[The 21st Presidential Election and Korean Democracy: Crisis, Division, and Realignment] ② Political Orientations and Candidate Choices of Korean Youth

I. Introduction

Defining the youth demographic in Korean society is highly complex. According to dictionary definitions such as the Standard Korean Language Dictionary or Ourmal-saem, youth refers to 'all individuals in the stage of adolescence within society.' This tautological definition, which does not present specific characteristics of youth, only allows for understanding the youth demographic in contrast to middle-aged and elderly age groups. Consequently, Korean youth are categorized through various age ranges without a unified biological age definition for 'youth.' For instance, legally, the Youth Basic Act defines youth as ages 19 to 34, the Youth Employment Promotion Special Act as ages 15 to 29, and the Special Act on Support for Small and Medium Enterprise Human Resources as ages 15 to 34. Some local governments also consider individuals up to age 39 as youth through ordinances.

The situation is not significantly different in academic circles. Generally, there is a strong tendency to categorize age groups by decades based on a specific election period and consider those under 29, the youngest age group, as youth. Furthermore, this categorization of youth is becoming more confusing with recent attempts to develop a 'generational discourse' around young people. Concepts like Generation M, Generation Z, or the collective term 'MZ Generation' further complicate the understanding of who is included in the youth demographic.

Despite the imprecision in the conceptual definition of youth, this chapter aims to explore whether the political orientations and behaviors of Korean youth differ from those of middle-aged and elderly individuals by examining the processes of three South Korean presidential elections: 2017, 2022, and 2025. This research is important in two aspects, notwithstanding the terminological ambiguity. First, there is a growing interest in the political attitudes of Korean youth in recent years. The significant increase in youth voter turnout and political participation since the 2010s was expected to have a substantial impact on the changing landscape of Korean politics, and indeed, it has manifested in concrete outcomes (Lee Jeong-jin 2022; Cho Chang-deok 2022). Moreover, the internal division among youth, represented by '20-something males' and '20-something females,' which emerged in the late 2010s, had a significant effect in the April 2021 Seoul mayoral by-election and was observed to have a similar influence in the 2022 presidential election (Kim Eun-yi & Song Min-ho 2022).

Second, there is a suspicion that the political attitudes and behaviors of Korean youth in recent years differ significantly from those of the past, as well as from youth in Western countries. Generally, young people are known to reject authority or tradition (Gramsci 2011), oppose hegemony, and exhibit strong liberal tendencies (Mill 2015). Contemporary sociology and political science have frequently tested these tendencies among youth. Alwin and Krosnick (1991) confirmed through longitudinal data analysis that political attitudes become more conservative with age, and Glenn (1974) indicated that this conservatism is related not only to age but also to cohort differences. Recent studies on Generation M or Z in the United States and the United Kingdom also show that these theoretical predictions remain valid.

For example, the Pew Research Center in the United States (2018) categorizes Millennials as those born between 1981 and 1996, and Gen Z as those born from 1997 onwards, verifying that Millennials in the U.S. are the most liberal and democratic compared to other generations. Furthermore, the UK's National Centre for Social Research, in a 2024 report, defines Gen Z, born between 1997 and 2012 in the UK, as the first generation to grow up with the internet and be concerned about environmental and social issues. They are described as having strong liberal tendencies, opposing authority, favoring drug legalization and the abolition of the death penalty, and exhibiting strong progressivism with a high preference for social welfare (Lucas et al. 2024). In essence, current Western youth, Millennials and Gen Z, still align with the old motto that 'youth are progressive.' In contrast, recent Korean society is rife with perceptions of youth conservatism. Debates regarding the conservatism of Korean youth, particularly concerning national security and gender equality, are heated (Choi Jong-sok 2020). This study aims to expand the concrete evidence that complements these debates.

Despite the objectives and necessity of this study, there are significant methodological limitations in elucidating the characteristics of Korean youth by comparatively analyzing their political attitudes and behaviors longitudinally. This is also why this study is confined to an exploratory discussion. Firstly, analyzing the political attitudes and behaviors of a specific age group over a long period requires the accumulation of sufficient data. While this study has the advantage of comparing survey data from three presidential elections, the survey questions used in each period differ, thus limiting the variables available for comparative analysis. Secondly, analyzing a defined demographic group as youth presents challenges in identifying the independent influence of each variable due to the linear relationship between age, period, and cohort. Although methodological improvements such as hierarchical models have been proposed, they are also contingent upon the accumulation of sufficient data.

Therefore, this study attempts an analysis within the constraints of limited data that allow for a longitudinal examination of the changes in the political attitudes and behaviors of Korean youth. In this process, rather than attempting a priori age categorization of youth, it adopts an inductive reasoning approach to identify the age range and characteristics that define Korean youth by categorizing age groups in five-year intervals based on the 2017 presidential election and examining their political attitudes and behaviors observed across the three presidential elections.

II. Data and Variables

This study utilizes survey data collected by the East Asian Institute (EAI) immediately following the 2017, 2022, and 2025 presidential elections to analyze the political perceptions and behaviors of Koreans. The accumulation of reliable and abundant data over an extended period is essential for analyzing changes in political perceptions and behaviors over time. Unfortunately, the data used in this study only partially meets these requirements.

Firstly, the data used in this study is reliable. To accumulate reliable data, it is essential to collect it through methodologically similar processes. The post-presidential election surveys conducted by the East Asian Institute fulfill the methodological consistency requirement as they all utilized the same professional survey agency, Korea Research. On the other hand, the data used in this study is not abundant. The three post-presidential election studies conducted over an eight-year period focus on the specific issues of each time point, reflecting contemporary concerns, resulting in a scarcity of data continuously accumulated across the three periods.

Due to the aforementioned characteristics of the data used in this study, it ultimately analyzes three variables that appear in common across the three surveys conducted since 2017. These three variables are the type of media used to obtain election-related information, self-placement of ideological orientation, and the candidate chosen in each election. The survey questions used to measure each of these pieces of information are as follows. First, the question to ascertain the media through which respondents obtain election-related information was "Through which of the following channels do you obtain the most information related to the election?" This exact question was used in all three surveys. However, the response options differed: the 2022 survey added 'YouTube' and 'Election materials sent by the National Election Commission,' which were not present in the 2017 survey. The 2025 survey, compared to 2022, added 'Online communities using KakaoTalk, Telegram, and other online platforms.'

This study categorized respondents into 'traditional media or legacy media' users, 'new media' users, and other media users. The 'traditional media' user group included respondents who obtained election information through terrestrial TV, general programming TV, newspapers, and radio. The 'new media' user group included users of new media, reflecting the additions at each survey point, such as 'internet portals,' 'social network services,' 'YouTube,' 'podcasts,' and 'online communities.'

Next, respondents' self-placement of ideological orientation was measured by the question, "How would you describe your ideological orientation? Please provide a number between 0 and 10, where 0 is very progressive, 5 is moderate, and 10 is very conservative." Finally, respondents' presidential candidate choices were measured by presenting the major party candidates and third-party candidates with seats in order of their party affiliation, along with an option for other candidates. Accordingly, the third-party candidates presented to respondents included Ahn Cheol-soo, Yoo Seong-min, and Sim Sang-jung in 2017; Sim Sang-jung in 2022; and Lee Jun-seok and Kwon Young-guk in 2025.

To identify the age groups that can be collectively referred to as youth within Korean society and to ascertain their political attitudes and behaviors, this study categorizes birth cohorts in five-year intervals based on age at the time of the 2017 presidential election. It then categorizes youth based on the similarities in political attitudes and behaviors observed across the three presidential elections. For this purpose, the number of birth cohorts was set at nine for 2017, ten for 2022, and eleven for 2025. The nine birth cohorts in 2017 included those born in 1957 and earlier as a single cohort, followed by five-year age groups, with the cohort born between 1993-1997 being the youngest. In the 2022 presidential election, the cohort born between 1998-2002 was added to the nine cohorts from 2017, and in the 2025 presidential election, the cohort born after 2003 was added.

Table 1 summarizes the data used in this study. First, looking at the distribution of birth cohorts by time point, it can be observed that data collection for some birth cohorts is relatively sparse. For example, the birth cohorts of 1983-1987 and 1993-1997 in 2017, and the birth cohorts of 1993-1997 and those born in 2003 or earlier in 2025, constitute less than 6% of the total respondents, indicating relatively sparse data compared to other age groups. However, even the smallest group has data for over 50 respondents, so it is not expected to cause statistically significant errors in analysis. Another point to note is the relatively low response rate for information sources for election information and chosen candidates in the 2022 presidential election survey. However, these missing values are also not expected to cause systematic errors and thus are not anticipated to weaken the validity of the analysis.

<Table 1> Descriptive Statistics

Note: 1. Percentage; 2. Average

The content to be analyzed is whether the main media types for obtaining election-related information differ by birth cohort. The main media for obtaining election-related information are not direct information about the respondent's personal political attitudes and behaviors. However, the recent development of information and communication technology is creating an unprecedented new political environment and indirectly influencing the formation of individuals' political attitudes. While attitudes toward realizing direct democracy or expanding online spaces for deliberation through advanced information and communication technology may develop, political polarization and extreme political animosity may also develop. In particular, discussions on media literacy emphasize the close relationship between age and new media utilization (Shim Mi-seon 2022). Therefore, examining the differences in the ways traditional and new media are utilized by birth cohort, and whether such differences persist, will provide a clue to understanding the current status of the youth compared to middle-aged and elderly populations in Korean society.

<Table 2> Election information acquisition behavior through traditional media such as TV, newspapers, and radio by birth cohort

Note: * p<0.1, ** p<0.05, 1= (Usage rate 2022 - Usage rate 2017), 2=(Usage rate 2025 - Usage rate 2022)

<Table 2> shows the proportion of individuals who utilize TV, including terrestrial and general programming channels, as a primary source for election information, or who utilize newspapers and radio, broken down by birth cohort. According to the results in <Table 2>, it is reasonable to define the Korean youth cohort in 2017 as those born after 1983, corresponding to the age of 34 at the time. The proportion of individuals born after 1983 who obtained election information through traditional or legacy media such as TV, newspapers, and radio was less than 40%, at 35.9%, 31.9%, and 33.3%, respectively. In contrast, individuals born before 1982 showed a relatively higher proportion of using traditional media, ranging from the 50% to the 80% range. Notably, individuals born between 1963 and 1982 show a usage rate in the 50% range, and those born before 1962 show a usage rate above 70%, confirming a clear distinction among the three categories of youth, middle-aged, and elderly.

However, the validity of the aforementioned categorization of youth based on their primary sources of election information during the 2017 presidential election cycle is no longer applicable as the election cycles of 2022 and 2025 have passed. The 2022 presidential election cycle indicates a declining reliance on traditional media across all birth cohorts compared to 2017. Furthermore, fluctuations in the reliance on traditional media across birth cohorts make it difficult to establish a standard age for categorizing the youth demographic. If we consider a 50% reliance on traditional media as a benchmark, the youth cohort would include individuals born from 1968, who were 54 years old at the time of the 2022 election. Even if we lower the threshold to below 40% reliance, individuals born from 1973, who were 49 years old at the time, would need to be included.

The 2025 presidential election cycle presents a similar pattern. Compared to 2022, the reliance on traditional media has increased again across most birth cohorts, excluding two cohorts considered very elderly. Additionally, the fluctuations in traditional media reliance across birth cohorts allow for a division into groups: those born before 1973 with a reliance rate above 50%, and those born after 1973 with a reliance rate below 50%. However, it remains difficult to distinguish between the youth and middle-aged demographics among those born after 1973 with a reliance rate below 50%.

As demonstrated above, the temporal and cohort-specific variations in primary sources for election information suggest that current media usage patterns in Korean society cannot serve as a basis for distinguishing between youth, middle-aged, and elderly populations. Nevertheless, the validity of the 34-year-old benchmark identified in the 2017 election cycle can be re-examined in light of the variations observed in the 2022 and 2025 election cycles. Firstly, in the 2022 election cycle, the reliance on traditional media among individuals born after 1988, corresponding to the age of 34, remained very low, in the 20% or 10% range. The birth cohort of 1978-1982, corresponding to the age of 40-44 at the time, also showed a low reliance rate of 26.4%, which might be considered a margin of error within the context of significant shifts in media consumption behavior at that time.

In this regard, the reasons for the significant reduction in the use of traditional media for obtaining election information during the 2022 presidential election cycle are not clear. However, considering this alongside the trend in 2025 where media usage rates increased again across most birth cohorts, the impact of the COVID-19 pandemic can be inferred. Under the COVID-19 pandemic, with reduced face-to-face interactions and increased reliance on non-contact communication, various political knowledge and information were likely acquired through new types of media utilized for non-contact communication.

According to this interpretation, the period of 2025, after the COVID-19 pandemic has subsided, represents a time when traditional media usage has rebounded, moving away from such influences. The Korea Press Foundation's annual Media Usage Survey also supports this interpretation. Data from the 2024 Media Usage Survey indicates a gradual decline in news consumption through online video platforms, which had surged in previous years, during 2021 and 2022 (Korea Press Foundation 2024). Ultimately, the COVID-19 pandemic is associated with an increase in news consumption via new media and a decrease in news consumption via traditional media across society as a whole. Consequently, the gap in traditional media usage between the youth, middle-aged, and elderly populations, which was discernible based on traditional media reliance in 2017, has been significantly reduced.

If this interpretation is valid, then the conventional categorization of youth is likely to remain useful when controlling for societal changes such as the COVID-19 pandemic. Secondly, by 2025, there is little difference in the reliance on traditional media across almost all birth cohorts. Therefore, the utility of using reliance on traditional media to categorize the youth demographic in the 2025 presidential election cycle is very low. However, if we consider individuals up to age 34 as youth, based on the 2017 experience, it can be assessed that in 2025, the youth demographic exhibits no difference from the middle-aged and elderly populations in terms of their reliance on traditional media for obtaining election information.

<Table 3> Behavior of obtaining election information through new media such as internet portals, SNS, and YouTube by birth cohort

Note: * p<0.1, ** p<0.05, 1=(2022 usage rate-2017 usage rate), 2=(2025 usage rate-2022 usage rate)

<Table 3> reconfirms the declining value of traditional media as a source for obtaining election information from the perspective of new media utilization. Besides traditional media, sources for obtaining election information are diverse, including new media, acquaintances, and campaign materials from the National Election Commission. Therefore, this study aims to examine whether the decrease in the dependency of young people on traditional media is closely related to their usage patterns of new media, rather than other sources. The results in <Table 3> allow for interpretations and evaluations similar to those in <Table 2>. First, focusing on the 2017 presidential election, young Koreans can be considered those born after 1983, who were 34 years old at the time. Among them, 61.5%, 64.7%, and 64.8% utilized new media such as internet portals, social networking services, and podcasts to obtain election information. In contrast, the dependency on new media for those born before 1982 was below 40%.

The utilization of new media in the 2022 and 2025 presidential election phases also allows for interpretations similar to <Table 2>. Looking solely at new media utilization in the 2022 presidential election phase, differences are observed between the birth cohort after 1973 and the cohort born before that year. However, it is also true that the dependency on new media was relatively higher for the birth cohort born after 1988, who were 34 years old in 2022. Furthermore, in the 2025 presidential election phase, the dependency on new media converges to a certain level across all birth cohorts, making it difficult to distinguish between young, middle-aged, and elderly demographics.

Consequently, categorizing the youth in Korean society in contrast to the middle-aged and elderly based on media usage behavior for obtaining election information has low validity. However, during the 2017 presidential election, individuals aged 34 and under showed a clear difference from those aged 35 and above, characterized by low dependency on traditional media and high dependency on new media. Therefore, the birth cohorts from 1983 to 1987, corresponding to ages 30-34 in 2017, remain a potential starting point for categorizing youth in contrast to the middle-aged and elderly in Korean society. The validity of this categorization needs to be supplemented through various verification methods in subsequent research.

Nevertheless, if we categorize youth starting from the birth cohorts of 1983-1987 within limited validity, one characteristic of media usage behavior among Korean youth in 2025 for obtaining election information is that it does not significantly differ from that of the middle-aged and elderly. However, one point worth mentioning is that these results are based on the selection of primary information sources. In other words, these findings may differ when comparing youth, middle-aged, and elderly demographics based on more specific and abundant information regarding media usage, including the frequency and content of new media use.

III. Ideological Orientation of Korean Youth

Ideological orientation is the most representative of an individual's political attitudes. Ideological orientation serves as a cognitive framework for individuals to understand themselves and society and to make political choices (Rokeach 1973). The association between an individual's ideological orientation and age can be observed in the tendency for younger people to be relatively progressive (Glen 1974) or to respond more readily to new progressive values (Alwin et al. 1991), as discussed at the beginning of this paper. It is a social norm that relatively progressive young people tend to become more conservative as they age, emphasizing traditional values or order.

However, recent discussions suggest that the relationship between an individual's ideological orientation and age cannot be understood too simplistically, and require more meticulous verification through comparisons with age effects (age effect or life-cycle effect), cohort effects, and period effects. If the conservatism that comes with aging is referred to as the age effect, then the cohort effect implies that such conservatism is not simply related to age but varies depending on the historical and political experiences encountered during youth. For example, in the United States, not all young people are familiar with progressive values; rather, the Baby Boomer generation, born after World War II and experiencing the civil rights movement, anti-war movement, and feminist movement in the 1960s-1970s, exhibited progressive characteristics during their youth (Firebaugh and Davis 1988). Furthermore, the period effect refers to the effect where historically and politically significant events at a specific time influence all age groups and generations simultaneously.

When applying these discussions to the ideological orientation of Korean youth, it is necessary to pay attention to the following characteristics. If age or life-cycle effects are dominant, a tendency towards gradual conservatism is expected to be observed over time across all birth cohorts categorized in this study. Conversely, if cohort effects are dominant, a tendency for specific birth cohorts to maintain a stable ideological orientation irrespective of time, unlike other birth cohorts, will be observed. Finally, if period effects are dominant, a temporary shift towards progressivism or conservatism across all birth cohorts at a specific point in time can be anticipated.

<Figure 1> Average Ideological Orientation by Birth Cohort: At the 2007, 2017, 2022, and 2025 Presidential Election Periods

<Figure 1> presents the trend lines (local polynomial smoothed lines) using local polynomial regression, representing the average ideological orientation of 11 birth cohorts categorized in 5-year intervals across three presidential election periods, along with the 95% confidence intervals of the estimated means. In 2017, the mean was displayed as a solid line within the solid line representing the 95% confidence interval. In 2022, the mean was displayed as a solid line within the gray shaded area representing the 95% confidence interval. In 2025, the mean was displayed as a dotted line within the light green shaded area representing the 95% confidence interval.

The distribution of ideological orientation in 2017 in <Figure 1> shows, similar to what was observed in media usage behavior, that birth cohorts born after 1983, who were 34 years old in 2017, can be categorized as youth. Their average ideology is located between 4 and 4.5 points, and the upper limit of the 95% confidence interval does not exceed 4.5 points. Considering that the midpoint is measured as 5 on a scale of 0 to 10, they exhibit a progressive orientation compared to other age groups. In contrast, among those born before 1982, the birth cohorts from 1968-1972 maintain a progressive orientation on average but are more conservative than younger age groups. Those born before 1967 show a conservative orientation, averaging above 5 points. Therefore, the 2017 presidential election phase in South Korea observed an age effect where younger people are progressive and become more conservative with age, similar to Western societies. Furthermore, a progressive orientation is observed up to those born in 1968, who were 49 years old in 2017, and among these, the birth cohorts born after 1983 can be considered the youth group.

However, the distributions for 2022 and 2025 indicate that while a valid age effect between age and ideology no longer exists, they simultaneously imply the possibility of categorizing conservative youth. In other words, the U-shaped trend lines of average ideological orientation by birth cohort in 2022 and 2025 suggest that distinctions between young conservatives, progressive middle-aged individuals, and conservative elderly individuals are possible, focusing on the relationship between age and ideology. Notably, the categorization of youth based on age 34 in the 2017 presidential election also has high validity. The birth cohorts born after 1988, who were 34 years old in the 2022 presidential election, form the last cohort before the decline in ideological scores begins in the direction of progressivism from a conservative orientation in the U-shaped trend line. Similarly, in the 2025 presidential election, although not precisely categorized, the birth cohorts from 1988-1992, which include those born in 1991 who were 34 years old in 2025, show a decline in ideological scores comparable to 2022.

When categorizing Korean youth based on the age of 34, as described above, the Korean youth during the 2022 presidential election phase can be considered strongly conservative. As conservatism progressed across all birth cohorts in 2022, the birth cohorts born after 1988, who were 34 years old in 2022, all exhibited conservative orientations with average ideological scores above 5 points. Han Jeong-hoon (2022) argues that despite their self-identification as conservative, Korean youth are not forming a conservative generation.

This argument suggests that Korean youth under 34 in 2022 are not forming a self-identity and value system as conservatives, unlike past Korean youth or Western youth. Instead, they are temporarily exhibiting conservative tendencies due to the overall conservative shift across all birth cohorts in 2022. This claim is re-examined through the 2025 distribution. In the 2025 presidential election phase, Korean youth under 34 in 2022 maintain an average ideological score above 5 points, similar to 2022, but the intensity has weakened, as indicated by the decline in ideological scores compared to 2022. Furthermore, the youth who newly participated in the 2025 presidential election, those born after 2003, show a relatively strong progressive orientation in their average ideological scores.

<Table 4> reorganizes the information from <Figure 1> focusing on the average ideological scores for each birth cohort, to facilitate easier comparison of the characteristics of Korean youth with other generations. The bottom row of <Table 4> includes the results of a one-way ANOVA to show whether the differences in average ideological scores between birth cohorts are statistically significant, and the two rightmost columns present the results of pairwise comparisons between groups. First, examining the average ideological scores by birth cohort in 2017, youth can be categorized based on those born in 1983, who were 34 years old at the time. The birth cohorts from 1983-1987, corresponding to ages 30-34 at the time, exhibit the most progressive orientation with an average ideological score of 3.94. Including these, those showing a progressive orientation can be categorized as youth, those older than them with average ideological scores in the 4-point range can be categorized as middle-aged, and those with average ideological scores in the 5-point range can be categorized as elderly. Therefore, the 2017 presidential election phase was a period where the age effect, showing increasing conservatism from youth to the elderly, could be observed.

The emergence of a new youth demographic, or conservative youth, deviating from the prediction of progressive youth and the age effect, finds its basis in the changes in ideological orientation in 2022. At the time of the 2022 presidential election, the most progressive group is the birth cohort born between 1973-1977, with an average ideological score of 4.57. Both younger and older groups than this cohort exhibit relatively more conservative ideological orientations, leading to active social discussions about conservative youth or the youth of conservatism (Heo Seok-jae 2014).

Furthermore, the overall societal conservatism at the time complicates the discussion regarding the age threshold for defining youth. As indicated by the changes between 2017 and 2022 in <Table 4>, most significant ideological changes between birth cohorts represented a progression of conservatism. This societal conservatism makes it difficult to define youth based on the age of 34 in 2022. This is because the birth cohort of 1983, who were 34 years old in 2017 and are currently 39 years old in 2022, also exhibits an ideological orientation in the 5-point range, similar to subsequent birth cohorts. Moreover, those born after 1998, who newly participated in the 2022 election, also show a conservative orientation with a score of 5.46. Therefore, it appears reasonable to extend the definition of youth up to age 39 for the 2022 presidential election.

The reversal of the conservative trend observed in 2022 during the 2025 presidential election phase is also a noteworthy point. The impeachment of President Yoon Suk-yeol likely had a significant impact. However, the intensity of the progressive trend was not as strong as the intensity of the conservative trend in 2022. While changes in average ideological scores towards progressivism occurred in almost all birth cohorts, they did not reach statistically significant levels. However, the birth cohorts from 1983-1987 again became groups with an average progressive orientation. Consequently, the validity of categorizing youth based on the birth cohorts from 1988-1992, which include those born in 1991 who were 34 years old in 2025, has been strengthened. It is also worth noting that the birth cohorts born after 2003, who participated in the presidential election for the first time in 2025, exhibit a considerable progressive orientation compared to the 2022 presidential election phase.

<Table 4> Comparison of Average Respondent Ideology by Birth Cohort Across Three Presidential Election Periods

Note: * p<0.1, ** p<0.05

Ultimately, the ideological orientation of voters in Korean society provides a largely valid criterion for distinguishing between youth, middle-aged, and elderly populations. Furthermore, when utilizing ideological orientation, it appears reasonable to distinguish youth around the age of 34. Conversely, the Western age effect, which suggests that individuals become more conservative with age and young people are progressive, has low validity in current Korean society. Currently, Korean youth not only exhibit conservative tendencies amidst the overall societal conservatism of 2022 but are also not reverting to progressive tendencies similar to past youth cohorts amidst societal efforts to re-establish democracy following the impeachment of President Yoon Suk-yeol in 2025. Continuous research is needed to determine whether this phenomenon signifies the development of a generation of Korean youth with unique values and identities distinct from past youth cohorts, or if it represents a temporary temporal peculiarity.

IV. Changes in Candidate Selection among Korean Youth

Thus far, we have examined the difficulty in distinguishing youth from middle-aged and elderly populations in Korean society based on their primary media sources for election information and ideological orientation, and the fact that the most plausible criterion appears to be around 30-34 years of age. Here, we will reconsider the validity of this classification, focusing on candidate selection during presidential election campaigns.

<Figure 2> Changes in Support for Democratic Party Candidates by Birth Cohort

First, <Figure 2> shows the support rates for Democratic Party candidates in three presidential elections, broken down by birth cohort. Examining the 2017 presidential election, indicated by the thick black solid line, it appears reasonable to define the birth cohort born after 1988, corresponding to age 29 at the time, as Korean youth. This is because the 1988-1992 birth cohort showed the highest support for the Democratic Party candidate, implying a certain break from earlier birth cohorts. In contrast, the validity of the 29-year-old criterion is not high for the 2022 and 2025 presidential elections. In 2022, those aged 29 belong to the 1993-1997 birth cohort, and their support rate for the Democratic Party candidate is relatively high among younger demographics.

However, the birth cohorts adjacent to these, namely the 1988-1992 and 1998-2002 birth cohorts, show similar levels of low support, raising the question of whether these three birth cohorts need to be included in a single age group. In this case, the 34-year-old criterion becomes more valid than the 29-year-old criterion for the 2022 presidential election. The 2025 presidential election presents a similar situation to 2022. In the 2025 election, those aged 29 are born in 1996 and fall within the 1993-1997 birth cohort. However, to define the youth cohort, including the 1993-1997 birth cohort, based on support for the Democratic Party candidate, it is necessary to include the 1988-1992 birth cohort, which showed a similar level of support to the birth cohorts born after 2003. According to this, it seems reasonable to include those born after 1988, i.e., up to 34 years old, as youth for the 2025 presidential election as well. In contrast to this categorization of youth, it is noteworthy that the support of Korean youth for the Democratic Party candidate in the 2022 and 2025 presidential elections is relatively low compared to other age groups. In other words, unlike Western societies where young voters tend to strongly support progressive Democratic Party candidates, Korean youth exhibited contrasting political behavior in the 2022 and 2025 presidential elections.

<Table 5> further examines this candidate selection behavior of Korean youth by distinguishing support for Democratic Party candidates, People Power Party candidates, and third-party candidates in each presidential election phase by birth cohort. In the 2017 presidential election, there were three popular third-party candidates: Ahn Cheol-soo, Yoo Seong-min, and Sim Sang-jung. In the 2022 presidential election, Sim Sang-jung competed as a candidate from a party with seats in the National Assembly, and in the 2025 presidential election, Lee Jun-seok and Kwon Young-guk competed as candidates from parties with seats in the National Assembly. The support rates for third-party candidates at each point in time refer to their respective support rates.

First, based on the 2017 presidential election phase, the possibility of using age 29, corresponding to those born in 1988, as the criterion for defining youth is the same as explained in <Figure 2>. This is because over 70% of the 1988-1992 birth cohort, who were aged 25-29 at the time of the 2017 election, supported the Democratic Party's Moon Jae-in, while the support rate for Moon Jae-in among birth cohorts older than them (born before 1987) gradually decreased to 60% and 50%. In contrast, for 2022 and 2025, categorizing youth based on age 29 at each point in time cannot avoid the problem that the Democratic Party support rate for the birth cohort corresponding to 30-34 years old at each point is similar to that of the categorized youth. Based on this discussion, the validity of categorizing youth based on age 34 at each presidential election, considering candidate support behavior, can be considered quite high.

As described above, when defining youth based on age 34 at each presidential election, the candidate support behavior of Korean youth exhibits several characteristics. First, unlike Western youth who consistently support progressive party candidates, the support of Korean youth is not concentrated on progressive party candidates. In the 2022 presidential election, Korean youth supported the People Power Party candidate more than the Democratic Party candidate, and in the 2025 presidential election, they showed relatively lower support for the Democratic Party's Lee Jae-myung compared to other age groups.

Second, Korean youth are relatively more likely to support third-party candidates compared to major party candidates when viable third-party options are available. In the 2017 presidential election, the support rate for third-party candidates among Korean youth remained between 25% and 37%, showing no significant difference from that of middle-aged and elderly demographics. However, in the 2025 presidential election, they showed over 28% support for third-party candidates, distinguishing themselves from middle-aged and elderly demographics who showed support below 20%. A point of contention regarding this third-party candidate support behavior among Korean youth is whether this support is directed towards candidate Lee Jun-seok personally. Following the 2025 presidential election, claims emerged that the support for third-party candidates among Korean youth stemmed from the personal appeal of candidate Lee Jun-seok and that this support would extend to the Reform Party.[1]

However, this argument is not valid, given that Korean youth also showed strong support for third-party candidates in the 2017 presidential election, when Lee Jun-seok was not a competing candidate. Rather, the support of Korean youth for third-party candidates can be characterized as a protest against the major parties. In other words, Korean youth are consistently expressing their dissatisfaction with the lack of policy attention and effectiveness from the major parties regarding youth issues in the current political landscape, and are seeking third alternatives. While interest in third alternatives was high across all age groups in 2017, this interest was sustained only among the youth in 2025. Recent media reports suggesting that the reason young Koreans supported candidate Lee Jun-seok was not due to his personal integrity or morality, but rather because his youth-oriented policies were relatively well-presented, further support this assessment. [2]

<Table 5> Cross-analysis of Candidate Support at Three Presidential Election Points by Birth Cohort

Note: * p<0.1, ** p<0.05

V. Conclusion

This chapter aimed to expand the recent growing interest and discussion surrounding the political attitudes and behaviors of the youth in South Korean society by analyzing their media consumption patterns, ideological orientations, and candidate support behaviors. The starting point of this research lies in clearly defining the concept of 'youth.' This is because the characteristics of youth's political attitudes and behaviors can be revealed through comparison with other age groups only when the definition of youth is clear. However, the definition of youth in Korea is currently very diverse, both legally and academically. Consequently, there is a low level of consensus regarding the characteristics of the Korean youth demographic.

Within these constraints, this study attempted to categorize and describe the characteristics of the youth demographic by comparing the political attitudes and behaviors exhibited by Korean youth during the three presidential election processes since 2017. In this process, rather than defining youth a priori, we utilized birth cohorts in five-year increments to identify youth as the age group exhibiting political attitudes and behaviors distinct from middle-aged and elderly populations. Through this strategy, this study argues that it is reasonable to broadly categorize the youth demographic in South Korean society as those born after 1991, up to the age of 34, as of 2025. While these young people showed progressive tendencies and candidate support behaviors similar to Western youth up to the 2017 presidential election, they have become relatively more conservative through the 2022 and 2025 presidential elections. Furthermore, in terms of candidate support behavior, they have not solely focused on the Democratic Party candidate. Instead, these Korean youth have expressed disappointment and criticism towards the major two parties, represented by the Democratic Party and the People Power Party, by giving around 30% support to a strong third-party candidate whenever one emerged.

The findings of this study are expected to have significant implications for South Korean society, alongside the demographic shift where the proportion of voters under 39, which was 35% in the 2010s, decreased to 29.9% by the 2024 general election. Firstly, as the conventional wisdom that 'youth are progressive' weakens in Korean politics, competition among established parties to gain youth support will intensify. The youth support concentrated on candidate Lee Jun-seok in 2025 was due to the high validity of the youth policies he proposed. Unlike the major two parties that merely paid lip service to 'youth' during election periods, he was able to consolidate dissatisfaction with the major parties by presenting feasible and effective policy alternatives to the youth. Secondly, the continuous decline in the proportion of the youth electorate compared to middle-aged and elderly populations reduces the social influence of the youth. This implies that the need for communication and consolidation of political preferences within the youth demographic will become more pronounced. The current confrontation between young women and men is a serious obstacle to this mobilization of youth power. Therefore, the future of Korean youth will be closely linked to the interaction between the future necessity of youth mobilization and the current conflict of gender-based discrimination.

VII. References

Kim, Eun-i, and Min-ho Song. 2022. “The Impact of 20-Something Perceptions and Policy Evaluations on Voting Participation in the 20th Presidential Election.” *Journal of Social Science Research* 48(3): 1-32.

Son, Jeong-hee, Chan-seok Kim, and Hyun-sun Lee. 2021. “A Study on the Responses of Each Generation to the Unique Communication Characteristics of the MZ Generation.” *Journal of Communication Design Research* 77.

Shim, Mi-sun. 2022. “A Study on the Factors Affecting Media Message Literacy Competency.” *Journal of Korean Media Information Research* 116: 38-76.

Lee, Min-hyung. 2023. “A Critical Review of Generational Theory: Focusing on the 'MZ Generation'.” *Theology and Practice* 87: 667-691.

Lee, Jeong-jin. 2022. “Limitations and Possibilities of Youth Politics as Seen Through Local Elections.” *Journal of Korean Local Politics* 12(2): 27-51.

Cho, Chang-deok. 2022. “Does Youth Political Participation Affect Youth Policy Budgets? Focusing on Metropolitan Local Governments from 2009-2018.” *Journal of Korean Local Autonomy* 34(1): 119-142.

Choi, Jong-sook. 2020. “Revisiting the '20s Male Phenomenon': Focusing on the Ideological Orientation and Gender Consciousness of those in their 20s and 30s-40s.” *Economy and Society* (Issue No. 125): 189-224.

Heo, Seok-jae. 2014. “Generations, Life Cycles, and Voting Choices: Analysis of Presidential Elections from 1992-2012.” *Korea and International Politics* 30(2): 16-199.

Korea Press Foundation. 2024. *2024 Media Usage Survey*. Korea Press Foundation.

Han, Jeong-hoon. 2022. “The Conservatization of Korean Youth? Focusing on Ideological, Policy Attitudes, and Voting Behavior in Presidential Elections from 2012 to 2022.” *Journal of International and Area Studies* 31(2): 285-318.

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

Firebaugh, Glenn, and Kenneth E. Davis. 1988. “Trends in Antiblack Prejudice, 1972–1984: Region and Cohort Effects.” *American Journal of Sociology* 94 (2): 251–272.

Glenn, Norval D. (1974). “Aging and Conservatism,” *Annals of the American Academy of Political and Social Science*, 415, 176–186.

Gremsci, Antonio. 2011. *Prison Notebooks*. Columbia University Press

Lucas, Olivia, Lovisa Moller Vallgarda, and Bernard Steen. 2024. *Society Watch 2024: Understanding the New Generation of Voters*. National Centre for Social Research.

Mill, John Stuart. 2015. *On Liberty, Utilitarianism and Other Essays*. Oxford University Press.

Pew Research Center. 2018. “The Generation Gap in American Politics” https://www.pewresearch.org/politics/2018/03/01/the-generation-gap-in-american-politics/?utm_source=chatgpt.com

Rokeach, M. (1973). *The nature of human values*. New York: Free Press.

[1] https://zdnet.co.kr/view/?no=20250603211650 (Accessed August 14, 2025)

[2] https://h21.hani.co.kr/arti/politics/politics_general/57622.html (Accessed August 14, 2025)

■ Author: Han Jeong-hoon _Professor, Graduate School of International Studies, Seoul National University.

■ Management and Editing: Im Jae-hyunEAI Research Fellow

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