East Asian approaches of activation: the politics of labor market policies in South Korea and Taiwan

ABSTRACT

This article compares the labor market policies (LMPs) of South Korea and Taiwan to explore the activation approaches in East Asia. These two countries took divergent paths to activation after the 1990s, and their decisive differences lay in the scale of direct job creation program as compared to vocational training and employment services. Diverse socio-economic contexts and democratic landscapes further shaped the activation repertoires available to the governments of both countries. We explain, however, this divergence chiefly through political competition employing not only programmatic but also nonprogrammatic measures, in which incumbent governments provide specific LMP measures in exchange for political support from their constituencies. Additionally, economic crises and political instability molded the diverse evolution of each country’s LMP. This study contributes to the understanding of the changing dynamics of East Asian social policies and highlights the circumstances that lead to policy changes.

KEYWORDS:

• South Korea
• Taiwan
• activating labor market policy
• programmatic competition
• nonprogrammatic competition

1. Introduction

South Korea (hereafter Korea) and Taiwan were both previously prototypes of the East Asian developmental state. Both countries embraced the principle of welfare through work, in which the employment regime aimed at high economic growth and functioned as the catalyst to successfully fight poverty and inequality. Meanwhile, labor market policies (LMPs) focused on generating a high-quality labor force, rather than protecting individual’s labor rights. Yet, the situation changed when the two countries experienced economic globalization and political democratization in the 1990s, leading to an increase in unemployment and precarious employment. In response, the two countries expanded their social policies and LMPs while dismantling the employment regime previously based on full employment.

The current literature classifies Korea and Taiwan as the same type of welfare regime (Goodman & White, 1998; Holliday, 2000; Kwon, 2005, 1998). The literature presents two opposing positions on the evolution of the welfare state in East Asia. One position contends that new welfare states are emerging in East Asia, while the other insists on the persistence of conventional developmental welfare. We attempt to address this incongruence by comparing LMPs in Korea and Taiwan. Since the LMPs are the main policy response to social turmoil and a key element of the employment regime in East Asia, they provide a suitable focus for the examination of the changing policy dynamics.

The present study explores how Korea and Taiwan responded to increasing unemployment resulting from globalization by paying particular attention to the activation in LMPs. LMPs have two common definitions. In a broad sense, it is the relationship between the active labor market policy (ALMP) and the passive labor market policy (PLMP). In a narrow sense, activation prioritizes employment services and occupational training over direct job creation (DJC) and unemployment benefits. In this study, we adopt the broad definition because DJC can also serve to promote access to work. We are interested in the factors which distinguish the patterns of activation in Korea and Taiwan.

The present study analyzes recent theoretical insights into the research on welfare states and the political economy. We highlight the role changes of actors and institutions resulting from democratization, with a particular emphasis on the roles of political competition in social policy development. We view both programmatic and nonprogrammatic competition as the mechanisms of political competition. Given the immature political systems in East Asia, domestic political forces tend to rely on nonprogrammatic measures. Additionally, we would like to note the impact of social unrest on social policy. While previous studies primarily examine economic crises, we also focus on the effects of political instability on policy-making. The data for analysis include government documents and statistics, newspaper articles, and expert interviews we conducted.

This article is structured as follows: Section 2 critically reviews the current research and presents a new discussion on East Asian case studies. Section 3 compares the divergence in LMP spending between Korea and Taiwan. Section 4 presents LMP development in both countries. The final section is devoted to an overall discussion and theoretical implications.

2. Literature review

Previous research on East Asian social policy delineates two theories: developmentalism and democratization. The developmentalism thesis (Kwon, 2005; Kwon & Holliday, 2007) outlines the limitations of welfare expansion after democratization and emphasizes that social policy remains undeveloped following the productivist welfare regime. In contrast, the democratization theory (Kim, 2008; Lee, 1999) stresses the profound shift in social policy development from low growth to compressed growth. Advocates of this theory claim that, after democratization, the welfare state – with emphasis on social citizenship – is emerging in East Asia. The paradigm shift stemmed from a social coalition of a center-left party, a democratic trade union, and civic movements (Lee, 1999; Peng & Wong, 2008).

Both theories have strengths and weaknesses. Developmentalism illustrates the limits of welfare reforms but fails to recognize the rapid expansion of East Asian social policies. Contrarily, the democratization theory explains the rapid welfare expansion but overlooks the limit of welfare reforms and the role of conservative governments. After all, conservative governments in Korea and Taiwan have remained in power longer than their progressive rivals, even after democratization. Furthermore, both approaches are insufficient in explaining the differences between these countries, and, instead, focus on the building of the East Asian typology of welfare state.

We seek to address these shortcomings by comparing the LMP development of both countries. We propose three theoretical innovations. The first aims to revise the generally accepted logic of democracy (Esping-Andersen, 1990). Instead of agreeing that democratization always results in the extension of social rights (Marshall, 1950), we contend that this process changes the institutional framework of competition between political forces – crucially impacting the development of the welfare state (Wong, 2004). Unlike in an authoritarian regime, which can retain power through state violence, political forces in a democracy have to vie to gain citizen support. Social policies that are conducive to welfare become a key issue in democratic competition. In this regard, we focus on the role of political forces, particularly governmental political forces and political parties, as major agents in the competition for political power in a democracy.

Second, we highlight the innovation of partisan theory (Hibbs, 1977; Huber & Stephens, 2014; Schmidt, 1996). This theory originally concentrates on the programmatic social policy competitions between political parties. Certain inventive studies (Kitschelt & Wilkinson, 2007, p. 21; Stokes et al., 2013) reveal that political competition assumes not only the programmatic policy competition mechanism but also the nonprogrammatic competition mechanism. In the former case, competing political parties present their policies concerning collective goods, such as economic growth, jobs, social policies, and health services (Kitschelt & Wilkinson, 2007, p. 2), and the electorate chooses those parties in line with their policy preferences. The politicians follow the criteria for public interest in the making of programmatic policies, thereby they are often accompanied by a public discussion (Stokes et al., 2013, p. 7–12). Moreover, the ruling party generally implements policies consistent with what it promised; the results of which affect voters’ decisions in future elections (Schmidt, 1996).

In contrast to programmatic policy competition, nonprogrammatic competition refers to a mutual mechanism in which political parties or politicians reward people who trade their votes for political favors (Stokes et al., 2013, p. 6). Here, office-seeking politicians distribute welfare benefits, such as non-social insurance programs, to supporters (Amenta & Tierney, 2014, p. 157). In their making, nonprogrammatic policies rarely follow the criteria for public interest and therefore seek to promote personal benefits of specific individuals or groups of individuals. Nonprogrammatic competition is often typical of economically underdeveloped countries or authoritarian political systems. This research strand further indicates that, even after economic development reaches a substantial level and democracy consolidates, nonprogrammatic competition persists to some degree in most countries (Lyne, 2007, p. 168; Kitschelt & Kselman, 2013).

Moreover, nonprogrammatic distributive policies include pork-barrel politics and clientelism (Stokes et al., 2013, pp. 12–13). In clientelism, patrons (politicians) provide clients (mostly individual supporters) with welfare provisions on condition that clients deliver votes or other forms of political support for the patron. Clients who do not reward their patrons in return for favors are often caught through strict monitoring and face the threat of penalty.

In contrast, pork-barrel politics provide differentiated welfare benefits to particular populations or local constituencies, who are likely to support patrons (politicians or political parties) in order to strengthen their own political constituency. This strategy does not require the strict monitoring of voter support so voters do not fear individual penalties (Stokes et al., 2013). The potential benefits can take the form of universal benefits for certain population groups (Brun & Diamond, 2014; Kitschelt, 2007; Piattoni, 2001). These universal benefits are still nonprogrammatic distributions since they do not aim to serve long-term public interest but, instead, the temporary interest of the ruling forces and their supporters (Skocpol, 1992). Office-seeking politicians not only use nonprogrammatic distribution to reinforce their constituencies’ loyalty (core voter hypothesis) but also to win the favor of those who support their opponents (buy-back hypothesis) (Takahashi, 2017, p. 17).

In the political science literature (Brun & Diamond, 2014; Keefer, 2007), the level of economic prosperity and/or the quality of democracy determines the prevalence of nonprogrammatic policies. Other literature highlights the role of political institutions, such as election systems and state bureaucracies, for maintaining nonprogrammatic competition in highly developed countries. For example, clientelistic competition is probably stronger in a country with unbridled political competition and a weaker bureaucracy – which means politicians have more discretionary power (Kitschelt & Wilkinson, 2007).

In contrast, since this study regards nonprogrammatic mechanisms as part of political competition, it draws on the approach of Croissant and Merkel (2004), who assume a close relationship between the institutionalization of the party system and the prevalence of nonprogrammatic mechanisms. As such, in a clientelistic party, where not the official party organization but the private network distributes economic resources and officials, the institutionalization of the party identity and organization is very difficult.

The above discussion bears important implications for our analysis: Korea and Taiwan are similar countries in terms of economic prosperity and democratization. They both have a strong state bureaucracy as a result of their pasts as developmental states. However, the consolidation of the party system is markedly different (Cheng, 2003; Croissant & Merkel, 2004, p. 10). Comparing the two countries allows for the verification of the effect of the party system consolidation on the occurrence of nonprogrammatic politics.

The final theoretical revision concerns the impacts of economic crises and political turmoil on social policy. Previous studies point to the impact of crises on social policy; for instance, during previous economic crises, it was not programmatic policy competition but rather nonprogrammatic competition that led to the expansion of fiscal and social policies (Armingeon, 2012; Higgs, 1987; Shughart, 2011). Crises created difficulties for large segments of the population and motivated political parties to introduce temporary pork-barrel pledges for political gains in the election, regardless of ideological colors.

Meanwhile, the impact of political instability on political competition receives less attention. After democratization, East Asian governments were more sensitive to political discontents. Civic movement organizations and opposition parties can take advantage of the insufficient institutionalization of partisan politics to create tough political challenges for the government. This is a feature of ‘counter-democracy’, in which social movements monitor elected politicians and participate in the political process through active protests (Rosanvallon, 2008). These circumstances can threaten the survival of the incumbent politicians, as demonstrated by the impeachment of the Korean President in 2016 (Lee, 2017). In this respect, this study argues that the differences in the extent and patterns of economic and political instability between Korea and Taiwan produce the differentiated composition of programmatic and nonprogrammatic political competition.

This study aims to track the development of, and clarify the differences in, LMPs after democratization in Korea and Taiwan respectively; and to identify the causes behind these differences. We consider both programmatic and nonprogrammatic competition together because the LMPs are a long-standing subject of nonprogrammatic competition. Nonprogrammatic competition is also relevant to the LMPs because of the significant discretionary power of bureaucrats (Hutchcroft, 2014).

3. Spending on the labor market policy in Korea and Taiwan

This section compares the trends in and the structure of the LMP in Korea and Taiwan.¹ The first section of Table 1 depicts spending on the LMP as a percentage of GDP. In both countries, the total expenditure has increased, while Korea has outpaced Taiwan in its allocation of resources to the LMP. LMP expenditure in Korea was 5 percent of the OECD average in 1997 and soared to around 60 percent in 2015. In contrast, the expenditure in Taiwan has stalled around 12 percent. In particular, the spending on ALMP in Korea was around the OECD average in 2015, but Taiwan’s spending remained at 20 percent. As a result, Korea spent almost five times more on the LMP than Taiwan did that year.

Table 1. LMP spending trends.

		1997	1999	2004	2007	2009	2010	2013	2015
LMP as % of GDP	Korea	0.08	0.73	0.27	0.35	0.79	0.59	0.64	0.65
	Taiwan	N/A	N/A	N/A	0.11	0.29	0.19	0.15	0.13
	OECD	1.57	1.39	1.24	1.03	1.52	1.53	1.35	1.12
ALMP as % of GDP	Korea	0.07	0.57	0.11	0.12	0.43	0.30	0.37	0.35
	Taiwan	N/A	N/A	N/A	0.07	0.12	0.11	0.10	0.09
	OECD	0.53	0.53	0.46	0.41	0.49	0.54	0.47	0.44
Activation ratio	Korea	4.53	3.49	0.65	0.52	1.21	1.04	1.35	1.22
	Taiwan	N/A	N/A	N/A	1.66	0.76	1.62	2.02	2.23
	OECD	0.51	0.63	0.59	0.65	0.48	0.54	0.54	0.66

LMP is the sum of ALMP and PLMP. Activation ratio is calculated by dividing ALMP by PLMP.

Sources: OECD ALFS (Annual Labor Force Statistics) in OECD Statistics; OECD SOCX (Social Expenditure) in OECD Statistics.

The other important dimension of LMP concerns its structures in terms of the activation ratio. Both Korea and Taiwan exhibit higher levels than the OECD average, albeit with some qualitative differences. Although Korea’s activation ratio has been increasing rapidly since 2009, Taiwan has also been faring well. The high activation ratio reveals that ALMP was much more extensive than PLMP in both countries. Hence, the rapid increase in spending on LMP is chiefly due to the expansion of the ALMP.

Table 1 tracks the major changes in LMP. In Korea, expenditures increased radically between 1997 and 1999 and between 2007 and 2009. In Taiwan, there was a surge in spending between 2007 and 2009. We argue that the global economic crises of these periods triggered the rapid expansion of LMPs in both countries.

Furthermore, trends in spending reveal that the patterns of LMP spending varied during each of the two global economic crises in these periods. The sharp increase in spending as a result of the 1997 financial crisis in Korea was temporary and decreased immediately after the crisis. In contrast, the LMP expenditures of both countries increased during the 2007 financial crisis and remained almost unchanged after the economic crisis. It is required to explore the factors which lead to these different patterns.

Next, we compare the expenditures of the sub-sectors of ALMP, which has a bigger share of LMP spending in both countries. Table 2 divides the ALMP into three areas, including training, employment assistance, and DJC. It then compares the average expenditures between 2010 and 2015 for each sub-sector of the two countries. Both countries spent similar amounts on training. Korea spent twice as much on employment assistance as Taiwan. However, the key difference between the two countries is DJC. Korea’s DJC expenditure was a whopping twenty times as much as Taiwan’s between 2010 and 2015.

Table 2. Expenditures on sub-areas of the ALMP as a percentage of GDP (averages based on the years from 2010 to 2015).

	Taiwan	Korea
Employment assistance	0.03	0.07
Training	0.03	0.04
DJC	0.01	0.2

Sources: OECD Statistics; Ministry of Labor Statistics, Taiwan, various years.

In conclusion, spending on LMPs in both countries grew significantly after democratization. We also ascertain that the ALMP played a much more significant role in the expansion of LMP spending than PLMP in both countries. However, the LMPs of these two countries deviate in terms of amount and structure: the lion’s share of which is determined by the divergent amount of DJC. This study attempts to address why the LMPs of both countries institutionalized in a highly activated form and why Korea has implemented its DJCP (direct job creation program) on an unusually large scale.

4. Politics of the labor market policy in South Korea

Despite large deficits in social protection, the authoritarian developmental state before the 1990s gained considerable public support thanks to the characteristics of its employment regime. First, high economic growth bolstered full employment centered on male workers. Second, Korean enterprises guaranteed lifelong employment for employees based on their skills and loyalty. Third, strict government regulations on employment also nurtured the emerging corporate culture of long-term service.

The employment regime operated as the functional equivalent to social policy: welfare improved through work. This approach helped to rapidly reduce the absolute poverty rate in Korea from 40.9 percent in 1965 to 5.3 percent in 1991 (Kim, Kwon, Lee, & Yi, 2011, p. 122). Under this employment regime, high economic growth led to growth in employment. The LMP was not institutionally differentiated, lacked income security for the unemployed, and consisted of occasional measures for supplying qualified labor. Occupational education programs were extended to train skilled workers. The DJC, targeted at the most vulnerable people, constituted the important program of the LMP.

4.1 LMP after democratization: globalization and the introduction of employment insurance

After democratization, the government expanded social policy by introducing the national pension scheme and the minimum wage system in 1988. In 1995, the government further expanded welfare with the introduction of employment insurance. Labor bureaucrats had tried in vain several times since the 1960s to introduce unemployment insurance. The government decided to introduce employment insurance only when welfare expansion became an important political issue after democratization. In the presidential election of 1992, all major candidates pledged to introduce employment insurance. Accordingly, the Ministry of Labor enacted the employment insurance law in 1993 (Ministry of Labor, 2016, p. 67).

The institutionalization of employment insurance focused very little on unemployment because full employment, based on high economic growth, continued. Instead, the conservative government utilized this insurance to promote the globalization of the Korean economy. The Kim Young-sam government (1993–1998) embraced globalization as the guiding principle for economic development and strengthening export competitiveness (Shin, 2001). In the 1990s, challenges from newly industrialized countries such as China hampered Korea’s low-wage export strategy, so the government aimed to reorganize the industrial structure and focused on high-value-added businesses. The government expected that employment insurance would promote industrial restructuring by establishing a human resource development system and relieving companies’ corporate layoff burdens.

Reflecting upon this motive, the active programs had priority over passive income protection in employment insurance. The welfare ideology of the conservative government – also known as productive welfare – which emphasized preventive welfare, reinforced activation of the LMP by encouraging productivity instead of ex-post and consumptive welfare (Committee for Promoting Segyehwa, 1995). On one hand, various rigid restrictions became effective to curb entitlement to unemployment benefits. Unemployment benefits were renamed job-seeking benefits and their coverage was confined to companies with more than thirty employees: payment was refused for voluntary layoffs, benefits were curtailed in cases of individuals refusing job placements and training, and, because of the low upper-ceiling, the actual income replacement rate in 1997 stagnated at 42 percent of previous wages (Seong, 2016, p. 28). On the other hand, employment insurance was equipped with various measures to compensate re-employment efforts: creating an employment promotion allowance, extending the job-seeking benefit period to two years when participating in vocational training, and establishing the vocational training program for all companies with fewer than 1,000 employees.

In summary, the government institutionalized employment insurance in Korea to facilitate industrial restructuring by supporting employment adjustment for companies. As a result, the government implemented a highly activated form of employment insurance, which focused on strict income protection and strong conditioning.

4.2 The first activation: economic crisis and LMP of the center-left government

The 1990s Asian financial crisis triggered high unemployment rates and mass poverty: a social and economic crisis that Korea had never experienced before. The government response began to break down the conventional employment regime: the practice of lifelong employment declined and the number of irregular jobs increased rapidly. Succumbing to the International Monetary Fund’s pressure, in 1998, the center-left government under Kim Dae-Jung introduced the layoff system and the Worker Dispatch Law to provide a legal basis for temporary work and job cuts. These measures undermined the foundation of full employment and welfare through work.

To cope with the rising unemployment, the government loosened eligibility requirements for, and extended the benefit duration of, unemployment benefits. The government also supplanted the former social assistance system with a new social assistance system granting all citizens, including non-disabled individuals with low incomes, the legal right to a minimum cost of living by raising benefit levels.

However, major innovations in the LMP came from the ALMP. The government established the DJCP as a major part of the ALMP, which aimed to provide temporary jobs for vulnerable groups, such as the elderly and the disabled people (Kwon, 2001). At the same time, the Kim government improved the institutions and organizations of the employment insurance – extending its legal scope of application and establishing vocational training programs as well as employment service system (Kwan, 2000, p. 8).

Increase in social provision owed much to the welfare ideology of the Kim government with a semi-social democratic orientation. It aimed to promote not only the balanced development of a market economy and democratization, but also expansive welfare including establishment of a comprehensive social security system (Presidential Committee for Quality of Life, 2002, p. 35). In this respect, the Kim government’s welfare model distinguished itself from the residual welfare model of the previous conservative government. However, the Kim government’s LMP resembled that of the predecessor in office, for it retained the principle of employment-friendly welfare. The government conditioned social protections while expanding employment maintenance and job creation. Labor unions and civic movements supported the welfare expansion of the Kim government. And in response, the government encouraged trade union and civil society participation in welfare legislation and execution through several informal and official networks, like the tripartite commission on labor, business, and the state (Kim & Choi, 2014).

Despite its extension, the LMP appeared to be an incomplete activation: the Kim government tried to improve institutions of the LMP. However, the major part of ALMP measures, including DJCP, was provisional and cut after the economic crisis. Improvements of PLMP were too short to secure the livelihood of the unemployed: strict requirements were maintained and the benefit level remained low, and the new social assistance followed the principle of the conditional benefits.

4.3. The second activation: economic and political crisis and the LMP of the conservative governments

During the 2008 global financial crisis, Korea’s economic growth rate fell to −4.5 percent in the fourth quarter, and the exchange rate soared rapidly. The conservative government of Lee Myung-bak (2008–2013) recognized that these conditions heralded a severe economic crisis and conceded that the originally planned economic policy, which promoted a small state and deregulations (Park, 2010, p. 14), was no longer appropriate. Instead, the government preemptively implemented expansive stimulation packages in 2009 and 2010. This was the third largest expenditure among the OECD countries (OECD, 2009, pp. 19–20). The ALMP accounted for 30.5 percent of the total stimulation package’s expenditure, of which almost 70 percent flowed to temporary job creation and 20 percent to vocational training. In addition, the government expanded the self-reliance program by emphasizing the principle of in-work benefit in social assistance (Kim & Nam, 2011).

After 2008, the conservative government contended with another crisis triggered by political challenges, such as protests over mad cow disease, election defeats – which included the 4–27 by-election and the 10–26 Seoul mayoral by-election – and the lowest job approval rating to date of a sitting president swamped with corruption scandals. Political instability intensified when progressive civic movements and labor unions allied with the Democratic Party: the major opposition party. In 2008, over one million people attended a candlelight demonstration over more than three months to protest against U.S. beef imports (Seoul Central District Prosecutors’ Office, 2009).

Throughout the term of Park Geun-Hye’s conservative government (2013–2016) – the successor of the Lee government – influence-peddling and corruption scandals triggered protests that reached unprecedented levels in terms of size and intensity. Between October 2016 and April 2017, a combined sixteen million people participated in twenty-three demonstrations that eventually led to the impeachment of the president (Sisajeoneol, 2017.03.17). The coalition of labor unions, social movements, and the opposition party constantly provoked protests that destabilized the Korean conservative governments in this period.

The conservative governments responded with an expansive LMP. They established an employment-welfare center to link job placements with those receiving job-seeking benefits and introduced new vocational training programs through the creation of a skills development account and vocational training vouchers. They also expanded subsidies for companies hiring socially disadvantaged people. In addition, in 2009, the Lee government introduced an earned income tax credit system, and the Park government expanded it substantially. Yet, the DJC remained the dominant program. The Lee government rapidly enhanced DJC as provisional measures during the financial crisis of 2008 and 2009 and extended it again in 2012. The Park government then improved the DJC despite the abatement of economic crisis by introducing new programs and enhancing payments.

The LMPs of these governments reflected their welfare ideologies: the Lee government’s ‘active welfare’ and the Park government’s ‘expansion of employment and welfare’. Both pursued an employment-friendly social policy (The Eighteenth Presidential Transition Committee, 2013, p. 35). Nevertheless, these programmatic welfare concepts appeared insufficient to account for the dramatic expansion and pattern of LMP under these governments because the first conservative government, the Lee government, originally held a passive and market-friendly position in regards to employment and welfare.

This study maintains that the conservative governments’ LMP shift cannot be understood without taking account of nonprogrammatic competition. We argue that the nonprogrammatic competition led to a time-limited stimulation package in response to the 2008 economic crisis. The expansive LMP of conservative governments since 2010 was a pork-barrel policy to combat political instability.

According to a welfare expert of the Park campaign (interview on 01.16.2019), by confronting the stubborn protest coalition, the conservative government sought to extend and strengthen its constituencies in order to maintain political power and win elections. The government utilized the LMP, particularly the DJC, to mobilize the most loyal members of the conservative party’s constituency – the elderly.

The government targeted the elderly based on their historical voting behavior. Cheong (2018, pp. 36–52) contends that ideology, region, and generation mattered in previous Korean presidential elections. The most significant contributor to election outcome was generational differences. Younger generations were swing voters between the center-left and conservative parties. In contrast, elderly voters constantly cast their ballots for conservative parties. In the presidential election of 2007, 71.7 percent of voters between 60 and 64 years old and 64.6 percent between 65 and 74 years old voted for the conservative candidate. This trend persisted in the 2012 presidential election: 76.1 percent of voters between 60 and 64 and 80 percent of voters between 65 and 69 voted for the conservative candidate (Seong, 2015).

The expansion of DJCP also complied with the request of elderly interest groups (interview with a manager of KORDI on 09. 26.2019). Among them, the Korean Senior Citizens Association (KSCA) was most influential because of the large number of members (an estimated Three million) who had a long term patron-client relationship with the conservative party. This association reaffirmed their influence by effectively attacking the center-left party during the 2004 general election and the 2012 presidential election regarding the scandal of ‘demeaning comments about the elderly’ (Hankyoreh21, 2004). Based on its political clout, this association tenaciously requested the expansion of DJCP at the election camp of the conservative government during the 2012 presidential election (interview with a welfare expert of the Park, Geun-Hye campaign on 29.09.2019). After the presidential election in 2012, the campaign realized that this DJC was a decisive commitment, leading to victory in the election (interview with a welfare expert of the Park, Geun-Hye campaign on 01.16.2019).

The political function of the DJC went beyond the election. The KSCA was also involved in implementing the DJCP. The government controlled such pro-government forces in order to mobilize political support (interview with a welfare expert of the Park, Geun-Hye campaign on 01.16.2019). In 2017, 15.9 percent of 1,224 organizations belonged to the KSCA (KORDI, 2017, p. 95).

The Korean structure of policymaking induced the large-scale implementation of nonprogrammatic policies. In Korea, the political parties – key players in programmatic competition – did not play a significant role in the policy-making process because they were under-institutionalized compared to other developed countries, even Taiwan (Lee, 2011, pp. 83–86; Cheng, 2003). Political parties had a very short lifespan due to the continued strategic split and recombination. There were few party members and their social linkages stayed loose. In these circumstances, policy-making was not the result of official political party organizations’ activities, but mainly the work of election campaigns or presidents and their close associates. The policy contents served to achieve a short-term election victory rather than pursue the party’s ideology and platforms (Kim, 2010). The characteristics of this policy decision-making process provided fertile institutional soil for the growth of nonprogrammatic policies.

As a result, the proportion of the DJC in the LMP – mostly accounted for by senior citizen job programs – increased excessively. In 2016, the expenditure on DJC accounted for 57 percent of the total ALMP expenditure (OECD stat). The proportion of DJC participants over the age of 65 was 66.5 percent in 2013 and increased to 78.7 percent in 2015² (Gang, 2015, p. 28–29). In addition, the lack of positive effect on employment shows that the DJC’s focus was to reward the elderly’s loyalty, rather than promote employment. In 2015, only 1.2% of participants aged 65 years of age or older succeeded in getting jobs in the private sector (Gang, 2015).

In conclusion, the conservative governments utilized the LMP as a pork-barrel measure to survive during political instability and to compensate political supporters for their electoral victories (core voter hypothesis).

5. Politics of the labor market policy in Taiwan

In Taiwan, the authoritarian Kuomintang (KMT) regime established different social insurance programs targeted at various occupational groups. Whereas generous benefits went to certain privileged occupational groups (military personnel, civil servants, and public-school teachers) in exchange for their political support, the Labor Insurance program introduced in 1950 was much less generous to other workers (Ku, 1997; Lin, 2006). This social insurance program extended benefits to the job injury, old age, medical care, disability, death, and maternity. Yet, only those workers loyal to the KMT regime enjoyed access to these broader social insurance benefits (Fu, 1993; Lee, 1992).

The social insurance’s occupational divide reflected the dual nature of the employment regime: while core clientele of the authoritarian regime worked in the public sector, the majority of the workforce worked in the private sector. This was largely due to the predominance of the ethnic divide between ‘Mainlanders’ and ‘local Taiwanese,’ which marked a pronounced political cleavage in Taiwanese identity politics and often superseded class politics (Wang, 1989).³ This political cleavage further characterized the LMP partisan politics of the democratization era. In developmentalism’s heyday, public employment creation seemed redundant and pricey. It further offered the regime little political gain because most ‘Mainlanders’ and supporters of the regime worked in the public sector with employment security. Instead, the authoritarian regime privileged vocational training as a key element of their LMP to support ‘welfare through work;’ this strict condition mainly affected ‘local Taiwanese’ who were the major beneficiaries. Already, by the 1960s, ‘human resource development’ emerged as one of the guiding programmatic principles of the employment regime. The government established public vocational training centers nation-wide, which became the major national agency for skill development (Lin, 2006). During the 1970s and 1980s, public employment services further gained significance by offering vocational training for unemployed workers. Active state involvement in the provision of these training programs reflected the fact that most industrial enterprises could not afford costly investments in internal staff training programs.

5.1. The LMP after democratization: labor market flexibility and the introduction of employment insurance

In 1984, amid rising social movements and labor protests, the government promulgated the Labor Standard Act: the centerpiece of labor legislation in Taiwan. Initially envisioned as the statutory protection of labor rights, the Act was one of the authoritarian regime’s strategies to placate disgruntled labor groups (Hung, 2006; Lee, 1992). Based on the premise of full employment and regular employment relationships, the Act aimed to strengthen employment protection by calling upon employers’ responsibilities for workers’ welfare. Together with the long-existing Labor Insurance legislation, this Act wove a basic safety net for workers.

The 1990s witnessed economic globalization’s structural transformation of labor markets. The first challenge to Taiwan’s economy occurred when China launched ‘reform and opening-up.’ In the aftermath of the Tiananmen Square incident in 1989, major Western countries imposed sanctions on China, and Taiwanese labor-intensive manufacturing enterprises took advantage of this opportunity by relocating their factories to China. Meanwhile, the remaining industrial employers requested the introduction of foreign guest workers to meet labor demand. In 1989, the government allowed companies to employ foreign guest workers for public construction projects. In 1992, foreign domestic caregivers were allowed to assist families in need (cf. Castles, 1998; Tseng, 2007). An increasing supply of foreign manual workers inevitably exacerbated the labor market situation and social inequality (Lin, 2009; Tsai, 2004). In response to these events, social activist groups called for public intervention in labor protection and social security (Wong, 2004). The conservative KMT government remained in power throughout this period and retained certain bureaucratic authority in dictating the expansion of social provisions. The government began the institutional expansion with pension and healthcare policies.

However, the conservative government refrained from providing generous support to the unemployed. In 1999, the government responded to rising unemployment with the incorporation of unemployment benefits within existing labor insurance programs to resist inducing welfare dependency. This mediocre policy failed to do justice to the scale of unemployment but dovetailed with the developmental doctrine that emphasized the mobilization of the inactive sector of the workforce rather than compensatory unemployment benefits. Moreover, as the economic downturn barely affected the KMT’s core clientele concentrated in the public sector, the conservative government had little incentive to seriously consider the unemployment problems of other occupational groups. Resorting to the conventional policy tools of employment services and vocational training characterized the initial years of democratization.

5.2. The first activation after democratization: center-left DPP government reforms

When the candidate of Democratic Progressive Party (DPP) Chen Shui-bian won the presidential election in 2000, public expectation for progressive reforms surged. Soon after, reform-minded welfare scholars and experts assumed key government positions, opening a new avenue for pro-welfare ideas to enter the policymaking process (Hsiao & Lin, 2000; Lin, 2006). Yet, in 2000, Chen publicly proclaimed, ‘economic development first, social welfare later,’ signaling his concession to pressures from employers. This dealt a blow to the progressive values that the DPP embraced in its alliance with the civic groups. From the very beginning of the DPP government, the LMP was an eclectic attempt to please working-class voters without fundamentally eroding the economic competitiveness of industries. The result was a policy focused on employment activation.

During the DPP’s reign (2000–2008), the government answered the demands of its clientele – mainly private labor force – by advancing the programmatic principles of ‘human resource development and technology innovation.’ The LMP reforms featured a turn towards employment-centered policies that differed significantly from the developmental era’s ‘welfare through work’ with very few LMP elements (interview with an official in Ministry of Labor on 10.02.2019). Major legislative initiatives of the time revolved around employment activation and unemployment protection, such as regulating mass dismissals, activating unemployed workers into re-employment, and strengthening trade unions. In view of rising unemployment and in the wake of the economic slump, the government strengthened unemployment protection. The resulting three closely-related labor acts advanced workers’ rights to associate and bargain collectively in labor disputes with employers – all essential measures to mitigate the impact of the economic downturn on unemployment rates.

The government introduced unemployment benefits for the laid-off (60% of an individual’s monthly wage for six months). Initially conceived in 1999 as an auxiliary regulation attached to the Labor Insurance program, in 2001, the government further extended unemployment benefits to more workers unemployed due to various causes, and, finally, formalized these expansions in the separate Employment Insurance Act of 2003 (Lee, 2010; Lin, 2006). This Act adopted an unequivocal activation doctrine and made access to unemployment benefits conditional on the beneficiaries’ readiness to participate in programs arranged by public employment service agencies (Art. 11–12). Under the rubric of ‘three in one employment services’ (sanheyi jiuye fuwu), the overall policy framework combined passive unemployment benefits with the provision of occupational training opportunities and support finding jobs.

After the start of the integrative employment services, they offered further vocational training and employment services to more unemployed workers (including part-time and voluntarily unemployed). The statutory regulation and monitoring of job search behaviors legalized the pronounced activation ideology that it was the responsibility of job seekers to return to employment as soon as possible. Each unemployed individual received personalized case management services including employment counseling and vocational training (Lee, 2010; Lin, 2006). The government devoted financial and staff resources to the Bureau of Employment and Vocational Training to offer training courses for various skills, such as home care, construction, and so forth. The Employment Service Act (promulgated in 1992 and amended in 2009) introduced public employment service centers to provide clients with employment consultations and relevant information about job opportunities.

In 1997 and 2008, the government introduced public employment and wage subsidies supplementary to employment training and services, but these subsidies were provisional and only for extraordinary circumstances (Chen, 2010). The government also offered temporary employment in the public sector – funding subsidies for self-employed individuals to start their own businesses and providing wage subsidies to enterprises that would otherwise sack redundant workers. Conditional subsidies also went to private firms that had plans to hire young college graduates.⁴ Once the crisis ebbed, temporary subsidies were also rolled back.

5.3. The second activation by the conservative KMT government

Taiwan encountered a severe challenge from the 2008 financial crisis when the conservative KMT candidate Ma Ying-jeou became president. The new conservative government adopted various measures to tackle the rising unemployment rate (Lee, 2010). Both central and local governments unveiled temporary programs to stimulate public-sector employment and provided wage subsidies to incentivize employers to fill job vacancies. In addition, the government created a three-year employment promotion program in collaboration with industrial employers to enhance employment opportunities, especially for younger workers.⁵ However, given the tremendous financial burden incurred from these kinds of public employment programs, they served merely as a temporary buffer against the abrupt unemployment shock.

At the same time, the conservative government embraced neoliberalism, which appealed to business and upper-middle income groups – the KMT supporters. During his presidential terms from 2008–2016, Ma endeavored to build closer ties to the Chinese economy. This strategy yielded political gains for those Taiwanese individuals working and residing in China, who typically voted for the KMT (Ku & Chang, 2017). Out of the neoliberal ideology, the government further deregulated the labor market, which allowed for an increase in irregular employment. Official discourse advocated for the neoliberal tenet activation, reflected in the president’s programmatic rubric for ‘employment activation’ to empower various workforces. This core idea favored the macroeconomic integration with the Chinese economy, along with the increased labor market flexibilization to generate economic momentum. The president called for ‘employability’ which placed emphasis on ‘human resource development’ and highlighted conventional developmental welfare ideology (interview with a labor economist on 09.25.2019). The slogan, ‘government supports banks, banks support enterprises, and enterprises support workers,’ emphasized the official belief in a pro-business and employment-centered activation policy approach.

Despite the government’s efforts to support industrial development by linking its economy with the Chinese prosperity, Taiwan’s economic growth remained sluggish over the last decade, creating unfavorable circumstances for job growth and wage increases. Irregular employment rates continued to rise (Chen, 2011; Lee & Ku, 2007), which unevenly affected young people with low education levels and middle-aged job seekers due to their lack of marketable skills. To address these challenges, the conservative government reinforced its policy orientation on vocational training and enterprise subsidies in terms of the employability doctrine. The policy priority drew on job training programs and employment services: each local job service center was obligated to achieve the goals set by their supervisory Labor Bureau. In addition to employment services, the government granted short-term financial subsidies to those employers who were willing to offer jobs to unemployed individuals. Subsidies also flew to enterprises with difficulties, mandating that they would refrain from mass layoffs. Ma’s administration differed from his predecessor’s in one significant regard: Ma’s government-supported (neo-liberal) managerialism with an emphasis on quantitative indicators to measure policy outcomes (Lee, 2009). However, the lack of coordination between government bodies of labor affairs and social affairs undermined the potential effects on available employment services.

In summary, the LMP in Taiwan demonstrates the government’s clear preference for activation in the areas of employment services and vocational training. Subsidies to private enterprises and public work creation functioned as short-term countermeasures in times of grave economic crisis. The introduction of employment insurance provided unemployed workers with better safety nets, although insurance was conditional on workers’ participation in training programs or cooperation with public placement agencies. Activation gained momentum after the neoliberal-orientated conservative KMT returned to power in 2008, and they made full workforce employment and labor force participation a priority. The deregulation of labor markets went hand-in-hand with the centerpiece of the LMP in Taiwan: employment services and vocational training.

Conclusion

Prior to democratization, the developmental states of Korea and Taiwan stood for the principle of ‘welfare through work’ underpinned by full employment. In the 1990s, globalization and stagnant economic growth led to the decline in full employment and the increase in non-regular jobs and unemployment. Korea and Taiwan institutionalized and expanded the LMPs to cope with increasing employment insecurity by focusing on active measures, such as employment services, job training, and DJC. They also made income protection for the unemployed strictly conditional on the recipients’ participation in active programs. However, the LMP expansion in both countries exhibit diverse dynamics: there is a vastly different scale in spending due primarily to the massive implementation of DJC in Korea.

Our analysis demonstrates the strong influence of political competition – including programmatic and nonprogrammatic mechanisms – on LMP development in both countries. Although various governments promoted LMP for realizing welfare ideologies of ‘employment-friendly welfare,’ they actually used LMP as pork-barrel measures to (re)gain electorate support diminished by economic and political unrest. Nevertheless, the Korean LMP was far more nonprogrammatic in nature than Taiwan’s counterpart, largely due to the Korean conservative government’s intensive pork-barrel politics. This explains why the government developed DJCPs with a particular focus on the elderly – a cohort that could hardly affect employment rates – on such a large scale since 2007.

Comparison of the LMPs between both countries bears important implications for understanding East Asian social policy. It supports the democratization thesis that democratization is an important factor of the extensive LMPs. At the same time, this study demonstrates that the constellation of programmatic and nonprogrammatic competition allowed by democratization determines the difference of LMPs between the two countries rather than the democratization shift itself. In contrast, the impact of the policy legacy of ‘welfare through work’ on the development of LMP is ambiguous. The approach of policy legacy helps us understand why LMPs in both countries follow the path of activation. However, it hardly explains why two countries with a similar institutional heritage trace distinct patterns in LMPs after democratization. In our view, three factors determine the pervasiveness of nonprogrammatic competition in the LMPs: social cleavages, institutionalization of the political parties, and economic and political instability. First, the primary reason why pork-barrel politics worked well in Korea’s ALMP was that the majority of the elderly – beneficiaries of these policy programs (e.g. the DJCP) – were loyal supporters of the conservative party. Cheong (2018, p. 36–52) demonstrates that generation heavily influenced previous Korean presidential elections. Elderly voters constantly cast their ballots for conservative parties, which provided them with the rationale for cementing their political support of DJCPs for the elderly.

This situation is different from Taiwan, where political cleavages took shape in tandem with the ethnic divide between the ‘Mainlanders’ and ‘local Taiwanese.’ This cleavage explains why activation in terms of employment services and vocational training gained ground because these policies proved cost-efficient and did not jeopardize the electoral base of each party (DPP and KMT). This kind of activation was also consistent with the prevalent programmatic ideology of human resource development. Consequently, although unemployment woes resulting from recent financial turmoil gave rise to public employment policies, support for them weakened once the crises receded.

Second, the degree of party institutionalization determines the extent of the mix of programmatic and nonprogrammatic measures. Taiwan is an exception in East Asia because it established programmatic parties after democratization according to which they created and implemented policy programs that reflected the ideological preferences of their constituencies. The stable party system permitted the relatively steady production and implementation of programmatic policies. In contrast, Korea lacked the ingredients for programmatic competition since the political cleavage was regional rather than ideological. As a result, political leaders’ personal networks, rather than official party organizations, implemented a wide range of pork-barrel policies to gain political support.

Third, multiple economic crises and political instability also contribute to diverse nonprogrammatic policies in Korea and Taiwan. The former experienced economic crises in 1997 and 2008 which fundamentally transformed its economic, employment, and social systems. In contrast, Taiwan weathered the economic crisis in 1998 but faced domestic unemployment arising from China’s economic rise in the 2000s and the global financial crisis of 2008. Korea’s conservative governments suffered from constant political instability due to a persistent protest coalition. In contrast, in Taiwan opposition to the conservative government was relatively moderate. As a result, Korea’s governments had an urgent need to secure political support through pork-barrel spending.

Finally, while highlighting the importance of democratization, this article contributes to the explanatory power of the partisan theory by considering programmatic and nonprogrammatic political competition. Our analysis also broadens the current understanding of the role of crises in social policy development by considering the effects of political instability as well as economic crises. When examining the development of East Asian welfare states, the democratization account can elucidate these social policy dynamics if combined with sophisticated partisanship theories that take into account political and economic constraints.

List of Interviewees

1. A manager of Korean Labor Force Development Institute for the aged (KORDI), on 09. 26. 2019.

2. A welfare expert of the Park, Geun-Hye Campaign, on 01. 16. 2019 and on 09. 26. 2019.

3. A labor economist from the National Taiwan University, on 09.25.2019.

4. An official responsible for vocational training in Ministry of Labor, Taiwan on 10.02.2019.

Acknowledgements

Previous draft of this paper was presented at the 15th Annual Conference of the East Asian Social Policy Network, University of Bristol, UK, 5-6 July 2018. The authors would like to thank Young-Jun Choi, Evelyn Huber, and the two anonymous referees of Policy and Society for their very constructive comments. Thanks also go to Yoonchul Nam and Suetgiin Soon for their assistance in collecting and compiling the data for the labor market policy of Korea and Taiwan. The usual disclaimer applies.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the National Research Foundation of Korea [2017S1A2A2041808].

Notes on contributors

Won-Sub Kim

Won-Sub Kim is a Professor of Sociology at Korea University in South Korea. He previously taught at Bielefeld University in Germany and at Kyung-Sang National University in South Korea. His research work centers on the theory of the welfare state, old age income security systems, and the East Asian social policy. He has published books and articles in German, English, and Korean languages, including “Institutionalisierung eines neuen Wohlfahrtsstaates in Ostasien? Eine Fallstudie über Südkorea” (2006), and several journal articles in Government and Opposition, Zeitschrift für Soziologie, International Journal of Social Quality.

Shih-Jiunn Shi

Shih-Jiunn Shi is a Professor of Social Policy in the Graduate Institute of National Development, National Taiwan University. He serves currently as Chair of the international organization “East Asian Social Policy Network”. His fields of research include comparative social policy with particular focus on East Asian social policy. He has published papers in major policy journals including the Journal of Social Policy, Social Policy & Administration, Policy & Politics, International Journal of Social Welfare, Public Management Review, Ageing & Society; Journal of Asian Public Policy; Environment & Planning C: Politics and Space.

Notes

¹ Due to Taiwan’s diplomatic isolation, the official definition of labor market statistics occasionally deviates from those of the OECD, in which Korea is a member state. Data collected from the Taiwanese government can thus underestimate the real budget inputs.

² In contrast, the percentage of participants aged nineteen years or under was very small (about 9 percent) in 2016 (Joint Ministries, 2016).

³ ‘Mainlanders’ (waishengren) refers to those immigrants who moved with the émigré KMT regime from China to Taiwan, while ‘local Taiwanese’ (benshengren) refers to residents whose ancestors did so long before 1949.

⁴ Information on the Bureau of Employment and Vocational Training: retrieved from http://www. evta.gov.tw.

⁵ Information on The Labor Council of the Executive Yuan: retrieved from https://www.mol.gov.tw/announcement/27179/13474/.