From Authoritarianism to Democratic Corporatism ? The Rise and Decline of Social Dialogue in Korea

This study analyzes the sudden rise, quick success, and gradual decline of democratic corporatism (DC) in Korea. The analysis and discussion in this paper is based on qualitative approach, combining semistructured interviews and a review of archival evidence on the sustainability of social dialogue in Korea. In addition, we also provide supplemental quantitative evidence based on interview results. The present study indicates that an event such as a sudden economic crisis can lead to short-term outcomes such as a swift experiment with DC, but determination of the very nature of a tripartite agreement and the long-term sustainability of DC hinges on structural elements and path dependence formed and built up over decades. Events in Korea show that a direct transition from state authoritarianism to DC is a difficult, if not impossible, task.


Introduction
Democratic corporatism (hereafter, DC) can be described as an industrial relations (IR) system in which labor and employer representatives are independent of the state and voluntarily participate through tripartite mechanisms in centralized wage bargaining and/or policy-making.DC has received widespread academic attention since the 1960s.Its key ingredient in the 1960s and 1970s was national-level collective bargaining that had macroeconomic impact through the creation of high-level union and employer associations and brought DC to the attention of the government and its eventual involvement [1].DC has received renewed interest since the 1990s in developed, newly developed, developing, and transitional economies.Unlike the shape of DC in the 1960s and 1970s, corporatist arrangements during the 1990s took diverse forms and served various purposes such as policy concertation for reforming institutions, labor markets, and macroeconomic policies [2].
Experiments with DC in South Korea (Korea, hereafter) since the late 1990s have not been closely examined, although the militancy of its labor movement, its confrontational IR, and its recent financial crisis of the late 1990s are well-known.These experiments likely escaped notice because Korea generally has been considered an inhospitable environment for DC because of its tradition of state authoritarianism, labor militancy, and under-institutionalized collective bargaining process [3].The most noteworthy of these experiments was the formation in January 1998 of the Korea Tripartite Commission (KTC) in reaction to the nation's financial crisis.The KTC soon produced a social compact that helped Korea return to economic stability.Since the social compact of 1998, however, DC has continued mostly in form only with no meaningful results.
The focus of the present study is to document and analyze the sustainability of DC in Korea and to critically reevaluate, in light of the Korean experience, the previous theoretical arguments concerning what elements are prerequisites for its success.Data for this research came from semistructured interviews conducted in August 2005 and April 2015 with labor representatives, employer representatives, government representatives, and neutral interviewees.
The present paper is unique in three aspects.First, because previous studies of the new DC dealt mostly with successful situations [1,4,5], they may have failed to identify significant obstacles to the success of DC.Unlike this previous research, the perspective of this article reflects examination of a much less successful attempt at the new DC and explains roles of various variables at different stages of DC.Second, the study provides an in-depth empirical case study based on interviews with participants who were involved in KTC and various pieces of archival evidence on the sustainability of social dialogue in Korea.Third, almost all the research on the best environments for DC has been conducted in Western Europe, despite widespread experiments with corporatist arrangements elsewhere.Although authoritarian corporatism in Latin America and Asia is well documented, DC has rarely succeeded outside of Western Europe [6,7].This lack of success outside of Western Europe raises the question of whether findings about success in that arena can be generalized to experiments with DC in a broader range of countries.Extending this research beyond Western Europe will broaden our understanding of recent experiments with corporatism by identifying other factors that may influence the success of the corporatist system under different historical and situational circumstances.

Literature Review and Propositions
2.1.DC Literature: Old and New DC in the 1960s and 1970s was based upon Keynesian economics that emphasized the demand side of the economy.Within this context, the main purpose of DC was to control inflation by preventing excessive wage increases and to contain unemployment by pursuing policies to achieve full employment.Countries such as Sweden, Austria, and Germany were found to be more capable of achieving both goals simultaneously, and the successes within these countries were attributed to their corporatist IR systems [8][9][10][11].
The factors commonly associated with the success of DC in this period were (1) high union density and a centralized union federation, (2) a centralized employers' organization with strong oligarchic leadership, and (3) the existence of a social democratic or labor party [8,9,12].These authors argued that a central union federation should be able to represent the genuine interests of its members and should be able to control them so that there were few strikes, particularly unofficial strikes.Likewise, the central employers' confederation should represent the interests of its members and exercise tight control over individual employers.In such situations, a high level of union density and a high level of oligopolistic leadership in the various employers' branch associations aided representation and control capability on both sides [9,11].Having a social democratic or labor party in place also could improve the effective functioning of the tripartite system by balancing power resources among IR actors.Also, political allies could guarantee the implementation in the national legislature of the agreements that the IR parties reach [8].In addition, a political culture of negotiating compromises [13] and extensive legal regulations supporting DC [10,13] were suggested as ways to facilitate the development and sustainability of DC.Other scholars such as Lehmbruch [14] and Van Den Bulck [15] considered the existence of policy networks as a prerequisite for successful DC.According to them, problems were restricted to the network, thereby preventing any serious interruption to the success of DC by confining flare-ups to small networks and stabilizing them through collective action.
Fierce international competition and globalization in the 1990s generated pressures in many countries for the state to become more active in improving national economic competitiveness.This adjustment created a new type of DC in which governments must respond to transnational developments by eliciting cooperation from IR actors to reform both labor market regulations and welfare systems to gain improved flexibility and competitiveness.The new literature on DC in the 1990s sought to understand in what ways the new pacts differed from the old.It focused on two issues in particular: (1) the type of political exchange involved and (2) whether the old institutional preconditions still applied.
First, DC in the 1960s and 1970s focused on interest intermediation, usually through central wage bargaining among IR actors.In contrast, DC in the 1990s and 2000s emphasized social concertation by incorporating IR actors into policy-making.In the 1960s and 1970s, DC focused on mediating among the various IR parties involved in a country's domestic economy.On the other hand, the later version of DC was mostly concerned with securing social concertation among these parties in the development of policies to ensure a country's position in international trade.Rhodes [16] called this type of DC "competitive corporatism." Second, the renewed interest in DC in the 1990s provided an opportunity to retest the validity of previous arguments on the preconditions necessary for the success of DC.This revisiting was spurred by the absence in a few countries-such as Italy, Ireland, and The Netherlands-of the previously cited preconditions for DC despite these countries' successful experiments with corporatism.In other words, many nations have been able to achieve some form of DC without a centralized labor movement; this suggests a problem with traditional theories of corporatism.Indeed, relatively low union density, fragmented labor movements, and/or a history of confrontational labor relations characterized IR in these countries [1,2,4,[17][18][19].According to these authors, the form that DC took in the 1990s suggests that DC can succeed without the conditions previously considered essential.
Others regard union democracy as a kind of favorable institutional arrangement of the newer DC in Italy and Ireland [1,4].For example, in centralized collective bargaining over pension reform in 1993 and 1995 in Italy, the Italian union confederations benefited from their weak centralization because grass-roots democracy within the unions gave legitimacy to the national organizations' controversial decisions [20].
The traditional corporatism literature mostly focused on the structure, strategies, and strength of trade unions.Employer interests and state traditions seem to be underemphasized in both the old and the new corporatist literature.A large and growing body of new institutionalist literature, especially in political studies and comparative political economy, addresses both problems [21].Essentially, this literature argues that (1) policies tend to be path dependent and constrained by established routines and norms, especially those policies institutionalized within state bureaucracies and (2) policies favorable to labor often occur not just because labor has power but also because these policies serve the interests of employers.
First, the new institutionalist literature emphasized the concept of path dependence in explaining the formation and continuance of political systems [22][23][24].Path dependence implies the dynamics of self-reinforcing processes in a system [24].That is, outcomes at a critical juncture are expected to provoke feedback mechanisms that reinforce the reoccurrence of a pattern in the future.According to these authors, events occurring during and shortly after critical junctures emerge as essential.Moreover, "once established, patterns of political mobilization, the institutional rules of the game, and even citizens' basic ways of thinking about the political world will often generate self-reinforcing dynamics" (p.700) [23].Thus, the dynamics triggered by an event or process at one point in time are likely to recreate themselves, even in the absence of the reoccurrence of the original event or process [22,23].
Second, this literature highlights the constitutive power of employers and employer organizations in shaping political-economic institutions and outcomes.The literature that "brings capital back in" forms two broad variants [25].One variant emphasizes the role of employers as essential partners in cross-class coalitions [26,27].For example, the fate of corporatist arrangements in Sweden and Denmark in the 1980s hinged on employers' strategies; when employers are intent on decentralization, even the strongest labor movements cannot resist.The evaporation of employer support for traditional bargaining arrangements has precipitated the breakdown of labor movements in some strong labor countries such as Sweden [27].Elsewhere employers' continued willingness to negotiate compromises within existing centralized institutions, as in Germany, has contributed to greater resiliency despite overall weaker labor strength [28].The other variant emphasized the concept of employer coordination, the capacity of employers to coordinate among themselves.Scholars have paid increasing attention to explaining that the degree of employer coordination and its existence (or absence) is related to variation in social institution, policy, and unemployment [29,30].Studies showed that employer coordination is at least as important as the centralization of unions in achieving many of the economic outcomes associated with corporatism [31,32].Essentially, they substituted the concept of employer coordination for the corporatist literature's emphasis on union centralization [28,31].

Propositions
The present study attempts to offer explanations for the early success and subsequent decline of DC in Korea.DC is a complicated and long process involving several stages, each with diverse issues and priorities.For example, issues at the beginning stage of DC (for example, those that motivate the parties involved to start a social dialogue) differ vastly from those at the sustaining stage (for example, those forces necessary to maintaining altered behavioral patterns), which usually lasts for several years.Thus, it is unlikely that a single set of independent variables consistently affects the several different processes of DC over a prolonged period.We propose that different groups of factors affect DC at different stages.The present study classified independent variables into three distinct categories: (1) Factors that mainly serve to motivate IR actors to accept the idea of DC and reach a consensus; (2) factors that affect the nature and content of a tripartite agreement; and (3) factors that sustain DC for a prolonged period.In other words, we propose that different groups of independent factors play different roles during successive stages of DC.
We first examined the role of a national economic crisis as a motivator for IR parties to change their behavior by accepting DC.Indeed, most classical literature on social and organizational change mentions the role of a crisis as a catalyst for discontinuation of existing behavioral routines in favor of experimentation with new ones [33].Put differently, old DC literature suggested that DC is most likely to be formed where the structural preconditions did exist even without crisis consciousness.This can explain the experiences in Sweden, Austria, and Germany in 1960s and 1970s.However, if adversarial relationships are the custom in IR, union leaders, employer representatives, and government officials may be reluctant to undertake joint efforts dedicated to any experiment outside the existing relationship.When this is the case, initiation, and completion of change among social parties may only happen as a reaction to great pressure such as a national economic crisis that the existing process is incapable of responding to [34].
Regini explained the role of crisis in the context of the adoption of DC [35].According to him, if a nation is in deep crisis, labor may find it very risky to exploit to the full its organizational and market power in its negotiations with employers [35,36].If it uses its strength without restraint in pursuing workers' short-term interests, the resultant economic disruption may imperil its future ability to defend its members' economic interests and eventually endanger the basis of its own power.Political exchange through DC may then be seen as a useful tool, because union self-restraint in the industrial arena is compensated for by access to resources the state can offer.
Indeed, a careful reading of the new DC literature suggested that since the 1980s, a sense of crisis has played a role in stimulating DC in nations such as Ireland, Italy, and the Netherlands that had little history of successful DC.For example, the first social concertation agreement in Ireland was the Programme for National Recovery (PNR) (1988)(1989)(1990).When this agreement was signed, budget deficit and government debt were escalating, investment was sluggish, and despite growing emigration, unemployment was soaring [37].This sense of crisis played an important role in creating DC.Since the late 1980s, five consecutive partnership agreements between labor, employers, and government have been reached [1].
After a few short-term experiments at DC in the late 1970s and early 1980s, the social concertation in Italy began in earnest in 1992 in the wake of deep economic and political crises.The macroeconomic and financial conditions in Italy had declined markedly from the late 1980s to the early 1990s.The country's economic woes were accompanied by an equally severe political crisis that initiated the so-called "Mani Pulite" investigation and dismantled two major governmental parties.In 1992, a shared sense of crisis caused the government, the three main unions (CGIL, CISL, and UIL), and the employer association Confindustria to sign an agreement that eliminated wage indexation (i.e., scala mobile), a heavily compressed wage distribution within firms which had had a significant meaning in Italian society [38,39].
The success of DC in the Netherlands in terms of increasing job growth and lowering unemployment in a turnaround from the "Dutch disease" to the "employment miracle" occurred in the early 1980s in the middle of an economic crisis [18].Unemployment had reached a record 13% in 1981, and the state supported more than a million people.A sense of crisis led three parties to produce central accords in 1982 and 1983 that led to wage moderation, drastic welfare reform, and enhanced flexibility in working hours, particularly in part-time jobs.The success of DC helped to weaken 'the pathology of welfare without work' and brought the nation into economic prosperity in the 1980s and 1990s [5].
The successful experiences of Ireland, Italy, and the Netherlands with DC in the 1980s and 1990s have similar characteristics: (1) relatively low union density and fragmented labor movements that are indicative of a lack of the structural factors that the old DC literature identifies as necessary to facilitating DC and (2) a lack of a successful tradition of DC.These cases suggest that in these nations, a sense of crisis was needed to force IR actors to try a concept of DC, which was a very different course of behavior from their accustomed routines that had led the nation into crisis.The present study proposes that when DC is not a familiar policy choice, crisis consciousness can facilitate DC by changing existing behavior and provoking the embrace of a new course of action to survive the crisis.In other words, in a nation without structural preconditions for, and a history of, successful DC, the acceptance of, and experimentation with, DC by three IR actors needs a powerful motivator, such as a nationwide crisis, that can significantly alter their behavioral patterns.Proposition 1.In a nation without structural preconditions for, and a history of, successful DC, the acceptance of, and experimentation with DC by three IR actors needs a powerful motivator, such as nationwide crisis, that can significantly alter their existing patterns of behaviors.
In the present study, the structural factors of DC identified by the old DC literature such as a centralized union federation, a centralized employers' organization with strong oligarchic leadership, and the existence of a social democratic or labor party were expected to explain the content and nature of a tripartite agreement in a country like Korea that lacked such structural factors.The old DC literature argued that structural factors facilitate the formation of DC.In those instances, without most of the structural factors but where a sense of crisis nevertheless brought about DC as proposed earlier, the lack of structural factors was hypothesized as the determinant of the content of the tripartite agreement instead.That is, the lack of structural factors in unions such as low union density, a fragmented labor movement, and a very weak labor party was interpreted as putting labor at a disadvantage relative to employers, and this imbalance of power relations was expected to influence the content and nature of the tripartite agreement.Thus, the outcomes of DC tend favor employers over labor.Furthermore, enterprise unionism, which is observed in East Asian countries such as Korea, Japan, Taiwan, and Singapore, is likely to weaken the power and leadership of the labor federations [40].Korean unions are generally organized at the individual enterprise.Enterprise unions have members including only full-time blue-collar and some white-collar workers, excluding part-time or contingent employees [41].In the unionized establishments, wages and working conditions are determined by collective bargaining between enterprise unions and individual employers.Enterprise unions align individual unions with their firms' labor relations policy, have less bargaining power than employers, and limit the incentives and resources for current unions to organize workers across firms [42].Under decentralized enterprise unionism, labor federations are likely to have difficulties representing and monitoring their member unions in their strikes against restructuring.These labor federations showed scant ability to represent and control their rank and file members [43].This weakness further undermined the power of labor in Korean society.Under these circumstances, any agreements in corporatist arrangements that could possibly hurt the economic competitiveness of employers, which is one of their key concerns, are unlikely because of employer opposition and possible government interference.We propose that in a nation without structural preconditions for DC, the outcomes of DC are more likely to reflect employer preferences than labor interests.Put differently, because of the imbalance of power relations caused by the implicit and explicit backing of the government, the outcomes of the DC in Korea were more likely to reflect employer preferences than labor interests.
Proposition 2. Because of the imbalance of power relations through the implicit and explicit backing of the government, the outcomes of the DC in Korea are more likely to reflect employer preferences than they are labor interests.
This study argues that the concept of path dependence developed in the new institutionalist literature explains the fragile nature of DC in a country with a long history of authoritarian labor policy [23,44].As Deyo showed, in nations with long history of labor suppression and marginalization, independent trade unionism is typically suppressed and marginalized by successive governments that co-opted labor groups and subordinated them to the government [7].In most cases, government labor policies in these nations can best be considered as a combination of marginalization and subordination.Governments were less likely to allow autonomous unions to participate in decision-making bodies at either the macro or micro levels [45].Therefore, these subordinated labor leaders lacked genuine support from their rank and file members and were even further removed from their wider constituencies.
Some external shocks such as nationwide crises, as stated above, may force the parties to try a more consensual approach to labor relations [46].Then the crisis recedes or is resolved, the urgent need for compromise may vanish, and all three actors are likely to regress to their old and familiar habits of confrontational relationships, and the government is apt to revert to its authoritarian labor policy.The concept of path dependence comes into play here.The long legacies of past authoritarian policies that shape the goals, expectations, and preferences of key actors may not be easily ended by a short experiment with DC.Instead, the long history of authoritarianism, which has long shaped the parties' basic ways of thinking about labor relations, will soon reassert itself and regenerate the self-reinforcing dynamics of labor suppression and an overly reactive labor movement.We propose that path dependence on authoritarianism caused the failure of DC in Korea.
Schmitter [9] (pp.126-127) specifically mentioned that "the transition to societal corporatism seems to depend very much on a liberal-pluralist past.Countries locked into authoritarian corporatism at an earlier stage of development are likely to find it much more difficult to evolve toward such a consensual solution."We believe that the path dependence of Korean labor relations, such as the legacy of the state's authoritarian labor policy that in turn confronts labor's legacy of a reflexively and overly hostile stance, will be quite contrary to the compromising and consensual nature of DC.In the long run this incompatibility will result in the malfunctioning of DC.That is, path dependence in a state tradition that was formulated under authoritarian governments and reinforced by confrontational relationships among IR actors will inhibit the successful long-term functioning of DC.Proposition 3. Path dependence in a state tradition that was formulated under authoritarian governments and reinforced by confrontational relationships among IR actors will inhibit the successful functioning of DC in the long run.

The Brief History of Korean DC
To provide a contextual background, we will describe the trajectory of the Korean IR system by focusing on the relationship between the state and labor, with emphasis on the recent experience with DC.Korea has shown impressive economic progress since the 1970s based upon low value-added products and low wages.However, the nation's economic competitiveness reached a near-crisis in the early 1990s because Korean products cost more than those of low-wage competitors (i.e., China, Thailand, and Malaysia), and were of admittedly poorer quality than those of high value-added competitors (i.e., Europe, Japan, and the United States) [47].The financial crisis of 1997-1998 was the culmination of this slowdown that began early in the decade.The crisis was sudden and shocking: the foreign exchange rate (won/$) soared from 921.85 in October 1997 to 1706.8 in January 1998 [48].
On 2 December 1997, the Kim Young Sam government (1993-1998) filed for national economic bankruptcy and asked the International Monetary Fund (IMF) for a finance package.The IMF bailout program provided an extensive finance package of about US$65 billion.In return, the IMF required the Korean government to (1) contain inflationary pressure through tight monetary and fiscal policy, (2) restructure the financial sector to make it more transparent, (3) speed up further trade liberalization, and (4) improve the transparency and timely reporting of economic data.In addition to these general conditions, the IMF explicitly requested the Korean government to take steps "to facilitate the redeployment of labor, in parallel with further steps to improve labor market flexibility" [49].
In response to the urgent need, the KTC, which included the Federation of Korean Trade Unions (FKTU), the Korean Confederation of Trade Unions (KCTU), and the Korean Employers Federation (KEF), was hastily formed in January 1998.After two weeks of heated debate under the leadership of newly elected President, Kim Dae Jung (1998-2003), the KTC agreed on 90 major items and publicly declared its social compact on 9 February 1998.The popular media called this "the Great Compromise".Among the 90 points of agreement, the most important items are (1) Agreement to allow immediate layoffs of redundant workers.
(2) Improved income security programs for the unemployed.
(3) Approval for government employees and teachers to organize.(4) Guarantee of trade unions' right to engage in political activities.
(5) Recognition of the right of unemployed workers to join trade unions organized at the trans-enterprise level [50].
The rank and file members of the KCTU, however, immediately defied the agreement.During the negotiations, KCTU members had demanded job-wage bargaining (e.g., saving jobs in exchange for lowering wages or saving jobs by reducing work hours); the agreement, which focused on the implementation of a legal layoff system, greatly disappointed them.Consequently, delegates from KCTU-affiliated enterprise unions refused by a large margin to ratify "the Great Compromise".After ratification failed, the leaders of the KCTU, including the chairman, immediately resigned [51].The government ignored the lack of KCTU ratification and swiftly enacted the layoff system.On the other hand, the FKTU successfully ratified the Social Compact in a meeting of its executive committee.
The newly elected KCTU leadership refused to rejoin the KTC.This refusal led on 3 June 1998, to the formation of a second KTC without the participation of the KCTU.However, the chairman of the KTC (Kim One-ki, a politician affiliated with the ruling party) persuaded the KCTU to return by making an exclusive agreement with it that the KTC agenda would include discussion of restriction of layoffs and worker leasing systems, reduction of work hours, and strict punishment of unfair labor practices by employers.
In July 1998, however, the government, without consulting the KTC, announced plans to restructure the banking and public sectors, a move that had the potential for large-scale layoffs.This unilateral action was considered a violation of the spirit of DC.Consequently, the FKTU and the KCTU decided to withdraw from the KTC, and the KCTU called a general strike.As a result, the chairman of the KTC was forced to reach another agreement with the FKTU and the KCTU on 25 July 1998.This agreement promised to discuss the issue of public hearings on the financial crisis, to establish a special committee in charge of strict punishment of employers' unfair labor practices, and to set up a special commission for prior consultation with labor in the event of public sector restructuring.The two labor groups agreed on 27 July 1998, to rejoin the KTC.In turn, this agreement angered the KEF, which then withdrew from the KTC; however, the government demanded that it return; this return occurred 29 July 1998 [50].
In August 1998, the government again unilaterally announced, in direct violation of the agreement of 27 July 1998, its plan to restructure the banking industry and privatize public enterprises.Consequently, labor leaders were increasingly suspicious of the authority and credibility of the KTC and publicly called it "a rubber stamp" for the government's neoliberal economic policy.These incidents eventually resulted in the withdrawal of the KCTU from the KTC on 24 February 1999, and a general strike by members of the KCTU [50].
The KCTU has not returned to the KTC, which is not functioning effectively and has not produced any significant agreements since February 1999.Although the KTC reached an agreement in February 2004 on creating more jobs, the agreement was reached without the participation of the KCTU and was considered merely a symbolic gesture with no real impact on the parties' subsequent behavior.
After Lee Soo Ho, a member of a less hostile group of the KCTU, became chairman of the KCTU in January 2004, the KCTU softened its outright opposition to the tripartite discussion.However, the intense and violent opposition of an aggressive faction of the KCTU made its return to the KTC impossible.Faced with this impossibility but recognizing the value of including the KCTU in the social dialogue, labor representatives of the KCTU and FKTU, employer representatives, and government officials devised in June 2004 an alternative tripartite forum, the Tripartite Commission (TC) to revamp the malfunctioning KTC and discuss pressing issues.The TC, however, functioned no better than the KTC.After the initial meetings, in August 2004, the KCTU again refused to participate in the TC and only returned in June 2006 [50].In the interim, neither the KTC nor the TC accomplished anything of substance.
The strategic considerations that have driven the actions of the parties involved in the KTC and TC since the Great Compromise deserve examination.Currently, the KTC are composed of representatives from government, employer groups, and the FKTU (excluding the KCTU).Despite the damage done to the credibility and effectiveness of the KTC by its unilateral actions, the unpopular government led by President Roh Moo-hyun (2003)(2004)(2005)(2006)(2007)(2008), seemed to want to keep them in existence to retain the semblance of DC in the formulation and implementation of important and delicate labor policies, and hoped the KTC or the TC to continue to operate to obtain some kind of consensus regarding labor policies.However, since the beginning of the pro-enterprise conservative governments led by President Lee Myung Bak (2008-2013) and President Park Keun Hye (2013-2017), the KTC has functioned poorly, with little tangible support from the government, and therefore, has failed to produce any significant agreements.In January 2018, the KCTU expressed its willingness to participate the KTC and increased the possibility of bringing both national trade unions back into social dialogue in 19 years [52].Attention has been paid on what the new body for social concertation advocated by the current President Moon Jae in (2017-current) and supported by both labor and employer will appear, as well as what issues will be on its agenda.

Methods
The primary methodology of this study includes interviewing many key participants as well as archival evidence through yearly periodicals and daily newspapers such as KTC reports over the last two decades.The research is based on semistructured interviews in two waves conducted from 2 August to 26 August 2005 and 1 March to 30 April 2015 with 36 key participants and experts of DC in Korea.In wave 1, the interviews were conducted face-to-face in almost all cases.In wave 2, interviews were completed via telephone in March 2015 and supplemented by via e-mail in April 2015.Each interview took approximately one hour on average.The semistructure interview is one where "More or less open-ended questions are brought to the interview situation in the form of an interview guide" [53] (p.94).It is flexible, but generally an established set of questions is covered with varying levels of standardization.The objective of our semistructured interview is not to aggregate data across interviewees, but rather explores similarities and dissimilarities of ranges across voices [53].These interviews can offer a detailed picture of social dialogue in KTC.Specifically, to verify propositions, it is necessary not only to review various social agendas dealt with in the course of activities, but also to examine positions of actors, the actors' point of views involved in social concertation, the rationales behind such positions, and coordination of disagreements among actors.We did not use a general public survey because it typically measures individual perceived preference of social dialogue and the use of a general survey may not be appropriate for evaluating the activities of the tripartite committee.
Our qualitative data is a unique and representative sample of social dialogue in Korea.The interviewees were either principal participants in the social dialogue or keen observers who wrote newspaper reports or research projects on the social dialogue in Korea.Interviews included nine labor representatives (five from the KCTU and four from the FKTU), nine employer representatives (three from the KEF, three from the Korea Chamber of Commerce and Industry, two from the KFI, and one from the Korea Federation of Small and Medium Business), nine government representatives (five from the Ministry of Labor, two standing committee members from the KTC, a Labor secretary to the President, and a National Assembly member affiliated with the Labor and Environment Subcommittee), and nine neutral interviewees (seven newspaper reporters and two researchers from the Korea Labor Institute).Confidentiality was promised to all interviewees regarding their individual responses.
Interviews covered from eight to nine semistructured questions regarding the perceived performance of the social dialogue, and prospects for the social dialogue.Some examples of questions are included in the following.How do we assess the achievement of the Tripartite Commission for years from 1999 to 2005 (in wave 1) and from 2006 to 2015 (in wave 2), respectively?How do you think the Tripartite Committee will be able to achieve in terms of the degree to which it contributed to economic, social, and labor relations in Korea?What are the benefits and costs of each party during the Tripartite Commission?What are the factors that constrain Tripartite Committee activities?We asked the interviewees the reasons for their perceived outcomes and prospects of social dialogue.We also allow informants the freedom to express their views in detail in their own term.
The descriptive results of the interviews regarding the evaluation of DC are listed in Table 1.However, the reasons the four groups gave for their perceptions of the outcomes of the social dialogue were too complex to summarize in a table and have been synthesized and described in the next section.First, most respondents (69%, 25 out of 36) in 2005 believed that the social dialogue had early success but later failed.Employer and government representatives responded more positively, such as perceiving success in the early period (e.g., during the financial crisis), but a majority of labor representatives (66%, six out of nine) considered the social dialogue a total disappointment or failure.Interviews representing public interests split into two groups: Five of nine respondents (55%) considered the social dialogue an early success that later failed and four of the nine respondents (44%) regarded it as an outright failure.This trend is similar in 2015.Notably, no one considered it a success.Thus, the perceived outcomes of the social dialogue were at best mixed and varied significantly depending on who was being interviewed in both waves.Second, most respondents (61%, 22 out of 36 respondents in wave 1; 45%, 16 out of 36 in wave 2) were relatively pessimistic about the prospects for the social dialogue; few respondents were optimistic about the outcome of the social dialogue.A sizable minority (31% in 2005 and 33% in 2015, respectively, from both waves of survey) believed that its outcome would be mixed (e.g., pessimistic in the near future and optimistic in the long run, and vice versa).Government representatives, who included employees of the KTC, were the most optimistic in 2005 while neutral interviews (interviews representing public interests) were the most optimistic in 2015.In sum, the respondents believed that the outcomes of DC were at best mixed or a failure; the future prospects for the social dialogue are most likely pessimistic.Third, it appeared that there was a significant difference in the views on the prospects of the outcomes of social dialogue between labor and employers in two waves of survey.Thus, we performed additional empirical analysis to treat the second interview as a related but separate research encounter, allowing for changes that may have occurred over time.We are particularly mindful of temporal change in the research process, especially when collecting data over an extended period.
Table 2 presents results of analysis based on comparisons of the proportion of pessimistic prospect of DC outcomes between labor and employers.The proportions of pessimistic prospect of DC for the two groups were computed by the ratio of the number of respondents who answered 'disappointment and failure prospect' to total from Table 1.We are interested in the difference in average outcome, defined as the perceived proportion of pessimistic prospect of DC in the labor group before and after crisis consciousness minus the difference in the average outcome in the employer group, a 'differences in differences' estimator [54,55].This estimator can be seen as taking the difference between two pre-versus-post (crisis consciousness vs. no crisis consciousness) estimators, subtracting the employers' group estimator, which captures the time trend from the labor group's estimator.Careful interpretation of the results has been described in the next section.

Factors Initiating DC (Proposition 1)
The Korean case shows that a sense of crisis among social parties was a key element in the establishment of DC in the late 1990s.The Asian financial crisis of the time brought a keen sense of urgency and impending disaster to Korean society.In November 1997, the value of the Korean won fell by approximately 50% against the U.S. dollar, and the stock market also declined ~60% [56].The IMF responded with a loan of US$65 billion that came with several conditions: the Korean government was required to impose a tight monetary and fiscal policy, liberalize capital and trade transactions, and improve labor market flexibility.If the government failed to implement these changes, the IMF package could be discontinued, an act that would result in national bankruptcy.
It seemed that the 1997-1998 economic crisis was so overwhelming that structural and traditional constraints hardly affected the strategic choices of the IR actors.Indeed, the 1997 economic crisis was a case in which the IMF held the nation at gunpoint until its demands for the Korean economy were met.One interviewee from the KTC said, "We were so desperate at that time.We just wished that three representatives in the KTC-labor, business, and government-could be locked in a room until they agreed upon the most important issue, the allowing of layoff."The sense of urgency of the three parties was severe enough to make them change their behavioral patterns such as their vestiges of authoritarian corporatism and to provoke a new course of action calculated to survive the crisis.These changes were forged as an act of national survival without reference to, or concern for, the presence or absence of the structural factors associated with DC.
This crisis that led to the creation of the KTC also forced all three parties to make several strategic decisions during its formation.First, at the government level, the economic crisis and the IMF intervention forced the state to implement bold economic and employment restructuring programs.International pressure and the likelihood of strong resistance by labor kept the Korean government from resorting to its traditional authoritarian and oppressive labor policy [41].Newly elected President Kim Dae Jung's political position was weak because his party was neither the majority in the National Assembly nor did it hold the most seats.Consequently, the government had to consider that resistance by organized labor could spoil the schedule of economic restructuring.In terms of these factors, the government turned to more consensual policy-making through the creation of the KTC.In other words, the government favored DC because it could contain workplace militancy.Moreover, to help entice labor into the KTC, the state urged big business to share the hardships of economic restructuring through corporate restructuring and the improvement of corporate governance.
Second, employers were reluctant to join the KTC at the outset.Korean employers subscribe to a neoliberal economic philosophy, and they viewed DC as an unfamiliar and unfriendly concept close to social democratic ideas.As the KTC sessions proceeded, however, employers began to realize that the KTC might be advantageous to them.That is, the main employer groups, the KEF and the FKI, realized the KTC could serve to implement programs promoting flexible utilization of labor.They thought that the work of the KTC would result in layoff systems that would weaken labor's resistance to the restructuring of business operations.Although the national employer associations recognized that participating in the KTC might force them to accept the need for chaebol reform, they also hoped that their participation might secure amelioration of reform efforts and gain concessions to compensate them for the costs involved [2].
Finally, the newly elected president's invitation to organized labor to join the KTC presented labor with a strategic dilemma.On the one hand, its acceptance of the invitation could be construed as acceptance of the introduction of a formal layoff system; on the other hand, refusal could be failure to accept its social responsibility to help overcome the economic crisis.The FKTU and the KCTU both opposed participation in the KTC at first and threatened to engage in general strikes if provisions on labor flexibility were legalized and implemented through it.Labor groups argued that the IMF conditions hurt labor more than employers and unfairly so because big business had caused the crisis [2].However, both labor groups eventually joined the KTC because (1) enormous public pressure made it impossible for labor to shun it and (2) failure to participate in the KTC could open the door to a bilateral agreement between the government and employer groups for a system of layoffs much worse than what might happen with labor present.These strategic considerations indicate that the economic crisis of 1998 was severe enough to significantly alter the existing behavioral patterns of all three parties from a three-way adversarial relationship to a tripartite experiment with DC.
This social compact of February 1998 was the first agreement in Korea between labor, management, and the state that was reached in an autonomous manner.The elements of this situation support our proposition that IR parties can accept and experiment with DC when confronted with a national crisis severe enough to require them to alter existing patterns of behavior and that this acceptance and experimentation with DC can occur even without the structural preconditions and success stories previously considered necessary for DC.
We argue that the sense of crisis shared by the three IR actors initiated DC particularly in instances in which most of the structural factors did not exist.We conclude that only the 1997-1998 economic crisis was enough to motive DC.It seems that the 2008 global financial crisis has had a limited impact on the Korean economy compared to the 1997-1998 economic crisis.Many interviewees in Wave 2 also claimed that "social dialogue and social compromise began largely in a 'crisis phase' and that the 2008 financial crisis had no significant impact".
Relatedly, as Table A1 in Appendix A presents, key macroeconomic variables such as GDP growth, inflation, and employment have been less affected by the 2008 global crisis, especially when compared to the indicators from the 1997-1998 economic crisis.For instance, the initial adjustments in the GDP, and the consumer price index during the 2008 crisis were approximately one third of those in 1997.In addition, the duration of the 2008 crisis was much shorter [57].It took six months for the GDP to recover to its pre-crisis level, compared to over 18 months during the 1997-1998 crisis.Additionally, the postcrisis performance of the Korean economy in 2008 has been relatively resilient compared to other crisis-inflicted advanced nations such as the U.S and the U.K [58].One notable characteristic is job loss-perhaps a direct indicator of society's suffering from an economic crisis-reflects the difference in severity of the two crises.The job-loss rate during the 2008 global crisis was within 1% of the year-on-year basis at its peak, May 2009, which was substantially lower than 7.4% in August 1998 [59].Overall, it appears that the 2008 global economic crisis incurred less pain than the 1997-1998 economic crisis, and therefore it did not lead to a sharp sense of crisis for the employment relations parties.
Repeated withdrawals from and rejoining the KTC by both the labor and the business over the past two decades have determined its functioning of DC [47].The best option for the two parties was to decide solely on the agenda to maximize the benefits and minimize costs but is almost impossible for each party to do so under the crisis consciousness.Then the next strategic options under crisis consciousness would be that labor and employers participate in the social concertation in the Tripartite Committee and share certain costs through the concession bargaining.Therefore, due to such concession negotiations, both the labor and the employers in this period will have a pessimistic view about the prospect of the outcome through social dialogue.In other words, when economic crisis is a credible threat, both sides are unsatisfied with a concession that sacrifices part of their self-interests, but the similarity of recognition and expectation about DC can at least lead to social dialogue under crisis.
However, after the sense of crisis disappears, the cost-benefit evaluation of the labor and the employers on the achievement of DC can vary, and thus the prospects for the DC performance of each labor and business can diverge.In particular, the employers have a greater bargaining power than the unions due to the imbalance of power relations in Korean IR and will be more powerful if there is no sense of crisis.Specifically, it seems plausible that the employers feel no need to negotiate a concession that sacrifices parts of their self-interests, thereby potentially making prospects of social dialogue less negative than those of the unions.The difference in difference estimator, −0.67 (p-value: 0.01) in Table 2 shows a decrease in the proportion of pessimistic employers under no crisis consciousness (Wave 2).The results show that no crisis consciousness may lead the two parties to having different expectations of DC and the employers become less negative of DC than the unions.It implies that divergence about the prospect of DC outcomes between the labor and the employers, caused by the absence of crisis consciousness can be detrimental to initiating and the rejoining of KTC, being consistent with Proposition 1.
The above evidence supports Proposition 1 that IR parties can accept and experiment with DC when confronted with a national crisis severe enough to require them to alter existing patterns of behavior and that this acceptance and experimentation with DC can occur even without the structural preconditions and success stories previously considered necessary for DC.

The Nature of the Tripartite Agreement (Proposition 2)
The history of social dialogue in Korea shows that the outcomes of DC are more likely to reflect employer preferences than labor interests.In other words, issues serving Korean employer preferences were more likely to result in agreements, but issues favoring labor interests typically failed to bring agreement.These results contrasted with the outcomes of social dialogue in some European countries such as Italy, the Netherlands, and Ireland, where agreements reflected the interests of both employers and labor rather evenly [1,4,5].This discrepancy in the European and Korean outcomes appears to reflect characteristics of Korea society that led to power imbalances among the parties and to tripartite agreements that favored employers over labor interests.These characteristics of Korean society are (1) the power employers gain through government support, (2) a decentralized and fragmented labor movement, and (3) a very weak labor political party.
Employers and the Korean government have long had a special relationship.Undergirding this relationship is the widespread belief in government circles that Korea, because it lacks enough natural resources, must rely on management and innovation to advance its economy.Because of this belief, the government has almost always supported employers in various ways, from heavy trade tariffs and low-interest, long-term bank loans to restraints on the labor movement [47].The tendency for the government to support employers' preferences over a neoliberal economic policy has intensified since the 1997-1998 economic crisis.When the IMF explicitly asked the Korean government to improve its labor flexibility in 1997, many government officials agreed with the IMF that a rigid labor market was one of the reasons for the financial crisis [41].
The state has both guided and supported the chaebols such as Samsung, Hyundai, LG, and SK that have led the country's economic growth.The Korean economy has been heavily dependent upon these chaebols.In the 1990s, the top 50 chaebols were responsible for nearly 20% of the gross national product (GNP) and employment and 40% of sales in manufacturing [41].The share of top three chaebols (Samsung, Hyundai, and LG) is getting bigger and bigger in recent years: The ratio of top three chaebols' sales of national GDP increased from 33% in 2007 to 57% in 2012 [47].The powerful chaebols' interests are represented by the FKI.Given the philosophical bent of government officials and the sheer size of the chaebols, it is little wonder that Korean employers have been able to lobby the government and persuade it to protect their interests.
The power of organized labor in Korean society is in sharp contrast to that of employers.Low union density, a decentralized and fragmented labor movement, and a very weak labor party leave most of the power on the side of capital.Union density in Korea has remained ~10% since the late 1990s.Moreover, since 1995, the labor movement has been divided into two federations, the FKTU and the KTUC.The FKTU was established in 1960 and received financial support from the government, and its policies and activities mostly were cooperative to the government.In the early 1990s, a militant and illegal national federation, the KCTU was organized and became a legitimate union federation in 1999 [41].The FKTU represents a more conservative economic unionism with a keen interest in "bread and butter" issues; the KCTU stands for more progressive and aggressive unionism with a heavy emphasis on broad political and social agendas.Both federations have competed bitterly for union members since the mid-1990s and maintain a very tense and adversarial relationship.This division further undermines the power of labor in Korean society.
A third reason that tripartite agreements in Korea favored employers is the absence of a major political party in Korea that supports the interests of labor.Korean political parties largely reflect regional interests and show little concern for labor's interests.In January 2014, the DLP occupied only eleven of the 300 seats in the National Assembly.None of the laws governing labor's political activity-the Trade Union Act, the Elections Act, or the Political Funds Act-extend full legal protection to their political activities.Thus, the labor movement has minimal formal political influence.Despite political democratization, little has changed in the power relationship between capital and labor in Korean society.Labor remains a weak counterpart of capital and state; this weakness allows the government to travel the familiar path of its authoritarian tradition.
Because of this power imbalance, the outcome of DC in Korea reflects employer interests over labor interests.First, the 1998 compact focused on employment reduction by allowing a legal layoff system that was a goal of employers.However, most workers did not see as a fair exchange the expansion of trade unionism's legal rights that accompanied introduction of the layoff system.Some improvements in basic labor rights were too abstract for most union members who feared the imminent loss of their jobs.Indeed, unemployment rates immediately after the financial crisis (7% in 1998 and 6.3% in 1999) were more than three times those recorded before the financial crisis (2.1% in 1995 and 2% in 1996) [59].Moreover, the unemployed faced particularly severe hardships because the unemployment insurance that became effective in July 1995 was limited in amount and in the number of workers it covered.In addition to workers' fears about job security, many labor leaders considered acceptance of the layoff system to be misguided and self-defeating because it ultimately would shrink the pool of potential union members.
Furthermore, several of the 1998 agreements that labor pushed for and secured through DC were never enacted or were enacted but substantially distorted in the legislative process.The most significant examples of this bad faith were (1) the failure of the National Assembly to enact the agreement to recognize the right of unemployed workers to join trade unions and (2) although the KTC agreed to give public employees the right to organize, the legislation subsequently enacted allowed only the formation of workplace associations, effectively prohibiting collective bargaining by their members [3].
The financial crisis and the subsequent corporate restructuring decreased the bargaining power of labor [41].For instance, unions lost about 50,000 (or 3%) of their members after the crisis.Therefore, it seems plausible that the labors' prospect of the DC performance would be even more negative whereas the employers would be more positive due to the imbalance of power relations.In Table 2, the result shows the proportion of employers who are pessimistic of the DC outcome is significantly lower by 0.67 (p-value: 0.001) than that of labor group under no crisis consciousness, which is consistent with Proposition 2.
The foregoing discussion supports Proposition 2, which holds that largely because of an imbalance of power, the outcomes of DC are more likely to reflect employer preferences than labor The issues favored by employers, such as the legalization of the layoff system in 1998 and the postponement of multiple unions in workplaces, were enacted with the implicit assistance of the government; in contrast, agreements were never reached by either the KTC or the TC on the issues favored by labor, such as legalization of government employees' unions and the reduction of legal working hours.

Sustainability of DC (Proposition 3)
The state in Korea has long history of suppressing and marginalizing labor.The government used the nation's economic success, substantial wage increases, and the military threat from North Korea to justify an authoritarian labor policy [41].The advent in 1993 of democratically elected presidents has been accompanied by gradual political democratization.However, political democratization did not necessarily bring industrial democracy in a strict sense.Specifically, the Korean government did not abandon elements of an authoritarian labor policy.The current labor law still includes restrictions on the labor movement.Among these are a very strict definition of legal strikes, especially in the public sector, and a prohibition on union activities by some government employees and university faculty.
The creation of the KTC in January 1998 has made no difference in the legacy of the government's authoritarian labor policy.Its continuation is evident in the number of labor leaders imprisoned for their activities.Under the presidency of Roh Moo-hyun, 204 union leaders were arrested in 2003.This number exceeds the annual average of 178 union leaders arrested during the presidency of Kim Dae Jung (1998-2002) and the annual average of 126 under the presidency of Kim Young Sam (1993-1997) [3].Thus, the number of union leaders arrested did not decline even after political democratization began to appear in the 1980s.Ironically, there were instances in which key KCTU negotiators participating in the KTC were arrested and imprisoned for activities related to technically illegal strikes.During an interview, an official of the KTC mentioned that "no doubt the imprisonment of labor leaders who participated in the KTC has caused deterioration in the atmosphere of the discussion among IR actors of the KTC".
The interviewees from the KCTU, the FKTU, and the neutral parties asserted clearly that the legacy of state authoritarianism severely inhibited the continuation of a social dialogue.Although the president, presidential secretaries, and politicians affiliated with the ruling party initially pursued DC, most cabinet members and bureaucrats did not share their goal.As time went by, these bureaucrats accustomed to traditional state authoritarianism became dominant as support for DC waned within the government.
Even immediately after the formation of the KTC, government officials treated labor as an object to control and contain instead of as a serious partner.The interviewees from the KCTU, the FKTU, and especially the neutral interviewees emphasized that the state's intention in the formation of the KTC seemed to be to contain labor militancy so that implementation of large-scale restructuring could proceed without provoking waves of violent strikes.Indeed, this intention was repeatedly revealed in such statements by government representatives during the operation of the KTC as in this comment that "the main function of the KTC is to prevent violent strikes during restructuring drives.That is all we want from the KTC" (cited from Lee [60]).This tendency led outside observers such as the International Labour Office to state that "Government's commitment to social dialogue is sometimes faltering and social partners doubt whether government has a political commitment to social dialogue" [61].The government was unaccustomed to sharing its authority with labor.The tradition and legacy of state authoritarianism seemed to make government representatives and officials hesitant to discuss with labor leaders important policy matters such as macroeconomic policies and industry restructuring plans.Thus, state and labor had vastly different expectations of DC.
The tensions and inconsistencies in government policies between DC and economic liberalism caused labor leaders to question the state's intentions and to lose confidence in the future of DC.Korean corporatism presents the trade union involvement in public policy but is dominated by policymakers who subscribe to a neoliberal economic policy.In turn, economic liberalism is opposed to the interests of trade unions and their members [62].What happened in Korea can be considered a sort of paradox-unions accepted industrial restructuring in the form of DC, which ultimately worked against their members.When Korean labor leaders realized the state had moved to deregulate the labor market in the guise of DC, the fragile working relationship among the IR parties fell apart.This breakdown was exacerbated by the way the government ignored some elements of the social compact and violated others.
The legacy of authoritarianism was expressed in the form of state dominance even in the membership and operation of the KTC.Labor and employer representatives constituted only a minority of the KTC membership (on average 31%), whereas the percentage of members representing the state and general public was almost 70%.The day-to-day operation of the KTC reflects the same tendency.An interview participant from the KTC mentioned that "Because almost two-thirds of employees of the KTC (25 out of 40 employees) are public servants belonging to the Ministry of Labor, the KTC is like a branch of the Labor Ministry, and its operation does not seem to be really independent".
The authoritarian tradition of government's labor policy was again expressed clearly when about 5000 South Korean police entered using force the headquarter of the KCTU, one of the two biggest national centers of Korean unionism, in downtown Seoul on 22 December 2013, and arrested 135 KCTU members who tried to block their entry.The background is that the South Korean government recently announced a unilateral decision to set up a separate company covering a part of the Korean railway system, KORAIL.When the Korean Railway Workers' Union (KRWU) went on strike 9 December 2013 to show their opposition to the decision, the government immediately laid-off some 7000 strikers, declared the strike illegal, and issued arrest warrants for 194 KRWU leaders and members [63].The police believed leaders of the strikers to be hiding in the KCTU headquarter and attempted to enter the building by force.Consequently, the FKTU again decided to withdraw from the KTC [64].
The more we get away from the economic crisis, the more likely the employers will have more advantageous outcomes and not be willing to give up in negotiations for the sake of social dialogue.The labor interviewees in Wave 2 emphasized that the concessions most important to the labor were often rejected by employers.Moreover, they added "both labor and business totally had different views on the true characteristic of the KTC.We hope that KTC is democratic corporatism, but they saw the KTC as a weaker and extended version of its old authoritarian corporatism."It also implies the path dependence of hostile and authoritarian IR in Korea and the discrepancy in the evaluation of prospects between the labor and the employers can be harmful to promoting social dialogue through the Tripartite Committee.
Overall, the lack of an ideology of social partnership and mutual trust, coupled with differences in expectations shaped by a long history of authoritarianism, appears to be deadly to the long-term effective functioning of DC.In such circumstances, DC appears to be more of a respite than a reshaping of relationships.Put differently, path dependence in a state tradition that was formulated under authoritarian governments will soon repeat itself, reinforced by confrontational relationships among IR actors, and regenerate the dynamics of labor suppression and a reactive labor movement.The above discussion supports our Proposition 3 that path dependence in the authoritarian government tradition inhibited the continuation of DC after its initial success.

Conclusions
The present research is a single case study focusing only on the experience of Korea since the late 1990s.This singularity suggests a generalizability problem from a methodological point of view.Despite this limitation, the three propositions formulated in the present study seem to be supported.Specifically, the present study showed that because DC is a long and complicated undertaking dealing with diverse issues in several stages, different groups of variables affect DC at different stages.
First, when there is neither a history of successful DC nor structural preconditions that favor its development, acceptance of DC requires a powerful motivator such as nationwide crisis.In a sense, this finding is consistent with the new literature on DC that documents the success stories of DC in countries such as the Netherlands, Italy, and Ireland, where structural preconditions for DC did not exist.
Second, with the implicit and explicit assistance of the government, the outcomes of DC in Korea favored employers over labor.The absence of structural preconditions seemed to result in a power imbalance in Korean society, which led to disadvantageous outcomes for labor in the tripartite agreement.Whereas the old DC literature identified structural factors as preconditions for DC, this study shows that their absence largely determined the content of the agreement.
Third, path dependence in state authoritarianism kept DC from functioning successfully even after democratization in Korea.The path dependence of the authoritarian government and its subscription to a neoliberal economic policy seemed a powerful deterrent to the sustainability of DC.Korea illustrates the difficulty (maybe because of path dependence), if not impossibility, of a direct transition from state authoritarianism to DC, a difficulty already noticed by Schmitter [9].
Although some countries with fragmented labor movements such as Italy, the Netherlands, and Spain have demonstrated impressive outcomes in social dialogue since the 1990s, a newly developed country like Korea, which shared some of the same conditions (such as a divided labor movement and low union density), has been unable to make the same progress [61].The present study shows that the path dependence of an authoritarian state tradition, which suppresses labor and favors employer interests, is an influential factor in determining the fate of social dialogue in a country like Korea.
The good news from the present study may be that an experiment with DC can have a limited degree of success if there is a powerful motivator such as an economic crisis.The bad news may be that path dependence is the legacy of an authoritarian labor policy and seriously hampers the long-term continuation of DC.In sum, the present study indicates that a temporal event such as a sudden economic crisis leads to short-run outcomes, such as a swift experiment with DC (as shown in Proposition 1); however, structural elements and path dependence accumulated over decades determine the very nature of the tripartite agreement forged to handle the crisis (as shown in Proposition 2) and of long-term outcomes such as the sustainability of DC (as shown in Proposition 3).
Our study also has limitations that suggest refinements and future research.First, our use of a 36 sample of an in-depth interview was a limitation.However, we believe that qualitative interview method used in this study is highly useful to obtain information on the actors' strategies of participation in the Tripartite Committee, and the changes in actions among three actors.Moreover, the supplemental quantitative evidence presented in this study needs caution of interpretation.We present that empirical evidence is consistent with predictions from our propositions, not a direct verification of claims.Thus, it calls for a quantitative analysis that can be generalized by variables that have influenced the activities of tripartite committee over past 20 years.A few important questions remain.For instance, future research would extend to evaluating the effectiveness of Tripartite Committee.Moreover, researchers need to examine the role of government in encouraging responsible cooperation among the three IR actors.These are just a few of the examples that researchers attempt to answer for the benefit of social concertation.

Table 1 .
Descriptive statistics: two waves of interviews with key participants of DC in Korea a .
a List of Respondents: Nine Labor Representatives, Nine Employer Representatives, Nine Government Representatives, and Nine Interviewees Representing Public Interests.

Table 2 .
Comparisons of the proportion of pessimistic prospect of social dialogue between labor and employers.
p-value in parentheses.