South Korea, one of the leading democracies in Asia, has undergone democratic backsliding sparked by the declaration of martial law by President. As the President was impeached and the new government has established in accordance with constitutional procedure, this episode demonstrated South Korea’s democratic resilience. Nevertheless, it is still necessary to investigate backgrounds and implications of this political turmoil, thereby providing suggestions to avoid further deterioration of democracy in South Korea and other Asian countries.

With the critical awareness that effective solutions addressing democratic backsliding require a proper assessment of this phenomenon, ADRN conducted this research project, extending the scope from institutional reforms to behavior of political elites and citizens’ perceptions.

The report investigates contemporary questions such as:

● Does South Korean presidential systеm allow president overwhelming power?

● What is the negative synergy effect of hardline tactics by the president and legislature?

● Does political elites exacerbated conflicts, thereby provoking top-down democratic crisis?

● How has citizens’ confidence in democracy changed during political upheaval?

Drawing on a rich array of resources and data, this report offers country-specific analyses and suggests policy recommendations to enhance democratic resilience in South Korea and the larger Asia region.

Global Spread of Democratic Backsliding and South Korea

Sook Jong Lee

Senior Fellow, East Asia Institute; Distinguished Professor, Sungkyunkwan University

I. Global Spread of Democratic Backsliding

Since the late 2000s, the world has entered an era of democratic backsliding. Two global reports on democracy have drawn the attention of scholars and the media for their classification of regime types. The first report is the annual publication released every March by the Varieties of Democracy (V-Dem) Institute at the University of Gothenburg in Sweden. The 2025 report, which examined 179 countries, assessed that the quality of democracy experienced by individuals worldwide has regressed to the level of 1985, while the democratic status of individual countries has reverted to that of 1996 (V-Dem Institute 2025). Researchers at the V-Dem Institute describe the past 25 years as the “Third Wave of Autocratization.” Backsliding has been particularly evident in the areas of freedom of expression, freedom of association, clean elections, and the rule of law. It has been especially pronounced in Eastern Europe, South Asia, and Central Asia. Consequently, the number of democratic countries (88) fell below that of authoritarian regimes (91) for the first time in 2024—72% of the global population lives under authoritarian rule. Among democratic countries, only 29 are classified as liberal democracies, making this the smallest group among the four regime types: liberal democracy, electoral democracy, electoral autocracy, and closed autocracy.

The second report is the Democracy Index, published also every March by The Economist in the United Kingdom. Its 2025 report indicated a continuous decline in the average global democracy index since 2010, with backsliding particularly marked in civil liberties, the electoral process and pluralism, and the rule of law (Economist EIU 2025). Owing to this trend, only 25 countries—accounting for 15% of the 167 studied countries and territories—are classified as “full democracies,” representing 6.6% of the studied population. A total of 46 countries, accounting for 27.5% of the studied countries, were “flawed democracies,” representing 38.4% of the population. Regardless of whether flawed or not, these two categories together comprise 43% of the studied countries and 45% of their population. Meanwhile, 60 countries were categorized as “authoritarian regimes,” accounting for 35.9% of the countries studied and 39.2% of the population. The remaining 36 countries were classified as “hybrid regimes,” representing 20.6% of the countries and 15.7% of the studied population.

This global decline in democratic countries has revitalized academic interest in democratic backsliding. Some researchers have focused on the gradual nature of this process, which contrasts with the past abrupt collapses of democracy through military coups. Such gradual backsliding typically erodes democratic systеms slowly through legal frameworks and electoral mechanisms. A few key studies exemplify this transition. Bermeo (2016: 10) identifies three forms of democratic backsliding. The first is the “promissory coup,” in which elected governments are overthrown under the pretense of protecting the democratic order. The actors behind these promissory coups pledge to restore democracy through future elections. Notable examples include the military coups in Thailand and Myanmar. In Myanmar, the military staged a coup and declared a state of emergency on February 1, 2021, promising elections set to take place. But the military government has not kept its promise so far.

The second type is “executive aggrandizement,” wherein an elected government incrementally dismantles institutional checks on its power without changing the regime type, weakening opposition forces. Typically, these governments seize control of legislative or judicial bodies and use them to suppress the opposition. Most democratic backsliding in electoral democracies, such as in Turkey and Hungary, falls into this category. The third type involves the strategic “manipulation of elections,” which employs subtle tactics rather than outright rigging. These include denying candidates equal opportunities, limiting access to media, allocating public funds solely to the ruling party’s campaign, obstructing voter registration, or embedding loyalists in national electoral commissions. These tactics are often deployed in combination.

Scholars affiliated with the V-Dem Institute are publishing research focused on democratic backsliding and the restoration process. Lührmann, a leading scholar in this field, outlines the stages of autocratization. The first stage involves structural and contextual challenges that increase the risk of autocratization. At this point, citizens’ discontent grows due to economic crises, inequality, immigration tensions, political polarization, and the influence of social media. In the absence of democratic parties or processes, and when citizens’ commitment to democratic norms deteriorates, the second stage—anti-pluralism—emerges. If anti-pluralist parties and leaders successfully mobilize voters and win elections, this marks the “onset of autocratization.” Lührmann argues that backsliding can be reversed if checks and balances remain in operation and opposition—whether from citizens, civic groups, or political parties—against autocratization. Without such resilience, the final stage of democratic breakdown begins (Lührmann 2021).

Resilience—efforts that counteract backsliding—can be categorized into “onset resilience,” which prevents autocratization, and “breakdown resilience,” which prevents the collapse of democracy once autocratization has begun. Boese et al. (2021) analyzed 4,372 episodes in 64 democratic countries between 1900 and 2019 and found that the vast majority (98%) did not end in autocratization. However, they also found that once autocratization begins, the probability of democratic survival sharply declines, with only 23% avoiding the collapse of democracy. Notably, 60% of the 64 episodes occurred after 1993, following the Cold War. This study statistically analyzed the factors contributing to democratic resilience and found that judicial checks is key in preventing backsliding—validating the notion that the judiciary is “the last bastion of defense.” Economic development helps prevent the onset of backsliding. Nevertheless, economic development made little difference once backsliding had already begun. Furthermore, the study found that geographic surroundings of having other democratic countries as neighbors and a long history of democratization serve as protective factors against backsliding and collapse (Boese et al. 2021).

The case of South Korea offers a particularly valuable reference point for global research on democratic backsliding and resilience. The relationship between backsliding and resilience is inherently oppositional. South Korea will be a significant case for understanding how direction changes and how one trajectory gains momentum over the other.

II. Democratic Backsliding in South Korea

At the end of last year, democratic backsliding, a topic of global concern, occurred in South Korea. As one of Asia’s leading democracies, the declaration of martial law on December 3 by South Korean President Yoon Suk Yeol shocked the international community. Korean National Assembly nullified the martial law within a few hours and passed the bill impeaching President Yoon on December 14. Until the Constitutional Court judges upheld the bill unanimously on April 4, the country had experienced months of serious conflicts driven by intensified political polarization. A new government led by President Lee Jae Myung started right after June 3 snap presidential election. This half year grand episode showed South Korea’s democratic resilience by returning to democratic order through constitutional processes and the rule of law. Nevertheless, the whole process inflicted significant damage on its democratic systеm. This included mass protests from a deeply divided public, legal disputes, and the first instance of mob violence to the court house in the nation’s history. The reports referenced earlier did not overlook this development. The V-Dem Institute reclassified South Korea—previously consistently ranked as a liberal democracy since the early 1990s—as an electoral democracy (41st worldwide). The Economist also downgraded South Korea from a full democracy to a flawed democracy (32nd globally, 5th in Asia).

Multiple factors likely contributed to democratic backsliding in South Korea. At the institutional level, these include the excessive power of the presidency, legislative paralysis due to inter-party conflict, the judicialization of politics, and the politicization of the judiciary. At the societal level, contributing factors include political polarization, escalating social conflicts, the spread of disinformation, and the rise of extremist minority factions. Nonetheless, certain forces have contributed to resilience and have served as checks on democratic decline. Active civic engagement has consistently been a source of resilience during political crises, and adherence to constitutional order has functioned as a constraint on democratic erosion. This deterrence would not have emerged without the national pride Koreans place in their democratization achievements. Nevertheless, the imposition of martial law and the ensuing efforts to restore order served as a wake-up call to scholars of Korean democracy.

The research series “Democratic Backsliding in South Korea” is an academic response to these concerns. This study was initiated with the critical awareness that effective solutions require a proper assessment of democratic backsliding in South Korea. Most discussions have focused on institutional reforms, such as altering the power structure or addressing political polarization by modifying electoral laws. However, it remains uncertain whether institutional reforms alone can alter the behavior of political parties or politicians or improve political culture. With such a critical lens, four scholars analyzed institutional structures and political behavior together.

In the first paper of the series, Professor Jin Seok Bae analyzes whether the recurring signs of democratic backsliding in South Korea originate from the structural features of its presidential systеm. Professor Bae argues that the declaration of martial law in 2024 by then-president Yoon Suk Yeol and the resulting constitutional crisis were caused by a combination of characteristics of presidential systеm, asymmetrical development between political parties and civil society, and the prevailing political culture in South Korea. From a comparative perspective, South Korean constitution does not give president very strong power officially. However, the Korean president can enjoy informal power beyond legal boundaries. He or she can exert control over the legislative and executive bodies through influencing nomination of the ruling party candidates, budgetary allocation authority, appointments of vast positions in the public sector, and public opinion maneuver. Professor Bae also identifies the influence of the President Office over cabinet ministries, submission of ruling party to President authority, legislative gridlock, the rigidity of the single-term presidency, and conflicting dual legitimacy as contributing factors to democratic backsliding. These dynamics have weakened accountability mechanisms, reduced checks on presidential power, and turned political parties into mere electoral tools, while citizens have become objects of emotional mobilization. This pattern is becoming increasingly entrenched.

The author proposes a dual strategy for institutional reform. First, to reform the vertical power structure without requiring a constitutional amendment, he advocates internal democratization of political parties, increased transparency in the nomination process, and expanded citizen participation. Second, for reforms requiring a constitutional amendment, he suggests redistributing power between the president and the National Assembly, limiting emergency powers, and aligning electoral cycles. He emphasizes that as institutions, structures, and political culture are deeply intertwined, institutional redesign must be pursued alongside tangible behavioral changes and the expansion of democratic culture to eradicate the drivers of backsliding.

In the second paper, Professor Jung Kim contends that the constitutional crisis surrounding the declaration of martial law stemmed from the failure to uphold informal norms derived from constitutional provisions—specifically mutual tolerance and institutional forbearance. Professor Kim explains that President Yoon’s declaration of martial law was prompted by escalating tensions between an opposition-controlled National Assembly, which pursued multiple impeachment motions against senior executive officials, and the president, who repeatedly exercised his authority to request reconsideration of legislative bills. Kim argues that the prolonged use of “constitutional hardball tactics” by both sides led President Yoon to adopt a “constitutional beanball tactic” to break the political deadlock. Kim asserts that this choice was shaped by a politically polarized “national narrative” formed over decades of intense electoral competition between conservative and progressive blocs. Over the past ten years, overlaps between the supporter bases of the two major parties have declined, while affective polarization has intensified, increasing the emotional distance between the supporters of two different parties. Accordingly, electoral strategies have shifted from persuading moderate voters to mobilizing core supporters. This environment has rendered the oppositions and the president’s constitutional strategies viable for electoral gains. These developments have substantially eroded the democratic norm of institutional forbearance. President Yoon’s declaration of martial law effectively lowered the political cost of violating the norm of mutual tolerance. Therefore, Kim concludes that democratic backsliding in South Korea is, for the time being, inevitable.

In the third paper of the series, Professor Sunkyoung Park identifies the nature of South Korea’s democratic crisis as a “top-down democratic crisis.” The crisis is not driven by a change in public perception or behavior but by political elites who have exacerbated political conflicts and lost the capacity to resolve them. To explain elite behavior, Park draws on Juan Linz’s concepts of “loyal democrats” and “semi-loyal democrats.” Several politicians in the People Power Party (PPP) promoted claims of electoral fraud. Moreover, a few PPP politicians defended the rioters who broke into the Seoul Western District Court. In this sense, he ruling PPP at the time included a larger proportion of semi-loyal democrats who are responsible for the country’s democratic backsliding.

Park identifies three causes for the emergence of the “top-down democratic crisis.” The first is a phenomenal cause: the party’s democratic self-correction ability weakened due to the diminished influence of moderate conservatives, triggered by repeated electoral defeats in the Seoul metropolitan area and the rise of regionally-based hardliners within the party. The second cause is a decline of exchanges and dialogues between politicians of different parties. Accordingly, opportunities for mutual understanding and political learning have been decreased. Analysis of legislative research group activities across different parties confirms this decrease. The third cause is a change in the incentive structure. Politicians have become increasingly responsive to the voices of extreme minority groups, driven by pressure from aggressive supporters and certain biased new media, rather than representing the broader citizenry. Park argues that this shift has enabled semi-loyal democrats to strengthen their footholds within political parties.

In the fourth paper, Professor Woo Chang Kang asserts that recent democratic backsliding in South Korea is better characterized as “top-down backsliding.” Rather than “bottom-up backsliding.” While the top-down backsliding involves the consolidation or expansion of power by political elites, particularly executive leaders, the bottom-up backsliding occurs when citizens voluntarily withdraw normative support for democracy or reject its values. Kang analyzes data from seven surveys conducted between 2003 and 2025 to assess South Korean citizens’ attitudes toward democracy. He finds that support for democracy in South Korea has steadily increased and remained steadfast despite recent political turmoil. Democracy has essentially become “the only rule of the game” in South Korean society.

However, the 2025 survey revealed a decline in support for democracy and a rise in support for authoritarianism among the males of industrialization generation, millennials, and Gen Z. Conversely, it showed the increased support for democracy among Generation X men, millennial women, and Gen Z women, resulting in minimal change in the overall response ratio. Although support for democracy among millennial and Gen Z males was relatively lower, and a significant decline occurred during the time of post- martial law turmoils, still 60–70% of these groups remained supportive of democracy. Kang argues that this continued South Koreans’ support for democracy will serve as a crucial asset in overcoming any top-down democratic backsliding.

III. Future Directions for Research

This study serves as a foundation for more in-depth investigations into democratic backsliding in South Korea. Two presidents from the conservative party have been impeached within eight years, a reality that undermines the international image of “K-democracy.” The conservative party must undertake painful self-reflection, acknowledging that both presidents were from its ranks, and initiate significant internal reform and reconstruction. The progressive, now ruling, party must also reflect on criticisms that it has engaged in hardball politics by taking advantage of its supermajority in the legislature. Without power-sharing and cooperation with opposition parties, the combination of a dominant ruling party and a strong executive can risk testing the resilience of Korean democracy once again.

Judicial independence is widely regarded by scholars as the final safeguard against democratic backsliding. Therefore, legal and institutional reforms must be carefully examined to ensure the judiciary is free from partisan influence. Public protests, which have historically served as engines of democratic restoration during crises, must now evolve from reactive street mobilization to proactive institutional political engagement to prevent democratic backsliding in advance. As the roles of the ruling and opposition parties inevitably alternate, and public support is inherently fluid, the political arena must transcend short-term partisan interests. Rather, politicians must pursue long-term, bipartisan political reform to make democracy working for citizens.

Furthermore, greater attention must be given to the global trend of democratic backsliding. Comparative studies are academically necessary to identify similarities and differences in democratic backsliding between countries. It is equally important to determine the conditions under which democratic recovery succeeds or fails. Democratic backsliding is contagious, just as democratic resilience can inspire others. When more countries succeed in halting backsliding, the prospect of a global democratic renewal becomes more promising. ■

References

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

Boese, Vanessa A., Amanda B. Edgell, Sebastian Hellmeier, Seraphine F. Maerz, and Staffan I. Lindberg. 2021. “How democracies prevail: democratic resilience as a two-stage process.” Democratization 28, 5: 885-907.

Economist EIU. 2025. “Democracy Index 2024: What’s wrong with representative democracy?” https://www.eiu.com/n/campaigns/democracy-index-2024/ (Accessed May 14, 2025)

Lührmann, Anna. 2021. “Disrupting the autocratization sequence: towards democratic resilience.” Democratization 28, 1: 22-42.

V-Dem Institute. 2025. Democracy 2025: 25 Years of Autocratization - Democracy Trumped? March 2025. https://www.v-dem.net/documents/61/v-dem-dr__2025_lowres_v2.pdf (Accessed May 14, 2025)

Democratic Backsliding in South Korea:
Institutional Root of Crisis in a Presidential Systеm

Jin Seok Bae

Professor, Gyeongsang National University

I. Introduction: Did the Crisis Stem from the Power Structure?

The declaration of martial law by President Yoon in December 2024 cannot be explained merely as a governance failure by a single adm?nistration or a temporary episode of political instability. This event quickly escalated into a series of constitutional crises, from the impeachment of the president to early presidential elections. It also indicated that the political systеm of South Korea was vulnerable to structural threats that could undermine the foundations of democracy. Subsequently, did South Korea’s democratic crisis stem from the power structure of the presidential systеm, as political leaders and the media unanimously suggested? Or was it the result of President Yoon Suk Yeol’s unique personal traits?

This article does not address this issue as a simple dichotomous choice. Rather, it comprehensively analyzes how the institutional design of the presidential systеm, the organization of political parties, the nature of political culture, and the governing behavior of political actors—who operate within that political culture—interact to paralyze democratic functioning. In particular, the recurring patterns of a “strong president–weak party” structure and confrontational politics rooted in populist affective mobilization have compounded the risks inherent in the presidential systеm, continuously weakening democratic foundations.

This article will examine (i) how structurally concentrated presidential power operates institutionally, (ii) why political parties and the National Assembly fail to function as independent political actors, and (iii) how these institutional operations manifest as democratic backsliding in political practice.

II. Democratic Backsliding or Crisis?

As many researchers point out, democratic backsliding differs from democratic breakdown. Democratic backsliding is not a violent overturn but a systеmic change characterized by the gradual erosion of democracy from within. This process occurs when politicians, who obtain office through democratic elections, undermine democratic values and principles while operating within the formal boundaries of the legal systеm. During this process, executive power expands, harassment of the opposition increases, and suspicions arise regarding state interference in the electoral process (Bermeo 2016).

South Korea has also been discussed within the framework of democratic backsliding. As South Korea experienced the impeachment of President Park Geun-hye and the subsequent adm?nistrations of Moon Jae-in and Yoon Suk Yeol, signs of democratic backsliding were detected. Although the power of the executive body grew, the responsiveness of the executive and the president to legislative oversight by the National Assembly declined. The use of prosecutorial powers to investigate the opposing party further destabilized the political world. Consequently, executive power inevitably influenced the democratic electoral process (Kwon 2023).

Most importantly, the entire backsliding process followed political bloc lines. Emotional polarization has replaced ideological division, and elections have become emotional proxy wars between opposing factions. Citizens were reduced from active political agents to objects of mobilization, rendering participatory democracy virtually paralyzed (Song 2025). Furthermore, the fierce collision between the president and the opposing party over presidential vetoes on legislative bills and impeachment eroded the two unwritten rules of South Korean politics: institutional forbearance and mutual respect. Both sides exercised their legal authority without restraint. As politicians and civilians became entrapped in emotional polarity, the political environment turned confrontational and hostile, leaving no room for compromise or cooperation.

President Yoon’s declaration of martial law elevated South Korea’s democratic crisis, previously discussed under the framework of democratic backsliding, into a new dimension. Historically, discourse on democratic backsliding has emphasized the gradual erosion of democracy, mobilized through legal means. However, the prevailing view is that this incident was an attempted self-coup—more aligned with democratic breakdown. As it involved an effort to overturn democracy through illegal and violent means, it clearly crossed the definitional threshold of democratic backsliding. Ultimately, citizens and the National Assembly thwarted the attempt at martial law. Nevertheless, the episode revealed that even elected leaders in democratic states may contemplate a self-coup to expand their power. Alarmingly, South Korean democracy, once regarded as a model of democratization, came close to not merely democratic backsliding but outright democratic breakdown due to a violent self-coup.

The circumstances confirmed during the impeachment proceedings following the declaration of martial law only deepened South Korea’s democratic crisis. Emotional polarization effortlessly transgresses democratic and constitutional norms. Although the ruling party deliberately incited this sentiment, public aversion to opposition politicians was also evident in survey data, which showed a refusal to accept the impeachment of a president who had clearly violated the Constitution and the law. Some emotionally charged extremist groups even resorted to violence, attempting to seize a courthouse during President Yoon’s investigation and arrest. These extremist groups did not infiltrate the ruling People Power Party; rather, several politicians actively incited them. The emergence of extremist forces—who physically attacked constitutional institutions and are now approaching the center of power through the mainstream conservative party—is a stark warning for South Korean democracy on multiple levels.

III. Institutional Vulnerabilities of the Presidential Systеm and Democratic Backsliding

Is the democratic crisis in South Korea an institutional or structural problem, or does it stem from the flaws of political leaders? Opinions on this question are sharply divided. While some attribute the crisis to the general characteristics of the presidential systеm or the specific nature of the South Korean presidential systеm, others emphasize the lack of governing competence or leadership deficiencies in individual presidents.

First, some prioritize a structural approach. Choi Kwang-Eun (2025) argues that the structural dysfunction inherent in the presidential systеm poses the key threat to democracy. In this view, President Yoon’s attempt at a self-coup cannot be interpreted as merely a personal or exceptional act of a president with unique traits. Instead, critics highlight the excessively concentrated power in the presidency and the weakness of the legislative and judicial branches in checking that power as structural vulnerabilities of the South Korean Constitution. In this context, the political establishment and the media have described the systеm as an “imperial” presidency. The proposed remedy has consistently been the dispersion of presidential power.

Conversely, Yoon Yeojoon and Han Yoonhyung (2025) present a different approach. They argue that responsible and competent leadership can overcome institutional constraints. According to them, the core problem lies in the systеm that allows political leaders lacking in governance philosophy and adm?nistrative competence to become presidents. The solution, they argue, should not be sought in institutional reform but in the selection of qualified leaders. The repeated emergence of “unqualified leaders” has contributed to ongoing democratic instability in South Korea.

Song Ho-Keun (2025) proposes a third, integrative approach. He notes that the reformist overreach of the Roh Moo-hyun adm?nistration, the authoritarian tendencies of the Park Geun-hye adm?nistration, and the attempted declaration of martial law under President Yoon Suk Yeol occurred under the same presidential systеm. He contends that institutional dysfunction and presidential behavior are not separate factors but mutually reinforcing. Similarly, Park Sanghoon (2018) acknowledges that these events may stem from individual presidents’ conduct but argues that the structural conditions enabling repeated patterns of such behavior—presidential office-centered governance and the subordination of political parties—must be addressed through comprehensive reform.

Interpretations of the martial law declaration also differ between actor-based and structural perspectives. These perspectives diverge in identifying the source of the crisis and prescribing reform priorities and strategies for prevention. From the actor’s perspective, President Yoon’s poor governance and ineffective communication skills are identified as the primary causes, with solutions focused on cultivating and selecting political leaders who demonstrate leadership and effective communication skills. Conversely, the structural perspective attributes the crisis to inadequate constitutional provisions on martial law and presidential office-centered governance. In this view, solutions include constitutional amendments focused on decentralizing executive power and enacting legal reforms.

The “imperial presidency” is also debated within this framework. Critics who regard presidential power as excessively concentrated argue that such centralization impedes democratic consolidation. According to this view, the South Korean president possesses excessive constitutional authority compared to presidents in other systеms. However, others argue that the issue lies not in the extent of presidential powers but in the weakness of party politics. They contend that the constitutional powers of the South Korean president are not unusually extensive. Paradoxically, despite the difference in perspectives, South Korean politics often demonstrate a pattern in which the president appears “imperial” at the beginning of the term but “vulnerable” at its end (Bae and Park 2018).

Is the South Korean president’s power exceptionally strong compared to the situation in other countries, justifying the label “imperial”? The index developed by Shugart and Carey (1992) remains one of the most widely used measures of presidential constitutional power. This index classifies presidential power into legislative and non-legislative dimensions and quantifies each item on a four-point scale. According to this index, the South Korean president scores 9 points for legislative powers and 2 points for non-legislative powers, totaling 11 points. Conversely, Brazil and Argentina each scored 19 points, and Chile scored 14 points. The United States scored 12 points, slightly higher than South Korea’s 11 points. Based on this comparison, the constitutional powers of the South Korean president are not sufficiently strong to be characterized as “imperial.” There is little justification for this label from a strictly constitutional perspective.

Nevertheless, the perception of the South Korean presidency as “imperial” persists due to its actual political (de facto) influence, which exceeds its formal constitutional limits. The president exercises wide-ranging informal authority, including control over candidate nominations, leadership in budget formulation, management of key appointments, and influence on public discourse. Especially in relations with the ruling party, the president wields substantial authority surpassing that of a party leader, with the ability to intervene in both the nomination of National Assembly candidates and the overall structure of party primaries. Furthermore, the “Blue House Government,” in which key aspects of governance are handled not by ministers but by “senior presidential secretaries,” also supports the concentration of power in the president, even though it is not stipulated in the Constitution. While constitutionally moderate, South Korea’s presidency enables highly centralized governance in practice, fueling continued criticism of the systеm as “imperial.”

Fundamentally, the problem of the South Korean presidential systеm extends beyond the scope defined by the Constitution and is amplified in practice. Rather than attributing the cause solely to the institution or leadership, the structural vicious cycle generated by their interaction must be recognized. The combination of the president’s legal and extra-legal authority, the institutional inability to check this authority, and the strategies of political actors who exploit this environment places South Korean democracy in a state of recurring crisis.

IV. Relationship Between the President and the Ruling Party

South Korean president has maintained control over the ruling party through constitutional authority and various unofficial means. Even after democratization in 1987 and throughout the “Three-Kims” era, the president continued to exert significant influence over the ruling party and the broader government as its leader. This influence contributed to the perception of an “imperial president.” Although the principle of separation between the party and the government was reinforced during the Roh Moo-hyun adm?nistration, the president’s dominance over the ruling party remains structurally embedded. In South Korea, presidential dominance over the ruling party is primarily exercised through three channels: authority over the nomination of National Assembly candidates, control over appointments, and influence in budget formulation.

First, the authority over candidate nominations is not a formal presidential power. However, in practice, it has become a powerful tool for dominating the ruling party. As regionalism deepened in South Korea after democratization, regional political elites came to regard party nominations as a key mechanism for maintaining their political influence. Consequently, National Assembly candidates prioritized securing party nominations over competing in the general elections. This dynamic created a structure in which the president’s preferences are naturally reflected in the nomination process.

Second, the president’s appointment authority is another critical means of controlling the ruling party. South Korean president holds unilateral authority over numerous appointments, including ministers, vice ministers, and heads of public institutions. The institutional framework permits National Assembly members to serve concurrently as cabinet members. Therefore, the president’s appointment power substantially undermines the fundamental principle of checks and balances between the legislative and executive branches.

Finally, the budget formulation authority serves as a vital channel for the president’s control of the ruling party. Unlike in a pure presidential systеm, the South Korean president exercises significant power in the formulation of the national budget. Thus, the president can influence individual district representatives by allocating funds to their districts. For these representatives, securing a regional budget is directly connected to re-election prospects. Pork barrel spending, also known as “note budgets,” is evidence of the direct influence of the president’s political power. These structural dynamics were apparent when the party and the presidential office were integrated during the Lee Myung-bak adm?nistration and when non-mainstream figures were excluded from nominations under the Park Geun-hye adm?nistration (Hur 2017).

While the institutional framework of the South Korean presidential systеm incorporates the principle of separation between the party and the government, in practice, the president controls the ruling party through nomination, appointment, and budgetary authority. This creates a structural mechanism in which power is concentrated in the individual president.

Such concentration of power distorts the function of elections in South Korean politics, intensifying political tension and hostility. The president’s unilateral control over appointments generates expectations and anxieties that an electoral outcome will result in a wholesale redistribution of state resources and key positions. Consequently, elections have evolved from arenas of compromise and competition into zero-sum contests for political survival. Political parties are increasingly prioritizing the pursuit of power over cooperation. In this context, coalition-building disappears from the political landscape, and political logic shifts toward emotional mobilization and structural hostility.

The recurring pattern of labeling previous adm?nistrations as sources of “deep-rooted evils,” followed by investigations and punitive measures, perpetuates a cycle of political retaliation. This reinforces an adversarial perception wherein opposing political camps are viewed not as competitors but as criminal entities, entrenching a pattern of loyalist and confrontational politics. According to research by the Pew Research Center, South Korea and the United States rank among the lowest in terms of perceived social cohesion between party supporters (Silver 2022). This can be understood as the outcome of institutional power structures and entrenched political polarization.

V. Democratic Backsliding and Recurrent Leadership Crises

All South Korean presidents experience a significant decline in political momentum and approval ratings during the latter half of their terms. This “lame duck” phenomenon is an inevitable product of the institutional framework. Owing to the single-term systеm, presidents struggle to establish a long-term political foundation or maintain stable, cooperative relationships with the National Assembly. Frequent general and regional elections further dampen national governance momentum. Presidential power peaks in the early months of the term but inevitably diminishes over time. While ambitious reform agendas may be pursued at the outset, presidents often become politically isolated in later stages, leaving governance defensive. This imbalance grows from the structural disharmony between the presidential term and the legislative election cycle.

The South Korean presidential systеm frequently results in a divided government, as presidential and legislative elections are held separately. Although this resembles the United States model, South Korea experiences more pronounced political conflicts and legislative gridlock due to its weak party systеm and internal factionalism. Clashes between presidential initiatives and legislative checks are routine. When compounded by internal divisions within the ruling party or declining presidential approval ratings, the president becomes increasingly politically isolated.

In response, presidents adopt strategies to circumvent or bypass legislative oversight. They issue executive orders, implement presidential decrees, or mobilize public opinion to bolster support (O’Donnell 1994). These approaches erode the democratic principle of separation of powers, weaken the legislature’s role, diminish horizontal accountability, and exacerbate the concentration of power in the president. Ultimately, the balance between executive and legislative authority collapses, and political conflict and social polarization intensify.

South Korean presidential systеm centers on a president who is directly elected by the people and endowed with substantial authority, assuming a high degree of accountability. However, this accountability is largely vertical between the president and voters, while horizontal mechanisms of oversight—such as the National Assembly, judiciary, and media—remain weak. When backed by high public approval, the president’s unilateral and authoritarian policy decisions face little effective resistance from the legislative branch or political parties. In effect, the separation of powers—fundamental to a functioning democracy—is severely undermined, putting democratic governance at risk (Park 2018).

Contemporary democracy depends on rational and transparent policy debates that transcend mere electoral participation. Nevertheless, in South Korea, such debates have become increasingly emotional and confrontational. When presidents use the media to pressure opposition forces or neutralize the authority of the National Assembly, political discourse degenerates from rational deliberation into polarized emotion. Song Ho-Keun (2025) describes this phenomenon as “the politics of emotional mobilization,” warning that the foundations of deliberative democracy are eroding and the democratic crisis is accelerating.

VI. Institutional Reform and the Task of Restoring Democracy

The democratic crisis in South Korea is complex, with actors, institutions, structures, and culture deeply intertwined. Approaches to institutional reform must avoid overly simplistic perspectives. In addressing the weaknesses of South Korean democracy, it is critical to distinguish between areas that can be reformed without constitutional revision, those that require constitutional amendments, and those that are unlikely to be resolved even through such changes.

Some areas of political reform can be achieved through the revision of existing laws without altering the current constitutional power structure. A representative example is electoral law reform. Restructuring the electoral systеm could be vital in improving proportionality between vote share and seat allocation, enhancing the accountability of elected officials, stabilizing the party systеm, and mitigating the winner-takes-all dynamic that currently dominates politics in South Korea.

Amending the Political Parties Act is also crucial. The excessive regulations embedded in this Act constrain the activities of political associations representing diverse regional stakeholders. Establishing a systеm that permits the creation of regional parties for participation in local elections could significantly contribute to political decentralization. If the Public Official Election Act and internal party regulations were revised to ensure more democratic nomination procedures and party governance, a vertically subordinate relationship between the president and the ruling party may be mitigated.

Meanwhile, certain structural issues require constitutional revision. The power distribution among the president, the prime minister, and the National Assembly is defined by the Constitution and cannot be altered without amending it. Similarly, core powers—such as authority in emergencies, military command, and budget formulation—are constitutionally mandated and require amendment for change. The misalignment of presidential and legislative election cycles is another issue that can only be solved through constitutional revision. Decisions about whether to hold these elections simultaneously or to treat legislative elections as a midterm assessment have a significant impact on power dynamics within the president’s affiliated party. Simultaneous elections would likely strengthen presidential authority while clarifying accountability between the president and the National Assembly. Midterm elections, while dispersing power, could increase the risk of prolonged political deadlock due to ambiguities in responsibility under a divided government. If the power structure were to shift significantly—toward a parliamentary, semi-presidential, or four-year presidential systеm with reelection—restoring public trust and achieving societal consensus would be essential.

Ultimately, institutional reform aimed at restoring democracy cannot be a linear process. Constitutional amendments and legislative reforms must proceed concurrently across multiple dimensions.

VII. Conclusion: A General Assessment of Democratic Crisis and Response Strategies

The crisis facing South Korean democracy results from a complex interaction of factors and cannot be attributed to any single cause. It cannot be fully understood through analyses focused solely on individual failings, such as the moral shortcomings or leadership deficiencies of specific presidents. Rather, the current situation is the product of the complex interplay between structural flaws in the presidential systеm, unique cultural traits of South Korean politics, and the capabilities of individual political leaders.

This article began with the question, “Is the democratic crisis facing South Korea stemming from the power structure or the incompetence of a leader?” In summary, this cannot be answered through a dichotomous framework. Although the South Korean presidential systеm formally adheres to the principle of separation among legal, adm?nistrative, and judicial powers, in practice, it produces a concentration of power in the presidency. This structure undermines the core democratic principle, weakens checks and balances, and results in isolated presidential decision-making and rigid governance.

This analysis has examined multiple layers of the crisis. These include the excessive concentration of power in the presidency, the breakdown of institutional oversight, the immaturity of party systеms oriented around individual presidents, recurring leadership crises, the misalignment of electoral cycles, and the expansion of emotional, political mobilization. Collectively, these elements illustrate the extent to which South Korean politics have diverged from democratic ideals.

A three-pronged strategy must be pursued. First, for areas that do not require constitutional amendments, democratizing political parties and institutionalizing citizen participation must be accomplished through voluntary reform efforts and legislative change. Second, for structural challenges necessitating constitutional reform, long-term social consensus and political leadership must guide a reconfiguration of the power structure. This includes redefining the relationship between the president and the National Assembly, evaluating alternative systеms such as parliamentary or semi-presidential models, aligning electoral cycles, and introducing mechanisms such as a vote of no confidence. Third, long-term strategies—such as civic education, the restoration of public deliberation, and political education—are necessary to address entrenched cultural bottlenecks. These include the absence of institutional forbearance and mutual respect, political polarization driven by negative partisanship, and emotionally charged political mobilization.

At its core, democracy is not secured by institutions alone but realized through political culture, civic behavior, and a continuous attunement between systеms and social reality. The democratic crisis in South Korea reflects the failure of such alignment in institutional and cultural terms. Overcoming this crisis requires comprehensive reform across structural, institutional, and cultural dimensions. ■

References

Bae, J. S., and Park, S. 2018. “Janus face: the imperial but fragile presidency in South Korea.” Asian Education and Development Studies 7, 4: 426-437.

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

Dahl, R. A. 1989. Democracy and Its Critics. New Haven: Yale University Press.

Ganghof, S. 2023. “Justifying types of representative democracy: a response.” Critical Review of International Social and Political Philosophy 27, 2: 282-293.

Hur, Mi-yeon. 2017. “South Korea’s Blue House Scandal.” E-International Relations. February 21. https://www.e-ir.info/2017/02/21/south-koreas-blue-house-scandal/ (Accessed April 30, 2025)

Kellam, M., and Benasaglio Berlucchi, A. 2023. “Who’s to blame for democratic backsliding: populists, presidents or dominant executives?” Democratization 30: 1-21.

Linz, J. J. 1990. “The perils of presidentialism.” Journal of Democracy 1, 1: 51–69.

Mainwaring, S., and Shugart, M. S. 1997. Presidentialism and Democracy in Latin America. Cambridge: Cambridge University Press.

O'Donell, G. A. 1994. “Delegative democracy.” Journal of Democracy 5, 1: 55-69.

Samuels, D. J., and Shugart, M. S. 2010. Presidents, Parties, and Prime Ministers: How the Separation of Powers Affects Party Organization and Behavior. Cambridge: Cambridge University Press.

Shugart, M. S., and Carey, J. M. 1992. Presidents and Assemblies: Constitutional Design and Electoral Dynamics. Cambridge: Cambridge University Press.

Weyland, K. 2020. Populism: A Political-economic Approach. Oxford: Oxford University Press.

Choi, Jang-Jip. 2003. Democracy after Democratization (in Korean). Seoul: Humanitas.

Choi, Kwang-Eun. 2025. The End of Presidential Systеm: Beyond Rebellion and Toward the 7th Republic (in Korean). Pyeongtaek: Jeongjikhan Mosaek.

Kwon, Hyeok Yong. 2023. “Democratic Backsliding in South Korea (in Korean).” Korean Political Science Review 57, 1: 33-58.

Park, Sanghoon. 2018. Blue House Government (in Korean). Seoul: Humanitas.

Song, Ho-Keun. 2025. The Anthology of Hostility Politics (in Korean). Paju: Nanam Publishing House.

Yoon, Yeo-joon, and Yoon-hyung Han. 2025. Qualification of the President (in Korean). Seoul: MG Channel.

The Crisis of South Korean Constitutional Democracy Before and After Martial Law

Jung Kim

Professor, University of North Korean Studies

Why did President Yoon Suk Yeol invoke the power to declare martial law? What are the implications of the series of impeachment motions initiated by the National Assembly and the series of requests for reconsideration exercised by the president before the declaration of martial law in relation to the constitutional order of South Korea? How did the National Assembly’s impeachment motion and the Constitutional Court’s ruling on the impeachment following the declaration of martial law affect democracy in South Korea?

I. Surface of the Constitutional Crisis: Collapse of the Norms of Mutual Toleration and Institutional Forbearance

The Constitution does not guarantee the proper functioning of democracy. Owing to the inherent conceptual gaps and ambiguous meanings common to all legal texts, constitutional provisions alone cannot prevent the transition from democracy to authoritarianism. In successful democracies, informal norms—although not stipulated in constitutional provisions—emerge from them and regulate political conduct. “Mutual tolerance” and “institutional forbearance” are two such informal norms essential to the stable functioning of democracy. Mutual tolerance refers to “the acknowledgment that political rivals exist as long as they abide by the Constitution and have the right to compete for power and govern society.” Institutional forbearance is defined as “the discreet exercise of legal authority” and reflects the recognition that “political power, even within legal boundaries, can render the existing systеm unstable.” When democracy is operating smoothly, the importance of these norms is less visible. However, once democracy begins to falter, violations of these norms become more evident. The failure of mutual tolerance and institutional forbearance to operate as regulatory norms of political behavior signals that democracy is in danger (Levitsky and Ziblatt 2018).

South Korean Constitution includes provisions that support the norm of institutional forbearance among political parties, such as the National Assembly’s right to initiate impeachment motions against high-ranking executive and judicial officials (Article 65 of the Constitution) and the president’s right to request reconsideration of legislation passed by the National Assembly (Article 53 of the Constitution). High-ranking officials in the executive and judiciary, including the president, are expected to exercise restraint in the use of their powers, aware that abuse or misuse could lead to impeachment. This expectation constitutes the constitutional norm of institutional forbearance. Similarly, the National Assembly is expected to wield legislative power with caution, understanding that the president may respond with a request for reconsideration when legislation introduces radical policy changes. These rights must not be invoked frequently if they are to fulfill their constitutional purpose (Helmke, Kroeger, and Paine 2022).

When either mechanism is overused, the inherent constitutional norm of institutional forbearance collapses, impeding the smooth functioning of democracy. If political parties breach this norm, they are said to have adopted a “constitutional hardball tactic.” Constitutional hardball tactic describes the weaponization of constitutional means to pursue partisan advantages, undermining institutional forbearance (Tushnet 2025).

South Korean Constitution also presupposes the norm of mutual tolerance. Violations of this norm include deferring National Assembly elections, which are scheduled every four years (Article 42 of the Constitution), or presidential elections, held every five years (Article 70 of the Constitution), as well as refusing to accept the results of such elections. Other violations include invoking the right to declare martial law without satisfying all required conditions (Article 77(1) of the Constitution), obstructing the National Assembly’s right to demand the lifting of martial law (Article 77(5) of the Constitution), or rejecting the Constitutional Court’s ruling on impeachment (Article 111 of the Constitution). Each of these actions implies a rejection of mutual tolerance, which is indispensable for South Korean democracy to function.

When political parties engage in behavior that breaches the norm of mutual tolerance, they are said to have adopted a “constitutional beanball tactic.” Constitutional beanball tactic refers to the weaponization of constitutional tools for partisan purposes, destroying mutual tolerance in the process (Shugerman 2019).

Figure 1. National Assembly Impeachment Motions and Presidential Vetoes since Democratization

Source: “Impeachment cases in the Republic of Korea” under the category “Impeachment,” Wikipedia. https://ko.wikipedia.org/wiki/탄핵#대한민국 (Accessed March 24, 2025); “Presidential Requests for Reconsideration” under the category “Reconsideration,” Wikipedia https://ko.wikipedia.org/wiki/거부권 (Accessed March 24, 2025)

Figure 1 presents a bar graph depicting the number of impeachment motions initiated by the National Assembly and presidential requests for reconsideration of legislation from 1988 to 2024. The dark gray bars indicate the number of impeachment motions, while the light gray bars represent the number of presidential reconsideration requests.

In the 33 years preceding President Yoon Suk Yeol’s inauguration, the National Assembly filed 20 impeachment motions against high-ranking executive and judicial officials, averaging approximately 0.6 per year. The highest annual count was three, recorded in 2007 and 2020. Conversely, during the two years and six months after President Yoon assumed office in 2022, the National Assembly submitted 29 impeachment motions against high-level executive and judicial officials. This represents a 20-fold increase compared to the previous annual average, rising to approximately 11.6 motions per year. Notably, there were 11 motions in 2023 and 18 in 2024.

Before President Yoon’s term, there had been 16 presidential requests for reconsideration of legislation. During his tenure, this number rose to 33. The previous average of approximately 0.5 requests per year surged to about 13.2 per year, marking a nearly 30-fold increase. Previously, the highest annual record was four cases in 1989; this record was surpassed with eight cases in 2023 and 25 in 2024.

As demonstrated by the dramatic increase in impeachment motions and reconsideration requests, it is difficult to deny that the opposition party and the president engaged in clear violations of institutional forbearance, resorting to what may be described as “over-utilization” of constitutional powers. By the time President Yoon invoked the power to declare martial law, one of the core constitutional norms essential to the healthy functioning of liberal democracy—institutional forbearance—had already significantly deteriorated.

In summary, President Yoon’s declaration of martial law resulted from an escalating confrontation between the opposition party, which controlled the National Assembly and initiated a series of impeachment motions against senior executive officials, and the president, who repeatedly exercised his authority to request reconsideration of legislative bills passed by the National Assembly. In response to the opposition’s “constitutional hardball tactics” involving excessive use of its impeachment power, the president resorted to “constitutional hardball tactics” by excessively invoking the right to request reconsideration. This stalemate continued over an extended period. The president’s declaration of martial law constituted a “constitutional beanball tactic” intended to break the prolonged political deadlock. In hopes of “restoring” democracy, President Yoon made a paradoxical decision to suspend constitutional order.

II. Deeper Layer of the Constitutional Crisis: Polarization of the National Narrative

In his address declaring martial law on December 3, 2024, President Yoon justified his “constitutional beanball tactic” as follows:

I declare martial law to protect the Republic of Korea from the threats of North Korean communist forces to immediately eradicate the unscrupulous pro-Pyongyang anti-state forces that pillage the freedom and happiness of our people and protect free constitutional order. For this, I will certainly eradicate such anti-state forces and the culprits of the country’s ruin who have committed evil acts up until now. It is an inevitable measure to guarantee the people's freedom, safety, and national sustainability against the actions of anti-state forces seeking to overthrow the systеm.

President Yoon labeled the opposition party as “anti-state forces” and targets to be eradicated. The norm of mutual toleration—one of the constitutional principles essential to the effective functioning of democracy—had already collapsed (Levitsky and Ziblatt 2018: 8).

A longer-term perspective is required to identify the root cause of the collapse of the constitutional norms of mutual toleration and institutional forbearance. Since democratization in 1987, South Korea transitioned from a political systеm that suppressed social conflict to one that exposed it. The discontent among citizens that had been previously suppressed began to surface while political elites selected issues to maximize votes. As repeated elections aligned public grievances with elite vote-maximization strategies, the party systеm gradually came to reflect society’s core political division. South Korea lacked traditional sources of political conflict found in premodern societies, such as ethnicity, race, language, or religion. Its modern political conflicts were also relatively moderate, including class, the rural–urban divide, environmental issues, and human rights. The traditional social cleavages around which political parties typically organize were largely absent. That allowed democratization to take root without major societal upheavals such as civil wars or large-scale unrest. However, the cost has been the current “pernicious polarization” of political parties (Song 2025).

Since the Roh Moo-hyun adm?nistration in 2003, partisan polarization between progressive and conservative parties has deepened to unprecedented levels. In the conservative narrative, citizens who supported reconciliation with North Korea were marginalized. In the progressive narrative, those who pursued rapprochement with Japan were excluded. Emotionally charged antagonism between the two camps became widespread.

When the discursive frameworks of political elites project emotionally charged narratives in which the progressive and conservative blocs refuse to recognize each other as legitimate, partisan competition in South Korea descends into pernicious polarization. When a progressive party assumes power, hostility among conservative supporters intensifies and vice versa. In such a democratic context, elections become emotionally driven confrontations rather than contests over policy. Consequently, when democratic norms clash with partisan interests, many political elites tend to adopt a “partisanship before democracy” approach. This trend signals the onset of democratic backsliding in South Korea (Kim 2023).

The conservative and progressive blocs in South Korea promote political narratives that combine a nationalist structure—pitting “in-group” against “out-group” with a populist structure that casts political elites as adversaries of the people. This fusion of populism and nationalism fragments the public into supporters of the “People Power Party” and the “Democratic Party,” reinforcing a partisan polarization of the nationalist narrative that turns citizens against one another. President Yoon’s decision to employ a constitutional beanball tactic occurred amid intensifying constitutional hardball tactics between the two blocs, rooted in polarized nationalist narratives that have solidified over more than half a century since the presidency of Roh Moo-hyun (Cho and Hur 2025).

III. Democratic Retrogression in the Republic of Korea: Partisan Divergence and Sorting

A more accurate understanding of national narrative polarization requires an analytical distinction between partisan divergence and sorting.

First, partisan divergence refers to growing heterogeneity between two ideological or affective blocs. Ideological partisan divergence indicates a widening policy gap between voters who support progressive parties and align with progressive values and voters who support conservative parties and align with conservative values. Affective partisan divergence refers to a widening emotional divide between voters who support progressive parties while holding a strong aversion toward conservative ones and voters who support conservative parties and feel a strong aversion toward progressive ones.

Conversely, partisan sorting refers to increasing homogeneity within a party that was previously internally divided ideologically or affectively. Ideological partisan sorting occurs when a growing proportion of progressive (or conservative) party voters also support corresponding ideological values. Affective partisan sorting occurs when a growing proportion of progressive (or conservative) party voters also hold a strong aversion toward the opposing party (Kim 2022).

Figure 2 depicts the kernel density estimates of ideological and affective partisan divergence among South Korean voters in 2012 and 2022. In the diagram for ideological partisan divergence on the left, the horizontal axis ranges from 0, representing the highest level of agreement with progressive values, to 10, denoting the highest level of agreement with conservative values. In the diagram for affective partisan divergence on the right, the horizontal axis ranges from 0, indicating maximum aversion to the conservative party, to 10, indicating maximum aversion to the progressive party. The progressive party favorability score (0–10) is subtracted from the conservative party favorability score (0–10), and the resulting “partisan emotional score (-10–10)” is rescaled to a 0–10 scale.

Figure 2. Partisan Divergence among South Korean Voters in 2012 and 2022:
Kernel Density Estimation

Source: Ideological Partisan Divergence: East Asia Institute’s 2012 General Election and Presidential Election 7th Panel Study, Background Item 1, Question 1 & East Asia Institute’s 2022 General Election and Presidential Election 2nd Panel Survey, Background Item 6. Affective Partisan Divergence: East Asia Institute’s 2012 General Election and Presidential Election 7th Panel Study, Question 6-1-3 and 6-1-4 & East Asia Institute’s 2022 General Election and Presidential Election 2nd Panel Survey, Question 9-1 and 9-2. 2012 data https://kossda.snu.ac.kr/ (Accessed March 24, 2022)

Note: Ideological Partisan Divergence: 0 represents maximum agreement with progressive values; 10 indicates maximum agreement with conservative values. Affective Partisan Divergence: The party favorability score (0–10) is subtracted from the conservative party favorability score (0–10), and the resulting partisan emotional score (-10–10) is rescaled to a 0–10 scale. A score of 0 indicates maximum aversion to the conservative party, while 10 indicates maximum aversion to the progressive party.

The ideological partisan divergence shows that compared to 2012, the number of progressive-leaning voters increased slightly in 2022, while centrist voters decreased and conservative-leaning voters declined modestly. The affective partisan divergence shows that compared to 2012, the number of voters with negative sentiment toward the conservative party increased slightly in 2022, while the number of neutral voters decreased significantly, and those with negative sentiment toward the progressive party increased slightly. Statistical analysis confirmed that ideological and affective partisan divergence advanced over the decade. However, both distributions remain closer to a unimodal than a bimodal form. Although ideological and affective partisan divergence is visually evident among South Korean voters, the data do not indicate full-scale polarization.

Figure 3. Partisan Sorting among South Korean Voters in 2012 and 2022:
Kernel Density Estimation

Source: Party identification: East Asia Institute’s 2012 General Election and Presidential Election 6th Panel Study Question & East Asia Institute’s 2022 General Election and Presidential Election 1st Panel Survey, Question 9. Others are identical to those in Figure 2. 2012 data: https://kossda.snu.ac.kr/ (Accessed April 24, 2022).

Figure 3 illustrates kernel density estimates of ideological and affective partisan sorting among South Korean voters in 2012 and 2022. In the diagram for ideological partisan sorting at the top, the horizontal axis ranges from 0, indicating maximum agreement with progressive values, to 10, indicating maximum agreement with conservative values. In the diagram for affective partisan sorting at the bottom, the horizontal axis ranges from 0, representing maximum aversion to the conservative party, to 10, representing maximum aversion to the progressive party.

The ideological partisan sorting shows that compared to 2012, the proportion of progressive-leaning voters within the progressive party increased in 2022, while the proportion of conservative-leaning voters declined. Among conservative party supporters, the proportion of conservative-leaning voters remained relatively stable, while the share of centrist-leaning voters increased and that of progressive-leaning voters decreased. Despite this compositional shift, there remains considerable overlap in the ideological distribution of voters supporting the progressive and conservative parties. Thus, ideological partisan sorting has not advanced significantly over the past decade.

The affective partisan sorting a more significant shift. Compared to 2012, the proportion of voters with a strong aversion to conservatives within the progressive party increased substantially in 2022, while the proportion of neutral voters and those with an aversion toward progressives declined markedly. Among conservative party supporters, the proportion of voters with a strong aversion to progressives also increased significantly, while neutral voters and those with an aversion toward conservatives declined. As a result of this shift, the overlap in emotional distribution between the two party supporter groups has decreased. This suggests that affective partisan sorting has progressed considerably over the past decade. Although ideological partisan sorting among South Korean voters seem to be far from polarization, affective partisan sorting is visibly nearing that threshold.

Ultimately, this analysis demonstrates that pernicious polarization is most clearly observed in affective partisan sorting. In other words, the primary targets of polarized national narratives projected by political elites are voters who support their respective parties and who have developed increasing animosity toward the opposing party over the past ten years. As the overlap between partisan voter groups diminishes and the emotional divide widens, party electoral strategies have shifted from persuading centrist voters to mobilizing their base supporters. The “median voter theorem” is no longer relevant, as parties diverge toward the ideological extremes instead of converging toward the center. This explains why the opposition’s constitutional hardball tactics and the president’s constitutional beanball tactics serve as effective vote-gathering strategies (Merrill III, Grofman, and Brunell 2024).

IV. Constitutional Democracy in South Korea after the Impeachment of President Yoon

Table 1. Public Support for Impeachment during the Trials of President Park Geun-hye and President Yoon Suk Yeol

	Impeachment trial period for President Park Geun-hye			Impeachment trial period for President Yoon Suk Yeol
	Dec 2016, 2nd week	Feb 2017, 2nd week	Mar 2017, 1st week	Dec 2024, 2nd week	Feb 2025, 2nd week	Mar 2025, 3rd week
Overall	81%	79%	77%	75%	60%	58%
Conservative	66%	63%	50%	46%	25%	26%
Centric	86%	85%	86%	83%	60%	64%
Progressive	96%	95%	95%	97%	96%	95%
Ruling Party Supporter	34%	27%	14%	27%	10%	13%
No Party Preference	72%	71%	69%	79%	63%	51%
Supporter of Opposition Party	99%	96%	97%	97%	South Korean Political Elites and Democratic Backsliding   Sunkyoung Park Professor, Korea University     I. Introduction   Where does the current democratic crisis in the Republic of Korea originate? Is it rooted in institutional flaws within the power structure, such as the presidential or electoral systеms? Is it a result of shifts in citizen preferences, such as affective polarization or the erosion of democratic norms? This article argues that the primary focus should not be institutional flaws or citizen preferences but the preferences and behaviors of political elites and the incentive structures that constrain them to examine the causes of the democratic crisis in South Korea that emerged following the declaration of martial law on December 3.   This argument does not deny the existence of institutional flaws in the power or electoral systеms. Such flaws have functioned as structural constants in the politics of South Korea since 1987. However, the current situation is best understood as a crisis precipitated not by these flaws but by the failure to overcome them or by political elites who have exploited them.   This article also contends that no significant shift in citizen preferences precipitated President Yoon’s declaration of martial law on December 3. Although there are minor indications of crisis at the citizen level—such as pressure from a small group of hardline supporters or the influence of biased media—Woo Chang Kang’s analysis in Paper 5 of this working paper series finds no meaningful changes in citizen preferences. Indicators such as ideological distribution, affective polarization, support for authoritarian regimes, and policy preferences in 2024 remain largely consistent with previous years. Therefore, there is little basis to argue that democratic backsliding originated from citizen demand.   Where, then, does the current democratic crisis stem from? This article asserts that the crisis in South Korea’s democracy represents a “crisis from above,” whereby the December 3 declaration of martial law constituted a sudden rupture at the political elite level. Political conflict and the inability to resolve it destabilized the constitutional order.   This perspective aligns with recent studies on democratic backsliding in other countries. Levitsky and Ziblatt (2018) argued that democracy collapses when political elites within established parties fail to serve as gatekeepers against extremists. Bartels (2023), in his book Democracy Erodes from the Top, described democratic erosion in Europe. Druckman (2024) summarized the ideological dynamics of democratic backsliding, titling one section “Elites as Agents of Erosion.” Similarly, Kneuer (2021) emphasized the role of elected leaders as drivers of erosion processes (Kneuer 2021: 1447).   This article maintains that South Korea is experiencing a comparable form of democratic backsliding initiated by political elites and aims to examine why this elite-driven crisis erupted. The article is structured as follows: Chapter 2 uses Linz’s (1978) distinction between loyal democrats and semi-loyal democrats to systеmatically compare the behaviors of key politicians during the December 3 declaration of martial law and the impeachment process. It is essential to distinguish partisan behavior under a stable democratic constitutional order from actions that directly threaten the regime itself. Once this distinction is clear, it becomes possible to identify who engaged in anti-democratic or extremist behaviors and why such actions contributed to the democratic crisis from above.   Chapter 3 examines the emergence of semi-loyal democrats through three analytical frameworks: changes in the temporal political structure, weakened inter-party communications, and shifts in the incentive structure. The first framework refers to internal dynamics within the conservative party. Owing to repeated general election defeats in the Seoul metropolitan area, the number and influence of centrist politicians in conservative parties declined. Consequently, hardline members from regions with lower possibility of electoral punishment assumed leadership roles, reducing the likelihood of restraining extremist behavior.   The second issue involves deficits in training, education, and communication among politicians after entering the National Assembly. Parliamentary democracy relies on a process of political learning and dialogue, through which politicians engage with members of other parties to understand differing political philosophies and policy perspectives. If inter-party exchanges or learning opportunities have diminished compared to the past, building solidarity across party lines during democratic crises becomes more difficult. To verify this hypothesis indirectly, I rely on the National Assembly Research Groups, as a proxy to measure inter-party communications and cooperations. These groups, formally established in 1994, increased annually until a decline began in 2016. Furthermore, compared to the 20th National Assembly, the 21st and 22nd Assemblies exhibit less diversity in party affiliation among members of these groups. This suggests that inter-party communication and cooperations have declined.   Finally, the incentive structures that constrain politicians’ behavior are examined. In recent years, party politics have come under direct pressure from hardline supporters and the influence of biased media outlets. Within a context in which extreme voices are disproportionately amplified within party discourse, semi-loyal democrats who prioritize partisan advantage over democratic principles have gained greater support. This development is identified as a key factor that has aggravated the current crisis.   II. Crisis from Above: Loyal and Semi-loyal Democrats   In his 1978 book, Juan Linz analyzed the factors associated with the collapse of democratic regimes from multiple perspectives. Linz explained that parties and politicians who violate the fundamental principles of democracy pose a threat to democratic systеms. He argued that democratic crises and regime collapses can be understood by distinguishing political elites as either loyal democrats—those committed to democratic norms—or semi-loyal democrats—those who appear to support democracy but, in practice, undermine its principles.   First, loyal democrats respect the outcomes of free and fair elections regardless of their victory or defeat. Second, they unequivocally reject violence (or the threat of violence) as a means of achieving political objectives. The difficulty lies in distinguishing loyal democrats from semi-loyal democrats under normal democratic conditions. Both generally adhere to democratic procedures and compete based on partisan interests, making it difficult to identify their true democratic commitment. Linz argued that political crises resolved this issue of a priori unobservability. During political crises, their stance becomes evident in how they respond to violent or anti-democratic acts by members of their parties or supporters. Loyal democrats denounce such actions openly and unequivocally, even when the perpetrators are political allies. Conversely, semi-loyal democrats adopt an ambiguous position—avoiding criticism, remaining silent, or even endorsing such actions.   Levitsky and Ziblatt (2023) employed Linz’s framework to illustrate loyal and semi-loyal democrats during democratic crises in various countries. For example, in the 1930s, Sweden’s conservative party expelled youth members of the Swedish Social Democratic Youth League who promoted fascist ideology. In 1981, following a coup attempt in Spain, members across the political spectrum united in opposition. These are examples of loyal democrats. Conversely, members of France’s Republican Federation supported or remained silent about the 1934 riots, exemplifying semi-loyal democrats.   The democratic crisis surrounding the December 3 declaration of martial law and President Yoon’s impeachment presents a critical moment for distinguishing loyal democrats from semi-loyal democrats in South Korea. The first criterion is the acceptance of the results of free and fair elections, regardless of the outcome. Despite the absence of credible evidence of electoral fraud in South Korean elections since democratization, some politicians either alleged fraud or offered implicit support of fraud with ambiguous stances (Kim 2025; Han 2025). Their behaviors are the key characteristic of semi-loyal democrats who violate a core democratic principle.   The second criterion is tolerance for violence. Advocating, justifying, or condoning violence as a means to resolve political disputes—particularly when perpetrated by their party—is a hallmark of semi-loyal democrats. A clear instance was the raid on the Western District Court. On January 19, 2025, far-right groups supporting then-President Yoon Suk Yeol stormed the Seoul Western District Court to protest his arrest and the subsequent warrant review. They vandalized property and instigated a riot. Equally alarming was the fact that a politician had made remarks that appeared to encourage violence and that other politicians and supporters defended or minimized the incident (Yoo 2025).   Another criterion is the response to the vote to lift martial law following the December 3 declaration. That declaration was issued in clear violation of constitutional procedures and requirements. The deployment of military helicopters and troops to the National Assembly and the National Election Commission, as well as the prohibition of all political activity, were unconstitutional and illegal actions. These measures represented an undemocratic and violent attempt to suppress politics through military force. Under the Constitution, the National Assembly bears the responsibility of ending such violent confrontations and preventing further crises. Failing to participate in the vote to lift martial law without a valid justification signals behavior aligned with semi-loyal democrats.   Importantly, the classifications of loyal and semi-loyal democrats are not fixed or inherent traits. Politicians may shift between the two categories depending on their political circumstances. A legislator who previously acted as a loyal democrat may adopt semi-loyal behaviors during a crisis within their party. Conversely, a politician who once displayed semi-loyal tendencies may evolve into a loyal democrat through deliberate efforts and a process of learning.   How do politicians, who are constitutionally obligated to uphold democratic norms, become semi-loyal democrats? What drives individuals who once appeared to be loyal democrats to act otherwise? The next chapter will explore these questions using three analytical frameworks: changes in the temporal political structure, weakened inter-party communications, and shifts in the incentive structure.   III. Assessment of Causes   1. Change in the Phenomenal Political Structure   Following the December 3 declaration of martial law, President Yoon and several members of the People Power Party exhibited behaviors characteristic of what Linz describes as semi-loyal democrats. These included claiming that the 21st and 22nd general elections were fraudulent, encouraging the riot at the Western District Court, and making statements that downplayed the gravity of the riot. The proximate phenomenal causes of such extreme behavior may be traced to shifts in the internal power dynamics of the conservative party following repeated defeats in general elections. From the 20th through the 22nd general elections, the conservative parties consistently lost at districts in Seoul and Gyeonggi province. This trend diminished the influence of centrist politicians representing Seoul and Gyeonggi constituencies and allowed hardline members from regions with lower possibility of electoral punishment to assume leadership.   Figure 1 illustrates vote shares of the two major parties in the Seoul and Gyeonggi districts during the 20th, 21st, and 22nd general elections. In the 20th election, the average vote margin between the Democratic Party and the Saenuri Party was 3.63% (approximately 4,020 votes). In the 21st election, the average vote margin increased to 11.77% (approximately 13,943 votes) between the Democratic Party and the United Future Party. In the 22nd election, the gap was 8.99% (approximately 11,492 votes).   Figure 1. Vote Shares in the Seoul and Gyeonggi Districts of the Two Major Parties   Figure 2. Number of Legislative Seats in Seoul and Gyeonggi Regions Held by the Two Major Parties   Figure 2 shows the number of legislative seats secured by the two major parties in Seoul and Gyeonggi districts. Out of 108 seats allocated in these districts, the Democratic Party secured 75 seats in the 20th general election, 92 in the 21st, and 90 in the 22nd. Owing to the low proportionality of the current electoral systеm, the gap in seat counts was disproportionately larger than the actual vote margins. The electoral outcomes in Seoul and Gyeonggi areas, less influenced by regionalism, are influenced by national public opinion and election trends. As conservative parties experienced repeated defeats in these regions, centrist politicians either disappeared or saw their political base significantly diminished, while hardline conservatives rose in prominence. Voices advocating for reform within the party weakened, and even when a centrist leader was elected, their leadership was often challenged or replaced by hardliners (Jeong et al. 2022; Cho and Min 2024). In effect, the proportion of loyal democrats within conservative parties—those capable of restraining extreme or anti-democratic behaviors—has declined.   2. Decline in Inter-party Communications and Political Learning   The second cause concerns the political learning of elites and inter-party communication. The current decision-making systеm of the Korean National Assembly, made as a mixture of consensual and majoritarian models, is designed to inefficiently pay both a high decision-making cost and a high acceptance cost (Moon 2016; Jeon 2015). Under this structure, it is not easy for an indiviual legislator to make opportunities for inter-party communication and cooperation. Nonetheless, legislative outcomes are ultimately unattainable without communication and negotiation across party lines. Effective parliamentary politics require that politicians engage in inter-party communication when forming policy decisions. Through these interactions, politicians undergo high-level political learning, gaining experience in achieving compromise even amid conflict. They also develop skills such as log-rolling, exchanging support for mutually desired bills or policies. These experiences deepen their understanding of opposing viewpoints, reduce conflicts arising from misunderstandings, and lower communication costs.   However, leading up to the December 3 declaration of martial law, inter-party dialogue was unstable, and intra-party communication was lacking. In the absence of communication among legislative members and across parties, proposals to cooperate temporarily to address the democratic crisis and restore political stability were reduced to symbolic gestures.   Then, did legislative members truly engage in less inter-party dialogue than in the past? Did this decline lead to insufficient trust necessary for cooperation? While the level of trust is difficult to quantify, the frequency of inter-party exchanges can be assessed through available data, providing indirect insights.   The frequency and diversity of official study groups—referred to as National Assembly Research Groups—were examined to compare the extent of inter-party dialogue and communication quantitatively. These groups were established in 1994 under the Regulations on the Support for National Assembly Research Groups. Members can form such groups freely, regardless of party affiliation, to pursue research in areas of shared interest. Each group must comprise at least 10 National Assembly members from at least two parties and must include at least one member from a different party. The groups are designed to foster inter-party dialogue and collaboration. Furthermore, as each member may join no more than three research groups, participation reflects personal, political, and policy interests to some extent. Information regarding the number, types, and membership of these groups is publicly available on the Open Assembly’s Information Disclosure Portal.[1] However, information about research groups with which each member is affiliated is available only from the 16th National Assembly onward.   Figure 3. Number of National Assembly Research Groups by Year   Figure 3 shows the annual number of National Assembly Research Groups established between 1994 and 2024. Although the number of research groups increased each year after 1994, it peaked in 2016 with 75 groups and subsequently declined during the 20th and 21st Assemblies. The number of research groups tends to increase in the latter half of a legislative term compared to the first one. Nevertheless, the total number in the 21st and 22nd Assemblies remained lower than in the 20th and 19th Assemblies.   Figure 4 displays the rate of participation in research groups by political parties. During the 20th National Assembly, the participation rates of the Saenuri Party and the Democratic Party were similar. Conversely, during the 21st and 22nd Assemblies, the participation rate of the United Future Party/People Power Party declined slightly. The distribution of the number of members within research groups indicates that most groups comprised 11 or 12 members, slightly above the minimum requirement of 10 members.   Figure 4. Rate of Participation in National Assembly Research Groups by Political Party   A key factor in National Assembly Research Groups is the diversity of political party affiliations among their members. To quantify this, each term was assessed using the Shannon Diversity Index. This index, commonly used in biology to measure species diversity, captures the diversity of components within a group; a higher index reflects greater diversity.   Figure 5 displays a histogram of the calculated Shannon Diversity Index for party affiliation within research groups across legislative terms. Compared to the 20th Assembly, the 21st and 22nd Assemblies had fewer groups with high diversity indexes. The average diversity index values declined over time: 0.25 in the 20th Assembly, 0.23 in the 21st, and 0.21 in the 22nd.   Figure 5. Level of Diversity in Research Groups   Based on the analysis of various indicators—including the temporal distribution of research groups, participation rates by party, and diversity levels measured by the Shannon Diversity Index—the 21st and 22nd Assemblies experienced reduced inter-party dialogue and exchange compared to the 20th Assembly.   3. Change in the Incentive Structure   Finally, this section examines the incentive structure that constrains or enables political behavior. Politicians’ behavior is ultimately shaped by incentives that either restrict or encourage specific actions. The categorization of individuals as loyal or semi-loyal democrats reflects fluid responses to political circumstances, not fixed dispositions. Therefore, semi-loyal democratic behaviors of certain politicians are likely influenced by evolving incentive structures within their parties.   A frequently cited recent shift in incentives involves the rise of fandom politics, driven by hardline supporters and biased media. As Shin notes (Shin and Lee 2023: 116), few academic studies have addressed this phenomenon, as terms such as fandom politics, political fandom, or hardline support are not yet academically defined but have emerged in media and political discourse. However, if fandom politics is broadly understood as civic engagement by a politically motivated minority group seeking systеmic change, this concern becomes a longstanding question in democratic theory: how should minority groups that exert political influence be interpreted?   In a democracy—defined as a systеm in which representatives are elected by majority participation—active public engagement is a core virtue. Nonetheless, many theorists caution that democracy may be endangered by the nature and substance of such participation. Classical theorists such as Dahl (1956) warned of the majority’s potential to infringe on minority rights. Conversely, more recent studies cite the Tea Party and the Make America Great Again movement as cases where strong participation by exclusive and dogmatic minority groups posed a danger to democratic stability (Eisenstadt 2002; Levitsky and Ziblatt 2023).   In South Korea, the media and various critics have highlighted the risks posed by fandom political groups such as Moonpa, Gaeddal, and Taegukgi-budae. These critiques focus on the direct pressure exerted by fan clubs or aggressive party members against politicians whose speech or actions diverge from group expectations, often through mass text messaging or social media comments. Such politics are described as emotionally biased (Oh 2021), dismissive of institutional processes (Park 2023), and grounded in a culture of aversion (Kim 2024). Fandom politics has emerged due to the declining influence of traditional intermediaries such as political parties, labor unions, and social movements (Park 2018). However, it also exhibits creative and productive potential (Lee 2021), and some view unconditional criticism as complacency (Chun 2017). Nevertheless, most observers remain concerned about its negative implications.   Theoretically, active civic participation should be encouraged within representative democracies, as it can serve to limit elite capture by bureaucratic or privileged interests. The active involvement of minority groups with strong preferences is not inherently problematic. However, democratic dysfunction arises when two specific conditions coexist.   The first is a significant gap in participation intensity. While most modern citizens engage only in low-cost political actions such as voting, hardline supporter groups frequently participate in high-cost political activities. If this participation gap is narrow, hardline supporters are unlikely to exert disproportionate influence. Nevertheless, when ordinary citizens are minimally engaged and hardline groups are highly active, the imbalance may distort democratic responsiveness. Politicians may become overly exposed to the latter and increasingly interpret their views as representative of the majority.   The second condition involves the exclusivity and extremity of hardline supporters’ positions or their advocacy for similarly aligned policies. If such groups are willing to sacrifice democratic norms to advance political objectives or protect preferred politicians, their influence may lead National Assembly members to adopt semi-loyal democratic behaviors.   The presence of biased new media amplifies the negative impact of hardline supporters. Some media outlets repeatedly circulate unverified, extreme narratives for profit. In such an environment, where the voices of a small group of extreme supporters are overrepresented within political parties and amplified by biased media, it becomes difficult for politicians to resist the pressures created by this incentive structure. When anti-democratic hardline groups attack loyal democrats for prioritizing democratic principles over partisanship, other political elites may be easily drawn toward semi-loyal behavior. In some cases, semi-loyal democrats may receive greater intra-party support than loyal ones. For example, some supporters of former President Yoon Suk Yeol labeled legislators who voted for impeachment as traitors. Speakers at related rallies reportedly expressed as much hostility toward People Power Party members who supported impeachment as they did toward Democratic Party members (Jeong 2024). This suggests that the semi-loyal democratic attitudes of certain People Power Party members may have been shaped by incentive structures driven by minority hardline supporters.   IV. Conclusion   This article identifies the root of the democratic crisis in South Korea—illustrated by the December 3 declaration of martial law—not in institutional design or citizen behavior but in the preferences and actions of political elites, shaped by evolving constraints and incentive structures. Linz’s distinction between “loyal democrats” and “semi-loyal democrats” illuminates how, during crises, elites prioritized partisan interests over democratic principles, contributing to democratic and constitutional crisese . Chapter 3 identifies the restructuring of the power structure within conservative parties, declining inter-party communication, and pressure from extremist minority bases as key factors promoting anti-democratic behavior.   Nevertheless, this study has several limitations. First, as the current crisis originated within the conservative party, the analysis of semi-loyal democrats has focused primarily on internal dynamics within that party. Consequently, the roles of parties other than the People Power Party remain underexplored. Second, the causal analysis in Chapter 3 relies predominantly on indirect indicators. For instance, Section 2 uses the current status of National Assembly Research Groups to suggest a decline in inter-party political learning but lacks in-depth analysis of the actual extent of intra-party dialogue within those groups. Moreover, other forms of inter-party engagement were not considered in this analysis.   Most importantly, this study does not address additional significant factors, such as voter’s affective polarization or organizational mobilization outside political parties. Future research should comprehensively consider these factors to identify the mechanisms driving democratic backsliding. ■   References   Dahl, Robert. 1989. Democracy and Its Critics. New Haven: Yale University Press.   Druckman, James. 2024. “How to Study Democratic Backsliding.” Advances in Political Psychology 4S (Supple. 1): 3-42.   Eisenstadt, Shmuel. 2002. Paradoxes of Democracy: Fragility, Continuity, and Change. Baltimore: Johns Hopkins University Press.   Kneuer, M. 2021. “Unravelling democratic erosion: Who drives the slow death of democracy, and how?” Democratization 28, 8: 1442–1462.   Levitsky, Stevent and Daniel Ziblatt. 2018. How Democracies Die. New York: Crown.   ______. 2023. Tyranny of the Minority. New York: Crown.   Linz, Juan. 1978. The Breakdown of Democratic Regime. Baltimore: Johns Hopkins University Press.     Cho, Mi-deop, and Seo-young Min. 2024. “Han Dong-hoon forced to resign as representative… PPP to transition to interim committee systеm.” (in Korean) Kyunghyang Shinmun. December 16. https://www.khan.co.kr/article/202412162051015 (Accessed May 14, 2025)   Chun, Jeong-hwan. 2017. “Changes in Civic Politics and Public Discourse after the Candlelight Protests: ‘Moon Fans’ vs. ‘Han Kyung Oh,’ Fandom Politics and Anti-Intellectualism.” (in Korean) History Criticism 120: 386-406.   Han, Ye-seop. 2025. “Kwon Young-se criticizes Constitutional Court and defends ‘election fraud theory’?” (in Korean) Pressian. February 6. https://www.pressian.com/pages/articles/2025020614341509573 (Accessed May 14, 2025)   Kim, Do-hyung. 2025. “New spokesperson for People Power Party supported allegations of election fraud: “Martial law was a landing operation in Gwacheon… Yoon showed his true colors.” (in Korean) Hankook Ilbo. January 6. https://www.hankookilbo.com/News/Read/A2025010614370003904 (Accessed May 14, 2025)   Jeon, Jin Young. 2015. “Does the Revision of National Assembly Reform the19th National Assembly?” (in Korean) Journal of Contemporary Politics 8, 1: 99-125.   Jeong, Dae-yeon, Gyo-hyung Gu, and Hong-geun Lee. 2022. “PPP pushed for Lee Jun-seok’s expulsion and ended up in chaos.” (in Korean) Kyunghyang Shinmun. August 26. https://www.khan.co.kr/article/202208261746011 (Accessed May 14, 2025)   Jeong, Seong-sik. 2024. “Vote is invalid, traitors passed impeachment… impeachment motion pass caused devastation.” (in Korean) Gyeonggi Ilbo. December 14. https://www.kyeonggi.com/article/20241214580113 (Accessed May 14, 2025)   Kim, Joo Hyung. 2024. “Democratic Citizenship Beyond the Politics of Hatred.” (in Korean) Horizon of Knowledge 36: 18-30.   Lee, Seung-won. 2021. “Fandom Politics and Populism: Planning for an Alternative Political Culture.” (in Korean) Culture Science 108: 105-124.   Moon, Woojin. 2016. “Political Institutions in Korea and Directions for Their Reform.” (in Korean) Journal of Contemporary Politics 9, 1: 41-74.   Oh, Hyun Chul. 2021. “Dual Characteristics of Moon Jae-in’s Fandom.” (in Korean) Civil Society and NGO 19, 1: 3-38.   Park, Gwon-il. 2018. “The Symptom of Political Fandom: On Criticism of ‘Philosophical Arguments Against Moon Jae-in Fans’.” (in Korean) Jaeum & Moeum 38: 189-200.   Park, Sanghoon. 2023. Hate Democracy. (in Korean) Seoul: Humanitas.   Shin, Jin-wook, and Se-young Lee. 2003. Reboot Korean Politics. (in Korean) Seoul: Medici.   Yoo, Ji-woong. 2025. “The ruling party defended the court raid… democracy trampled.” (in Korean) News Tomato. January 19. https://www.newstomato.com/ReadNews.aspx?no=1251346 (Accessed May 14, 2025)     [1] https://open.assembly.go.kr/portal/infs/cont/infsContPage.do?cateId=NA21000 (Accessed May 13, 2025)     South Korean Democratic Crisis and 'Bottom-up Backsliding'?   Woo Chang Kang Professor, Korea University     I. Introduction   Research on democratic erosion has classified its causes and processes into two categories: “democratic erosion from the top” and “democratic erosion from the bottom.” Studies focusing on “democratic erosion from the top” interpret erosion as the result of strategic decisions made by political elites—particularly those in the executive branch—who seek to consolidate or expand their power. These studies highlight instances in which democratically elected leaders, while maintaining the façade of constitutional and legal systеms, gradually expand executive authority and weaken institutional checks through measures such as promulgating executive orders to neutralize legislatures, dominating the judiciary, controlling the press, suppressing dissent, and modifying electoral systеms (Bermeo 2016; Levitsky and Ziblatt 2018). These changes frequently occur in a gradual manner and are often enacted under the guise of legality, thereby making it difficult for citizens to perceive them as direct threats. By accepting the process as part of normal democratic functioning, citizens may become desensitized to its dangers, which carries the risk of dulling civic resistance (Bartels 2018; Ginsburg and Huq 2018).   Meanwhile, studies focusing on “democratic erosion from the bottom” emphasize the pivotal role of citizens’ voluntary acceptance and normative support in sustaining democracy. Democracy’s political legitimacy is derived from citizens’ recognition of it as the most just and appropriate form of governance (Lipset 1959). The preservation of democratic norms is not achieved through coercion; rather, it is sustained by the voluntary adherence to these norms (Dahl 1971). The concept of legitimacy becomes more resilient when it is founded not solely on ephemeral performance, often termed “specific support,” but also on principles of endorsement and commitment, referred to as “diffuse support” (Easton 1965). In the absence of diffuse support, democratic systеms are susceptible to internal collapse during periods of economic crisis or political turmoil. The viability of democracy is contingent upon citizens’ belief in the capacity of democracy to address political challenges through institutional frameworks (Classen 2020). Democracy becomes consolidated when members of a society accept democracy as the only legitimate form of governance and believe that all political conflicts and problems must be resolved within democratic procedures and norms. In this sense, democracy is viewed as “the only game in town.”   Following the 2022 presidential election, significant developments in South Korean democracy have been observed, suggesting a persistent “erosion from the top.” Examples of such controversies include the procedural issues surrounding the Korea Communications Commission’s dismissal of public broadcasting board members and the privatization of YTN. Furthermore, the president’s repeated exercise of the veto (requests for reconsideration) against legislation passed by the opposition—such as the Grain Management Act, Nursing Act, Trade Union and Labor Relations Adjustment Act (also known as the “Yellow Envelope Act”), Three Acts regarding Broadcasting, Kim Keon-hee Special Prosecutor Act, and Special Act on the Itaewon Crowd Crush—can be interpreted as attempts by the executive branch to undermine legislative oversight and curtail media scrutiny. Controversies surrounding selective or targeted investigations by state institutions, including the Board of Audit and Inspection and the Prosecutors’ Office, have also raised concerns about the erosion of political neutrality. These developments are analogous to the actions of “elected autocrats,” as described by Levitsky and Ziblatt, who have demonstrated a capacity to undermine institutional mechanisms to maintain checks and balances. Simultaneously, the opposition’s attempts to resolve legislative deadlock through impeachment motions have drawn criticism for breaching the norm of “institutional forbearance.” These trends are indicative of a pattern of “erosion from the top,” whereby political elites gradually erode the integrity of democratic institutions from within the legal framework.   International assessments of South Korean democracy corroborate these developments. The 2025 Democracy Report by the V-Dem Institute at the University of Gothenburg in Sweden identified South Korea as a country undergoing a process of “autocratization” in 2024, marking the second consecutive year this classification has been applied. Although South Korea ranked 17th as a liberal democracy in the 2021 report, it was reclassified as an electoral democracy in 2025. The Economist Intelligence Unit, the research arm of The Economist (UK), issued a similar judgment in its 2024 Democracy Index. In this index, South Korea received a score of 7.75 out of 10. While categorized as a “full democracy” from 2020 to 2023, South Korea was downgraded to a “flawed democracy” in 2024 and was identified as one of the 10 countries experiencing the most precipitous decline in democratic scores.   More recently, concerns about “erosion from the bottom” have intensified. On January 19, 2025, individuals opposed to the execution and issuance of an arrest warrant for President Yoon Suk Yeol entered, occupied, and vandalized property at the Seoul Western District Court, eliciting widespread shock throughout South Korean society. Persistent allegations of election fraud have coalesced with the opposition to President Yoon’s impeachment, and are now being articulated in public square. The increasing appeal of civil resistance and the criticism of constitutionalism and the rule of law have contributed to a heightened sense of crisis, indicating that South Korean society is not impervious to the influence of far-right forces, a phenomenon that is spreading across the United States and Europe.   Erosion from the top and that from the bottom are not mutually exclusive but rather interconnected phenomena. The likelihood of anti-democratic attempts by elected autocrats succeeding is amplified in environments where public support for democracy is deficient. Conversely, a robust civic commitment to democratic legitimacy can function as a critical bulwark against elite-driven erosion. In this context, this article conducts a time-series analysis to examine the shifts in public support for democracy in South Korea in the context of recent rapid political developments.   II. General Time Series Trend   The following survey item was used to assess South Korean citizens’ attitudes toward democracy: Respondents were asked to indicate which statement best reflected their view. 1) “Democracy is always better than any other systеm.” 2) “In certain situations, dictatorship can be better than democracy.” 3) “For me, there is no big difference between democracy and dictatorship.” The first response reflects the belief in the absolute superiority of democracy over other systеms and may be interpreted as a measure of diffuse support (Easton 1965). Those who opt for the second response demonstrate conditional support for democracy; during periods of crises, they may potentially forsake democratic principles, thereby legitimizing autocracy or authoritarianism. The selection of the third response is indicative of a pervasive sense of cynicism towards the political systеm in its entirety. This report analyzes the period spanning from 2003 to 2025. Data from 2003 to 2022 are drawn from the Asia Barometer Survey, while the 2025 data are derived from the “Polls on Polarization and Democracy in South Korea” conducted by the East Asia Institute on January 22–23, 2025.   Figure 1. Overtime Changes in Attitudes Toward Democracy, 2003–2025   As demonstrated in Figure 1, South Korean citizens have demonstrated increasing support for democracy over the past two decades. In the 2003 survey, only 49% of respondents stated that “democracy is always better than any other systеm.” This figure increased to 66% in 2011, 71% in 2019, and 76% in 2022. In the surveys conducted following the period of extreme political polarization that ensued after the 2022 general elections, the declaration of martial law in 2024, and the subsequent impeachment proceedings, 75% of respondents expressed their preference of democracy as the optimal systеm. Concurrently, in 2006, 36% of respondents stated that “in certain situations, dictatorship can be better than democracy.” This figure underwent a steady decline, reaching 20% in 2011, 25% in 2015, 17% in 2019, and 16% in 2022, respectively. In 2025, the proportion remained at 16%. Additionally, the proportion of individuals who believed “there is no big difference between democracy and dictatorship” decreased from 33% in 2003 to a mere 9% in 2025. Overall, Figure 1 suggests that the belief in democracy’s absolute superiority is firmly entrenched in South Korean society. Despite the persistent erosion from the top, citizens have come to internalize the tenets of democracy.   III. Analysis by Generation   This trend raises the question: what factors enabled such a transformation? One potential explanation may be found in political learning. Following the democratization of the nation in 1987, South Korea has sustained a democratic systеm for a period exceeding three decades. During this period, citizens repeatedly witnessed peaceful transfers of power through elections. Moreover, South Korean democracy demonstrated resilience during the Asian financial crisis in 1997 and the global financial crisis in 2008. A review of significant events reveals that substantial political change can be achieved through civic participation within the democratic framework. For example, the 2008 protests against imported U.S. beef and the 2017 candlelight demonstrations for impeachment demonstrated this. Through these diverse political experiences, South Korean citizens have come to recognize and internalize the value and functionality of democratic governance. Another potential factor is the “generational replacement effect.” Millennials and Generation Z, born subsequent to the democratization in 1987, have been raised in a political environment that is markedly distinct from that experienced by preceding generations. These generations have not experienced military authoritarianism; they were educated to view authoritarian rule negatively and have been exposed to a discourse that emphasizes democratic principles and values. With expanded press freedom, citizens gained access to political information through a variety of media platforms, developing a heightened sensitivity to democratic norms. As these cohorts integrated into society, they likely contributed to the overall increase in public support for democracy in South Korea.   As illustrated in Figure 2, there is discernible shift in attitudes toward democracy across different generations. The classification of generations by birth year is as follows: Industrialization Generation (1940–1959); 86 Generation (1960s); Generation X (1970s); Millennial Generation (M Generation; 1980s); Generation Z (post-1990). As illustrated in Figure 2, there was a consistent increase in support for democracy across all generations. Among the Industrialization Generation, support for democracy increased from 43% in 2003 to 76% by 2022. A decline in support for dictatorship under certain conditions was observed, with the percentage of respondents decreasing from 38% in 2006 to 17% in 2022. In 2025, support for democracy among this group exhibited a slight decrease of approximately three percentage points, while support for dictatorship in specific situations increased by a similar margin, reaching 21%. The 86 Generation exhibited a comparable trend. In 2003, 50% of respondents expressed support for democracy, a figure that rose to 79% in 2022 but decreased to 74% in 2025. The level of support for dictatorship reached its zenith in 2006 at 40%, but subsequently underwent a steady decline, reaching 16% in 2022 and showing a modest uptick to 18% in 2025. A consistent pattern of support for democratic norms was exhibited by Generation X. In 2003, 49% of respondents expressed a preference for democracy, and this figure increased to 71% in 2022, marking an approximate improvement of 22 percentage points. In contrast to the trends observed in Industrialization and 86 Generations, Generation X exhibited an increase in support for democracy, reaching 80% between 2022 and 2025. Concurrently, support for dictatorship under certain conditions exhibited a decline, from 18% to 13%, marking a decrease of approximately five percentage points. The proportion of respondents indicating no preference between democracy and dictatorship also underwent a decrease during these years.   Figure 2. Generational Trends in Attitudes Toward Democracy, 2003–2025     The trends for Millennials and Generation Z reveals a high degree of similarity. While 53% of Millennials selected democracy in 2003, this figure increased to 80% in 2022. In that year, Millennials demonstrated a higher level of support for democracy compared to Generation X. This ratio underwent a slight decrease in 2025, reaching 76%. At the initial appearance of Generation Z in surveys in 2011, approximately 57% expressed support for democracy, which was considerably lower than the support levels observed among other generations. However, as Millennials recorded a similar 53% support in the 2003 survey when they were in their 20s, and that support for dictatorship among Generation Z never exceeded 21%, a generational replacement effect can be inferred. This effect indicates that Generation Z exhibits a stronger inclination toward democracy compared to previous generations. A modest decline in support for democracy was observed between 2022 and 2025. However, there was no substantial rise in support for dictatorship. In contrast to the Industrialization and 86 Generations, which exhibited a withdrawal of support for democracy and an escalation in favor of dictatorship following political events such as the declaration of martial law and presidential impeachment, Millennials and Generation Z did not replace democratic support with autocratic preference.   IV. Changes by Generation and Gender   In recent years, considerable attention has been directed toward the tendency of South Korean politics to react sensitively to the political conservatism of Millennial and Generation Z men, especially regarding gender issues and the discourse on “fairness.” A clear tendency to support the People Power Party (conservative party) was demonstrated by men in their 20s in the 2020 general elections, the 2021 by-elections, and the 2022 presidential elections. In the exit poll conducted immediately after the 2022 presidential elections, 59% of men aged 20-29 and 53% of men aged 30-39 expressed support for candidate Yoon Suk Yeol, while 58% of women aged 20-29 and 50% of women aged 30-39 supported candidate Lee Jae Myung. This gender disparity between men and women in their 20s manifested during the impeachment proceedings that ensued following President Yoon Suk Yeol’s declaration of martial law in December 2024. According to the findings of a recent survey, the proportion of male participants in their twenties was minimal, constituting a mere 3% of the total. In contrast, 18% of the participants identified as female within the same age bracket. This phenomenon stands in stark contrast to previous instances, such as the 2008 candlelight demonstrations against imported U.S. beef and the 2016 impeachment rallies, where the participation rate of men was reported to be in the low to mid-10% range (BBC News Korea 2025-02-14).   Within this context, Figure 3 illustrates how support for democracy varies by gender across each generation. Among the Industrialization Generation, the gender gap remained minimal, with the exception of the 2025 survey. In 2025, compared to 2022, female respondents demonstrated negligible change, whereas male respondents exhibited a six-percentage point decrease in support for democracy, from 75% to 69%, and an 11-percentage point increase in support for dictatorship, from 17% to 28%. In contrast, the 86 Generation exhibited a contrasting trend. Among the male, support for democracy exhibited a slight decline, from 78% to 76%, representing a two-percentage point reduction. Conversely, among women, support exhibited a more precipitous decline, from 80% to 72%, marking an eight-percentage point decrease. In contrast, support for dictatorship increased by four percentage points, from 16% to 20%. In this generation, women demonstrated comparatively greater shifts than men. Among Generation X, there was a notable shift in the characteristics of men. In the 2022 survey, 67% of Generation X men expressed support for democracy; this figure surged to 84% in 2025, marking a 17-percentage point rise. Concurrently, support for dictatorship decreased by 12-percentage points, from 22% to 10%. The most significant shift in attitudes toward democracy and dictatorship was observed among Generation X men between the two surveys.   Figure 3. Attitudes Toward Democracy by Generation and Gender, 2003–2025     Concurrently, Millennials and Generation Z exhibited gender disparities in their attitudes toward democracy even prior to the 2025 survey, deviating from the trends observed in other generations. For Millennials, this gender gap emerged in 2015. The 2011 survey revealed that 67% of men and 68% of women expressed support for democracy, indicating no significant disparity between the two groups. However, following 2015, there was a marked increase in support for democracy among women, while the increase among men was more modest, resulting in a widening gender gap. The gap reached 6%p in 2015, 8%p in 2019, and 12%p in 2022. In the 2025 survey, 84% of Millennial women expressed support for democracy, while only 69% of Millennial men did so. This resulted in a gender gap of 15 percentage points. A mere 7% of Millennial women expressed support for dictatorship, constituting the group with the lowest level of endorsement for dictatorship among all categories.   The most salient gender disparities were exhibited by Generation Z. In 2011, when this cohort first appeared in surveys, the gender gap was already evident: 64% of women supported democracy, compared to 52% of men—a 12 percentage point difference. Notably, the 2019 survey revealed a non-existent gender gap, with 76% of men and women expressing support for democracy and 14% expressing conditional support for dictatorship. This convergence may be indicative of the influence of the 2017 impeachment experience. However, since that time, support for democracy among Generation Z women has steadily increased, reaching 79% in 2022 and 81% in 2025. Conversely, support among men in the same generation exhibited a decline from 73% in 2022 to 63% in 2025, resulting in a 18-percentage point disparity in 2025.   These trends among Millennial and Generation Z men are consistent with concerns about the “conservatization of men in their 20s.” Nevertheless, it is noteworthy that the decline in support for democracy among this demographic has not been accompanied by a proportional rise in support for dictatorship. From 2022 to 2025, support for democracy among Millennial men decreased by five percentage points, while support for dictatorship increased by only three percentage points. A similar trend was observed among Generation Z men, who exhibited a 10-percentage point decline in support for democracy, while support for dictatorship increased by a mere six percentage points. This finding is in sharp contrast to the 11-percentage point increase in support for dictatorship observed among Industrialization Generation men in the 2025 survey.   V. Conclusion   Citizens’ principled support and attachment to democracy itself, namely diffuse support, constitute the foundation for democratic stability. A decline in public conviction and trust in democracy signals an erosion from the bottom and a weakening of resistance against erosion from the top. Currently, concerns are growing regarding the democratic erosion in South Korea due to political turmoil, including the declaration of martial law, the impeachment of the president, and two by-elections within a span of ten years. This study examined the evolution of democratic support among South Korean citizens over the past two decades.   An analysis of seven surveys conducted between 2003 and 2025 reveals that support for democracy has deepened among the public. In 2006, only 43% of respondents stated that democracy was better than any other systеm; however, this figure increased to 76% in 2022, marking a 33-percentage point rise. Concurrently, support for dictatorship under certain circumstances diminished from 36% in 2006 to 12% in 2022, signifying a decline to one-third of the original level. The survey conducted in January 2025 survey revealed no substantial shift in public attitudes towards democracy or dictatorship, even conducted in the aftermath of the declaration of martial law and during the presidential impeachment proceedings. However, the data indicated generational and gender-based disparities in responses to the crisis. The 2025 survey revealed a decline in support for democracy and an increase in authoritarian sympathy among men from the Industrialization Generation, Millennials, and Gen Z. In contrast, Generation X men, Millennial women, and Gen Z women exhibited an increase in support for democracy, leading to a minimal overall change in aggregate figures.   The lower support for democracy among Millennial and Generation Z men—and its decline during the martial law situation—corresponds with the discourse around the “conservatization of men in their 20s.” Nonetheless, it is crucial to acknowledge that 68% of Millennial men and 63% of Generation Z men regard democracy as “the only game in town,” and the proportion who prefer dictatorship remains merely one-third that of those who support democracy. A notable distinction emerges when South Korea is juxtaposed with the United States and Western Europe, where democratic erosion from the bottom is a matter of increasing concern. In contrast, the decline among younger men in South Korea remains modest. For example, Foa and Mounk (2016) reported that while approximately 60% of Americans born in the 1940s (the Industrialization Generation) agreed that residing in a democratic country is essential, only about 30% of those born in the 1980s (Millennials) concurred with that perspective, thereby unveiling a considerably pronounced generational decline.   Despite the prevailing circumstances of martial law and impeachment, South Korean democracy has exhibited remarkable resilience. This resilience is bolstered by the unwavering commitment of South Korean citizens to democratic values. Since the democratization that took place in 1987, the public has gradually come to accept democracy as not only a systеm of governance, but also a fundamental societal value. This shift in perspective has been the result of a prolonged process of political learning. At present, in the year 2025, South Korean society largely accepts democracy as “the only game in town”—the benchmark for democratic consolidation defined by Linz and Stepan (1996)—despite some variation across generational and gender lines. These findings suggest that recent democratic erosion in South Korea is driven less by grassroots disaffection and more by elite strategies from above. The public’s robust support for democracy will prove to be a pivotal asset in countering and reversing top-down democratic erosion in the forthcoming years. ■   References   Bartels, Larry M. 2023. Democracy Erodes from the Top: Leaders, Citizens, and the Challenge of Populism in Europe. Princeton, NJ: Princeton University Press, 185–215.   BBC News Korea. 2024. “Why Have South Korean ‘Men in Their 20s’ Become Conservative?” (in Korean) 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 the Public Respond to Government Performance? The Asymmetric and Dynamic Nature of Legitimacy Beliefs.” American Journal of Political Science 64, 1: 182–200.   Dahl, Robert A. 1971. Polyarchy: Participation and Opposition. New Haven, CT: Yale University Press.   Easton, David. 1965. A Systеms 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.   목록 6대 프로젝트 Center for Democracy Cooperation Related Publications Working paper [ADRN Working Paper] South Korea's Approach to Democratic AI Governance 2026-07-20 Asia Democracy Research Network Working paper [ADRN Working Paper] Mapping East Asian Initiatives on AI Governance for Democracy (Part 7~11) 2026-07-01 Asia Democracy Research Network Close 오늘은 이창을 열지 않습니다. 닫기