Between Fragmentation and Centralization: South Korean Industrial Relations in Transition


  • Joohee Lee is at the Department of Sociology, Ewha Womans University.


This study investigates profound changes in South Korean industrial relations after the Asian financial crisis of 1997. Korea's neoliberal labour reforms have produced a large number of non-standard workers, deepening the union representation gap. Realizing that the fragmented enterprise unions could not adequately protect workers from this degradation of labour, trade union leaders began a major organizational drive at the industry level and tried to institutionalize sectoral bargaining. A political space for union centralization was partially opened because the state needed labour's co-operation to implement neoliberal reform packages. However, disorganized centralization in Korea, where important decisions on wages and working conditions have been negotiated mainly at the company level, has faced limitations in achieving meaningful changes in the dualistic structure of the labour market. This study concludes with a review of changes in Korea's labour law in 2010 and a discussion on the effects of the law on bargaining rights of non-standard workers and the incipient industry-level bargaining. This trend towards union centralization may continue, but the notable gap between the formal bargaining structure and actual practice is expected to widen.

1. Introduction

South Korea (hereafter ‘Korea’) has long been known for its rapid economic growth led by the strong developmental state. Since the Asian financial crisis of 1997, however, the country's developmental state began to acquire more neoliberal characteristics (Lim and Jang 2006). In some ways, the neoliberal state was more harmful to labour than the developmental state had ever been. The developmental state had at least increased the number of wage labourers, and its severe legal and political oppression gave them a strong incentive to organize and defend basic labour rights. The profound neoliberal reforms and structural adjustments after the financial crisis produced a large number of non-standard workers, which critically divided and weakened labour. With globalization, unionism has waned. Few unions now command the same degree of representative capacity that they did in the past. The representation gap is even more acute for countries such as Korea, where the union density is barely 11 per cent, and most of the non-standard workers are not protected by trade unions and collective bargaining. The fragmented enterprise-level collective bargaining structure has also deepened the representation gap.

However, the financial crisis has also presented possibilities for change. Labour needed industry-wide solidarity to secure employment protection, and the government required labour's co-operation to implement structural adjustments. Realizing that enterprise unions could not adequately protect their members' employment security, union leaders began launching major organizational drives at the industry level (Lee and Lee 2004). As a result, Korea structure of industrial relations has been centralized for the last decade. The Korean case is a clear anomaly because the dominant trend in nearly all advanced industrial countries has been towards the greater decentralization of collective bargaining. This article investigates profound changes in recent Korean industrial relations, which were the result of this double movement: neoliberal reforms that fragmented labour were accompanied by centralization of collective bargaining structure. As Polanyi (1957 [1944]) aptly demonstrated, whenever markets dominate a society, powerful institutions emerge to check the action of the market. Neoliberal policies can certainly be detrimental to equitable distribution and employment protection, but they are frequently accompanied by a larger political space that can induce more revitalization efforts by labour.

Despite the establishment of formal procedures, the actual practice of centralized collective bargaining has been deeply conflicted. In addition to employers' persistent resistance to participate in industry-level bargaining (Jeong 2001), organizational characteristics of industry unions have imposed severe difficulties on the process of centralization. Local unions have been under the strong influence of the structure and practices of enterprise bargaining, such as ‘enterprise consciousness’ and financial autonomy. As a result, a unique two-tier model of bargaining has developed in Korea, where industrial unions wield little power over the decisions of locals. Thus, the transformation of the bargaining structure in Korea can best be described as disorganized centralization as opposed to organized decentralization in some co-ordinated market economies (CMEs) of Europe. Whereas organized, decentralization in some CMEs has satisfied bargaining parties' flexibility and security needs (Andersen 2009), disorganized centralization in Korea has created the opposite situation. Employers have complained about the rigidity of bargaining practices, and because of a notable gap between the formal structure and actual practice, non-standard workers have not been properly protected under the new collective bargaining framework.

The purpose of this article is therefore twofold. The first is to better comprehend the nature of neoliberal labour reforms that simultaneously created the daunting pressure for increasing flexibility and an outlet for union centralization. The second is to evaluate the extent and outcome of collective bargaining centralization, and whether this process has led to meaningful changes in the dualistic structure of the labour market by a comparative analysis of the metal, finance and hospital sectors. The metal sector includes large conglomerates (hereafter chaebols) in heavy industries, such as autos and steel. The finance sector is representative of white collar unions that experienced severe structural adjustments after the Asian financial crisis of the late 1990s. The hospital sector trade union movement was able to establish the first democratic industrial union after the transition to democracy. The data for this study were drawn from primary and secondary document analyses and interviews with key union leaders and rank-and-file members. The primary documents included trade unions' internal documents, bargaining agreements and minutes of executive committee meetings. This study concludes with a review of changes in Korea's labour law in 2010, which may severely constrain the bargaining rights of non-standard workers and weaken unstable industrial unions and their industry-level bargaining. It appears that Korean labour's desperate attempt to overcome the neoliberal paradigm and restore solidarity is likely to face yet another new challenge.

2. Industrial relations under the developmental state and after democratization

The basic configuration of Korean industrial relations was formed by the legacies of the developmental state and democratization. Early state-led industrial strategy in the 1960s and 1970s took advantage of cheap and abundant labour, which ignited the struggles of women workers in export-oriented light industries (Koo 2001). However, labour movements in these labour-intensive manufacturing industries were too weak and decentralized to mount an effective collective action. As heavy and chemical industrialization in the 1980s encouraged monopolistic production by large chaebols, the number of skilled male workers in these sectors increased drastically. In the early 1980s, wage differentials by firm size were practically non-existent in Korea. Even workers in small firms earned as much as workers in large firms, despite the fact that these firms differed significantly in their ability to pay after the growth of heavy and chemical industries, which created and favoured chaebols. The fact that workers in large firms failed to secure wage premiums up to the late 1980s demonstrates how successful the repressive labour control policies of the previous authoritarian governments were.

The legal framework of industrial relations under the former authoritarian regimes was specifically designed to discourage workers' concerted collective action. The amendment of the labour laws in 1963 banned trade unions from participating in political activities and establishing a second union at both the plant and national level. The Chun Doo Hwan regime in the 1980s proved even more restrictive in its labour policies than its predecessor. Under 1980 revisions of the labour laws, national and industry-wide collective bargaining was formally eliminated, and all intervention by third parties in the collective bargaining process, including that by the national federations, was forbidden. In addition, the changed labour laws in 1980 completely decentralized the Korean collective bargaining structure by reorganizing industry-based unions into enterprise unions. As a result, most trade unions were organized on an enterprise basis, and collective bargaining is conducted generally at the enterprise level.

Core workers in these strategically important sectors of the economy forcefully demanded their long-lost rights, when the rupture of the authoritarian rule opened the political space to do so in 1987. Korean workers have indeed defied common expectations associated with East Asian labour quiescence. During the Workers' Struggle of July–September 1987, over 3,000 strikes took place. It was a spontaneous outbreak of long repressed demands rather than a well-orchestrated political action. Nevertheless, it heralded the beginning of the new democratic movement, which aimed at progressive social reforms. As shown in Table 1, the number of legally recognized unions and union members grew rapidly since 1987: from 1987 to 1989, the number of local enterprise unions almost tripled, and that of union members doubled.

Table 1.  Union Density, Union Membership, Number of Unions, 1985–1992
YearUnion densityUnion membership (in thousands)Number of unions
  1. Source: Ministry of Labour, internal documents, various years.


After peaking at 18.7 per cent in 1989, however, union density has steadily declined since 1990. The shift of employment from manufacturing to services, aggressive managerial opposition to unionization and the rapid expansion of non-standard workers have all contributed to the ever-decreasing union density. Union density alone cannot fully capture the power of the trade union movements. The degree of centralization is an important measure of working class political power especially for a country such as Korea, as a strongly centralized labour movement can negotiate for an inclusion of labour in the new democratic politics and induce major concessions from the state, which temporarily has to discard its usual anti-labour stance. Unfortunately, this crucial aspect of working class power has always been absent in Korea. Table 2 demonstrates that the decentralized collective bargaining structure has not changed during the transition period. Given the fragmented bargaining structure, sectoral imbalances in Korean industries began to generate a dualistic labour market structure, mostly thanks to vigorous enterprise union activities, which pushed for wage increases for their own members. Core workers in large firms did not want to move towards industry-level bargaining and to engage in political struggles that might only benefit workers in secondary labour markets, and such interests and preferences of rank-and-file members affected the platform of the union leadership via the mechanism of democratically contested union election (Lee 1998).

Table 2.  Collective Bargaining Structure: 1987 and 1988
 Collective bargaining (%)Wage bargaining (%)
  1. Source: Kim (1994).

Enterprise bargaining77.485.466.874.2
Multi-employer bargaining14.210.421.222.6
Labour-management consultation8.
Number of respondents155144184190

In addition, national-level labour federations in Korea were extremely weak and divided. The older and larger national labour organization, the Federation of Korean Trade Unions (FKTU; Hankuk Nochong), has lacked effective leadership since its birth. In 1961, the military junta, which took over power, formed 15 nationwide industrial unions in a top-down manner, and united them at the top in the FKTU. Since then, the FKTU has been under the direct control of the state authority and its intelligence agency, the Korean Central Intelligence Agency (KCIA). The FKTU also inherited anti-leftist ideology from the Confederation of Korean Trade Unions (CKTU; Daehan Nochong), the right-wing paramilitary youth group established in 1946 for the sole purpose of countering the revolutionary National Council of Korea Labour Unions (NCKU; Chunpyung). As a result, the opportunistic and collaborative FKTU became a political supporter of each authoritarian regime, urging local unions to cooperate with state policies. The FKTU's lack of independent political interests perpetuated an institutional vacuum in workers' interest representation. The FKTU wanted to rebuild its tainted public image after 1987 when dissident labour leaders seriously challenged its leadership by discarding old anti-communism and criticizing the state's unfair intervention in labour affairs.

Since the Workers' Struggle of July–September 1987, there have been vigorous attempts by labour activists to create an alternative national-level organization, which could eventually replace the pro-business FKTU and provide a political base to initiate revising and/or eliminating anti-labour laws. The Korean Trade Union Congress (KTUC; Chunnohyup), established on January 1990 with the participation of 14 regional union councils and two occupational/industrial councils, was the first outcome. The KTUC categorically rejected the FKTU's collaborative business unionism and pursued comprehensive social reforms. Another national-level organization, the Korean Council of Trade Unions (KCTU; Chunnodae) was founded in 1993, in a continuing effort to establish the second national centre based on industrial unionism. The KCTU is an association of several union federations, including one non-manufacturing/white-colour union federation, two chaebol enterprise union councils and the KTUC. The KCTU eventually established what was informally referred to as the Second Nochong, the Korean Confederation of Trade Unions (KCTU, Minju Nochong) on 11 November 1995. The reasons that the KCTU transformed itself into the more formally organized KCTU were basically twofold. First, the KCTU wanted to restructure enterprise-based unions into industry-based unions through the strengthened leadership. Second, the KCTU sought to establish the democratic labour movement as the major social democratic force in Korea, capable of participating in the formal policy-making process. It was true that the KCTU aimed clearly at militant and political unionism, but the KCTU-affiliated large enterprise unions was not always supportive of its political agenda. The newly found strong bargaining power at the enterprise level with decentralized collective bargaining structure and divided national level labour movements precipitated the premature decline of the political labour movement.

3. The financial crisis and the fragmentation of labour

The Growth of Non-Standard Employment

The Korean economy started to become more tightly integrated into the increasingly globalized world economy immediately after its transition to democracy in the late 1980s, when the pace of trade liberalization accelerated. Facing stronger competition, Korean employers began to demand labour flexibility. They were burdened with strong and militant trade union movements, which spread rapidly since the democratic transition and the Great Workers' Struggle1 of 1987. Core workers in protected heavy industries were organized by large enterprise unions, and were able to demand higher rates of wage increases by using militant bargaining strategies. During this period of economic prosperity, higher wage rates spread to small- and medium-sized firms, legitimizing the enterprise union system coupled with a decentralized collective bargaining structure.

With increased competitive pressure from financial liberalization, the government and employers actively pursued various strategies to promote labour market flexibility. The amendment of the Labour Standard Act in December 1996 reflected their desire to eliminate rigid rules and regulations governing internal labour markets in core sectors of the economy. The pivotal event that signalled the start of massive neoliberal labour reforms in Korea was the Asian financial crisis of 1997. After the crisis, there was a general consensus among Korean state elites that the major features of the Anglo-American economic system, that is shareholder-oriented corporate governance, open and competitive financial systems, and flexible labour markets, are the ‘global standards’ (Lim and Jang 2006).

Almost a decade of the constant emphasis on the need for a flexible labour market2 has produced a large number of non-standard workers in Korea. In general, non-standard employment includes those jobs with temporary or short-term contracts, part-time jobs and dispatch work. As displayed in Table 3, the increasing number of non-standard workers was conservatively estimated by the government at approximately 35 per cent of all wage-earners between 2003 and 2009, which discredits the assumption that Korean labour market still lacks flexibility. The Korean Contingent Workers' Center estimated that there were more than 8.5 million non-standard workers in 2009, accounting for approximately 52 per cent of all wage-earners. The reason that the Center's and other pro-labour research institutes' estimates are substantially higher than that by the government is because they considered low-wage workers in small shops employing less than five workers as non-standard workers. These workers are mainly day or temporary workers who do not receive any lump-sum retirement pay, and their wages are sometimes lower than those of average non-standard workers. Workers in non-standard employment arrangements earned less than regular workers, and the wage gap has been widening. In August 2000, non-standard workers earned 53.5 per cent of the average monthly wage of regular workers. In March 2010, they earned just 46.8 per cent (Korean Contingent Workers' Center 2010). The size of the non-standard workforce has raised concerns because these jobs pay low wages, provide few social insurance benefits and do not have implicit guarantees of long-term employment.

Table 3.  Non-Standard Employment, 2001–2009
YearNational Statistics OfficeKorean Contingent Workers' Center
  1. Note: These figures are for August of each year, Supplementary Survey on Economically Active Population, National Statistics Office.

  2. Sources: The Korea Labour Institute (2010), KLI Labour Statistics; and Korean Contingent Workers' Center (2010), The Contingent Workers.


The Protection of Fixed-Term and Part-Time Employees Act (hereafter ‘the Nonstandard Employment Protection Act’) was legislated in 2007 in this context. This new law was supposed to provide non-standard workers with more rights by prohibiting an undue preference for regular workers over non-standard workers based on their employment status as long as they perform the same or similar jobs and limits the employment period for temporary workers to a maximum of two years, after which employers are required to convert non-standard workers into regular ones. The Nonstandard Employment Protection Act, however, has not operated in a vacuum. With the neoliberal government hostile to labour, and employers eager to lower labour costs, the effects of the Act have been quite harmful to non-standard workers. The fact that the law lacks some fundamental measures that restrict reasons for hiring non-standard workers has not helped either. First, it is easy to evade the article prohibiting discrimination against non-standard workers: employers can simply make a few changes to the job description for non-standard workers to make it slightly different from that for regular workers. Employers have also refused to turn non-standard workers who worked for more than two years into regular workers. Instead, they have been making pseudo-standard employment contracts that are not very different from those for non-standard workers in terms of wages and working conditions. However, most employers have found better, more flexible ways, including outsourcing and the termination of non-standard employment contracts. As a result, the Act has facilitated the proliferation of extensive outsourcing and massive layoffs of short-term contract workers. A KCTU official stated that ‘this law is not about protecting non-standard workers, but about institutionalizing and mass-producing them . . . as long as it exists, why would a company hire expensive regular workers when they can easily hire and fire cheap temps?’ (The New York Times, 23 July 2009).

The Nonstandard Employment Protection Act has thus been far less effective than expected, or to be precise, it may have worsened the situation by further dividing Korean workers. The Korean labour market is now composed not only of regular and non-standard workers but also of semi-regular workers and workers with extremely short-term contracts. Previous studies analysing changes in non-standard employment before and after the Act have reported that although the number of temporary workers expecting contract renewals declined slightly, the number of part-timers, in-house subcontracted workers, on-call workers, day workers and other forms of low-wage and short-term non-standard workers increased (Kwon 2009; Lee and Chung 2008). The 2007–2008 strike by E-Land workers, which lasted more than 500 days, was set off by the implementation of the law, and it epitomized the struggle against the neoliberal labour stance of both the state and employers. Clearly, the neoliberal reforms prescribed by the International Monetary Fund (IMF) and carried out by the Korean government led to the deregulation of the labour market, which weakened labour and divided the labour movement. Efforts to centralize trade unions have largely been ignited by this dramatic degradation of labour.

Tripartite Commission and the Partial Inclusion of Labour

It was ironic that President Kim Dae Jung, the former anti-authoritarian opposition leader who took office in 1998 with a mission to cure the ailing economy, consolidated the political coalition that supported neoliberal reforms. He was the first Korean president who was not overtly hostile towards labour. The Kim administration attempted to enhance the legitimacy of neoliberal labour policies by creating the Tripartite Commission (Nosajung Wiwonhoe, renamed the Economic and Social Development Commission in 2007). Although its role was limited because of the lack of co-ordination and the absence of the militant KCTU, the first social pact drafted by the Commission in 1998, ‘Tripartite Joint Statement on Fair Burden-Sharing in the Process of Overcoming the Economic Crisis,’ made it possible to revise the Labour Standard Act to ease the procedure for laying off workers due to managerial reasons. After the revision, the IMF imposed structural reforms and the resultant employment adjustment produced a large number of unemployed and workers employed under non-standard arrangements.

The first social pact also unfolded a political space for the labour movement to change the collective bargaining structure. Pledging to establish a ‘democratic market economy’ and needing the active co-operation of trade unions to implement the neoliberal reforms prescribed by the IMF, the Kim administration actively invited the participation of labour. According to the tripartite joint statement of 1998, the government promised labour that in exchange for labour's consent to employment adjustments, it would guarantee and support free collective bargaining arrangements between labour and management and expand public expenditure for social safety net programmes. Later that year, labour also demanded industry-level councils that could prepare for industry-level bargaining, and the Kim administration promised to support the formation of such councils. For the first time in Korea, labour politics began to revolve around the important issues of unemployment, employment protection and welfare expansion (Noh 1999). The participation of labour representatives in the social dialogue allowed these representatives to realize the increasing need for reorganizing their union structure towards more centralized forms.

Since assuming office in 2003, former human rights lawyer and labour activist Roh Moo-Hyun continued, at least in the early stages of his presidency, the policy of partial labour inclusion. When Roh was elected, his transition team affirmed that the new government would institute a multi-level collective bargaining system; endorse industrial unions and industry-level bargaining pushed by trade union leadership; reduce the increasing gap between regular workers in large firms and workers in non-standard employment arrangements; and refrain from interfering in conflicts between labour and management. During the early stages of his term, labour officials were willing to participate in meetings arranged by industrial union leaders and encouraged employers' representatives to participate. As a result, employers could no longer ignore industrial unions' demand for multi-employer bargaining.

However, policies promoting the partial inclusion of labour do not necessarily mean that the Korean government fully supported the progressive goal of the democratic labour movement. As Kong (2006a: 379) stated, the authoritarian inclusion of labour meant ‘being selectively pro-labour without being anti-business.’ Actually, Roh's presidency was marked by a relatively large number of labour disputes because of the tougher measures aimed at facilitating labour market flexibility and structural adjustments. The priority of the government's labour policy switched back to enforcing laws to prevent illegal strikes, and the state gradually withdrew its lukewarm support for industrial unions. The government also refused to institute extension mechanisms,3 which would have been critical to stabilizing the incipient industry-level bargaining, and reducing the gap in wages and working conditions between small and large firms. Instead, President Roh began following his predecessor's neoliberal labour policies, emphasizing the need for improved labour flexibility in meeting global standards.

At the same time, the Roh administration managed to capitalize on the existing framework of social dialogue to mitigate trade unions' opposition. The second social pact for job creation was forged by the Tripartite Commission in February 2004. Labour agreed to cooperate in stabilizing wages over the next two years to create jobs and reduce the wage gaps between regular and non-standard workers. Firms promised to refrain from downsizing, and in cases in which this would be inevitable, to minimize the number of dismissed workers through good-faith consultations with trade unions. In return for the compromise between labour and management, the government would facilitate corporate investment for job creation by deregulating business activities, and expanding tax and financial benefits to firms. All social partners agreed to cooperate to improve working conditions for the non-standard workforce. To prevent a repeat of the failure of the first pact, the representatives also agreed to establish the ‘Joint Public-Private Sector Committee on Job Creation’ to monitor the implementation and progress of the promises made in the pact (Tripartite Commission 2004).

However, the second social pact also failed to achieve its proposed goals. The Tripartite Commission remained crippled by the absence of the more militant KCTU, whose member unions included the largest enterprise unions with the most bargaining power, such as that at Hyundai Motor Company (HMC). In fact, the success of the social pact was critically dependent on whether large enterprise unions would comply with the promises made at the national level. Because of the deepening gap between the primary and secondary labour markets, trade union members in large enterprises had a strong preference for enterprise-level bargaining, which remained as an important institution in which pay levels and working-time arrangements were decided. Because of the long history of confrontational labour relations and the lack of co-operation between and within social partners, Korea did not have an effective mechanism for co-ordination of collective bargaining.

4. ‘Disorganized’ centralization: a cross-sectoral comparison

Industrial unionism has been a long-term goal of democratic labour movement in Korea. The threat of high unemployment and redundancy have made some enterprise union leaders realize that the enterprise union system cannot adequately protect their own employment security, let alone represent the interests of all workers. As discussed in the previous section, the centralization of collective bargaining structure in Korea has also been facilitated by changes in the political sphere, which ended a long period of labour exclusion. The liberalization of Korea's labour market has been accompanied by minimal and limited inclusionary labour politics (Kong 2006b), which nonetheless provided labour with exceptional opportunities to reorganize enterprise unions into industrial ones.

Table 4 shows that despite the declining union density, the trade union structure has been centralizing since the early 2000s. The Ministry of Employment and Labour estimated that a majority (52.9 per cent) of all organized workers belonged to industrial or regional trade unions in 2009. Among the two national labour federations, the KCTU, a militant organization, has been more successful in transforming its affiliated unions into industrial ones than the FKTU, a moderate organization. About 35 per cent of union members belonged to the KCTU, compared to the FKTU, which organizes 45 per cent in 2009. The remaining 20 per cent of union members belonged to neither federations. Currently, over 70 per cent of KCTU-affiliated union members belong to industrial unions, whereas it is only 35 per cent for FKTU-affiliated members. These new industrial unions have adamantly demanded industry-level bargaining. According to Lee (2009), approximately 25 per cent of unionized workplaces in the manufacturing sector, 45 per cent in the finance and insurance sector, 56 per cent in the health and welfare sector, and 21 per cent in the public administration sector were covered by industry- or region-wide bargaining.

Table 4.  Union Density, Union Centralization and the Unionization Rate for Non-standard Workers
YearUnion density% of Industrial/regional union members% of Non-standard workers among union members
  1. Note: ‘—’ means that the data are not available.

  2. Sources: The Ministry of Employment and Labour (2010), Analysis of Trade Union Structure, 2009; the Korea Labour Institute (2010), KLI Labour Statistics.


However, it remains unclear whether Korea's collective bargaining structure can be stabilized at the industry level. Although employers failed to prevent the changes in the formal collective bargaining structure, many refused to participate in sectoral bargaining. The disorganized industry-level bargaining structure was not generating any collective goods such as standardized employment terms and reductions in transaction costs, and the value of such goods themselves was diminishing in the context of globalization. Threatened by global competition, employers in general have adapted to economic turmoil by decentralizing the bargaining structure. This decentralization has allowed individual firms greater flexibility in negotiating wages and working conditions to suit their unique needs (Marginson et al. 2003; Poulsen 2006). It should be noted, however, that employers might differ in terms of their participation in the multi-employer bargaining process. In Korea, those most successful in blocking centralization were larger employers in the heavily export-oriented metal sector. Employers in protected public service sectors maintained their interest in sectoral bargaining because it helped them to avoid powerful enterprise unions' leapfrogging wage claims.

The decentralized structure of industrial unions also complicated the process of centralization. Industrial union leaders have operated within two major constraints. They have had to redistribute ‘upward’ the organizational resources that had been owned by enterprise unions while catering to the needs of rank-and-file members. These members have not always been supportive of their leadership's efforts to establish sectoral bargaining arrangements. They have been concerned that sectoral bargaining would be used as an instrument for wage restraints. There has been a low degree of co-ordination between bargaining levels, and sectoral agreements have seldom produced a co-ordinated framework for supplementary agreements at the enterprise level. Thus, organizational strategies of industrial unions mattered. When industrial unions were willing to accommodate employers' and rank-and-file members' needs for greater flexibility, the process of centralization was much smoother.

The substantive outcomes of the new bargaining structure in Korea can be evaluated by a variety of factors, but the present study focuses on collective bargaining agreements for non-standard workers. With the deepening gap between the primary and the secondary labour markets, regular workers in large enterprises were not always eager to defend non-standard workers' basic rights because these temporary workers were the first to be eliminated when needed. Thus, one of industrial unions' most pressing goals was to reduce the representation gap and protect non-standard workers via collective bargaining. However, because wage rates and working conditions were negotiated mainly at the enterprise level in the disorganized sectoral bargaining, a large number of unorganized non-standard workers were not properly protected by these newly emerging centralized collective bargaining arrangements.

To investigate the mechanisms and outcomes of this particular case of bargaining centralization in Korea, this study conducts a cross-sectoral comparison. Since the mid-1990s, the relevance of national-level industrial relations system as a unit of analysis has been questioned (Hollingsworth et al. 1994; Locke and Kochan 1995), and many researchers have focused on the sector as the key unit for comparative analysis (Wailes 2007). That has been mainly because economic globalization has placed enormous pressure on highly exposed sectors, making them more similar across national borders. At the same time, the diversity within national systems has been rising because of different levels of exposure to trade and international competition. The three industrial sectors chosen are the metal, finance and hospital sectors. Table 5 summarizes the major characteristics of the three sectors. These sectors were selected not only because they had the most advanced sectoral bargaining arrangements, but also because they reflected diverse market situations and the strategies of trade unions' and employers'. Presentation of these sectors begins with the metal sector, the sector with the most disorganized centralization, and ends with the hospital sector, the sector least disorganized.

Table 5.  Sectors and Basic Characteristics
Trade exposureHighMediumLow
Union strategyRadical/militantModerate/collaborativeModerate/militant
Affiliated national centreKCTUFKTUKCTU
No. of union members147,70186,52636,500
Date of union establishmentFebruary 2001March 2000(1961)February 1998
% Sectoral bargaining14%a100%88%b

The Metal Sector

Affiliated with the KCTU, the Korean Metal Workers Union (KMWU) inherited a tradition of democratic labour movement dating back to the late 1980s. Its membership, which was just over 40,000 at first, grew to almost 150,000 in 2006 when enterprise unions in the auto industry voted to join the KMWU (KMWU 2007a). The membership of large enterprise unions, however, turned out to be highly problematic. The union members at HMC (43,000) and Kia Motors (28,000), a subsidiary of the Hyundai-Kia Automotive Group, represented approximately 50 per cent of the KMWU's entire membership. Because these largest automakers frequently engaged numerous small- and medium-sized firms for subcontract work, 78 per cent of its members belong to the auto industry in 2007. In addition, enterprise size varied considerably in the metal sector. Although the vast majority of firms were those employing less than 300 workers (76 per cent), the number of union members employed by these small- and medium-sized firms accounted for approximately 7 per cent of the membership (KMWU 2008). The KMWU's unique internal structure made it extremely difficult to co-ordinate conflicting interests within the union.

Rank-and-file members in larger firms preferred to maintain the relative autonomy of company-level bargaining. On the contrary, those in small firms insisted on the centralization of the bargaining level to at least the regional level. In an effort to undermine the ‘enterprise consciousness’ of its members in large firms, the KMWU decided to disintegrate local unions associated with a specific firm and reorganize them as regional locals by September 2009, but this did not materialize. Rank-and-file members associated with enterprise locals were vehemently against the decision, and it has created much discord inside the KMWU. The president of the KMWU summarized the situation as follows:4‘Enterprise locals and the KMWU do not recognize and listen to each other. It's a daunting problem that holds back the progress of industrial unionism.’

The Korean Metal Industrial Employers Association (KMEA), the official employers' association in the metal sector, reflected the structure of the industrial union. The KMEA consisted of mainly small- and medium-sized firms, covering just 14 per cent of the KMWU's membership (Shin 2008). Large employers were categorically opposed to sector-wide bargaining. The most explicit example of this view can be found in Hyundai's official reply to the KMWU's request to participate in sectoral bargaining:5‘We would like to remind you that Hyundai Motor Company does not have a duty to engage in sectoral bargaining. It cannot reflect the great variety of industries and firm sizes, working conditions, and ability to pay in the metal sector. In addition, with overlapping bargaining issues at different levels, it is hardly an efficient way to conduct collective bargaining.’ Some employers demanded negotiations at the single-industry level, such as the automotive or shipbuilding industry, instead of bargaining at the sectoral or regional levels. They were searching for ways in which their own collective action problems could be overcome. Creating smaller and more homogeneous bargaining units, however, had not been acceptable to the KMWU because it could have eroded the principle of one unified industrial unionism.

As a result, efforts to institutionalize a centralized bargaining structure in the metal sector were in disarray, and collective bargaining took place in many forms. Sector-level central agreements have been forged since 2003, but these agreements did not prescribe basic employment standards for the metal sector, except for setting minimum wages. Occasionally, lower-level bargaining began even before a sectoral agreement was forged. Because of the unique structure of the KMWU and the limited scope of sectoral bargaining, non-standard workers in this sector have suffered from a severe representation gap. There were a total of 70,895 non-standard workers in workplaces affiliated with the KMWU in 2007. However, non-standard workers in a majority of these workplaces (52.7 per cent) were not allowed to join the same local unions, which are composed mainly of regular workers (KMWU 2007c). Even when non-standard workers are unionized under the KMWU, they have to join a regional local union because enterprise locals for regular workers have resisted the formation of unified local unions representing both regular and non-standard workers. For example, union officers at HMC and Kia Motors, the two largest enterprise locals in the KMWU, failed to change their unions' internal regulations, which would have turned non-standard workers into regular members of their local unions (The Hankyoreh, 12 October 2007). According to a survey by the KMWU in 2007, compared with regular workers, non-standard workers did the same (66.1 per cent) or more difficult work (19.4 per cent) while receiving only 60.2 per cent of regular workers' wage rates (KMWU 2007c). Because sectoral bargaining in the metal sector has not regulated wages and working conditions, it has been difficult to standardize wages and working conditions for regular and non-standard workers.

The union-bargaining power of the metal sector has been declining because employers have been leaving the country in search for low-cost production facilities, particularly those in China. In addition, the extremely heterogeneous internal composition of the metal sector has not been helpful. Because export-oriented chaebol auto companies have been imposing further cost-cutting measures on their supplier chain, the wage gap between large and small firms and between regular and non-standard workers has been worsening. Both employers and local union members in large firms have wanted to maintain the flexibility and autonomy of the enterprise-level bargaining. It is ironic that the more the KMWU emphasizes the solidarity of the metal sector, the more the collective bargaining structure gets disorganized as regular workers in large firms become more concerned with sectoral bargaining, which can be used as an instrument of wage restraint. The radical strategies of the KMWU have thus far been unsuccessful.

The Finance Sector

As a result of the Asian financial crisis of 1997, the extent of restructuring in the finance sector was deeper than that in any other sectors. The government impelled the reduction of the number of large banks by allowing profitable banks to acquire financially weaker ones and encouraging mergers between banks that were large and solvent (Lee and Park 2007). The Korean Financial Industry Union (KFIU) was thus established amid threats of massive layoffs and drastic employment adjustments. Compared with the metal sector, the finance sector is smaller and more homogeneous because it is composed mainly of banks and financial institutions. The KFIU does not have regional organizations,6 and each bank constitutes a local union.7 Thus, the transition from an enterprise union to a local union of the KFIU has been orderly and rather uneventful. However, the peaceful transformation of the union structure implies that former enterprise unions did not have to change their company-centred operational practices. In addition, the KFIU's leadership has been accommodating, recognizing the strategic importance of local unions.

The KFIU (affiliated with the moderate FKTU) originated from the first industrial union in the financial sector, the Korea Finance Workers Union (KFWU), which was founded in 1961 (Kim 2003). In addition to the institutional memory of previous sectoral bargaining, the finance sector has been blessed with collaborative employers.8 The Korean government has long controlled the entire credit system, and the banking industry had remained highly regulated until very recently. Despite its privatization and deregulation after the financial crisis, the finance sector has been subject to unofficial interventions by the government (Lee and Park 2007). Because of this ‘quasi-public’ characteristic of banks and financial institutions, the government has directly or indirectly influenced the appointment of chief executive officers (CEOs). As a confrontational approach to labour relations can harm their career, these officers have generally avoided taking an anti-union stance.9

The government's regulatory efforts have also resulted in similar wages and working conditions across the locals. The sector's employers, already accustomed to pattern bargaining, have been less antagonistic towards sectoral bargaining arrangements. Since 2003, the Korea Federation of Banks (KEB), a trade association of Korean banks, has taken the role of employers' association, and has been helpful in consolidating the centralization process. However, because the KEB is not an official employers' association, its ability to co-ordinate and enforce collective agreements has been quite limited (Park 2007), and thus, the KFIU has persistently demanded the formation of an association for employers, which was finally established in 2009. Sectoral bargaining has taken place without any interruption since the formation of the industrial union in 2000.

Unlike in the centralized formal system, however, the actual practice in this sector has been much more decentralized. For example, increases in wage rates are recommended at sector-level negotiations, but individual banks are free to adjust the rates at the local bargaining table. There are many ‘opening clauses’, so sectoral agreements constitute only general guidelines that have to be renegotiated at the company level. Frequently, collective bargaining issues have been identical in both sectoral and local bargaining arrangements. Although employers have complained about the inefficiency of bargaining twice at each level, it could be this kind of flexibility that has actually maintained a formal system of sectoral bargaining. Employers' desire for managerial flexibility has been well preserved under the current system.

The KFIU, just like the industrial union in the metal sector, has failed to recruit non-standard workers. The finance sector lost almost 60,000 union members because of the Asian financial crisis, and by 2007, the sector as a whole employed approximately 80,000 regular workers and 40,000 non-standard workers, which indicates that one-third of its workforce was composed of non-standard workers. Despite the fact that some non-standard workers are former regular employees who were dismissed because of the financial crisis, the wage gap between regular and non-standard workers in the finance sector has been wider than that in the other sectors. After the financial crisis, Korean banks have developed more individualized and performance-based compensation systems, which have allowed the wage gap to widen even further.

Because local unions, which have been composed mainly of regular workers, have been reluctant to accept non-standard workers as their members, the KFIU established a separated local union solely for organizing non-standard workers in 2004. However, fewer than 100 non-standard workers joined the local union by 2007, and its activity has been quite limited. After the president of the local union retired, nobody was appointed or elected to represent the union. Instead, one local officer assumed the role of acting president.10 In addition, sectoral bargaining has not included employment arrangements and working conditions for non-standard workers. Thus, the acceptable ratio of non-standard labour, the extent of non-standard jobs and wage rates for non-standard workers have been negotiated at the company level by local unions (KFIU 2006).

The Hospital Sector

It was only after the democratic transition in 1987 that unionization in hospitals began to take place. The Korean Health and Medical Workers Union (KHMU) was also established to counter the casualization of labour as a result of neoliberal labour policies. Approximately 90 per cent of hospitals in Korea are private, whereas national or public hospitals account for just 10 per cent of all medical institutions (Lee 2007). Among private hospitals, private university hospitals have been the dominant players because they represent approximately 50 per cent of the KHMU's entire membership. They have also been the largest and had the greatest ability to pay. There have been huge wage gaps and differences in working conditions between large and small- and medium-sized hospitals. The KHMU has not been immune to disputes regarding proper organizational and bargaining units. Unlike the KMWU, however, the KHMU has pragmatically decided to accept enterprise locals and made consistent efforts to encourage local unions to participate in sectoral agendas and activities (Cho 2004).

Employers in the hospital sector have adamantly resisted sectoral bargaining at first, and thus, it took six years for the KHMU to forge the first sectoral agreement in 2004. Labour relations in the hospital sector had been highly contentious as militant KCTU-affiliated enterprise unions were frequently in conflict with the top management of hospitals who typically did not have adequate expertise in labour relations. Since then, however, centralization efforts in the hospital sector have progressed well, and bargaining rules and regulations have been institutionalized rapidly. Unlike the industrial unions in the other two sectors, the KHMU has had control over the bargaining of its locals. Wage rates negotiated through sectoral agreements have generally been respected at the local bargaining table.

Because sectoral bargaining in the hospital sector has been relatively successful in the provision of collective goods, such as taking wages and other working conditions out of competition, some employers, particularly those in small hospitals who were not adequately equipped to address strong enterprise unions, have started to realize the benefits of a unified response.11 Hospitals have been protected from global competition, so employers with militant enterprise unions have been pressured to increase wages more rapidly than their unorganized counterparts. Employers in large hospitals have enjoyed reductions in labour costs because the KHMU has restrained the wage demand in its effort to curb the widening gap between large and small hospitals. The KHMU has also been flexible in accommodating employers' demands to allow differential increases in wages depending on hospital types. Based on the framework agreements of 2004, which required the formation of an employers' association, the Korean Health and Medical Employers Association (KHMEA) was established in 2007 with 96 member hospitals.

Of the three sectors, the hospital sector has displayed the least amount of discrepancy between the formal system of centralization and the actual practice. The KHMU, through sectoral bargaining, has been able to improve worker solidarity. In the 2007 sectoral agreement, the KHMU and the KHMEA agreed on a 4.0∼5.3 per cent of wage increase rates, and on spending a certain portion (1.3∼1.8 per cent) of these increases to improve working conditions for non-standard workers and eliminate discriminatory practices against them. Since the agreement, the proportion of non-standard workers has decreased from 20.4 per cent in 2007 to 16.8 per cent in 2008 (KHMU 2008). In 2009, there were 7,247 non-standard workers in KHMU-affiliated hospitals, and 8.9 per cent of them were members of the KHMU, which were well above the national average (KHMU 2009). Unlike those in the other two sectors, organized non-standard workers in the hospital sector have been allowed to join enterprise locals along with regular workers. There is no separate local union just for non-standard workers in the hospital sector. The wage gap between regular and non-standard workers was narrower than the national average. In 2007, non-standard workers earned 59.5 per cent of regular workers' wage (KCTU 2009).

However, the KHMU has not been free from the general organizational dilemma facing other industrial unions in Korea.12 Compared with union members in small private hospitals, those in more profitable private university hospitals whose ability to pay is greater have been less sympathetic to the KHMU's demands for a more centralized collective bargaining structure (Lee 2005, 2006).

5. Labour law revision in 2010 and the future of Korean trade union movements

Then, how sustainable are the newly centralized collective bargaining structures in Korea? Because industrial relations in Korea are inextricably connected to politics, President Lee Myung-bak, the leader of the conservative Grand National Party (GNP) who took office in 2008, has adversely affected the progress of sectoral bargaining. The Lee administration has placed greater emphasis on employment policies13 that could promote labour market flexibility, and paid little attention to labour relations. For example, in 2008, the government and the ruling GNP announced a plan to extend the employment period for non-standard workers from two years to a maximum of four years, arguing that a four-year term would protect non-standard workers better than a two-year term. However, both labour and the public have been vehemently against the amendment, arguing that it would produce more non-standard workers. In addition, after creating the ‘National Employment Strategy Meeting’ in early 2010, the government announced the ‘Measures to Spread a Flexible Work System’, which dictates that the public sector should introduce more part-time jobs and more flexible work hours (Korea International Labour Foundation 2010). While emphasizing job creation and the need for strictly enforcing law and order regarding trade union activities, the Lee administration has made its anti-union stance very clear to employers. As a result, recent sectoral bargaining sessions in the metal and hospital sectors have ended in stalemates because of uncooperative employers.

The revision of Korea's labour law in January 2010 by this pro-business administration was one of the most important changes to the legal framework of Korea's labour relations since its democratization. The revised law, which had been delayed for 13 years, finally permits multiple unions at a single workplace from 1 July 2011, and prohibits employers from paying full-time union officials. Workers, however, may take time off from work to carry out trade union functions. The maximum time-off limit would be set by the Time-Off System Deliberation Committee. This revised law is expected to further promote organizational change in trade union movements and induce unions to go beyond individual enterprises, particularly those in small firms that cannot survive on their own.

One major problem with respect to the revised law is that it forces trade unions to determine the representative bargaining union if there are two or more trade unions. If trade unions cannot determine the representative union within the prescribed period, and the employer does not consent to dividing bargaining units, the trade union representing the majority of members automatically becomes the representative bargaining union. In the process, trade unions can organize a ‘joint representative bargaining team’ instead of delegating the bargaining authority to the majority union. However, a union that organizes less than 10 per cent of the members of all the trade unions is not eligible to participate in the joint team (Article 29-2). Because most trade unions composed of non-standard workers are not likely to organize more than 10 per cent of trade union members, the bargaining rights of non-standard and/or minority workers are likely to be seriously impeded by this restriction. The procedure for determining the representative bargaining union can also be detrimental to the burgeoning industry-level bargaining. Because the law requires all the trade unions at a workplace, including members of an industrial union, to participate in the procedure for determining the representative union, a branch or local of the industrial union that does not have a majority status would lose the right to participate in the industry-level bargaining process. This may, at least temporarily, strengthen enterprise unionism and enterprise-level bargaining, thereby weakening industrial unions and trade union movements based on broader membership categories.

6. Conclusion

Industrial relations in Korea after the Asian financial crisis of 1997 demonstrate that neoliberal reforms can be a double-edged sword for a country with dynamic trade union movements. Despite the inescapable pressures for labour flexibility imposed by the neoliberal state, union leaders have aspired to develop the capacity to intervene in the regulation of labour markets to prevent the persistent marginalization of non-standard workers. As a result, the fragmented enterprise-level bargaining structure, invented by the previous authoritarian state, has become considerably centralized in the last decade. Considering that the last three decades have witnessed the decentralization of collective bargaining in nearly all of the advanced industrial countries (Katz 1993), the Korean case is an exception. The Korean case is different from some European countries, for example Italy and Ireland, where centralized bargaining has re-emerged and survived (Baccaro 2000; Baccaro and Simoni 2007). Unlike such countries, Korea has maintained the enterprise union system for most of its industrialized period, and thus, its bargaining structure has been extremely fragmented at the enterprise level. Japan had been similar to Korea in terms of its enterprise unionism, but prolonged recession in Japan has further decentralized its collective bargaining structure and weakened the power and the influence of labour (Suzuki 2004). The power and dynamism of Korea's trade union movements after the financial crisis (Kim and Kim 2003) have stood in sharp contrast to those of Japan's trade union movements after the Heisei recession in the 1990s. The Korean case demonstrates that even if a change in the economic context weakens the rationale for multi-employer bargaining, the collective bargaining structure can still be centralized.

In the centralization process, the strategies of trade unions that could maximize the gains through the social dialogue were critical. However, the political space for union reorganization has been opened precisely because the state has required the co-operation of labour to implement neoliberal reform packages. In Korea, where the legacy of a powerful interventionist state still resonates, the state's strong neoliberal turn may signify limited opportunities for the consolidation of the centralized collective bargaining structure. The relentless pursuit of structural and employment adjustments has deepened the dualistic tendencies in labour markets, adversely affecting the formation of internally cohesive industrial unions. Interest aggregation among organized workers has been difficult because there has been a clear divide between workers in large firms, who generally avoided sectoral bargaining, and those in small firms, who generally wanted to be included. Organizing employers has been even more troublesome. The disorganized internal structure of industrial unions and the collective bargaining structure have not been able to solve employers' collective action problems. Along with their own local unions, some large employers have resisted sectoral bargaining.

Nonetheless, the three industrial sectors analysed in this study suggest the importance of recognizing diversity within a national industrial relations system. As Behrens and Jacoby (2004: 118) superbly illustrated based on case studies of German industrial relations sectors, there is ‘hidden diversity inside “national models” as well as those model's impressive capacity for change over time.’ Sectoral bargaining occurred in all three sectors, but the degree of bargaining co-ordination differed substantially. The formal structure of centralized bargaining was established with relative ease in the finance and hospital sectors, where the industrial unions were more flexible in allowing the bargaining autonomy of enterprise locals. Compared to the metal sector, those sectors were also relatively sheltered from international competition, and thus, their industrial unions possessed stronger bargaining power. The metal sector was characterized by the diverse internal composition of firms, with a few dominant chaebol automakers and a large number of small-sized firms serving as their subcontractors. The industrial unions in the metal sector were known for the militant and radical strategies that aimed to weaken the power of former large enterprise unions to reduce the huge gap in wages and working conditions. However, the result ran counter to the original intention. The radical strategies had critically harmed the industrial union's capacity to co-ordinate diverging interests within the union.

The extent to which a collective bargaining structure is centralized can also be judged by the outcome of such centralization. Disorganized centralization in the metal and finance sectors, where important decisions on wages and working conditions have been negotiated mainly at the company level, has not led to substantial changes in the dualistic structure of the sector's labour market. Collective bargaining agreements in the hospital sector, which to a certain extent has succeeded in narrowing the wage gap between non-standard and regular workers, are important exceptions to the trend, but the sector's efforts have also been limited in reversing the overwhelming process of raising labour market flexibility. Most non-standard workers in Korea still do not have a mechanism for voicing their opinions in their workplaces.

Although the prospects are rather grim, institutionalized sectoral bargaining practices in Korea are not likely to be dissolved soon. Unemployment shocks and the casualization of work in the Korean labour markets after the Asian financial crisis, together with strong neoliberal stance of the current government and its relentless drive to create more non-standard jobs, have produced many problems that cannot be solved by enterprise unions and company-level bargaining alone. Further, the recent economic crisis in 2008 has reinforced the anxieties associated with insecure labour markets. Thus, the trend towards bargaining centralization may continue, albeit slow paced and different from the Western route. However, it remains highly uncertain whether such centralization could lead to a certain degree of co-ordination that could narrow the gap between employees working under different employment arrangements.


This work was supported by the National Research Foundation of Korea Grant funded by the Korean government (NRF-2010-330-B00133).


  • 1

    During the summer of 1987, when Korea underwent its democratic transition, there were 3,000 violent strikes throughout the country. The average number of strikes was approximately about 200 a year until 1986 (Lee and Lee 2004).

  • 2

    It is not clear, however, whether the Korean labour market needs more flexibility. Cho and Keum (2001) found that the job retention rate for the Korean labour market during the Asian financial crisis decreased to a level that the US labour market has never experienced. By comparing the two-year retention rates before and after the crisis, they also suggested that job stability has been eroding.

  • 3

    Many industrial union leaders in Korea wanted to introduce the German model of industry-level bargaining and labour legislation, which consider a collective agreement as generally binding when the agreement covers 50 per cent of the employees. They requested the inclusion of the extension mechanism in Korea's labour law, but the Roh administration never considered this request seriously.

  • 4

    ‘Interview with the President of the KMWU,’Labor Society (Nodong Sahoe), January–February 2010 (in Korean).

  • 5

    An official reply to KMWU document no. 17-05-001, quoted from KMWU (2007b).

  • 6

    Approximately 90 per cent of KFIU workplaces are in Seoul, so there has been no need for regional offices.

  • 7

    A local union for non-standard workers is the only exception.

  • 8

    The tradition of labour-management co-operation in the finance sector goes back to the foundation of trade unions in 1959. At the time, some employers actively endorsed the formation of unions because of their pro-business attitudes.

  • 9

    Interview with an officer at the KFIU headquarters (October 2008).

  • 10

    Interview with a local union officer at Kookmin Bank (September 2008).

  • 11

    Interview with an officer at the KHMU headquarters (January 2009).

  • 12

    The local union at the Seoul National University Hospital, together with those at several other national university hospitals, has left the KHMU in protest against the KHMU's control over wage rates after the 2004 sectoral agreement.

  • 13

    The Ministry of Labor was renamed the Ministry of Employment and Labor on 5 July 2010.