[Diagnosis Series on the Regression of Korean Democracy] V. Crisis of Korean Democracy and 'Regression from Below'?

I. Introduction

Research on democratic backsliding approaches the causes and processes of regression by dividing them into 'democratic erosion from the top' and 'democratic erosion from the bottom.' Studies focusing on 'erosion from the top' view regression as the result of strategic choices made by elites holding political power, particularly heads of the executive branch, to consolidate or expand their power. These studies focus on how leaders elected through popular vote, while maintaining the outward forms of the constitution and laws, gradually expand executive power and weaken checks and balances through methods such as neutralizing the legislature via executive orders, controlling the judiciary, suppressing the media, and persecuting critics (Bermeo 2016; Levitsky and Ziblatt 2018). Such changes may appear to follow legal procedures and occur gradually and subtly, making it difficult for citizens to perceive the weakening of democracy as a direct threat, and thus risking the normalization of the regression process as part of democracy itself, thereby blunting citizen resistance (Bartels 2018; Ginsburg and Huq 2018).

Meanwhile, those focusing on 'erosion from the bottom' believe that the stability and sustainability of democracy depend on whether citizens voluntarily accept democracy and provide normative support for the regime. Democracy can acquire political legitimacy when citizens accept it as the most just and appropriate political system (Lipset 1959), and it is maintained not by external coercion but by citizens' voluntary acceptance (Dahl 1971). This legitimacy is further strengthened when it is supported by principled support and attachment to democracy itself, known as diffuse support, beyond specific support based on temporary performance or particular policies (Easton 1965). In the absence of diffuse support, democracy can easily collapse from within during crises such as economic shocks or political turmoil, as citizens may lose faith in the system's ability to resolve problems through institutional channels (Classen 2020). Democracy is consolidated when societal members do not consider alternatives to democracy and believe that all political conflicts and issues can be resolved within democratic procedures and norms, meaning democracy is accepted as 'the only game in town.'

The major events experienced by Korean democracy since the 2022 presidential election show several signs of democratic erosion from the top. For example, the procedural legitimacy disputes surrounding the dismissal of public broadcasting board members by the Korea Communications Commission and the privatization of YTN, as well as the President's frequent use of veto power (right of recall) on bills passed by the opposition party (e.g., the Grain Management Act, the Nursing Act, the Yellow Envelope Act, the Broadcasting Act 3, the Special Counsel Act for First Lady Kim Keon-hee, the Special Act on the Itaewon Disaster, etc.), can be interpreted as attempts by the executive branch to weaken the legislative branch's oversight function and influence the media's critical monitoring role. Furthermore, controversies over selective or targeted investigations in the operation of power institutions such as the Board of Audit and Inspection and the prosecution raise concerns about the potential compromise of political neutrality of state institutions. This is similar to the methods described by Levitsky and Ziblatt where 'elected autocrats' neutralize the judiciary or other checks and balances. On the other hand, the opposition party's choice to resolve legislative deadlock with the executive branch through impeachment proceedings is also not free from criticism for undermining the norm of 'institutional forbearance.' These events align with 'erosion from the top,' where elites in power gradually undermine democracy within the framework of legitimate institutions.

International assessments of Korean democracy reflect these changes. In its 2025 Democracy Report, the Varieties of Democracy (V-Dem) Institute at the University of Gothenburg, Sweden, classified South Korea as a country undergoing 'autocratization' for the second consecutive year, following 2024.[1]While South Korea was classified as the 17th freest democracy in the world in the 2021 report, it was no longer classified as a liberal democracy in the 2025 report, but rather as an electoral democracy. The 'Democracy Index 2024' released by the Economist Intelligence Unit (EIU), a British economic analysis organization affiliated with The Economist, offered a similar assessment of Korean democracy. In that report, South Korea scored 7.75 out of 10 points and was classified as a 'full democracy' from 2020 to 2023, but fell to 'flawed democracy' in 2024. South Korea was one of the 10 countries with the largest score decline in the report. Recently, voices expressing concern about 'erosion from the bottom' have also been rising. In particular, the incident on January 19, 2025, where individuals who opposed the execution of an arrest warrant and its issuance against President Yoon Suk-yeol stormed, occupied, and vandalized the Western District Prosecutors' Office, sent shockwaves through Korean society. Furthermore, the claims of those who have consistently argued for election fraud are now being expressed in public squares, combined with voices opposing the impeachment of President Yoon Suk-yeol. Observing those who criticize the rule of law and constitutionalism under the banner of the right to resist, a sense of crisis arises that Korean society is not immune to the problems of the far-right movements spreading in the United States and Europe.

Erosion from the top and erosion from the bottom are not mutually exclusive but are closely interconnected. Anti-democratic attempts by elected autocrats are more likely to succeed when public support for democracy is weak. Conversely, strong public conviction in the legitimacy of democracy can serve as a primary means of controlling erosion from the top by elites. In this context, this paper aims to examine how public support for Korean democracy has changed over time amidst the rapidly evolving political landscape of South Korea.

II. Overall Time-Series Changes

To understand the attitudes of Korean citizens towards democracy, the following question was used: 'Which of the following statements best reflects your position? 1) Democracy is always better than any other system. 2) In certain situations, dictatorship may be better than democracy. 3) For me, there is no significant difference between democracy and dictatorship.' Response 1) can be interpreted as a measure of diffuse support (Easton 1965), reflecting the perception that democracy has an absolute advantage over other systems. In contrast, respondents who chose 2) express conditional support for democracy and may be interpreted as likely to reject democracy and justify dictatorship or authoritarianism in times of crisis. Finally, respondents who chose 3) exhibit a cynical attitude towards the political system as a whole. The analysis period for this report spans from 2003 to 2025. The Asia Barometer Survey is used for the period 2003-2022, and the data from the 'Polarization Perception Survey' conducted by the East Asia Institute from January 22-23, 2025, is used for 2025.

[Figure 1] Time-Series Changes in Attitudes Toward Democracy, 2003-2025

Figure 1 shows a continuous increase in diffuse support for democracy among Korean citizens over the past two decades. In the 2003 survey, only 49% responded that 'democracy is always better than any other system,' but this figure rose to 66% in 2011, 71% in 2019, and 76% in 2022. Even in the survey conducted amidst extreme political polarization following the 2022 presidential election and the subsequent period of martial law and impeachment in 2024, 75% of respondents indicated that democracy is better than other systems. Conversely, the response 'in certain situations, dictatorship may be better than democracy' has been steadily decreasing since reaching 36% in 2006. After recording 20% in 2011 and 25% in 2015, it fell to the mid-teens, recording 17% in 2019 and 16% in 2022, and remained at 16% in the 2025 survey. Meanwhile, the response 'there is no significant difference between democracy and dictatorship' also significantly decreased from 33% in 2003 to 9% in 2025. In summary, Figure 1 suggests that the perception of democracy's absolute superiority has taken root throughout Korean society, and despite the ongoing erosion of democracy from the top, the institutional legitimacy of democracy is becoming increasingly internalized in citizens' beliefs.

III. Generational Analysis

How has this change been possible? Firstly, it may be the result of political learning. Since democratization in 1987, South Korea has maintained a democratic system for over 30 years. During this time, Korean citizens have repeatedly experienced peaceful power transitions between ruling and opposition parties through elections. Moreover, Korean democracy remained stable through the 1997 IMF financial crisis and the 2008 global financial crisis. Through events like the mad cow disease protests in 2008 and the candlelight vigils during the 2017 impeachment process, citizens confirmed their ability to bring about substantial political change through participation within the framework of democracy. Through these diverse political experiences within the democratic system, Korean citizens have come to understand and internalize the value and function of democratic institutions. Secondly, there is the possibility of a 'generational shift effect.' The Millennial (M) and Z generations, born and raised after democratization in 1987, grew up in a qualitatively different political environment than previous generations. These generations, who did not experience the military authoritarian era, grew up in a social discourse that negatively evaluated past authoritarian rule and emphasized democratic principles and values through their education. Furthermore, with the expansion of press freedom, they have freely accessed political information through various media, developing a higher sensitivity to democracy compared to previous generations. As they have become established members of society, it is possible that support for democracy has generally strengthened in Korean society.

Figure 2 examines changes in attitudes toward democracy by generation. Based on birth year, individuals born between 1940-1959 are classified as the Industrialization Generation, those born in the 1960s as the '86 Generation,' those born in the 1970s as Generation X, those born in the 1980s as Generation M, and those born after 1990 as Generation Z. According to Figure 2, support for democracy has continuously increased across all generations. For the Industrialization Generation, the proportion of respondents supporting democracy was 49% in 2003, rising to 76% in 2022. The proportion of those who believed dictatorship might be better, recorded at 38% in 2006, has continuously decreased thereafter, reaching only 17% in 2022. In the 2025 survey, the proportion of those choosing democracy decreased by about 3 percentage points compared to the 2022 survey, and in some cases, the proportion of those who responded that dictatorship might be better increased accordingly to 21%. The '86 Generation shows similar changes to the Industrialization Generation. In 2003, the proportion of respondents choosing democracy was 50%, increasing to 79% in 2022. In the 2025 survey, it decreased to 74%, a drop of about 5 percentage points. In some cases, the proportion of those who responded that dictatorship might be better was highest at 40% in 2006, and has continuously decreased thereafter to 16% in 2022, before slightly increasing to 18% in 2025. Within Generation X, support for democratic norms has also continuously increased. In 2003, 49% chose democracy, and by 2022, this proportion had increased by approximately 22 percentage points to 71%. Interestingly, unlike the Industrialization Generation and the '86 Generation, support for democracy in Generation X further increased in the 2025 survey. In the 2022 survey, only 71% chose democracy, a lower rate than the Industrialization Generation or the '86 Generation, but in the 2025 survey, 80% chose democracy. Between the 2022 and 2025 surveys, the proportion of those who responded that dictatorship might be better decreased by about 5 percentage points from 18% to 13%, and the proportion of those who responded that they did not care whether it was democracy or dictatorship also decreased.

[Figure 2] Time-Series Changes in Attitudes Toward Democracy, 2003-2025: Generational Analysis

The situation for Generation M and Generation Z is not significantly different. For Generation M, 53% of respondents chose democracy in the 2003 survey, and this proportion increased to 80% in 2022. As of 2022, they showed stronger support for democracy than Generation X. This proportion slightly decreased to 76% in the 2025 survey. For Generation Z, it began to appear in the 2011 survey, and based on 2011, the proportion of respondents supporting democracy was about 57%, which could be considered somewhat lower than other generations. However, considering that Generation M recorded 53% in the 2003 survey when they were in their 20s, and unlike other generations, Generation Z has never recorded a proportion of those choosing dictatorship above 21%, it is possible to confirm to some extent the effect of a generational shift where Generation Z shows stronger support for democracy compared to previous generations. Furthermore, it is noteworthy that although support for democracy slightly decreased between 2022 and 2025, there was no significant change in the proportion of respondents who answered that dictatorship might be better in certain situations. This contrasts with the Industrialization Generation and the '86 Generation, where political changes including martial law and impeachment led to a withdrawal of support for democracy and an increase in support for dictatorship; in the MZ generation, the withdrawal of support for democracy did not translate into support for dictatorship.

IV. Changes by Generation and Gender

There is growing interest in the political conservatism of MZ generation men in recent Korean politics, particularly their conservative attitudes related to gender issues and their tendency to react sensitively to the discourse of 'fairness.' The tendency of men in their 20s to support the People Power Party (conservative party) was clearly evident in the 2020 general election, the 2021 by-elections, and the 2022 presidential election. Notably, exit polls conducted immediately after the 2022 presidential election showed that 59% of men in their 20s and 53% of men in their 30s supported candidate Yoon Suk-yeol. This stands in stark contrast to the 58% of women in their 20s and 50% of women in their 30s who supported candidate Lee Jae-myung. The so-called gender gap between men and women in their 20s was also confirmed during the impeachment process following President Yoon Suk-yeol's declaration of emergency martial law in December 2024. According to one survey, men in their 20s accounted for only 3% of all participants, while women in their 20s accounted for 18%. This is in contrast to the participation rates of men, which were in the low to mid-teens, during the mad cow disease candlelight vigils in 2008 and the impeachment rallies in 2016.BBC News Korea, February 14, 2025.

In this context, Figure 3 shows how support for democracy varies by respondent's gender within each generation. For the Industrialization Generation, the difference between male and female respondents is not significant, except for 2025. In the 2025 survey, there is no significant change for female respondents compared to 2022. Meanwhile, for male respondents, support for democracy decreased by about 6 percentage points from 75% to 69%, while support for dictatorship increased by about 11 percentage points from 17% to 28%. For the '86 Generation, the opposite change occurred. For male respondents, support for democracy decreased by only 2 percentage points from 78% to 76%, but for female respondents, it decreased by about 8 percentage points from 80% to 72%, and support for dictatorship increased by 4 percentage points from 16% to 20%. Relatively larger changes occurred among women compared to men. In Generation X, attitude changes among male respondents were significant. In the 2022 survey, support for democracy among men in Generation X was 67%, but in the 2025 survey, it increased by a remarkable 17 percentage points to 84%. Conversely, responses favoring dictatorship decreased by 12 percentage points from 22% to 10%. Men in Generation X showed the most significant change in attitudes toward democracy and dictatorship between the two surveys.

[Figure 3] Time-Series Changes in Attitudes Toward Democracy, 2003-2025: Generational and Gender Analysis

Meanwhile, Generation M and Generation Z show different patterns from previous generations in that gender differences among respondents appeared before the 2025 survey. For Generation M, the gap between men and women began to appear around 2015. In the 2011 survey, 67% of male respondents and 68% of female respondents supported democracy, with no significant difference. However, after 2015, support for democracy among female respondents increased sharply, while the increase in support among male respondents did not keep pace, leading to a gender gap in support for democracy. The difference between male and female respondents was 6 percentage points in the 2015 survey, 8 percentage points in the 2019 survey, and 12 percentage points in the 2022 survey, continuing to increase. In the 2025 survey, 84% of female respondents supported democracy, while only 69% of male respondents stated they supported democracy, a difference of about 15 percentage points. For female respondents in Generation M, only 7% responded that dictatorship might be better in the 2025 survey, the lowest figure among all surveyed groups.

Generation Z is the group with the most pronounced gender gap. A gender gap already existed in the 2011 survey when Generation Z first appeared. While 64% of female respondents supported democracy, the figure for male respondents was only 52%, a difference of about 12 percentage points between male and female respondents. Interestingly, the gap between men and women disappeared in the 2019 survey. Both men and women in Generation Z responded that democracy is better by 76%, while 14% responded that dictatorship is sometimes better. This change appears to reflect the experience of the 2017 impeachment process. However, thereafter, support for democracy among women in Generation Z continued to increase, reaching 79% in 2022 and 81% in 2025, while conversely, support for democracy among men in Generation Z decreased to 73% in 2022 and 63% in 2025. In 2025, the gap between men and women in Generation Z reached 18 percentage points. The changes observed among men in Generation M and Generation Z align to some extent with concerns about the 'conservatism of men in their 20s.' However, it is important to note that the decrease in support for democracy among these groups does not translate into increased support for dictatorship. Between 2022 and 2025, support for democracy among men in Generation M decreased by 5 percentage points, but support for dictatorship only increased by 3 percentage points. For men in Generation Z, support for democracy decreased by 10 percentage points, but support for dictatorship only increased by 6 percentage points. This clearly contrasts with men in the Industrialization Generation, where support for dictatorship increased by approximately 11 percentage points in the 2025 survey compared to previous surveys.

V. Conclusion

Principled support and attachment to democracy itself, known as diffuse support, form the foundation for maintaining democratic stability. Therefore, if citizens' confidence and belief in democracy decline, it is evidence of democracy experiencing erosion from the bottom and a loss of the momentum to resist erosion from the top. At a time when concerns about the regression of Korean democracy are escalating due to political turmoil, including the declaration of emergency martial law, the impeachment of the president, and the need for a second snap election within a decade, this study examined how Koreans' support for democracy has changed over the past two decades.

By analyzing the results of seven surveys conducted from 2003 to 2025, we confirmed that diffuse support for democracy has become established among Korean citizens. Based on the overall respondents, the proportion of those who responded that democracy is always better than any other system increased by 33 percentage points from 43% in 2006 to 76% in 2022. During the same period, the proportion of those who responded that dictatorship is better in certain situations decreased by two-thirds, from 36% in 2006 to 12% in 2022. Even in the January 2025 survey, conducted amidst the martial law crisis and ongoing impeachment proceedings against the president, there was no significant change in the general attitudes of Korean citizens toward democracy and dictatorship. However, when examined by generation and gender, differences in responses to martial law and impeachment were observed. For men in the Industrialization Generation, Generation M, and Generation Z, support for democracy decreased and support for dictatorship increased in the 2025 survey compared to previous surveys. Conversely, for men in Generation X and women in Generation M and Z, support for democracy increased, resulting in no major fluctuations in the overall response rates.

The relatively lower support for democracy among men in Generation M and Generation Z, and the significant decrease during the martial law period, are somewhat consistent with the discussion on the 'conservatism of men in their 20s.' However, we must remember that 68% of men in Generation M and 63% of men in Generation Z still consider democracy 'the only game in town,' and the proportion of respondents choosing dictatorship is only about one-third of those who chose democracy. Notably, compared to the United States and Western Europe, where concerns about erosion of democracy from the bottom are currently rising, the decrease in support for democracy among men in the MZ generation is not substantial. For example, according to Foa and Mounk (2016), in the United States, about 60% of those born in the 1940s (equivalent to the Industrialization Generation) agreed with the statement that living in a democracy is essential, whereas only about 30% of those born in the 1980s (equivalent to Generation M) agreed, indicating a much more pronounced decline in democracy among younger generations.

Despite the emergency martial law and impeachment proceedings, Korean democracy is demonstrating considerable resilience. This resilience is underpinned by citizens who express firm support for democracy. Over more than 30 years of democratic experience since democratization in 1987, Korean citizens have internalized democracy not merely as a system but as a fundamental societal value through political learning. As of 2025, while there are some differences in attitudes toward democracy based on generation and gender, democracy is generally accepted in Korean society as 'the only game in town,' as proposed by Linz and Stepan (1996) as a condition for democratic consolidation. This supports the assertion that the recent regression of Korean democracy is a 'regression from the top' caused by the strategic choices of political elites, rather than an 'erosion from the bottom' stemming from the weakening of democratic values among citizens. The robust support for democracy held by Korean citizens will serve as a crucial asset in the process of overcoming the regression from the top in the future of Korean democracy. ■

References

Bartels, Larry M. 2023. “Democracy Erodes from the Top.” In Democracy Erodes from the Top: Leaders, Citizens, and the Challenge of Populism in Europe, Princeton University Press, 185–215.

BBC News Korea. 2024. “한국 ‘20대 남성’은 왜 보수화됐나? [Why Have South Korean ‘Men in Their 20s’ Become Conservative?].” February 14. https://www.bbc.com/korean/articles/c159vendkl8o (Accessed: May 13, 2025.)

Bermeo, Nancy. 2016. “On Democratic Backsliding.” Journal of Democracy 27, 1: 5–19.

Claassen, Christopher. 2020. “Does Public Support Help Democracy Survive?” American Journal of Political Science 64(1): 118–134.

Dahl, Robert A. 1971. Polyarchy: Participation and Opposition. New Haven, CT: Yale University Press.

Easton, David. 1965. A Systems Analysis of Political Life. New York: Wiley.

Foa, Roberto Stefan, and Yascha Mounk. 2016. “The Danger of Deconsolidation: The Democratic Disconnect.” Journal of Democracy 27, 1: 5–17.

Ginsburg, Tom, and Aziz Z. Huq. 2018. How to Save a Constitutional Democracy. Chicago, IL: University of Chicago Press.

Levitsky, Steven, and Daniel Ziblatt. 2018. How Democracies Die. New York: Crown.

Linz, Juan J., and Alfred Stepan. 1996. Problems of Democratic Transition and Consolidation: Southern Europe, South America, and Post-Communist Europe. Baltimore, MD: Johns Hopkins University Press.

Lipset, Seymour Martin. 1959. “Some Social Requisites of Democracy: Economic Development and Political Legitimacy.” American Political Science Review 53, 1: 69–105.

[1]Among the 42 countries classified as undergoing 'autocratization' in the 2024 report are Hong Kong, where the de facto one-country, two-systems framework collapsed through the 2014 Umbrella Movement and the 2020 National Security Law; Cambodia, which effectively transitioned to a de facto one-party dictatorship after the Constitutional Court forcibly dissolved opposition parties in 2017; Myanmar, where civil resistance and bloody suppression have repeatedly occurred since the 2021 military coup; and the Philippines, where thousands were extrajudicially executed under the guise of the war on drugs during the Duterte administration, and concerns about historical distortion have heightened following the election of President Ferdinand Marcos Jr. in 2021.

■ Kang Woo-chang_Professor, Department of Political Science and International Relations, Korea University.

■ Editor: Park Han-soo_EAI Research Fellow

    Inquiries: 02 2277 1683 (ext. 204) hspark@eai.or.kr