Sage Journals: Discover world-class research

Abstract

The labour market dualisation thesis posits a clear division between low-skill, nonstandard jobs and high-skill, standard jobs, largely driven by firm demands for skill-based quality production. This study interrogates this assumed linkage by analysing firm-level employment patterns in South Korea. Using establishment-based surveys and workplace interviews, the findings largely support the dualisation thesis but reveal a deviation: high-skill jobs can also be assigned to nonstandard employment forms. Notably, ‘core work’ is frequently outsourced to subcontractors, reflecting the enduring influence of institutions initially structured for a low-cost export strategy. This study suggests that South Korea’s latecomer institutions, including hierarchical interfirm relationships dominated by conglomerates (Chaebols) and a history of adversarial industrial relations, complicate the dualisation scenarios observed in other advanced coordinated economies. This study highlights how South Korea’s latecomer institutions shape skill-based employment inclusivity, ultimately demonstrating limited labour market inclusiveness in a country known for its remarkable industrial growth.

Keywords

South Korea dualisation skill trade union latecomer economies

Introduction

This article investigates the relationships among skills, industrial relations institutions, and the use of nonstandard employment contracts in South Korea (hereafter, Korea) after the Asian Financial Crisis of 1997–98 (AFC). Despite continuous industrial growth in Korea, a trend of the use of nonstandard work, however, spells trouble in terms of labour market inclusiveness (OECD, 2020). There are ample challenges arising from the globalisation of economic activities, neoliberalism, and deindustrialisation, which have been heightened since the breakout of the AFC. This issue is central to academic debates about the sustainability of the changes in the Korean ‘developmentalist’ growth model, which, during the Golden Era of post-war development, enabled rapid industrialisation and pulled people out of absolute poverty by offering human capital development and job opportunities (Amsden, 1989; Haggard, 1990). As the Korean economy has transitioned toward a structure based on skills and knowledge, the flexibilisation patterns of the workforce, influenced by the institutional legacy of this growth model, require thorough investigation. While previous studies have documented the expansion of nonstandard work, particularly in low-skill jobs, the working arrangements of high-skill jobs remain underexplored. This study addresses this gap by posing two research questions: (1) What skill requirements do firms consider when determining working arrangements, and (2) how do latecomer institutions and trade unions influence the coordination of skill-based innovation?

According to Amsden, 1989, institutions that promote exports and industrial upgrading are at the centre of Korean firms’ export-led strategy. ‘Latecomer institutions’ in this case encompass hierarchical interfirm relations dominated by large conglomerates (chaebols), reliance on advanced foreign technology, and highly conflictual industrial relations. This set of institutions provides firms with relevant resources for quickly moving up value chains and competing in export markets (Yeung, 2016).

However, this institutional system has been, particularly since the AFC, eroded, and the export-led growth model has been adjusted to cope with deindustrialisation (Kalinowski, 2019; Pirie, 2006). For the state, liberalisation reforms have become key policy tools (Thurbon, 2001), and paradoxically, these reforms led to the expansion of the welfare state to mitigate rising labour market risk. On the corporate side, upgrading to quality-product segments, along with the adaptation of ‘best practices’ in human resources management (HRM), has become a prevalent method for mediating rising market pressure and labour costs (Kim and Bae, 2004).

The latecomer institutions designed for export-led growth, post-AFC liberalisation policy reforms, and firm strategies seem to have resulted in labour market inclusiveness in distinctive ways. Numerous studies have highlighted the distinction or ‘duality’ between standard and nonstandard worker groups in terms of job security, compensation, and training, which became more consequential after the AFC (e.g. Lee, 2011). Some scholars associate this divide with firm size, as is the case for chaebol firm employees versus employees of smaller firms (Song, 2014); other scholars link this change to gender gaps (Kim et al., 2020; Kwon and Doellgast, 2018; Lee and Kim, 2020; Peng, 2012).

The employers’ interest in skills for producing high-quality products is at the centre of the debate regarding the dualisation scenario in other industrialised coordinated market economies (CME) (Palier and Thelen, 2010). As political economists claim, the production of high-quality products is socially coordinated with institutions that are designed to foster the growth of high skills (Busemeyer and Trampusch, 2012; Streeck, 2011). In particular, the Varieties of Capitalism school (VoC) posits that in German and Japanese CMEs, the development of firm-specific skills among employees is seen as being in step with employers’ core interests as a way to support the permanent employment of their workers (Hall and Soskice, 2001; Hassel, 2014). These highly valuable and specific skills form the foundation of core work, which enables firms to differentiate their products in the export market.

Scholars in the fields of critical sociology and industrial relations have pointed to robotisation as a crucial factor that undermines skill formation, and they have more recently highlighted the political power of trade unions as a core factor behind dualisation (Kim, 2010; Park, 2014). Despite these insights, little is known about the links between skills and standard employment for innovation. How to produce innovative products and services is a particularly crucial managerial issue, as many Korean firms’ business strategies have come to emphasise a product differentiation approach that targets niche markets. Even though Korean firms are now competing with their European and Japanese counterparts, they design their skill strategies in response to institutions that are different: Korean latecomer institutions were originally designed in terms of a cost-leadership strategy to increase their level of technological proficiency (Whitley, 1999).

This article revisits labour market dualisation through the lens of skill-based innovation, exploring how latecomer institutions have influenced employment patterns for high-skilled workers engaged in core work. To address this inquiry, the study analyses the diverse relationships between skill and employment across firms in the post-AFC institutional environment. It employs a framework proposed by VoC, which situates the analysis of institutional coordination of business strategy, firm-specific skills, and trade unions at its core. Furthermore, this study draws upon the extant literature on human resources management (HRM) and industrial relations to examine the employment decisions of firms. As most research on dualisation tends to be macro-based, the distinction is rarely made from the strategic perspective of the firm. However, the HRM literature shows that the use of different employment modes by firms depends on the different strategic importance attached to certain skills, with the value and specificity of the skills being required differently in different firms (Green, 2013; Lepak and Snell, 1999; Stavrou et al., 2015). This firm-level perspective enables the differentiation of various nonstandard employment contracts, particularly fixed-term contracts and subcontracting contracts. We are particularly interested in these two types of contracts because subcontractors, unlike fixed-term contractors, are indirectly employed by the firm. Because of these different institutional characteristics, each nonstandard employment contract is associated with a different risk portfolio for the employer. Furthermore, trade unions are also included in the analysis, as the literature on both industrial relations and the political economy suggests that they play a crucial role in skill development and employment strategy (Hassel, 2014; Palier and Thelen, 2010).

The article combines a quantitative analysis of the Korean Workplace Panel dataset with qualitative analysis, such as interview findings. Employing mixed methods allows for a better illustration of the patterns and mechanisms of change by revealing cross-firm differences in workforce arrangements and considering how skills are managed within firms. In particular, Seawright (2016)’s integrative approach involving mixed methods is highly beneficial for capturing the distinct institutional coordination of the strategies and behaviours of firms. By leveraging this approach, we can effectively incorporate Korea’s latecomer institutional legacy into our analysis, providing valuable insights into the correlations identified by quantitative models.

This study contributes to the literature on labour market dualisation, with a particular focus on Korea’s position as a newly emerged economy. It offers new insights into the nature of skill specificity in such contexts, calling into question the mechanisms established in mature CMEs. The findings indicate that Korean firms have established a skill-saving innovation strategy driven by the country’s latecomer institutional features, such as interfirm relations and trade unions. This strategy suggests that a considerable degree of labour flexibilisation, coupled with limited in-house skill development, can still support the production of high-quality goods. Within these mechanisms, subcontractors, along with other nonstandard worker groups, play a pivotal role. Furthermore, by focussing on employment practices in Korea, this study also enriches the literature on inclusive industrial growth, suggesting that newly emerged economies may face greater labour market marginalisation than mature economies.

The present paper is structured as follows: the second section discusses the features of latecomer institutions, which underpins the behaviour of firms in terms of business strategy, firm-specific skills and trade union activities. In the third section, these firm characteristics are presented as propositions regarding the distribution of different employment types. The fourth section presents the methodological strategy employed. The fifth section presents the findings of the regression analysis, which is followed by a qualitative study examining the underlying institutional dynamics in the sixth section. The seventh section presents the conclusion.

Korea’s latecomer institutions and their legacies for the labour market and skill-based innovation

Although there are diversities in firms’ skill-employment strategies, their innovation activities for producing high-quality products are shaped by the institutional characteristics of the national business system in which firms operate (Hall and Soskice, 2001). Korea’s national business system, which emerged post-war as a sovereign economy in the global business landscape, distinguishes itself from other coordinated economies that initiated industrialisation earlier, such as manufacturing giants such as Germany and Japan (Streeck and Yamamura, 2001). Initially, designed for rapid poverty alleviation through export-oriented industrialisation, Korea’s latecomer institutions featured a strict hierarchy among the state, firms, and workers, aiming to catch up quickly with global leaders in advanced technology (Amsden, 1989; Whitley, 1999). In this hierarchical network, large business groups, Chaebols, rapidly expanded into global markets with extensive state support, primarily subordinating smaller-sized suppliers. Unlike the peaceful management–labour cooperation observed in post-war coordinated economies (Busemeyer and Trampusch, 2012), Korean workers were initially excluded and oppressed until democratisation in 1987. Despite this, a culture of labour resistance persisted, particularly among chaebol employees, due to the enterprise union institution (Kwon and O'Donnell, 2003).

Latecomer institutions crucially influenced the evolution of firms’ innovation strategies. The core interest of business was growth through cross-industry diversification and the continuous upgrading of target business segments (Haggard, 1990; Hemmert, 2018). In this process, expansion into ever-new industries was highly dependent on technological transfer through joint ventures or direct investment with foreign-based firms. Being subcontracted by foreign-based firms, mostly Japanese manufacturers, they assemble and export final products to advanced economies (Castley, 1997; Shim and Lee, 2008). As Kim (1997) pinpoints, Korean firms’ diversification strategies were built on their initial imitation of foreign innovative products and progressed towards in-house innovation.

The features of in-house innovation are closely related to latecomer institutions. Owing to Japan’s colonial legacy and post-war economic success, Chaebols actively adopted the Japanese model of trust-based subcontracting relationships, along with an internal labour market (ILM) that emphasises targeted skills development, and job security with gradual increases in wages (Bae and Form, 1986; Castley, 1997; Shin, 2001). However, this model has not been fully fledged due to the deeply rooted mistrust between management and labour and was later challenged by the increased labour costs resulting from the democratisation of 1987 and the AFC of 1997–98. In this development, although ILM was adopted for a high-performance work system (Bae and Lawler, 2000; Kim, 2008), benefiting from a more cooperative labour relationship (Kong, 2012; Lee, 1998), the legacy of labour oppression hinders extensive corporatism for skill-based innovation. Rather, product innovation is driven by college-graduate engineers, whereas production process innovation is achieved through extensive robotisation (Jo et al., 2016; Song and Yi, 2014). Furthermore, the strict interfirm hierarchy serves as useful for cost reduction since it allows the subcontracting relationship to be expanded, limiting suppliers’ investment in innovation activities (OECD, 2020). Consequently, compared with those in earlier industrialised coordinated economies such as Germany and Japan, the latecomer institutions resulted in only a small number of firms dominating innovation activities with the possession of thinner in-house innovation capacities. This has contributed to a notable reliance on cutting-edge products produced by external firms, which are frequently foreign-owned.

In light of these distinctive institutional characteristics, the subsequent section presents a series of propositions designed to examine the impact of post-AFC business strategies, firm-specific skill strategies, and trade unions on employment patterns.

The role of firms’ skill demand, trade unions, and latecomer institutions in workforce flexibilisation

While high-value, firm-specific skills and the cooperative role of trade unions are important for fostering innovation in mature CMEs, these features were initially less well established in latecomer CMEs. Notably, the process of upgrading market segments has led to more diverse skill demand across firms, in contrast with the more uniform skill requirements observed in mature CMEs. According to Porter (1985), variation in skill demand is tied primarily to firms’ strategic positions within the value chain. While firms in mature CMEs are predominantly positioned within the high-end segment, latecomer CMEs exhibit a more dispersed range of firms across different segments of the value chain.

The demand for valuable skills is shaped by product market strategies linked to firms’ positioning within different segments of the value chain (Miles et al., 1978; Porter, 1985). Product market strategies are classified broadly into quality (i.e. product differentiation) and cost leadership. The former usually targets high-end products and services by differentiating themselves from their rivals through higher-value creation. In contrast, the latter refers to low-value segments within the value chain, that is, the opposite of quality.

In this respect, firms with a quality-oriented product market strategy require high-value skills from a significant proportion of their workforce. In contrast, firms with a cost leadership strategy aim to become low-cost producers in their industry and, given the economies of scale, such as highly automated assembly and reduced overheads, require a lower share of high-skilled workers and a seemingly higher share of low-skilled workers. Thus, because high-value skills are critical, firms with a quality-oriented strategy are more likely to employ a high proportion of standard workers than those with a cost strategy (Lepak and Snell, 1999).

Proposition 1

Due to the higher value of skills, the more quality-oriented product market strategies firms’ use, the larger the share of workers on standard contracts that firms tend to employ.

Although product market strategy influences the firm’s demand for valuable skills in general terms, this approach says little about the nature of those skills, such as whether they are general or firm-specific. The resource-based view of firms (RBV) highlights that the targeted strategy can be realised only when paired with the capacity to mobilise unique resources (Wright et al., 2001). Product market strategy and strategic human resource management are closely linked since business decisions depend on internal capacities, capabilities, and competencies.

In that context, firm-specific skills or unique human resources and their internal development are deep-seated in achieving ‘core competence’ (Hamel and Prahalad, 1994) and encompassing ‘dynamic capabilities’ (Teece et al., 1997). The RBV contends that the workforce’s internalisation to develop firm-specific skills comes about at the point where respective labour costs meet future productivity or where the market scarcity of skilled workers is too high when those skills are sourced (Purcell, 1999). Thus, the scope for firm-specific skills is determined by core activities for business over which firms endeavour to maintain control and are unlikely to delegate to other organisations.

Proposition 2

For capability building, firms that provide firm-specific training tend to have a greater proportion of standard workers that they tend to employ.

Scholars of industrial relations and the political economy, alongside the strategy and capability perspectives, underscore the vital role that trade unions play in developing capabilities. Cooperative industrial relations are key in aligning employers’ interests with the stable acquisition of specific skills, which in turn supports their preference for providing long-term employment stability. This cooperation not only stabilises the workforce but also strengthens employees’ bargaining power and job security. The high-performance work system (HPWS) serves as a prime example of such an innovative workplace practice, fostering innovation and sustainability within the internal labour market through effective collaboration between management and trade unions (Appelbaum et al., 2000).

Moreover, according to Shin (2014), without taking on a quality-product strategy, trade unions’ effects become negative for firm competitiveness. The production of quality products requiring valuable and firm-specific skills resolves the management–union conflict concerning the adoption of HPWS. Kong (2012) highlights that the adaptation of HPWSs by some Korean chaebol firms enhances cooperative relations between unions and management, suggesting that trade union participation can be positively supported if firms target high-end products.

While trade unions have a significant effect on innovation, they also have a detrimental effect on employment. This is evidenced by the expansion of nonstandard work through the mechanism of employment rent (Sørensen, 2000). Nonstandard workers are regarded as a ‘buffer’ with an ‘incentive’ that serves to increase their bargaining power. Buffer effects serve to protect standard workers from the risk of unemployment. In contrast, the incentive effect – namely, the possibility of converting nonstandard contracts into standard contracts – increases productivity and efficiency. The combined effect is to substitute the hiring of standard workers with nonstandard workers, leading to a reduction in the number of new standard worker hires. The ‘incentive-buffer’ mechanisms reinforce each other until an equilibrium state in the segmentation process is reached (Polavieja, 2005).

Buffer effects can vary between different types of nonstandard contracts due to the varying application of the ‘logic of membership’ (Schmitter and Streeck, 1999). Since fixed-term contractors are directly bound to the firm through their contracts, unions may advocate that these workers transition into standard contracts, thereby increasing union membership. However, unions often struggle to negotiate on behalf of subcontractors, as these workers are employed by legally independent entities. This issue is particularly pronounced in enterprise-based bargaining systems, such as those in Korea, where the distinction between employees and outsiders is clearly defined. Consequently, the union’s inclusiveness approach to representing different worker groups (Doellgast et al., 2018) may vary significantly, with a stronger focus on fixed-term contractors than on subcontractors.

Proposition 3a

Due to trade unions’ bargaining power, unionised firms tend to retain more standard workers than nonstandard workers.

Proposition 3b

Due to buffer effects, unionised firms are more likely to employ nonstandard workers than nonunionised firms.

Proposition 3c

Due to the union’s inclusiveness approach, which is based on the logic of membership and enterprise-based unionism, unionised firms are less likely to use subcontractors than fixed contractors, regardless of the business strategy and the demand for firm-specific skills.

Mixed methods and data

A mixed-method approach was employed to address the following research questions: (1) what skill requirements do firms consider when determining working arrangements, and (2) how do latecomer institutions and trade unions influence the coordination of skill-based innovation? This approach integrates quantitative and qualitative studies rather than merely triangulating between them (Seawright, 2016). The qualitative component assists in identifying pathways to correlations identified in the regression analysis. Specifically, our qualitative inquiry focuses on discerning institutional dynamics that Korean firms navigate concerning skills for innovative product development. Since the integrative approach requires a sequential design, the regression analysis first aims to answer research question 1. To this end, the analysis examines the effects of the demand for valuable skills and specific skills, as well as unionisation, on the determination of various employment types. Second, deviant cases are selected on the basis of the statistical results (Seawright and Gerring, 2008), enabling the addressing of research question 2. In the present study, deviant cases refer to unionised firms that utilise high shares of subcontractors to produce high-quality products. Moreover, firms in the automobile, chemical, electronics, machine tools, steel and shipbuilding sectors were selected for analysis since these sectors represent Korea’s primary export sectors for which latecomer institutions were established (Amsden, 1989). Consequently, despite potential sector-specific variations in mechanisms, institutional factors have shaped the general level of skills, firm strategies, and trade unions in these sectors (Hall and Soskice, 2001). Therefore, this best captures the institutional mechanism, addressing how latecomer institutions and trade unions influence the coordination of skill-based innovation, as outlined in research question 2.

Quantitative study

Regression analysis was done using data from the Workplace Panel Survey (WPS), which was prepared by the Korea Labor Institute. This dataset includes four waves of data collected from 2005 to 2013. The WPS is a longitudinal survey, and its samples cover 1,770 unit firms drawn from all industries. The total number of observations is 8,922.

Random effects models were run via the xtreg command of STATA software. The dependent variable is the share of standard workers in the total workforce for Model 1; the share of fixed contractors in the total nonstandard workforce for Model 2; and the share of subcontractors in the total nonstandard workforce for Models 3, 4, and 5. Although the Hausman test results suggested choosing fixed-effect models, I chose random-effects models for the following reasons: first, my datasets do not carry large unit sizes (Clark and Linzer, 2015); second, my theoretical approach of neoinstitutionalism fails to follow linearity and makes a strict exogeneity assumption (Wooldridge, 2002). There are six models in total with four models featuring interaction terms: Models 1, 2, and 3 are basic; Models 4 and 5 have an interaction term added; Model 6 is a fully integrated interactive model. Propositions 1, 2, 3a, and 3b are tested by Model 1. The rest of the models are for Proposition 3c. The simple models and the interactive models look as follows:

I. Standard worker $= β_{0} + \sum β_{1} {CONTROLS}_{i t} + β_{2} {product market strategy}_{i t} + β_{3} firm specific {training}_{it} + β_{4} {trade union}_{i t} + ε_{i t}$

II. Fixed-term contractor $= β_{0} + \sum β_{1} {CONTROLS}_{i t} + β_{2} {product market strategy}_{i t} + β_{3} firm specific {training}_{it} + β_{4} {trade union}_{i t} + ε_{i t}$

III. Subcontractor $= β_{0} + \sum β_{1} {CONTROLS}_{it} + β_{2} {product market strategy}_{it} + β_{3} firm specific {training}_{it} + β_{4} {trade union}_{it} + ε_{it}$

IV. Subcontractor $= β_{0} + \sum β_{1} {CONTROLS}_{i t} + β_{2} {product market strategy}_{i t} + β_{3} {firm specific training}_{i t} + β_{4} {trade union}_{i t} {+ β}_{5} {product market strategy}_{i t} * {firm specific training}_{i t} + ε_{i t}$

V. Subcontractor $= β_{0} + \sum β_{1} {CONTROLS}_{i t} + β_{2} {product market strategy}_{i t} + β_{3} {firm specific training}_{i t} + β_{4} {trade union}_{i t} {+ β}_{5} {firm specific training}_{i t} * {trade union}_{i t} + ε_{i t}$

VI. Subcontractor $= β_{0} + \sum β_{1} {CONTROLS}_{i t} + β_{2} {product market strategy}_{i t} + β_{3} {firm specific training}_{i t} + β_{4} {trade union}_{i t} + β_{5} {product market strategy}_{i t} * {firm specific training}_{i t} + β_{6} {firm specific skill}_{i t} * {trade union}_{i t} + β_{7} {product market strategy}_{i t} * {firm specific training}_{i t} * {trade union}_{i t} + ε_{i t}$

The independent variables are business strategies, specific skill development, and trade unions. Firm size, ROA, debt ratio, and industry skill levels are taken as control variables. In addition, professional skill is also included, as the uniqueness of the firm is not only firm specificity but also specialised expertise such as lawyers (Lepak and Snell, 1999). Professional skill includes a variable that indicates an HR policy promoting professional certifications that are generally recognised within the industry and beyond. The codes are as follows: coded 2 if HR policy includes preferential treatment for relevant certification holders; coded 1 if there are relevant certifications but no preferential treatments; and coded 0 if there is no such HR policy. To verify the effects of the skill value (Proposition 1), I use the product market strategy variable. This proxy is tested to distinguish between a product differentiation strategy (for a niche market) and a cost-leadership strategy (towards a mass market). Concerning the product market strategy, I take a relevant question (A210) from the WPS survey, which measures product/service differentiation levels. Its four-point scale captures Porter’s strategic spectrum (market frontier with innovative products = Code 4; moderate emphasis on developing new products = Code 3; stable growth with a focus on efficiency of the existing operation = Code 2; cost-sensitive, only reacting to changes in the environment = Code 1).

For the effects of skill uniqueness (Proposition 2), I regress HR policy on firm-specific skills, the variables related to human resource management and development. I suggest a proxy for firm-specific skills, the dummy variable training type on priority (coded 1 if in-house training or on-the-job training are more important than external training or no training). Finally, trade union is used to measure the political context (Proposition 3). It is taken as a binary variable of having or not having trade unions. This binary variable is coded as 1 if there are trade unions and 0 if there are none.

All the models include control variables such as industry skill levels, return on assets (ROA), debt ratio, firm size, and year. Two variables for industry skill levels are constructed: industry high skill and industry low skill. I scaled these variables by their proportions in industries with high- or low-skilled workers. Industry high skill is measured by the ratio of the top 20% of occupations within industries in terms of skill level, and industry low skill is measured by the shares of the bottom 20%.

Qualitative study

Thirty-six semistructured interviews were conducted in person, emails, or text messages in 2016 and 2018, as were media interviews and descriptive statistics. The interview partners are HR managers (21), trade union leaders (9), and line managers and senior management (6). They are employed by chaebol firms (16) and non-chaebol supplier firms (11). The distribution of interviews by industry can be found in the online appendix (see Figure S1).

An analysis of the effects of skill value, specificity, and trade unions on employment type

Descriptive statistics

The patterns of the top three types of employment, that is, standard contractors, fixed contractors, and subcontractors, vary according to firm characteristics (see Matrix of Correlations in the Appendix). Product market strategy, firm-specific training, and trade unions show positive correlations with standard workers (0.078, 0.045, and 0.081, respectively). This underscores the crucial role of trade unions, given their greater correlation than the others’. However, these firm characteristics have varying effects on fixed contractors and subcontractors. While product market strategy and firm-specific training are negatively correlated with fixed contractors (−0.116 and −0.047, respectively), they exhibit positive correlations with subcontractors (0.073 and 0.078, respectively). Finally, the positive correlation between trade unions and subcontractors (0.091) is notably greater than that with fixed contractors (0.013) and even surpasses that with standard workers (0.081). See the online appendix for employment patterns by industry (see Table S1).

These complex employment outcomes result from the diverse ways in which firms assign strategic importance and thus make skill arrangements. As more firms move to higher segments, they demand high-value skills and unique skills that enable them to build distinctive competence in the market. As the matrix in Figure 1 captures, the concentration of firms becomes denser on the horizontal axis, implying that more firms produce quality products and services. However, among them, the demand for unique skills on the vertical axis is similarly distributed. This may imply that the development of unique capabilities is a sufficient condition for producing high-value products. All in all, this demonstrates that valuable and unique skills have become a more critical part of firms’ competitiveness, so firms demand more high-skilled workers in various working arrangements. (Table 1).

Figure 1.

The distribution of firms in HR architecture.

Table 1.

Random-effect regression model table.

	(1)	(2)	(3)	(4)	(5)	(6)
	Standard worker	Fixed contractor	Subcontractor	Subcontractor	Subcontractor	Subcontractor
	No interaction	No interaction	No interaction	Strategy*firm-specific	Firm-specific*trade union	Strategyfirm-specificunion
Prod Mkt strategy	.567***	−.028***	.029***	.005	.029***	.009
	(.2)	(.007)	(.008)	(.011)	(.008)	(.019)
Firm-specific training	.943**	−.031**	.061***	−.067	.044*	−.014
	(.416)	(.015)	(.017)	(.05)	(.025)	(.09)
Trade union	1.444**	.006	.131***	.131***	.112***	.137*
	(.645)	(.021)	(.025)	(.025)	(.029)	(.08)
Prod Mkt Str*Firm-				.041***		.018
Specific training				(.016)		(.028)
Firm-specific*Union					.033	−.089
					(.034)	(.106)
Prod Mkt Str*Union						−.01
						(.024)
Prod Mkt StrFirm- specificUnion						.041 (.034)
Professional skill	−.815**	.019	−.006	−.006	−.007	−.006
	(.393)	(.015)	(.017)	(.017)	(.017)	(.017)
Industry high skill	−.251***	.432***	−.183***	−.189***	−.18***	−.192***
	(.055)	(.08)	(.065)	(.065)	(.065)	(.064)
Industry low skill	−.081***	−.15*	−.058	−.064	−.058	−.068
	(.031)	(.087)	(.112)	(.112)	(.112)	(.113)
ROA	2.257	−.157*	−.003	.003	−.003	.002
	(1.901)	(.081)	(.099)	(.099)	(.099)	(.098)
Debt ratio	1.673**	.006	−.016	−.016	−.016	−.015
	(.825)	(.03)	(.04)	(.04)	(.04)	(.04)
Firm size	−2.692***	.024***	−.029**	−.029**	−.029**	−.029**
	(.308)	(.008)	(.014)	(.014)	(.014)	(.014)
Year	−.007	.009***	.023***	.023***	.023***	.023***
	(.08)	(.003)	(.003)	(.003)	(.003)	(.003)
Constant	118.656	−18.42***	−46.183***	−46.6***	−45.933***	−46.703***
	(160.975)	(6.087)	(5.957)	(5.965)	(5.954)	(5.981)
Observations	5101	3179	4056	4056	4056	4056
Pseudo R²	.z	.z	.z	.z	.z	.z

Standard errors are in parentheses.

***p < .01, **p < .05, *p < .1.

Results

Model (1) reports the share of standard workers in the total workforce. I report the results for each variable. Among the control variables, industry high skill and industry low skill alike are significant with a negative coefficient, suggesting that there is a wide variety of working arrangements in place for high-skilled jobs. While ROA is not statistically significant, the debt ratio is significant and positive. However, firm size is significant with a negative coefficient. Among the explanatory variables, product market strategy and firm-specific training are statistically significant, with positive coefficients. These results demonstrate that firms tend to develop more valuable or unique skills through the use of standard contracts. Notably, the coefficient of firm-specific training is far larger than that of the product-market strategy. This suggests that standard employment arrangements are more critical in developing firm-specific skills than in securing valuable skills. Moreover, the negative association of professional skill helps to demonstrate the criticality of firm specificity for building competence. Finally, trade union is positively correlated. Trade unions’ influence seems to be most critical for standard workers, as its positive coefficient is far larger than those of product market strategy and firm specificity. This means that of those variables, the role of industrial relations is the most critical for securing standard workers. Consequently, the results of Model (1) strongly support the internal labour market thesis, which claims the firm’s internal supply of labour with the active involvement of trade unions for building competence.

Models (2) and (3) report the shares of fixed contractors and subcontractors, respectively, in the nonstandard workforce alone. First, in Model (2), firm size, year, and industry high skill are significant with a positive effect, whereas the rest of the control variables are not significant. More fixed contractors are used by larger firms, and the size of that group has increased over time, although the change is very small. Both product market strategy and firm-specific training are significant, with negative coefficients. Trade union is not statistically significant. Firms with higher skills tend to use fewer fixed contractors. Second, Model (3) reports the share of subcontractors in the nonstandard workforce. Firm size is significant and negatively correlated with the share of subcontractors in the nonstandard workforce. However, year has a positive coefficient. This implies the evolution of the subcontracting workforce, indicating that subcontracting has become a more prevalent alternative to other nonstandard working arrangements over time. Product market strategy, firm-specific training, and trade union are significant and positively correlated.

Models (4) and (5) are interactive models¹ of subcontractors. While product market strategy*firm-specific training is significant, firm-specific training*trade union is not significant. Despite the statistical non-significance, the latter can also be interpreted from the table because the analytical interest lies ‘in the marginal effect of X on Y for substantively meaningful values of the conditioning variable Z’ (Brambor et al., 2006: 74; see also Ai and Norton, 2003). Following the recommendation of Brambor et al. (2006) and Williams (2012), the command margins is used to estimate the marginal effects of firm-specific training given a different type of product market strategy. In the same vein, Figure 2 shows the plot for the marginal effect of firm-specific training on the share of subcontractors in the nonstandard workforce conditional on different product market strategies. The marginal effect is positive, which means that the positive marginal effect of the firm-specific training strategy increases the share of subcontractors if firms use a differentiation strategy (at 3 or 4 in Figure 2). In the same vein, Figure 3 shows the plot for the marginal effect of trade unions on the share of subcontractors conditional on firm-specific training. This demonstrates the greater marginal effects of trade unions when there is a firm-specific training policy in place.

Figure 2.

Predictive margins of firm-specific training with 95% confidence intervals.

Figure 3.

Average marginal effects of trade union at different value of firm-specific training with 95% confidence intervals.

Model (6) includes two additional terms of product market strategy*union and reports a fully interactive model with a term of product market strategy*firm-specific training*trade union. Figure 4 shows the plot for the marginal effect of trade unions on the share of subcontractors conditional on the strategies. The positive effect of trade unions is greater for firms that have a firm-specific training policy and use a more quality-oriented product market strategy. The line in the right-hand box shows that the marginal effect of trade unions increases as firms use more of a quality-driven strategy (1 = cost leadership strategy). The overall plot indicates that the use of subcontractors is more prevalent for unionised firms that produce quality products with unique features.

Figure 4.

Predictive margins of trade union and firm-specific training at different product market strategy with 95% confidence intervals.

It should be noted, however, that these models capture the total number of fixed contractors and subcontractors, including those of all skill levels. A robustness check was therefore conducted by running additional models to analyse the utilisation of fixed contractors and subcontractors for core activities (see Table S2 in the online appendix). These random effect logit models examine the probability of hiring at least one fixed contractor or subcontractor for skilled positions or core activities, respectively. First, while product market strategy is significant with a negative effect on the probability of using a fixed contract, firm-specific skill and trade union, as well as professional skill, are not significant (Model S1 in the online Appendix). Second, product market strategy, specific-skill training, and trade union have significant positive effects on the probability of using a subcontractor for core work, whereas professional skill is not significant (Model S2 in the Online Appendix). The interactive models² of subcontractors for core work (Models S3 and S4 in the online appendix) have similar marginal effects as their counterparts in our main models (see the online appendix). These results serve to reinforce the findings of our main models, providing reassurance that trade unions and employers of niche firms tend to subcontract core work while being unlikely to employ fixed contractors.

The quantitative analysis in this section has shown that as firms become more interested in producing quality products and developing firm-specific skills, they are more likely to have a larger share of standard workers in their workforce. The industrial relations claim is also supported by the finding that trade unions play a positive role in the high proportion of standard workers. These findings confirm the ‘dualisation’ scenario. However, employers and unions are also more likely to use subcontractors to replace standard workers rather than fixed contractors to do core work, showing that ‘skills and industrial relations’ may not completely protect workers from casualisation. Firms that invest in specific skills tend to use a larger share of subcontractors, as they become more quality-oriented. Despite the need for firm-specific skills, unionised firms use a greater proportion of subcontractors than non-unionised firms do. Furthermore, the analysis has shown that, among other firms, unionised niche firms use the highest share of subcontractors. These findings suggest that there is scope for using subcontractors rather than standard workers to develop unique capabilities in the pursuit of product differentiation.

Qualitative study

This section presents a qualitative analysis that traces the institutional dynamics underlying the previously identified substitution of standard workers with subcontractors for core work. Specifically, it examines how the latecomer institutional legacy of interfirm relations shapes the design of skills and work systems and how trade unions manage subcontractors in the context of innovation. The analysis focuses on chaebol firms, given their differentiation strategies, substantial in-house training systems, and notably higher levels of union participation than other firms do.

What are the Impacts of the Latecomer Institutional Legacy of Interfirm Relations on Innovation?

Although Korean firms have grown significantly over time, the innovation system between chaebols and parts suppliers differs because the hierarchical interfirm relationships are reinforced by chaebols’ post-AFC business strategy. Similar patterns are observed across the sectors examined in this study, suggesting that sectoral differences may be less significant within these contexts. Chaebol firms drive innovation in these sectors by replicating this hierarchy with their suppliers as they expand into new sectors (Wang and Tsai, 2010). For example, a union leader from the automobile industry pinpoints that although chaebols have moved into the high-end export market, they have kept production at low costs (Union#3). Innovation in chaebol firms is entirely in the hands of university degree holders in the R&D department (HR manager#6, Amsden, 1989; Han and Downey, 2014; Jo, 2016; Song, 2000). This engineer-led R&D was the driver of technological upgrading that led to Korea’s rapid industrialisation, but it remained active as an innovation strategy afterwards to save costs and avoid conflict with militant trade unions.

Meanwhile, suppliers avoid internal innovation. Instead, they focused on actively setting up joint ventures with foreign technology firms and aimed to import ‘the state-of-the-art’ parts to meet the high-quality requirements and cost pressures of chaebols (Union#3; Union#4; Management#2 of automobile suppliers). To mediate the pressure from the chaebol to provide niche parts at lower cost, suppliers continued to rely on foreign technology, just as they did amid export-led industrialisation, when advanced foreign technology was a critical resource for assembling and exporting finished goods, as Amsden (1989) observed thirty years ago. The following quote from a trade union representative of an automobile component supplier illustrates this:

[Innovation] is made and sold by the headquarters of foreign origin. Chaebol X follows what [e.g.,] German competitors do. We should introduce [state-of-the-art technology from Germany] to them for approval. Indeed, few parts makers in Korea have their original technology. The U.S. head office controls technology, technology for the product, technology for the production... We bring the improvements that have already been made in advanced countries. We only make it with blueprints (Union#4 of a supplier in the automobile industry).

The core activities of the supplier firms are largely confined to customising the imported parts to meet Chaebols’ requirements at low costs as cost pressure has increased, particularly since the AFC. Chaebols have increased competition by bringing Chinese suppliers, as ICT advancement has expanded the available choices of global suppliers. A senior HR manager in the electronics industry pinpoints that, as exemplified by his firm, some suppliers completely shut down their factories and switched to a brokerage business model, acting as intermediaries between foreign parts suppliers and Chaebols (HR manager#7). Owing to deteriorated net profit margins, suppliers are little motivated to invest in R&D to develop original technology. For example, the overall R&D spending per sales of the top six automobile parts supplier firms is approximately one percent, far less than that of their global competitors; for example, Denso, 9.2%; Balleo, 9.0%; Bosch, 8.8%; and Continental, 6.2% (Park, 2018). As Figure 5 shows, imported technology simplifies the production of core products for Korean suppliers.

Figure 5.

Core products of major automobile parts suppliers. (Source: Hana Bank; author calculated based on the report).

Management and workers claimed that specific or substantial skills are not required of standard workers (Union#2 of an electronics supplier; Union#3 of an auto supplier; HR Manager#5 of a chemical supplier). In the recruitment of standard workers, the candidates are not necessarily required to have prior technical knowledge since work does not demand advanced knowledge. Union members even highlighted that only 3 months of on-the-job training would be sufficient to develop the relevant skills (Union#1; Union#4). In addition, the union leader of a chaebol steel firm highlights that the types of skills that college graduates bring are too narrow for production work (Union#4). In this interpretation, an engineering degree leads one to be ‘overskilled’ or provide ‘wrong’ qualifications due to the high degree of specialisation of such credential grants. Indeed, production work usually requires a skillset that equips one to engage in troubleshooting.

An HR manager of a chaebol electronics maker reported that even in chaebol firms, skills become obsolete within a few years, and workers’ skill levels do not improve after that point (HR manager#4). Further years of experience do not increase one’s skill level proportionally, and only a few experienced engineers are needed by chaebols. The following quote illustrates this point:

There are not many workers who are doing something very important. Roughly speaking, only a few people are needed, and the development of their skillsets is completed in five to seven years. After that, the work they do is almost the same (HR manager#4 of a chaebol electronics firm).

Differences in the importance of work and the required skill levels between standard and nonstandard workers are trivial. Both employers and trade unions stressed that standard workers do not exclusively do jobs requiring specific skills and that nonstandard workers, particularly subcontractors, do not only undertake peripheral or non-core work. In practice, jobs for subcontractors are as important as those for standard workers in contributing to the firm’s distinctive capabilities (Union#5 of a chaebol steelmaker; Lee, 2023).

The jobs for standard workers and subcontractors are rather different in nature. A union leader representing nonstandard steel workers stressed that in the divided work system, standard workers are not familiar with subcontractors’ jobs, and vice versa, as each worker group focuses on different areas of expertise (Union#5). Specialisation is enhanced by a lack of broader training for multiple skills since the law does not allow contractor firms to provide systematic training for subcontractors (HR manager#6 of a chaebol automaker). Contracting firms usually include training costs as part of their service fees. These professional divides do not necessarily depend on the importance of the work being undertaken.

Therefore, becoming a standard or nonstandard worker is determined without regard to one’s skills (HR manager#5 of a chemical supplier; Union#1 of a chaebol steelmaker). It depends on luck, such as timing, if one is hired by a contractor firm or by a subcontracted firm.

I'm not sure about the differences in the quality of work between standard and noncontract workers. I really do not know. Can a contractor do my job? Of course, he can. Probably, short trainings may be needed (Union#1 of a steel firm in a chaebol group).

Trade unions are not actively involved in the skill development of their members, but the patterns of skill formation avoidance vary according to the characteristics of industrial relations. The extent of cooperation is related to the post-industrial development of individual (chaebol) firms (Kong, 2013). In cooperative industrial relations, the trade unions only follow the management’s head-on initiatives on technical training (Union#4). Meanwhile, strong industrial relations have granted trade unions the autonomy to oversee the internal labour market. An HR manager of a chaebol car company with strong industrial relations reports that there are no teams in his HR development department that provide training programmes to production workers (HR Manager#2). Nevertheless, an HR manager from the same chaebol group stressed that trade unions are not desperate to provide relevant training in light of cutting-edge innovation. He illustrates:

We [management] never get involved in what they [trade unions] do. It is their territory. They keep up their ‘traditional’ way (HR Manager#3 of a machinery firm in a chaebol group).

How do trade unions manage nonstandard workers to facilitate innovation?

Both trade union leaders and HR managers suggested that the introduction of the protection law for nonstandard workers following the post-AFC partial liberalisation reform led to the expansion of the use of subcontractors over other nonstandard employment forms, such as fixed-term contracts. Standard and fixed-term contractors should now be paid equally for the same jobs. More notably, fixed-term contractors, who are directly hired, have to be switched to standard contracts if they remain in the same jobs after 2 years of service. However, fixed contracts rarely turn into standard contracts since works for fixed contractors are specifically designed for non-core work and fall into separate job categories. Instead, trade unions choose subcontractors over fixed contractors for the sake of managing costs and achieving flexible long-term relationships. It is now common for subcontractors to even perform core work because the subcontracting relationship usually lasts for years. A trade union leader from the steel industry stressed that different parts of work processes are assigned to each subcontractor and are usually established as long-term commitments (Union#5).

Both employers and trade unions suggest that trade unions act as managers at production sites, keeping production processes under control (HR Manager#1; Union#3; Park, 2014). Trade unions have gained authority on key managerial issues over production sites and exclusively supervise the production workforce. Employers are little involved in appointing the rank and file and directing HRM on the production sites and are only concerned with total labour costs. The management of the nonstandard workforce, for example, appointing subcontractors, is also performed at the discretion of the trade union. To meet the assigned budgets, trade unions implicitly agree on the extensive automation of production lines and on the use of subcontractors even for core work as a way to achieve an annual increase in wages. For instance, an HR manager from a chaebol automaker observed that newly established factories no longer employ production workers on standard contracts (HR Manager#6). These factories are staffed entirely by subcontractors, with a small number of trade union members serving as managers overseeing their work. The division of managerial roles between management and trade unions is illustrated as

Management is concerned only with the total gross amount of wages. The fluctuation [in wages] has not been high. I have been watching it for ten or more years. The firm manages on similar patterns. Wages increase, but, for example, there is no restructuring. Owing to continuous automation, there are no new replacements for those who retire. So, within that budget, there were wage raises, but [the total amount of wages] have been similar (Union#1 of an electronics supplier).

A retired CEO of a chaebol shipbuilder pinpoints that the extensive use of subcontractors is done to escape from hosting a single collective agreement, which represents the standard workers of all skill groups in the company (Management#1). Korean enterprise-based unions are skill-blind, as they have not evolved a tradition of craftsmanship (Frege and Kelly, 2004; Streeck, 2011). Union members generally believe that skill-based wages break down union member solidarity, as employers and the state had mercilessly done during the industrialisation period (Jo, 2016). The HR manager of a car parts supplier and the union leader of a chaebol steelmaker both emphasise that collective bargaining is concerned primarily with wages, with training and development rarely forming an independent agenda.

The findings suggest that there is considerable scope for trade unions to manage the workforce in the production of quality products. First, the data indicate that the innovation strategies of chaebols and parts suppliers may differ. For the former, innovation activities primarily take place in R&D labs, whereas for the latter, cutting-edge technology is often acquired from foreign-based firms, potentially restricting the need for the development of highly specific skills among the local population. Second, it appears that high skills are not necessarily required even for standard workers, who receive limited training; furthermore, there seem to be no significant differences in skill levels between standard workers and subcontractors. Third, partial deregulation and enterprise-based unionism pressure trade unions to secure current members’ interests, which might lead to the delegation of core work to subcontractors with long-term commitments and allow trade unions to supervise these subcontractors. As a result, among other nonstandard contract groups, subcontractors can potentially substitute for standard workers in extensive ways.

Discussion and conclusion: Flexible core employment and marginal inclusion in the dualising labour market

This study demonstrated the extent to which Korean firms employ nonstandard forms of employment and provided an overview of the underlying mechanisms influencing the patterns of flexible workforces. The findings confirm the dualisation scenario by showing the internal development of firm-specific skills to produce quality products, with trade unions playing a crucial role (Hassel, 2014), supporting propositions 1, 2, and 3a but not 3b. Consequently, firms tend to employ standard workers for core work while using nonstandard workers for non-core work. However, qualitative evidence presents an additional perspective, showing that trade unions are focused primarily on supporting existing union members. By avoiding the hiring of new standard workers, they can mitigate wage increases via the seniority wage system. This evidence is supported by recent media reports that Hyundai Motors, which has the most influential trade union in the country, has hired production workers on a standard contract for the first time in 10 years (Ahn, 2023).

The quantitative analysis also reveals that among various types of nonstandard workers, subcontractors can indeed substitute standard workers in performing core tasks, challenging the validity of Proposition 3c. The qualitative study offers further insight into this dynamic by exploring the effects of latecomer institutions and trade unions on the coordination of skill-based innovation, as addressed in research question 2: subcontractors offer skills that are as important to firms as those provided by standard workers. This is evident in firms reliant on imported cutting-edge technology, where innovation from workers is less critical. In supplier firms, the integration of imported technology grappled with Chaebols’ business strategy reduces the scope of core work, limiting specific skills to the customisation of this imported innovative technology. Consequently, low-skill requirements, the fragmentation of core tasks, and trade union oversight enable subcontractors to develop unique firm capabilities alongside standard workers. The findings corroborate the research suggesting that the importance of skill specificity for standard employment is overstated in the dualisation literature.

This new evidence at the firm level enhances our grasp of macro-level dualisation patterns. The evidence demonstrates that different nonstandard employment contracts contribute differently to the dualisation of workplace outcomes, as core work can be done by subcontractors replacing standard workers, whereas non-core activities are delegated to fixed contractors. This directly answers research question 1, which examines what skill requirements firms consider when determining working arrangements. The dualisation of the labour market is, therefore, not a simple institutional division between standard and nonstandard groups of workers according to their skills. This study has shown that dualisation patterns result from complex strategic employer behaviour (Lepak and Snell, 1999) and industrial relations (Schmitter and Streeck, 1999; Streeck, 2011), suggesting that different skill groups are engaged in different nonstandard working arrangements.

By providing a detailed mechanism between skill and employment, this study contributes to the Korean dual labour market literature, which largely neglects the role of skills in light of innovation. The findings confirm that Korean firms’ skill requirements are not as high as those documented in the literature on German and Japanese firms, corroborating the findings of previous research, which suggests that Korean firms have a skill-saving production system (e.g. Jo et al., 2016). As this study additionally pinpoints, the union’s low-skill strategy, which increases productivity by continuing to organise for the importation of the most innovative technology, helps us understand unions’ limited involvement in internal skill development and coalition with employers (Kim, 2010; Park, 2014) and explains why chaebols can focus mainly on improving process technology (Jo et al., 2016; Song and Yi, 2014). This institutionalised trade union strategy addresses research question 2, demonstrating the impacts of latecomer institutions and trade unions on the coordination of innovation. The skill-saving strategy for innovation shows that while core jobs are still better protected, flexibilisation is exposed to all jobs regardless of skill level. Moreover, the reliance on subcontractors in the skill-employment mechanism hints that dualisation can go beyond inequality by employment type but is fed into dualism by the capacities of firms, such as gaps between chaebols and the rest (Song, 2014).

Korea’s dualisation patterns offer insights into the extent of labour market inclusiveness in latecomer CMEs. The extensive flexibilisation of core workers suggests that skill-driven labour market inclusiveness is more constrained in latecomer CMEs than in mature CMEs owing to the resilience of latecomer institutions. These institutions enable employers to reduce core employment despite ongoing innovation, a trend increasingly observed in mature CMEs as well. However, latecomer CMEs do not fully align with the predicted ‘erosion’ of industrial relations as the sole driver of core workforce flexibilisation seen in mature CMEs (Benassi, 2016). The Korean case illustrates that latecomer institutions enable trade unions to push core work outside the firm’s boundaries in such a way that non-core work is permanently disintegrated (Doellgast and Greer, 2007; Rème-Harnay, 2023). Consequently, despite Korea’s notable economic advancements, the Korean labour market has become markedly more polarised and unequal in comparison to that of mature CMEs. This high degree of disproportionality can be attributed to the limited significance afforded to skill-based innovation, whereby the enduring influence of institutional legacies persists in the management of the challenges posed by deindustrialisation and globalisation. These pressures have led to the marginalisation of even highly skilled core worker groups from stable employment to a considerable extent.

To gain further insight into the relationship between innovation and labour market inclusiveness, it would be beneficial to continue focussing on newly industrialised or emerging economies. This approach would allow for the identification of mechanisms or variables that may have been overlooked in studies concentrated on mature economies. Although this study is limited in its focus on skill dynamics, dualisation can be driven by a variety of factors. The Korean case study, however, suggests that the role of subcontractors is critical in shaping the patterns of dualisation and requires further investigation to enhance our understanding of the expansion of labour market exclusiveness.

Supplemental Material

Supplemental Material - Skill-based dualisation in latecomer economies? Analysing core employment in South Korean firms

Supplemental Material for Skill-based dualisation in latecomer economies? Analysing core employment in South Korean firms by Francis K Yoon in Competition & Change.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: this work was supported by the Graduate School of East Asian Studies and German Research Foundation (DFG).

ORCID iD

Francis K Yoon

Supplemental Material

Supplemental material for this article is available online.

Notes

Appendix

Table 2.

Matrix of correlations.

Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)
(1) Standard workers	1.000
(2) Fixed contractors	−0.361	1.000
(3) Subcontractors	0.173	−0.474	1.000
(4) Product market strategy	0.078	−0.116	0.073	1.000
(5) Firm-specific training	0.045	−0.047	0.078	0.026	1.000
(6) Professional skill	−0.016	0.012	0.004	0.032	−0.010	1.000
(7) Trade union	0.081	0.013	0.091	−0.085	0.070	0.065	1.000
(8) Industry high skill	−0.268	0.125	−0.048	−0.042	−0.016	0.028	−0.031	1.000
(9) Industry low skill	−0.090	−0.045	−0.006	0.075	−0.010	0.010	−0.070	−0.048	1.000
(10) ROA	−0.016	−0.050	0.000	0.038	0.011	−0.016	−0.071	−0.011	0.007	1.000
(11) Debt ratio	0.069	0.019	−0.010	−0.016	0.028	−0.013	0.033	−0.056	−0.047	−0.338	1.000
(12) Firm size	−0.033	0.040	−0.002	0.054	0.144	0.141	0.365	0.012	−0.064	0.002	0.025	1.000
(13) Year	−0.011	0.056	−0.023	−0.052	0.073	−0.063	−0.048	0.038	0.076	−0.078	−0.038	−0.055	1.000

References

Ahn

(2023) Hyundai motor to hire production workers for the first time in 10 years. Hankyoreh.

Norton

(2003) Interaction terms in logit and probit models. Economics Letters 80(1): 123–129.

Amsden

(1989) Asia’s Next Giant: South Korea and Late Industrialization. New York, NY: Oxford University Press.

Appelbaum

Bailly

Berg

(2000) Manufacturing Advantage: Why High-Performance Work Systems Pay off. Ithaca, NY: Cornell University Press.

Bae

Form

(1986) Payment strategy in South Korea's advanced economic sector. American Sociological Review 51(1): 120.

Bae

Lawler

(2000) Organizational and hrm strategies in Korea: impact on firm performance in an emerging economy. Academy of management journal 43(3): 502–517.

Benassi

(2016) Liberalization only at the margins? Analysing the growth of temporary work in German core manufacturing sectors: liberalization only at the margins? British Journal of Industrial Relations 54(3): 597–622.

Brambor

Clark

Golder

(2006) Understanding interaction models: improving empirical analyses. Political Analysis 14(1): 63–82.

Busemeyer

Trampusch

(2012) The Political Economy of Collective Skill Formation. Oxford; New York: Oxford University Press.

10.

Castley

(1997) Korea's Economic Miracle: The Crucial Role of Japan. Houndmills, Basingstoke, Hampshire: New York: Macmillan Press; St. Martin's Press.

11.

Clark

Linzer

(2015) Should I use fixed or random effects? Political Science Research and Methods 3(2): 399–408.

12.

Doellgast

Greer

(2007) Vertical disintegration and the disorganization of German industrial Relations¹. British Journal of Industrial Relations 45(1): 55–76.

13.

Doellgast

Lillie

Pulignano

(2018) Reconstructing Solidarity: Labour Unions, Precarious Work, and the Politics of Institutional Change in Europe. Oxford, United Kingdom: Oxford University Press.

14.

Frege

Kelly

(2004) Varieties of Unionism: Strategies for Union Revitalization in a Globalizing Economy. Oxford; New York: Oxford University Press.

15.

Green

(2013) Skills and Skilled Work: An Economic and Social Analysis. Oxford, UK: Oxford University Press.

16.

Haggard

(1990) Pathways from the Periphery: The Politics of Growth in the Newly Industrializing Countries. Ithaca, NY: Cornell University Press.

17.

Hall

Soskice

(2001) Varieties of Capitalism: The Institutional Foundations of Comparative Advantage. Oxford [England]; New York: Oxford University Press.

18.

Hamel

Prahalad

(1994) Competing for the Future. Boston, Mass: Harvard Business School Press.

19.

Han

Downey

(2014) Engineers for Korea. San Rafael: Morgan & Claypool Publishers.

20.

Hassel

(2014) The paradox of liberalization - understanding dualism and the recovery of the German political economy: the paradox of liberalization. British Journal of Industrial Relations 52(1): 57–81.

21.

Hemmert

(2018) The Evolution of Tiger Management: Korean Companies in Global Competition. London; New York: Routledge, Taylor & Francis Group.

22.

(2016) Hyeon-Dae-Ja-Dong-Cha-Ui Gi-Min-Han Saeng-San-Bang-Sig: Han-Gug-Jeog Saeng-san-Bang-Sig-Ui Tam-Gu [the Agile Production System of Hyundai Motor Company]. Seoul: Hanoul Academy.

23.

Jeong

Kim

(2016) Unpacking the ‘black box’ of a Korean big fast follower: Hyundai motor company’s engineer-led production system. Asian Journal of Technology Innovation 24(sup1): 53–77.

24.

Kalinowski

(2019) Why International Cooperation Is Failing: How the Clash of Capitalisms Undermines the Regulation of Finance. Oxford, UK: Oxford University Press.

25.

Kim

(1997) Imitation to Innovation: The Dynamics of Korea's Technological Learning. Boston, MA: Harvard Business School Press.

26.

Kim

(2008) Han-Gug U-Lyang-Gi-Eob-Ui No-Sa-Gwan-Gye DNA [Labour Relations DNA of Korean Blue-Chip Companies]. Seoul: Bakyoungsa.

27.

Kim

(2010) Dae-Gi-Eob Seong-Jang-Gwa No-Dong-Ui Bul-An-Jeong-Hwa [the Growth of Large Corporations and the Precarisation of Labour]. Seoul: Baeksan Seodang.

28.

Kim

D-O

Bae

(2004) Employment Relations and Hrm in South Korea. Aldershot, England; Burlington, VT: Ashgate.

29.

Kim

Y-M

Kwon

(2020) Categorical matching as an organizational condition for gender inequality in the Korean labor market. Sociological Perspectives 63(1): 29–49.

30.

Kong

(2012) Cooperation in unlikely settings: the rise of cooperative labor relations among leading South Korean firms. Politics & Society 40(3): 425–452.

31.

Kong

(2013) Between late-industrialisation and globalisation: the hybridisation of labour relations among leading South Korean firms. New Political Economy 18(5): 625–652.

32.

Kwon

Doellgast

(2018) Women, employment and gender inequality in South Korea. In: Lee

Y-M

Kaufman

(eds) The Evolution of Korean Industrial and Employment Relations. Cheltenham, UK: Edward Elgar Publishing.

33.

Kwon

O'Donnell

(2003) The Chaebol and Labour in Korea: The Development of Management Strategy in Hyundai. London; New York: Routledge.

34.

Lee

(1998) Micro-corporatism in South Korea: a comparative analysis of enterprise-level industrial relations. Economic and Industrial Democracy 19(3): 443–474.

35.

Lee

(2011) Labour market risks in de-industrializing East Asian economies: the cases of Korea, Japan and taiwan. In: Hwang

G-J

(ed) New Welfare States in East Asia: Global Challenges and Restructuring. Cheltham, UK: Edward Elgar.

36.

Lee

(2023) Same boat, different destiny: subcontracted workers in the Korean shipbuilding industry. In: Varieties of Precarity: Melting Labour and the Failure to Protect Workers in the Korean Welfare State. Bristol, UK: Policy Press, 76–97.

37.

Lee

SS-Y

Kim

(2020) Female outsiders in South Korea’s dual labour market: challenges of equal pay for work of equal value. Journal of Industrial Relations 62(4): 651–678.

38.

Lepak

Snell

(1999) The human resource architecture: toward a theory of human capital allocation and development. Academy of Management Review 24(1): 31.

39.

Miles

Snow

Meyer

, et al. (1978) Organizational Strategy, Structure and Process. New York, NY: McGraw-Hill.

40.

OECD (2020) OECD Economic Surveys: Korea 2020. Paris: OECD Publishing.

41.

Palier

Thelen

(2010) Institutionalizing dualism: complementarities and change in France and Germany. Politics & Society 38(1): 119–148.

42.

Park

(2014) Hyeon-Dae-Ja-Dong-Cha-E-Neun Han-Gug No-Sa-Gwan-Gye-Ga Iss-Da [Hyundai Motor Company Has Korea’s Industrial Relations]. Seoul: Maeil Nodong News.

43.

Park

(2018) Han-Gug Ja-Dong-Cha-Bu-Pum-San-Eob-Ui Gyeong-Jaeng-Lyeog-Bun-Seog-Gwa Dae-Eung-Bang-an [the Competitiveness Analysis of Korean Automobile Industry]. Seoul: IBK Economic Institute.

44.

Peng

(2012) Economic dualization in Japan and South Korea. In: Emmenegger

Häusermann

Palier

, et al. (eds) The Age of Dualization: The Changing Face of Inequality in Deindustrializing Societies. Oxford; New York: Oxford University Press.

45.

Pirie

(2006) The Korean Developmental State: From Dirigisme to Neo-Liberalism. London, UK: Routledge.

46.

Polavieja

(2005) Flexibility or polarization? Temporary employment and job tasks in Spain. Socio-Economic Review 3(2): 233–258. doi: 10.1093/SER/mwi010.

47.

Porter

(1985) Competitive Advantage: Creating and Sustaining Superior Performance. New York: London: Free Press ; Collier Macmillan.

48.

Purcell

(1999) Best practice and best fit: Chimera or Cul-De-Sac? Human Resource Management Journal 9(3): 26–41.

49.

Rème-Harnay

(2023) Precarity and subcontracting relationships: the case of parcel delivery drivers in France. Work, Employment & Society 0(0): 09500170221142721.

50.

Schmitter

Streeck

(1999) The Organization of Business Interests: Studying the Associative Action of Business in Advanced Industrial Societies. Cologne, Germany: MPIFG. Report 99/1.

51.

Seawright

(2016) Multi-Method Social Science: Combining Qualitative and Quantitative Tools. Cambridge, UK: Cambridge University Press.

52.

Seawright

Gerring

(2008) Case selection techniques in case study research: a menu of qualitative and quantitative options. Political Research Quarterly 61(2): 294–308.

53.

Shim

J-S

Lee

(2008) The Korean Economic System : Governments, Big Business and Financial Institutions. Aldershot, UK: Ashgate.

54.

Shin

(2001) Gi-Eob-Nae-Bu-No-Dong-Si-Jang-Ui Hyeong-Seong-Gwa Jeon-Gae: Han-Gug Jo-Seon San-Eob-E Dae-Han Sa-Lye Yeon-Gu [development and transformation of internal labour markets: the case study of shipbuilding industry in Korea]. PhD Dissertation, Department of Sociology. Seoul: Seoul National University.

55.

Shin

(2014) Unions and the adoption of high-performance work systems in Korea: moderating roles of firms’ competitive strategies. International Journal of Human Resource Management 25(13): 1858–1880.

56.

Song

(2000) Technological catching-up of Korea and Taiwan in the global semiconductor industry: a study of modes of technology sourcing. Discussion Paper Series. No 15, APEC Study Center, Columbia Business School.

57.

Song

(2014) Inequality in the Workplace: Labor Market Reform in Japan and Korea. Ithaca, NY: Cornell University Press.

58.

Song

(2014) The Samsung Way: Transformational Management Strategies from the World Leader in Innovation and Design. New York, NY: McGraw-Hill Education.

59.

Sørensen

(2000) Toward a sounder basis for class analysis. American Journal of Sociology 105(6): 1523–1558.

60.

Stavrou

Parry

Anderson

(2015) Nonstandard work arrangements and configurations of firm and societal systems. International Journal of Human Resource Management 26(19): 2412–2433.

61.

Streeck

(2011) Skills and Politics: General and Specific. Cologne: MPIFG, Reportno. 11/1.

62.

Streeck

Yamamura

(2001) The Origins of Nonliberal Capitalism: Germany and Japan in Comparison. Ithaca, NY: Cornell University Press.

63.

Teece

Pisano

Shuen

(1997) Dynamic capabilities and strategic management. Strategic Management Journal 18(7): 509–533.

64.

Thurbon

(2001) Two paths to financial liberalization: South Korea and Taiwan. The Pacific Review 14(2): 241–267.

65.

Wang

Tsai

(2010) National model of technological catching up and innovation: comparing patents of Taiwan and South Korea. Journal of Development Studies 46(8): 1404–1423.

66.

Whitley

(1999) Divergent Capitalisms: The Social Structuring and Change of Business Systems. Oxford; New York: Oxford University Press.

67.

Williams

(2012) Using the margins command to estimate and interpret adjusted predictions and marginal effects. STATA Journal: Promoting communications on statistics and Stata 12(2): 308–331.

68.

Wooldridge

(2002) Econometric Analysis of Cross Section and Panel Data. Cambridge, Mass: MIT Press.

69.

Wright

Dunford

Snell

(2001) Human resources and the resource based view of the firm. Journal of Management 27(6): 701–721.

70.

Yeung

(2016) Strategic Coupling: East Asian Industrial Transformation in the New Global Economy. Ithaca; London: Cornell University Press.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.90 MB