From a dualized labor market to a dualized welfare state: Employment insecurity and welfare state development in South Korea

Health insurance

Mandatory health insurance was first introduced by the Park Chung Hee government, which was authoritarian in origin and nature and conservative in ideological orientation. Soon after seizing power in the 1960 military coup, Park’s junta government deliberated the introduction of social insurance for health care, unemployment, and industrial accidents as part of a social reform package to stabilize the economy. It decided to move forward with the industrial accident insurance program first and successfully instituted it. But the introduction and expansion of other social insurance programs was much more controversial.

As many welfare state scholars have noted, the specifics of social policy design are the real sources of distributional conflict among different groups (Mares, 2004). In this regard, it is highly suggestive that the Park government chose to implement a large-employer-centered system. This design received strong support from large businesses (Cho, 2008). The government mandated that large companies with more than 500 employees should provide workers with health insurance and that each employer should establish a separate insurance fund of its own, called a “health insurance society.” ² Each health insurance society was to set premiums according to the scope of coverage and to collect contributions. Although the system gradually grew to cover previously excluded groups—such as wage workers at smaller firms and non-wage workers in the rural and urban sectors—the initial institutional choice to develop a large-firm-centered system led to highly unequal health care coverage and wide variation in costs. It created a system in which the costs and benefits of health care insurance were determined by the insured’s labor market status. This segmentalist approach worked against workers at small firms, the self-employed, and those in the rural sector covered under “regional” health insurance societies and worked in favor of regular workers at large companies; thus, the approach generated political tension among social groups. Because risk was pooled within each insurer, insurance societies exhibited differential financial capabilities. Accordingly, those with the same level of earnings paid different amounts of contributions (Gauld, 2004). Wide differences by sector in terms of premiums and medical coverage also led disadvantaged groups to complain about unfairness and to demand major reform.

Although Korea’s transition from dictatorship to democracy brought the issue of health care reform to the fore, a truly national health insurance program is a recent political outcome. In 1989, President Roh Tae-woo vetoed a reform bill passed in the National Assembly, which was then dominated by opposition parties. However, in 2000, the Kim Dae Jung government finally put an end to the long, hard-fought battle over health insurance reform between those who benefited from the existing system and those who did not by integrating the separate insurance funds into a government-operated single payer system. This reform was intended to improve equity among the insured and health insurance system efficiency by virtue of risk pooling. The new, unified system still distinguishes between employees in the formal sector and non-regular workers, that is, the self-employed and those in the informal sector, in calculating premiums, but all the insured are entitled to the same coverage. Specifically, the former group contributes a fixed rate of their salary that is matched by employer contributions, and the latter pays different premiums based on property ownership, income, age and gender (SSA, 2011). The latter represents a highly heterogeneous labor market group, ranging from low-income, disadvantaged workers to high-income professionals. The unusually large Korean informal sector makes it difficult to levy a fair share of contributions on high-income groups that underreport their income and take advantage of administrative loopholes, and inequity between the formal and informal sectors persists.

Old-age pensions

The other two major social systems—old-age pensions and unemployment insurance—developed much later under democratic governments, which also raises an interesting question. In fact, the Park government considered introducing these two systems but did not push them through as with industrial accident and health insurance. With respect to old-age pensions, the Park government even passed a law and completed the necessary administrative preparation, but in the face of the first oil shock, it indefinitely postponed enforcement of the law. Evidence suggests that the main motivation for the Park government’s initial decision to establish a pension system was to secure a sizable and stable source of capital, which was needed for heavy industrialization (Yang, 2008). This plan, however, was met with strong opposition from businesses and workers who saw it as nothing more than another tax increase. ³ There are two main reasons why pensions were so strongly opposed while health insurance was so warmly welcomed. First, because pension benefits would be paid in the distant future and would be contingent on the government’s will and financial capacity to act on its promise, a new pension system was viewed with skepticism as an immediate burden rather than as reliable insurance for the future. ⁴

The second reason is equally important: the main beneficiaries of the new system—the employees of large firms—already had institutional mechanisms intended to provide retired workers with some measure of income security, that is, mandatory severance payments. The Labor Standards Act of 1953 stipulated that employers pay retiring workers a lump-sum “retirement benefit” in proportion to the length of their employment. Until the national pension system was adopted, this retirement benefits system was, to borrow from Margarita Estévez-Abe (2008), a “functional equivalent” of pensions for workers at firms regulated by the Labor Standards Act. However, because the Labor Standards Act always made exceptions for small firms, only workers hired by large and mid-sized firms were protected by the Labor Standards Act. To this day, small firms with fewer than five employees are exempt from the law and thus from the statutory requirement to provide retirement benefits. Essentially, this reflects division between the formal and informal sectors and demonstrates that firm size is a good indicator of labor market insecurity.

Going back to the discussion at hand, given the costs and risks associated with a government-operated pension system, employers in the formal sector did not have a reason to support it, particularly because they were already providing retirement benefits under the existing system. In fact, when a national pension was created later, the retirement benefits system caused a dispute between employers who demanded its repeal and workers who sought to retain it along with newly created pensions.

The postponed national pension was taken up by the Chun Doo-hwan government in 1986 and finally implemented in 1988 under the Roh Tae-woo government. Beginning with firms with 10 or more employees, it continued to extend coverage to smaller firms, rural areas, and finally, the self-employed in urban areas, following typical steps of expansion.

However, the long-term sustainability of the system became a politically controversial issue that was publicly questioned during the Kim Young Sam government. Consequently, the late 1990s were dominated by heated public debate over pension reform. The conservative camp argued for fundamental reform that would completely separate the earnings-related component from a flat-rate basic pension and would eventually privatize the former to reduce the government’s financial burden. The Kim Young Sam government decided on this option, but this decision was overturned in the midst of the Asian Financial Crisis. The more liberal incoming Kim Dae Jung government abandoned the radical two-tier plan and settled on a more moderate course of parametric reform, which reduced the replacement ratio from 70% to 60% and increased the age of pension eligibility from 60 to 65 (Yang, 2008).

Unemployment insurance

Previous studies in the welfare state literature have found that, among social insurance programs, unemployment insurance is the least common and tends to be adopted last (Flora and Heidenheimer, 1981; Mares and Carnes, 2009). Scholars attribute this tendency to the difficult political process through which unemployment becomes recognized as a social risk beyond individual responsibility (Mares, 2003). The establishment of unemployment insurance in Korea illustrates the process of overcoming concerns that unemployment insurance would create a disincentive to work. The bureaucrats who first proposed the idea and participated in designing unemployment insurance intentionally emphasized active labor market programs over unemployment benefits and named it “employment insurance” rather than unemployment insurance to stave off objections from conservative voices both inside and outside the government (Chung, 2008). It is also noteworthy that bureaucrats within the Ministry of Labor took the initiative to promote its adoption. Labor unions were not significantly involved in the initial process.

However, the issue became part of the national agenda in the early 1990s when it was co-opted by politicians and both workers and employers began to express their opinions. All major political parties incorporated unemployment insurance into their electoral platforms, and the labor confederation and the employers’ organization jointly called for early implementation of the system in 1993 (Lee and Moon, 2008). The newly elected Kim Young Sam government swiftly set to work, and the national employment insurance system was operational in 1995. Furthermore, the Asian Financial Crisis ushered in a series of major expansions due to the immediate need to cope with massive layoffs. Throughout this process, there was no serious political clash between labor unions, employers, and the government, unlike the development of health insurance and old-age pensions during this period. However, the established level of unemployment benefits was low, and the active labor market programs were largely ineffective.

Solving the puzzles

The first puzzle was why Korea’s authoritarian government, otherwise hostile to labor movements, launched social insurance programs. The proposed framework appears able to explain this puzzle. The authoritarian leadership, especially the Park government, adopted a variety of policy tools to develop capital-intensive, skill-intensive heavy industries in the pursuit of export-led industrialization. The production of heavy industrial goods required skilled workers in great numbers, which, of course, Korea was lacking in the 1970s. The shortage of skilled labor intensified with deepening industrialization, ⁵ and competition between manufacturing firms to secure skilled workers emerged as a major social concern by the mid-1970s (Yang, 2004).

Faced with the need to nurture skilled workers, the government pursued a number of measures, including the company-based provision of various welfare benefits. The introduction of mandatory medical insurance in 1976 is best understood as such an effort for the following reasons. First, the measure applied only to large firms with 500 or more workers. Second, the insurance was designed to be administered by each employer, guaranteeing discretion in administrative decisions. Therefore, the adoption of company-based medical insurance contributed to a tightening nexus between employment and medical benefits in the formal sector and reinforced skilled workers’ dependence on employers. Additionally, large employers, taking advantage of the oligopolistic market power granted by the government, were able to offer other fringe benefits to attract skilled workers. The divide between large and small firms would develop into full segregation of internal and external labor markets in the 1980s, consolidating a dualization of the economy. In short, it is not surprising that Korea’s authoritarian government promoted social insurance because this was consistent with their developmental strategy.

The second puzzle concerning employers’ compliant attitude toward social insurance is also addressed by my theoretical framework. Given that social insurance was designed to foster worker dependence on employers in order to help firms secure a much needed, high-quality labor force, large firms did not have particular reasons to be resistant. In contrast, small firms did have a financial reason to raise objections. Social insurance contributions would raise labor costs, and they were unable to shoulder additional costs. Accordingly, smaller firms were exempted from statutory requirements to join social insurance systems until a later stage of the expansion. Large firms, however, were very actively involved in the process of designing the health insurance system. The National Association of Medical Insurance Societies, a coordinating body of decentralized insurance corporations, was put into the hands of the employers’ peak organization, the Federation of Korean Industries (FKI), which worked in close consultation with the Ministry of Health and Social Affairs (Cho, 2008).

The employers’ position on unemployment insurance also lends strong support to my theory connecting employment insecurity for regular, skilled workers to social policy change. Unemployment insurance has a very different logic than health insurance. Unemployment insurance aims to provide income support in case of employment termination; thus, unlike medical benefits—which are contingent on the relationship with employers—unemployment benefits cannot be expected to facilitate workers’ dependence on employers. For this reason, Estévez et al. (2001) argue that whereas employment protection is an institutional mechanism of firm-specific skill economies, generous unemployment benefits tend to be found in economies geared toward industry-specific skills, such as Germany. Strong employment protection is useful in creating worker dependence on their employer. In contrast, unemployment insurance, which facilitates labor mobility between firms, is not as useful in that regard.

In Korea, unemployment insurance was introduced in the 1990s. Initially, employers were mildly resistant, but later, they expressed support for it because they expected unemployment benefits to help soften worker resistance to collective dismissal (Chung, 2008). Conversely, we may presume that unemployment insurance was not adopted earlier like other social insurance systems because it did not correspond to the company-based orientation of the social protection system.

The timing of unemployment insurance provides an insight into the third puzzle as well. In light of the globalization debate in the literature, the expansion of welfare reform in the late 1990s and early 2000s is puzzling. However, the employment insecurity-centered perspective can explain the expansion of social protection. This period saw the collapse of employment protection for regular, skilled workers and a concomitant breakdown of the work-based welfare system. A decline in security—of job security and the welfare benefits that come with it—led workers to seek a broader, more comprehensive insurance mechanism. Unemployment insurance represents such an institutional mechanism. Similarly, the integration of individually administered health insurance into a government-operated single payer system in 1999 was probably possible because insecure workers saw this option as providing better protection now that they could no longer expect guaranteed lifetime employment from a single employer.

With respect to this exchange of job security for social protection, two questions remain. The first question concerns why workers did not prevent the government from carrying out labor reform that weakened job security in the first place. For workers, however, it was not really a matter of choice. Workers, who had managed to veto liberal labor reform until the Korean economy was hit hard by the financial crisis in the late 1990s, had no choice but to swallow it when the Korean government decided to push through the liberalization of labor markets at the demand of the IMF. The second question is how workers weakened by job insecurity and labor reform were able to obtain a policy shift toward comprehensive social protection. Between employment security and a strengthening of social protection, the more preferred option for workers would have been employment security. The financial crisis and IMF made that option out of the question, forcing them to embrace the other option. Furthermore, the internationalization and tertiarization of the Korean economy intensified employers’ demand for a flexible labor market, putting workers on the defensive. It must also be noted that the expansion of social protection materialized when the government saw the benefits of it, rather than by workers’ organized movements. Although advocacy groups and trade unions demanded a strengthening of social safety nets whenever they could, it was not the political power of workers per se that resulted in the development of social policy.

Now, the last puzzle between the four concerned a sustained disparity in social insurance coverage between regular and contingent workers, large and small firms, and formal and informal sectors. These categories tend to overlap. Regular workers are likely to work at large firms in the formal sector, whereas contingent workers or workers under non-standard work arrangements are concentrated in the informal sector and small-sized firms. Indeed, the divide between labor market participants is substantial. Table 1 illustrates that significant differences in coverage between regular and non-regular workers and between firms of different sizes exist. Generally, regular workers are twice as likely to benefit from social insurance as non-regular workers, and the gap between small and large firms is even greater.

OPEN IN VIEWER

Table 1.

Social insurance coverage rates in Korea, 2008.

Firm Size	Social Insurance	Non-regular Workers (%)	Regular Workers (%)
Total	Old age	36.6	75.7
	Health	37.7	75.9
	Unemployment	34.5	63.8
Firms with fewer than five employees	Old age	7.8	25.6
	Health	8.3	25.0
	Unemployment	7.1	23.3
Firms with five or more employees	Old age	47.5	85.0
	Health	48.8	85.2
	Unemployment	44.8	71.2

Source: Quoted from Chang (2006: 191).

Why do the national social insurance systems exhibit such deep divides between labor market insiders and outsiders? The intertwined development of Korea’s employment and social protection system helps us understand the enduring segmentation of the social protection regime despite the expansion of social insurance. Korea’s concentrated industrial structure helped to solidify employment protection for regular workers and to maintain the division between internal and external labor markets. Social insurance was initially built for the industrial sector, which required a stable supply of skilled workers, and it was designed to complement, rather than replace, company-based welfare benefits.

The weakening of the stable employment system motivated regular, skilled workers to seek more comprehensive social policies. The expansion of social insurance, however, did not extend beyond the existing boundaries in the labor market; entitlements were dependent on contributions rather than granted on the universalistic ground. Because the disadvantaged were not able to make proper contributions to social insurance, they did not receive protection from the system. The underrepresentation of outsiders with vulnerable employment within trade unions made it hard for Korea’s fragmented labor movement to pay attention to outsiders’ employment insecurity.

In summary, the theoretical framework advanced above can offer persuasive explanations for the four puzzles of social policy development in Korea. Having confirmed that the patterns of Korea’s welfare state development are comfortably predicted by my theory, I now turn to the most recent phase of social policy change to provide further evidence for my argument. The following section analyzes how employment insecurity shaped insider preferences for or against change in broader social policies and demonstrates how threats to job security motivated them to respond to and to participate in the political debate surrounding the old age, health care and unemployment insurance systems.

Health care reform

The key issue in health care reform had long been the integration of decentralized health insurance societies. After a series of failures, such a reform finally came about under the Kim Dae Jung government. Why did the reform succeed in 1999 while earlier attempts had failed? Employment insecurity was an important factor in the success of this social policy reform. The preferences of pivotal workers over health care reform shifted in response to the greater probability of losing a stable job in the wake of changes in the employment protection regime in the late 1990s. Because employer-provided affordable health insurance was only accessible on the condition of stable employment, workers now faced a greater probability of losing their insurance as well. Workers thus developed renewed interest in national, integrated, and less employer-dependent health insurance. As a result, the balance between opposing voices over health care reform tipped in favor of reform.

As discussed earlier, the health insurance system continually grew throughout the 1980s, beginning with the employees of large firms and extending coverage to the rural areas and the self-employed in urban areas. The most important feature of the system, however, was an independently operated insurance society for each new group of people to be covered. This was an addendum to the initial system, which was designed to provide benefits only to workers at large companies. The expansion of coverage involved no risk pooling between insurance funds and required minimal government subsidization. Consequently, the system produced severe imbalances and inequities between “rich” and “poor” insurance societies.

Table 2 illustrates the disparities between public employees’, employed workers’, and regional insurance by comparing the ratios of benefits to contributions. Public employees’ insurance is managed by the government and covers all public employees. Employed workers’ insurance lumps all health insurance societies that were independently operated by employers in the private sector. Regional insurance refers to combined figures for the remaining health insurance societies. The most generous system between the three categories was public employees’ insurance, followed by employed workers’ insurance and regional insurance. Excluding public employees’ insurance managed by the government, the difference in generosity between the two remaining categories largely reflects the divide between labor market insiders and outsiders. Employed workers’ insurance, which covers larger firms, tended to have abundant funds due to economies of scale and relatively low-risk employees and, thus, provided relatively generous benefits to the insured. By comparison, the regional insurance representing small insurance societies was less generous. Members of these “poor” insurance societies paid higher premiums but, at the same time, did not receive equal benefits or services. This particular group of people started to voice their complaints and to demand the integration of insurance societies.

OPEN IN VIEWER

Table 2.

Comparison of benefits and contributions in Korean health insurance societies, 1998.

	Government Employees’ Insurance	Employees’ Insurance	Regional Insurance
Yearly Contribution per Person (in Korean Won)	127,002	137,412	138,345
Yearly Benefits per Person (in Korean Won)	186,180	153,372	144,514
Ratio of Benefits to Contributions	1.47	1.12	1.04

Source: Quoted from Wong (2004: 89).

Other drawbacks of fragmentation include economic inefficiency and overall inadequacy of benefits. Because a decentralized system incurs higher administrative costs economy-wide, the overall level of benefits was lower than it would have been in a centralized system. In addition, because the determination of premium rates and benefits was made by each insurance fund, the insured were at a disadvantage vis-à-vis their insurers and health care providers. If the societies had been integrated into a national system, their collective voice would probably have been better represented. Indeed, despite the progress in expanding insurance coverage to a greater portion of the population before the integration of insurance societies, both the average level of benefits and the scope of medical coverage remained very low. For example, 84% of the nation’s total health care bill was paid directly out-of-pocket, with only 16% paid by insurance in 1980, and these percentages did not change much throughout the 1980s (Wong, 2004). Therefore, the much-touted universalization of heath care was achieved only at a superficial level. Recognizing the economic efficiencies and structural weaknesses of the system, the government began considering integration as early as 1980, only to push it aside.

The most significant voice calling for health care reform first came from farmers not long after the expansion of health insurance into rural areas in 1988. Disgruntled farmers, allied with social movement groups, made three demands: integration of health insurance societies, increased government subsidies, and expanded range of benefits and health care services covered by insurance (Wong, 2004). The integration effort nearly succeeded, but it was ultimately thwarted when President Roh Tae-woo, pressured by opposition groups, vetoed the reform bill in 1989. Who opposed it? Obviously, large businesses voiced strong opposition because integration would deprive them of a sizable amount of funds. ⁶ Workers at these companies, moreover, were equally unenthusiastic about integration. Their ambivalence about the integration measure is evidenced by the position of the Korean Confederation of Trade Unions (KCTU).

The KCTU expressed its opposition in 1980 when integration was being discussed for the first time, and in the second attempt at integration under the Roh Tae-woo government, they tried to distance themselves as far as possible from the pro-reform camp. Then, in the third and final reform attempt, it articulated its complete opposition, claiming that an integration would be unfair for workers because their contributions would, in effect, be utilized to subsidize other high-risk groups (Wong, 2004). Although the government-sanctioned KCTU cannot be considered a legitimate voice representing workers at large during this pre-democracy period, its attitude paralleled the general skepticism on and disinterest of workers in moving toward universal social protection. As long as workers had stable, secure jobs, they were content with the existing system and felt that integration would not serve their interest.

The regulatory change in employment protection in the late 1990s, in conjunction with the rise of the democratic labor movement, changed worker preferences regarding integration and, subsequently, the balance of power between opposing voices. First, the democratic labor movement, represented by the Federation of Korean Trade Unions (FKTU), joined the pro-reform camp. ⁷ One of the most crucial issues from the workers’ perspective was the widespread underreporting of income by the self-employed due to concerns about potential unfairness for workers. The underreporting of income by self-employed professionals or those whose incomes were drawn from property ownership had always been an administrative challenge for Korea’s tax authorities, but with respect to the integration of insurance societies, it was particularly disconcerting for wage earners. Because premiums would be determined by income levels, high-income tax-evaders would likely be charged unfairly low rates at the expense of salaried workers. The opposition groups, including large businesses, waged public relations campaigns utilizing this argument to persuade workers and the public to oppose integration. This time around, however, the pro-reform groups gained widespread societal support. Now, given the potential risk of unfairness, why did the democratic FKTU come to favor the integration of insurance societies?

Unlike in the earlier period, many workers were now experiencing greater job insecurity. They faced a higher chance of falling victim to a layoff with a growing possibility of liberal labor reform being passed. Then, they would not benefit from the segmented system any more, and they began to understand that the decentralized system only produced economic inefficiency, decreasing benefits overall and suppressing the scope of medical coverage. In addition, it became increasingly clear that the system would no longer benefit workers at large firms. Insurance funds at large companies tended to record large surpluses that were not translated into expanded services or decreased co-payments. Given increased risk of job loss and unstable employment, workers at these large firms realized that integration could provide them with better protection if they fell ill and would reduce their medical costs. ⁸ In short, the success of the 1999 health reform was made possible by support from workers who, amidst heightened labor market insecurity, began to prefer more expansive social protection policies.

Pension reform

The first instance where we can detect a central role of employment insecurity in the reform process is the controversy over the relationship between severance payments and pensions. Mandatory severance payments have long been a key feature of Korea’s social protection system, even after the introduction of major social insurance systems. The 1953 Labor Standards Act established the statutory requirement that employers set aside at least one month’s salary per year of service for regular workers and pay this as a retirement benefit when employees leave a job, regardless of the cause of separation. This system is not common around the world, and in Estévez-Abe’s (2008) term, it is a “functional equivalent” of both unemployment insurance and an old-age pension system. A major difference is that the severance pay is employers’ liability, and it is paid out in a lump sum upon separation. These two features made the Korean system of severance pay an ineffective safety net for workers, especially during the financial crisis. The massive scale of bankruptcy in the late 1990s left dismissed workers empty-handed. Even before the financial crisis, one motivation for introducing pensions was to provide retirees with more reliable protection than that provided by severance pay.

However, the decision to introduce a pension system placed workers and employers at odds with each other. Employers, raising concerns of a double financial burden, argued that a newly created pension system should completely replace the severance pay system. Workers strongly resisted, claiming that severance payments, in the Korean context, are simply deferred wages rightfully due to employees. As discussed earlier, the developmental strategy of the authoritarian government encouraged large corporations to introduce a variety of employee welfare benefits to attract skilled workers, and in the absence of effective social insurance systems, the company-based welfare system provided social protection to a narrow segment of workers employed in the formal sector. The pension system was introduced because this firm-level provision of protection proved increasingly inadequate amidst increasing employment insecurity. In addition to their support for the pension system, workers sought a variety of insurance measures—including the preservation of the mandatory severance pay system—to cope with this heightened insecurity. To address the inadequacy of the severance pay system, in 1989, the government added a requirement that severance payments should take precedence over other financial obligations in the event of bankruptcy (KLI, 2000). However, the Constitutional Court ruled the requirement unconstitutional in August 1997, on the eve of the financial crisis. This decision led employers to reiterate their demands for complete elimination of the severance pay system, thus putting workers—who were increasingly concerned about their job security and income—on the defensive.

The pension reform of the Kim Dae Jung government was implemented against this backdrop. As health insurance finally became universal by including the self-employed in urban areas, the pension system was approaching the last stages of its expansion into a national system—extending its reach to urban, non-wage workers. However, preparatory studies revealed that with the addition of this vast group of people—estimated at 10 million—the generously designed pension system would be financially unsustainable (Yang, 2008). The fund, which had yet to reach maturity, was projected to go into deficit by 2022 and to be depleted by 2031 (OECD, 2000). Before the Kim Dae Jung government’s universalization reform, the preceding government had attempted to take drastic measures to improve the financial soundness of the National Pension System. The plan included both parametric and structural reforms: adjusting the levels of contributions and benefits ⁹ and contracting out the earnings-related components. Not surprisingly, this plan met with strong opposition from labor organizations and other advocacy groups. Workers were particularly upset about the curtailment of benefits included in the plan and blamed the government for a lack of accountability in the management of pension funds. In fact, the huge sum of accumulated funds was being appropriated by the government without any constraints because there were no effective mechanisms in place to hold it accountable to the insured. Before any reform that would lower the level of benefits, workers wanted to place the pension fund under democratic supervision.

Interestingly, workers did not object to extending pension benefits to the self-employed, unlike their opposition to the universalization of health insurance. The absence of workers’ resistance to the inclusion of the self-employed was the reason why the Kim Dae Jung government was able to implement universalization while rolling back a more drastic structural reform plan, against the World Bank’s forceful recommendation (Yang, 2008). The difference in worker attitudes reflects the distinctive nature of pensions. Because a statutory expansion of coverage in health insurance would lead to an actual expansion in the number of the insured—as basically everyone requires medical care—any redistributive effects between regular workers in the formal sector and those in the informal sector would appear immediately. In contrast, pensions are typically disbursed long after workers begin to contribute, and therefore, harmful effects supposedly caused by the inclusion of the self-employed would be realized in the distant future and not immediately felt by workers.

In addition, low levels of compliance from the self-employed, labor market outsiders and small firms made it uncertain whether universalization would indeed produce unfavorable redistributive effects for insiders. The pension system was designed to be “progressive,” that is, there would be redistributive effects from high-income to low-income groups. Unfortunately, the poor and labor market outsiders, who would benefit from the system, could not afford to pay social security contributions to the pension that they would collect benefits from 20 to 30 years later. When comparing the actual coverage rates for labor market insiders and outsiders, insiders are disproportionately covered, as Table 1 has shown. In other words, the national pension system was introduced to provide regular workers with a better shield against the changed economic environment.

The introduction of unemployment insurance

Finally, the case of unemployment insurance demonstrates the key role of employment insecurity in effecting policy change. As the political process of introducing and expanding unemployment insurance moved forward, labor organizations approached reforms cautiously because of concerns that employers might take advantage of the issue. Employers had long demanded that collective dismissals of workers be legalized. Labor organizations were wary of the possibility that employers might push for this change in exchange for agreeing to unemployment insurance. Accordingly, the democratic labor movement, now gathered under the National Congress of Trade Unions (NCTU), demanded that the introduction of unemployment insurance should be accompanied by statutory restrictions on collective dismissals (KLI, 2000). This concern was not unfounded. The Korea Employers Federation (KEF), an employers’ peak organization, indeed viewed the introduction of unemployment insurance positively, anticipating that it might mitigate worker resistance to collective dismissals (Chung, 2008: 182). However, the pro-government, cooperative KCTU was more proactive than the democratic labor movement, and as early as 1989, it submitted a petition for the implementation of unemployment insurance. It also tried to strategically use the issue in negotiations with businesses. Because of their mutual interest in unemployment insurance, the KEF and KCTU made a public announcement of their joint support for the government plan in 1993. This consensus between the KEF, the KCTU, and the government invited criticisms from the NCTU, the more democratic workers’ movement organization. Although the NCTU did not object to unemployment insurance per se, the government plan, in its view, fell short of an effective mechanism to protect workers: coverage was too narrowly targeted to large firms and benefit levels too low. The superficiality of the plan appeared to come from the government’s unwillingness to assume any substantial financial responsibility and was viewed as evidence that the real motivation was not to enhance protection for workers but to buy them off in exchange for a flexible labor market. Specifically, the NCTU demanded that coverage be extended to firms with five or more employees, rather than 30 or more, and that the government match employers’ and employees’ contributions to unemployment benefits as well as pay for any deficits that might be incurred from extending coverage to small firms (KLI, 2000: 254–255).

Despite this demand from the emerging democratic labor movement, the government implemented a much weaker version of insurance, with limited coverage and benefits. Later, coverage expanded when the financial crisis forced the government to take prompt action. The IMF also demanded the strengthening of the system. However, previously thwarted labor reforms—legalizing collective dismissals and dispatched worker agencies—were finally passed during this period. In fact, the IMF’s demand for a stronger safety net was premised on the implementation of labor market liberalization. In the context of impending liberalization of labor market protection, labor unions at the Tripartite Commission established by the government could not help but accept the offer of strengthening the system. In other words, workers had to pay a huge price—weakening of employment protection—for unemployment insurance. To be exact, unemployment insurance was a token compensation for the damaging compromise workers had to swallow. Despite the rapid expansion of unemployment insurance, disparities between types of workers and between firms and sectors persisted, following the distinct pattern observed in other social insurance systems.

In summary, the government and businesses promoted unemployment insurance to provide compensation for or as a bargaining chip in negotiations over the loosening of employment protection while workers were divided and ambivalent over the idea due to concerns about employment protection. Once labor reform became inevitable, exposing workers to heightened unemployment risk, labor unions finally embraced unemployment insurance.