bae Jin-seok (2017) and Bae Jin-seok (2022). Due to the perfect linear dependency (Period = Age + Cohort) between cohort effects, age effects, and period effects, including the three variables of period (survey year), age (age), and cohort (birth year) leads to the 'identification problem,' where the independent effect of each variable cannot be precisely distinguished (Heo Seok-jae 2014; Lee Nae-young and Jeong Han-ul 2013; Bhatti and Hansen 2012). To resolve this, attempts such as imposing constraints or applying the Intrinsic Estimator (I.E.) or hierarchical APC models are necessary, which are beyond the scope of this chapter (Koo et al. 2024).

[7]To compare with previous data, the scale was adjusted from a 0-10 scale by subtracting 5, resulting in -5 (very progressive) to 0 (neutral) to +5 (very conservative) for calculating the average. The horizontal axis represents the respondent's age group in 5-year increments at the time of the survey. Since presidential elections occur every 5 years, the axis moves one step to the right with each term. For example, individuals born between 1963 and 1967, who were '35-39 years old' in the first election in 2002, would be '40-44 years old' in the second election in 2007, '45-49 years old' in 2012, '50-54 years old' in 2017, and '55-59 years old' in 2022. In the final election in 2025, they will be 3 years older, reaching '58-62 years old.' It is important not to confuse this with the fact that the graph's 5-year interval on the axis is arbitrarily adjusted, resulting in the data point appearing at the '60-64 years old' tick mark, which represents an age 5 years older. The interval between the fifth data point (2022 survey) and the last, sixth data point (2025 survey) for each birth cohort is 5 years on the graph, but it is actually 3 years.

[8]This suggests that the progressiveness of the '386 generation' should not be assumed a priori. Empirically, while they are relatively more progressive than the previous authoritarian industrialization generation, they are not particularly more progressive than the generations born in the 1970s or 1980s. This aligns with Bae Jin-seok's (2017; 2022) argument refuting the 'progressiveness of the 86 generation.' However, it is clear that they possess a distinctly progressive identity compared to the majority of the previous 60s generation, born in the 1940s and 1950s, and have consistently voted for progressive or pro-Democratic Party candidates, with a few exceptions (e.g., in the 2007 presidential election, a significant number defected to support Lee Myung-bak or Moon Kook-hyun, in a move to judge the participation government).

■ Author: Jeong Han-ul_Director of the East Asia Institute.

■ Editor: Lim Jae-hyun_EAI Research Fellow

    Contact: 02 2277 1683 (ext. 209) | jhim@eai.or.kr

Pièce jointe : 정한울_4050 친 민주당 투표성향의 뿌리_250904_EAI 워킹페이퍼.pdf