이승윤
Sdrntopeosf112 525t:a9ig3631aacg0t5u0c9s75utl2h 3g t80f152hA ·
[논문소개] "한국 불안정노동 계급의 윤곽" [Beyond the precariat: Trajectories of precarious work and its determinants in South Korea]
0.
수년을 공들인 논문이 몇일 전에 게재되었다. 한 논문을 둘러싸고 이렇게 많은 토론을 했던 연구가 있을까 싶다.
마지막 결론장에 과감한 주장을 타이핑하면서, 나의 그간의 연구활동이 모두 바로 이 연구에 이르기 위한 공부였던 것일까 생각했다. 혼자 가만히, 조용히 돌돌 돌아가는 선풍기를 쳐다보며 왠지 모르게 눈물이 차올랐던 기억이 난다.
그때 작성했던 부분이, “우리는 전통적인 계급이론과 현대 프레카리아트(불안정노동) 담론을 종합하는 새로운 계급이론틀을 만들어내야 하며, 현재와 미래의 노동시장 불평등을 분석하는데 있어 계급 논의를 다시 부활시켜야 한다고 주장한다" 이다.
[내용 간략 소개]
1.
이 연구는 한국의 불안정노동자들이 노동시장에서 보이는 패턴을 통해 계급의 윤곽을 그려내고 그 결정요인을 조사했다. 한국노동패널연구(KLIPS) 데이터(2002-2021)와 집단기반궤적모델링을 활용하여, 노동시장에 진입한 19-34세 청년들의 이후 20년간 노동시장 경험을 우리가 구성한 불안정노동지수(Precarious work Index)를 활용하여 추적 분석하였다.
2.
분석결과, 각각의 특징적인 노동시장 여정을 나타내는 뚜렷한 불안정노동 궤적이 나타났으며, 특히 샘플의 34.4%가 이 불안정노동 궤적에 속했다.
이는 단순 이분법적 이중노동시장론을 포함한 기존의 노동시장 분절론에 대한 논의를 확대시킨다. 이어 진행된 다항로지스틱 회귀분석을 통해 저숙련 서비스노동자와 여성이 불안정한 그룹에 속할 가능성이 더 높다고 나타났다.
무엇보다, 이 연구는 새로운 불안정노동자 계급이 형성되고 있음을 시사하며,이는 전통적인 직업 경계를 초월하고 기존의 계급이론에 도전을 한다. 새롭게 부상하는 불안정노동 계급의 윤곽은 그 이질성으로 특징지어지며, 하위 그룹들은 다양한 정도와 형태의 불안정성을 경험했다.
3.
이 연구는 계급형성의 물질적 기반을 이해하기 위한 강력한 경험적 기초를 제공하지만, 우리는 계급 형성이 경제적 조건을 넘어선다는 것을 잘 이해하고, 이 연구의 한계로 다루었다.
다음 연구는, 정치적 정체성과 계급 의식과 같은 주관적 요소를 통합해야 하며, 이는 프레카리아트를 '즉자적 계급'에서 '대자적 계급'으로 변화시키는 데 중요하다. 이 과정에서 노동조합의 역할도 추가 조사가 필요할 것이다.
*한국 가톨릭대학교 지원으로 open access 논문이라, 누구나 다운 받으실 수 있습니다.
All reactions:438Yi San, Paul Ma and 436 others
54 comments
Author이승윤https://onlinelibrary.wiley.com/doi/10.1111/ijsw.12694
ONLINELIBRARY.WILEY.COM
Beyond the precariat: Trajectories of precarious work and its determinants in South KoreaBeyond the precariat: Trajectories of precarious work and its determinants in South Korea
4 d
4 days ago
Like
Reply
6
정승국
축하합니다.제 배열분석 논문과 약간 유사한 점이 있지만 훨씬 야심찬 기획을 가진 논문이군요. 읽어보겠습니다.
=====
Beyond the precariat: Trajectories of precarious work and its determinants in South Korea
Abstract
This study examines the evolving patterns of precarious work in South Korea over a 20-year period, challenging conventional understandings of labour market segmentation and class formation. Using group-based trajectory modelling on Korean Labour and Income Panel Study data (2002–2021), we identify five distinct trajectories of labour market experiences. Our findings reveal that 34.4% of the sample, categorised into persistently high precariousness and persistent precariousness groups, provides evidence for an emerging, heterogeneous precariat class. This new class formation transcends traditional occupational boundaries and challenges existing class theories. Multinomial logistic regression shows that low-skilled service workers and women are more likely to belong to precarious groups. Our study contributes by (1) providing empirical evidence for a diverse precariat class, (2) demonstrating the need for a multidimensional approach to precarious work, and (3) highlighting the dynamic nature of precarious work over time. We argue for bringing class discussion back into labour market inequality analysis through a new theoretical framework that synthesises traditional class theory with contemporary precariat discourse, offering insights for policy and theoretical development in understanding modern labour markets.
Abbreviations
- APPA
- average posterior probability of assignment
- BCW
- blue-collar worker
- BIC
- Bayesian information criterion
- CA
- capital accumulator
- GBTM
- group-based trajectory modelling
- IP
- increasing precariousness
- KLIPS
- Korean Labour & Income Panel Study
- LSF
- low-service functionary
- MSF
- mixed-service functionary
- PHP
- persistently high precarious
- PLP
- persistently low precarious
- PP
- persistently precarious
- PWI
- precarious work index
- SCP
- sociocultural professional
- SD
- steady decline
INTRODUCTION
This study examines the evolving patterns of precarious work and its determinants in South Korea over a 20-year period, with two primary objectives. First, we aim to identify latent groups with similar trajectories of labour market experiences from a longitudinal perspective, investigating whether a process of class formation based on precarity is occurring. Second, we analyse the relationships between these precarious trajectory groups, traditional occupational classes, and demographic factors, providing new insights into the precariat debate.
Precarious work has been extensively researched, focusing on its prevalence, characteristics, causes, and effects (Kalleberg, 2009; Rodgers & Rodgers, 1989; Vosko et al., 2009). It inherently involves a range of insecure, unstable, and unprotected job conditions that have implications for occupational classification, gender inequality, health, and welfare (Baek, 2014; Eichhorst & Marx, 2015; Johansson, 2018). Research in this area has traditionally focused on labour market dualism, bifurcating the workforce into secure and insecure employment, usually contrasting standard with non-standard jobs (Doeringer & Piore, 1971; Emmenegger et al., 2012; Reich et al., 1973).
Recent studies advocate for moving beyond this traditional dual segmentation towards a paradigm that includes various precarious employment patterns (Seo, 2021; Yoon & Chung, 2016). In South Korea (hereafter, Korea), research has confirmed multiple labour market segments (Ko, 2019; Lee, 2017), focusing on segmentation between standard and non-standard forms of work (Jang, 2017; Kim, 2022) and by company size (Jeong, 2007). Recent attention has shifted towards a comprehensive analysis encompassing company size, employment contract forms, union affiliation, and social security inclusion (Jeon, 2016; Ko, 2019; Lee, 2017), highlighting the multifaceted nature of precariousness in the Korean labour market. While these contributions are pivotal, they fall short of exploring the persistence of multidimensional precariousness over time or delineating distinct precarious work trajectories.
Standing's (2011) conceptualisation of the precariat as a new socio-economic class has sparked significant debate. While Standing argues that the precariat's unique conditions distinguish it from the traditional working class, critics like Wright (2016a,b) call for empirical evidence to substantiate this claim. Our study responds to this debate by investigating whether the stratification of precarious work can be understood through the lens of traditional class structures defined by occupational profiles (Erikson & Goldthorpe, 1993; Oesch, 2006; Wright, 1985).
To examine Standing's thesis and address its critiques, we employ a two-pronged analytical strategy. First, we develop and utilise a precarious work index (PWI) for group-based trajectory modelling (GBTM), investigating potential class formations that share long-term trajectories of labour market experience and precariousness. Second, through multinomial logistic regression, we analyse how these trajectories align with or diverge from traditional occupational class categories and other characteristics of the precariat identified by Standing.
Our data are drawn from the Korean Labour & Income Panel Study (KLIPS), spanning 2002 to 2021, the period following Korea's recovery from the 1997 financial crisis. The Korean labour market, characterised by a large proportion of non-standard and atypical employment arrangements (Lee, 2017; Lee et al., 2017), provides an ideal case for examining shifts in labour dynamics since the country's economic recovery in 2002.
This approach allows us to empirically examine the existence and composition of precariat groups while accounting for their potential diversity. By integrating socioeconomic factors, we present a nuanced picture of the influences shaping various work trajectories, contributing to the discourse on class stratification and the layered nature of precarious work within the evolving framework of the precariat. Our study advances the literature on precarious work by focusing on its longitudinal trajectories and extending the discourse surrounding the precariat through grounded empirical analysis. The findings offer crucial insights for both theoretical development and policy formulation in addressing labour market inequality and the changing nature of work in contemporary societies.
In the following sections, we first discuss defining precarious work with a multidimensional approach and then explore the shift from static labour market dualism to dynamic segmentation. We review the debate on the precariat and its critiques, examining the longitudinal dynamics of precarious work and potential class structure. In doing so, we engage with Standing's (2011) concept of the precariat and explore the relationships between precarious work trajectories and traditional class structures, gender, age, and other factors. Next, we describe our data, methods, and variables, including the construction of the PWI and the application of GBTM and multinomial logistic regression. Finally, we present our analysis and results, identifying distinct precarious work trajectories and their determinants, followed by a discussion of the findings and their implications.
PRECARIOUS WORK AND LONGITUDINAL DYNAMICS OF LABOUR MARKET SEGMENTATION
Defining precarious work with a multidimensional approach
Precarious work generally embodies uncertainty and unpredictability (Kalleberg, 2009), instability, lack of protection, insecurity and social or economic vulnerability (Rodgers & Rodgers, 1989). The scholarly consensus usually identifies three dimensions of precarious work: employment insecurity, income inadequacy and lack of rights and protection (Kalleberg, 2011; Kreshpaj et al., 2020; Lee et al., 2017; Murillo-Huertas et al., 2022; Olsthoorn, 2014; Rodgers & Rodgers, 1989; Seo, 2021; Yoon & Chung, 2016). This approach acknowledges the complexity of precarious work and provides us with a precise definition while focusing on the concrete and measurable aspects of precarious employment.
Additionally, some studies posit that subjective insecurity related to psychosocial work environments is another dimension (Seo, 2021); however, it is arguably more of a cognitive reaction resulting from precarious conditions rather than a defining attribute (Kreshpaj et al., 2020). Hence, this study did not consider subjective precariousness as an intrinsic dimension of precarious labour. Union membership could be considered another factor influencing job security; however, because of the very low rates of unionisation in Korea, it was excluded from this analysis.
Assigning appropriate weights to each precarious work dimension is critical. Equal weighting, commonly used in constructing composite indices, may overlook nuanced differences in significance (Kalleberg, 2000; Lee et al., 2017; Padrosa et al., 2021). Although equal weighting suggests a uniform approach to weighting, it can inadvertently result in a disproportionate emphasis on dimensions with more variables, thus distorting the structure of the composite index. Furthermore, equal weighting can introduce redundancy through double counting when highly correlated variables are combined, thereby indicating the necessity of a mechanism to identify and mitigate such overlap (Nardo et al., 2008:31). However, while adjusting weights to reflect the statistical robustness and coverage of data might seem advantageous, this method risks a bias towards more accessible indicators and undervalues data that are difficult to collect and quantify. In this context, recent studies have moved towards a hedonic weighting approach that links objective precarious work indicators with subjective well-being measured by self-reported life satisfaction (García-Pérez et al., 2020; Lee et al., 2022; Murillo-Huertas et al., 2022).
Hedonic weights have potential advantages in multidimensional precarious work assessments. However, considering that hedonic well-being is more validly measured through preference than self-reported life satisfaction, and given the distinction between preference and hedonic well-being (Fleurbaey et al., 2009), the application of hedonic weights necessitates careful consideration. Moreover, there is no gold standard for the application of weighting methods.
Therefore, this study employed factor loadings from confirmatory factor analysis as the basis for weighting. This approach not only enhances the robustness of our weighting methodology but also ensures that the weights are representative of the relative importance and interrelationships between the dimensions of precarious work. This methodological choice is particularly advantageous in socioeconomic research, as it allows for a nuanced and statistically sound understanding of complex multidimensional constructs (Nardo et al., 2008).
This study conducted an empirical investigation of the longitudinal patterns of precarious work in the Korean labour market. To measure the multidimensions of precarious work, we synthesised insights from recent studies that go beyond the dichotomy of standard and non-standard employment. The PWI, explained in the Data, Methods and Variables section, is informed by these insights and serves as a sophisticated tool for tracking and analysing the multifaceted nature of precarious work over an extended period.
From static labour market dualism to dynamic segmentation
Labour market segmentation theory argues that the labour market does not function as a single unified market but is divided into several sub-markets (Doeringer & Piore, 1971). These sub-markets are typically categorised into primary and secondary labour markets. The primary labour market is characterised by high wages and job stability, while the secondary labour market is marked by unstable employment and low wages. This labour market segmentation theory has led to a trend of classifying labour market precarity in a dichotomous form: standard employment for regular workers and non-standard employment for non-regular workers.
The dualisation approach, which divides the workforce into protected ‘insiders’ and unprotected ‘outsiders’, further develops the discussion on labour market dualism (Eichhorst & Marx, 2021; Emmenegger et al., 2012; Palier & Thelen, 2010). This institutionalised labour market dualisation often appears to be a fixed equilibrium, reinforced by insider politics and the strategic alignment with political entities like the Social Democratic Party, which has been instrumental in maintaining this dichotomy (Emmenegger et al., 2012). The inclusiveness of interest associations and the state's mediation ability in forming coalitions are also pivotal in perpetuating this stability (Thelen, 2014). Rueda (2005) further argues that the interaction between labour market institutions and social policies determines the degree of dualisation, with implications for the distribution of work-related risks and benefits.
This approach has merit in explaining how labour market precarity is structured and stratified, and why certain groups find themselves in precarious working conditions. However, existing research grounded in dual labour market theory has predominantly focused on the binary distinction between insiders and outsiders. This binary view fails to capture the complex and nuanced realities of contemporary labour markets, where various degrees and forms of precarity exist beyond a simple insider-outsider divide.
Recent scholarship has challenged the static view of labour market dualisation, recognising its dynamic nature in response to socioeconomic pressures, policy shifts, and evolving labour market strategies (Häusermann & Schwander, 2009; Palier & Thelen, 2010). This ‘dualism-as-dynamic’ approach acknowledges the multifaceted and heterogeneous nature of precarious work arising from labour market segmentation (Eichhorst & Marx, 2021). For instance, Seo (2021) identifies diverse segments within European labour markets, each with distinct characteristics and varying levels of perceived job security. Similarly, Yoon and Chung (2016) propose a ‘future insecure’ category in their analysis of the UK labour market, emphasising the importance of factors such as income disparity and social security provisions in differentiating labour market segments.
These studies suggest that precarious work itself may be stratified in various ways. However, while they have primarily focused on the phenomenon of labour market segmentation, they have paid little attention to the potential formation of class structures among precarious workers resulting from this segmentation. Furthermore, existing research is limited by its reliance on cross-sectional analyses. This approach fails to capture whether the diverse stratification of precariousness becomes entrenched over time from a longitudinal perspective. Consequently, these studies are unable to fully grasp the dynamic and evolving nature of precarious work across extended periods and the possibility of new class formations emerging.
Studies on labour market instability in Korea have traditionally been understood through dual labour market theory, focusing on the distinction between regular and non-regular employment contract types (Jang, 2017; Ko, 2019; Lee, 2016; Lee, 2017). These studies have primarily concentrated on identifying the phenomenon of labour market segmentation in Korea. Our study, however, examines the multiplicity of patterns within the labour market and re-examines the static view of dual labour markets by adopting a dynamic long-term perspective. This approach allows us to analyse labour market experiences that capture the evolution of precarious work trajectories in relation to labour market segmentation in Korea.
By adopting a longitudinal approach, we aim to capture not only the existence of labour market segments but also how new forms of stratification of precarious work emerge. This approach allows us to engage more deeply with Standing's concept of the precariat, examining whether it manifests as a distinct class or as a more complex, layered phenomenon within the Korean labour market. Our analysis thus contributes to a more nuanced understanding of precarious work, moving beyond static categorisations to explore the diverse trajectories and experiences that characterise contemporary labour markets.
Longitudinal dynamics of precarious work and the class structure
As discussed in the above section, if precarious work experiences are stratified in various ways, then how does this stratification relate to Standing's concept of the precariat as a new class theory and to the traditional class theory?
Standing's seminal work, ‘The Precariat: The New Dangerous Class’ (2011), introduces the precariat as a new socio-economic class navigating a landscape of informal, unstable, or low-paid work. The term ‘precariat’—a portmanteau of ‘precarious’ and ‘proletariat’—represents a group that challenges traditional class narratives and entrenched social structures. Standing (2011) argues that a distinctive set of precarious conditions such as low income, insecure employment and limited social protection, and so forth, sets the precariat apart from the traditional working class. He identifies several demographic groups as particularly susceptible to precarious conditions. Young people often face unstable employment and underutilisation of their skills, and women are disproportionately represented in part-time, temporary, and low-wage jobs, increasing their vulnerability to precarious employment. Elderly workers are frequently pushed out of stable jobs into insecure employment situations, lacking job security and adequate income. Highlighting the political alienation, also migrants and ethnic minorities are suggested as a precarious group experiencing insecure, low-paying jobs with limited legal protections or social security.
The conceptualisation of the precariat as a distinct class has sparked significant debate. While Standing contends that the precariat's unique set of material conditions distinguishes it from the traditional working class, critics like Wright (2015) argue that the precariat is a significant but not distinct part of the working class, emphasising the need for empirical evidence to substantiate the claim of the precariat as a separate class. This ongoing debate underscores the complexity of modern labour markets and the challenges in categorising emerging socio-economic groups, highlighting the need for empirical studies to examine the formation and characteristics of potential precariat groups within specific contexts.
In this study, we investigate whether the stratification of precarious work, identified as the precariat by Standing (2011), can be comprehensively understood through the lens of traditional class structures that are often defined by occupational profiles (Erikson & Goldthorpe, 1993; Oesch, 2006; Wright, 1997). Wright's (1997) analysis of traditional class structures centres on occupation as a determinant of class position. In discussing the role of occupation in class delineation, we incorporate Oesch's (2006, 2008) classification of ‘work logic’ and ‘skill level’, which is particularly pertinent, given the service sector's growing impact on class differentiation. Drawing on the methodologies of Schwander and Häusermann (2013), we categorised five occupational groups, each with distinct socioeconomic characteristics and associated security. These five occupational groups are capital accumulators (CAs), sociocultural professionals (SCPs), blue-collar workers (BCWs), low-service functionaries (LSFs) and mixed-service functionaries (MSFs).
While occupational boundaries may play a significant role in shaping precarious work experiences, research has shown that precariousness is also closely intertwined with demographic factors such as age, gender, education, and household income (Kalleberg, 2009). Seo (2021) and Yoon and Chung (2016) highlighted factors such as income disparity and social security provisions, which are critical for differentiating labour market segments. These studies suggest that occupational roles and the relative security they offer play crucial roles in shaping workers' experiences of precariousness. Furthermore, demographic factors such as gender and age have been identified as significant determinants of precarious work experiences (Kalleberg, 2009; Lee, 2017; Lee et al., 2022; Puig-Barrachina et al., 2014; Sapkal & Sundar, 2017). Gender has been identified as a critical determinant, with empirical evidence showing that women disproportionately experience higher rates of precarious work at higher rates (Häusermann & Schwander, 2009; Lee, 2017; Lee et al., 2017; Lee et al., 2022; Puig-Barrachina et al., 2014; Sapkal & Sundar, 2017). Youths are also widely recognised as major labour market outsiders, constrained from entering the labour market because of a decline in quality jobs and the prevalence of hiring practices that favour experienced workers (Häusermann & Schwander, 2009; Ko, 2019). Additionally, elderly people occupy a significantly weaker position in the labour market (Choi, 2018; Lee & Yang, 2017). Furthermore, many older adults in Korea are compelled to work because of inadequate retirement income.
In sum, this study employs a two-pronged analytical strategy to address these debates and critiques of the Precariat. First, we investigate potential class formations that share long-term trajectories of precarious work experience, which may represent different groups withing the precariats. This approach allows us to empirically examine the existence and composition of precariat groups while accounting for their potential diversity. Second, we analyse how these trajectories align with or diverge from traditional occupational class categories and other characteristics of the precariat identified by Standing and others. This strategy enables us to test whether different types of precarious experience indeed form a distinct class, or whether they represent a more complex stratification that transcends traditional class boundaries.
DATA, METHODS AND VARIABLES
Data
The data used in this study were derived from the KLIPS, an annual panel survey of a representative sample of 5000 households and household members (aged 15 years and older) living in urban areas in Korea. The first survey was conducted in 1998, while the 26th survey was underway as of 2023. The KLIPS encompasses comprehensive details on household characteristics, economic activities, labour market participation and social welfare engagement, which makes it the preferred dataset for scrutinising precarious work in Korea (Korea Labour Institute, 2022).
This study followed a consistent cohort for two decades, from 2002, the year the Korean government officially announced recovery from the 1997 economic crisis, until 2021. The 1997 economic crisis significantly affected the Korean labour market, making the post-crisis period relevant for assessing precarious work trajectories.
The analysis focused on individuals aged 19–34 years at the beginning of the study period, in line with Korea's legal definition of youth, a demographic vulnerable to shifts in labour market conditions, and essential for understanding the long-term impact of the economic crisis on precarious work in South Korea.
As the study ended in 2021, the participants originally identified in 2002 were aged 38–53 years. The analysis included employed individuals, those actively seeking work (unemployed), and the extended category of ‘potentially employable’, which covered the long-term unemployed (defined by a period of joblessness exceeding 6 months) and individuals marginally attached to the labour market. The latter group included discouraged workers who were not actively seeking employment due to various barriers or lack of success in finding work, but who still desired to work. Those who had completely disengaged from the labour market were excluded from the analysis.
After constructing the PWI for this sample, individuals with more than 5 years of missing values for the PWI score during the analysis period were excluded. The final analytical cohort comprised 1135 individuals who met these conditions and were consistently surveyed from 2002 to 2021.1
Methods and variables
The following analytical procedures were employed to examine the longitudinal trajectories and determinants of precarious work: The first step involved constructing the PWI, incorporating the elements of employment instability, income inadequacy, and lack of social protection (Baek, 2014; Lee et al., 2017). Next, GBTM was conducted to assess the longitudinal trajectories of PWI scores, aiming to identify distinct groups of individuals who showed similar patterns in their experience of in the labour market. Finally, a multinomial logistic regression analysis was conducted using the classifications obtained from the GBTM analysis as the dependent variable.
Constructing the PWI
The following steps were followed to construct the PWI using the index construction approach described by Nicoletti et al. (1999). The first step involved measuring three dimensions of precarious work: employment instability, income inadequacy and lack of social protection. Each of these three dimensions was classified as precarious if it met any of the criteria indicative of a precarious condition, shown in Table 1.
| Dimension | Indicator | Operational definition |
|---|---|---|
| Employment instability | Contractual relationship insecurity | <Wage worker> Fixed term: contract terms shorter than 1 year or not fixed but shorter than 1 year (= 1) Agency work, on-call work, dispatched work, dependent self-employment, domestic work (= 1) Part-time work (= 1) Otherwise (= 0) |
| Work insecurity | <Non-wage worker> Employers with <5 paid employees or self-employed without employees or unpaid family workers (= 1) Otherwise (= 0) | |
<Others> Unemployed (ILO standards) (= 1) Long-term unemployed (unemployed for more than 6 months) (= 1) Potentially job seekers classified as non-economically active population (= 1) Otherwise (= 0) | ||
| Income inadequacy | Low income | Wage worker: Less than either the median hourly wage (= 1) Non-wage worker: Two-thirds of the median income (= 1) Otherwise (= 0) |
| Lack of social protection | Low coverage of social insurance | Not enrolled in any of the following: public pension, employment insurance, worker's compensation, or national health insurance system (= 1) Otherwise (= 0) |
- Note: The operational definitions for the dummy variables in this study are as follows: a score of 1 indicates a precarious status, whereas a score of 0 indicates a non-precarious status. Any condition not explicitly classified as 1 was considered non-precarious.
In the assessment of employment instability, a preliminary differentiation was made among wage earners, non-wage earners and the unemployed. For wage earners, precariousness was evaluated based on the instability and temporariness of their employment contracts. Consequently, wage earners engaged in fixed-term contracts, agency work and part-time employment were categorised as experiencing precarious work conditions. For non-wage earners, employment instability was determined by small-scale employers with less than five employees, self-employed individuals without staff and unpaid family workers. Furthermore, the unemployed, long-term unemployed and potentially employable job seekers were classified under employment instability conditions.
Income inadequacy was measured based on the criteria for low wages set by the International Labour Organization. This assessment included both wage and non-wage earners who received less than two-thirds of the median hourly wage or income. Lastly, the lack of social protection was measured by the lack of enrolment in social insurance programmes, which is a pivotal component of Korea's social protection framework.
The second step involved constructing the weights for the three dimensions based on the principal factor method to extract common factors. In the development of the PWI, we assigned tailored weights to each of the three dimensions of precarious work: employment instability, income inadequacy and lack of social protection. We conducted confirmatory factor analysis to derive factor loadings and establish a robust and representative weighting system. This method allowed us to capture the relative importance and interconnections among the dimensions of precarious work, thus aligning with best practices for handling complex multidimensional constructs in socioeconomic research (Nardo et al., 2008).
The weights were obtained by squaring and normalising the estimated factor loadings for each of the three dimensions. The normalised squared loadings represent the proportion of the total variance of the PWI which is explained by each dimension (European Union & Joint Research Centre, 2008:90). These weights were carefully calibrated so that their total was equal to 3, providing a consistent basis for comparison across the PWI scale.
Finally, we allocated specific weights to each of the three dimensions and subsequently aggregated them to construct the PWI. Consequently, the PWI score ranged from 0 to 3, with a value of 0 indicating that all three dimensions of precarious work were absent. A PWI score greater than 0 was derived from various combinations of the three weighted dimensions (see Table 2).
| Year | Employment insecurity | Income inadequacy | Lack of social protection | Year | Employment insecurity | Income inadequacy | Lack of social protection |
|---|---|---|---|---|---|---|---|
| 2002 | 0.92 | 1.20 | 0.87 | 2012 | 1.07 | 0.97 | 0.96 |
| 2003 | 1.10 | 1.08 | 0.82 | 2013 | 1.19 | 0.83 | 0.99 |
| 2004 | 0.86 | 1.10 | 1.04 | 2014 | 1.16 | 0.91 | 0.93 |
| 2005 | 0.89 | 1.08 | 1.03 | 2015 | 1.19 | 0.77 | 1.05 |
| 2006 | 1.12 | 1.07 | 0.81 | 2016 | 1.20 | 0.93 | 0.88 |
| 2007 | 1.18 | 1.00 | 0.83 | 2017 | 1.11 | 0.93 | 0.96 |
| 2008 | 1.23 | 0.82 | 0.95 | 2018 | 1.06 | 0.93 | 1.01 |
| 2009 | 0.98 | 1.08 | 0.95 | 2019 | 0.88 | 1.07 | 1.06 |
| 2010 | 1.13 | 0.91 | 0.96 | 2020 | 0.99 | 0.91 | 1.11 |
| 2011 | 1.12 | 0.99 | 0.89 | 2021 | 0.87 | 1.07 | 1.06 |
In this study, we considered a PWI score above the annual average to be indicative of precarity. This threshold was determined based on the mean PWI score for each year; individuals with scores higher than the annual average indicates the group experiencing precarity.
GBTM and multinomial logistic regression
In this 20-year longitudinal study, we applied GBTM to identify the longitudinal trajectories of precarious work. The GBTM operates on the premise that the population is composed of distinct latent trajectory groups, which are a mixture of multiple distinct paths, as posited by Nagin et al. (2018). This approach is pivotal for delineating distinct clusters of individuals who exhibit analogous patterns of precarious work over the study period and enables an in-depth analysis of the characteristics notably associated with these groups (Nagin, 2005).
To estimate the trajectory function and determine the number of trajectory groups, we followed the procedural framework established by Nagin (2005). We began by fitting a single-trajectory model to the PWI score using a censored normal distribution, given the composition of the index as a composite measure. Our approach commenced with a third-order (cubic) polynomial to explore potential non-linear trends over an extensive period, and subsequently tested the quadratic and linear forms. This stepwise methodology was employed to ensure the most accurate representation of the PWI trajectory by transitioning to simpler polynomial forms based on the statistical significance of each component.
Our analysis involved refining the model by adding trajectory groups when their polynomial components were statistically significant (p < 0.05), ensuring that the complex changes in precarious work conditions over the past two decades were accurately represented. The Bayesian information criterion (BIC) was used to determine the best model fit for the GBTM of PWI. Model expansion stopped when the BIC improvements were marginal, indicating the capture of all distinct data patterns (Fraley & Raftery, 1998). We also monitored the model's entropy and average posterior probabilities, halting additional groups when entropy changes were minor, or when average posterior probabilities were below 0.7. These combined statistical criteria ensured that our model was both parsimonious and representative of the distinctive features in the PWI data over the 20-year period.
After identifying discrete groups using the GBTM, multinomial logistic regression analysis was applied, with the classification of group membership serving as the outcome of interest. Multinomial logistic regression facilitates the examination of datasets with one or more explanatory variables that predict an outcome, thus enabling us to inquire how different factors are related to the probability of belonging to each trajectory group delineated by the GBTM.
Occupational groups were reclassified into the following five categories. CAs include individuals such as senior executives, high-level managers and major business owners who accumulate significant capital through their positions. SCPs include teachers, healthcare professionals and others that typically require a high level of education in their profession and contribute to society's social and cultural capital. BCWs include skilled and unskilled manual workers involved in the physical aspects of production (e.g., factory workers, construction workers and mechanics). LSFs include customer service agents, retail clerks and hospitality personnel, who provide direct services to consumers. Lastly, MSFs include service workers who perform both technical and personal services, such as IT support staff, administrative assistants, and roles that blend services with other functions.
Additionally, control variables such as gender, age, years of education and equivalised disposable household income were included in the multinomial logistic regression. This approach fosters a comprehensive understanding of the determinants that shape individual paths in the context of the dynamic nature of precarious work. The values for age and years of education from 2002 were included in the analysis. Similarly, we also included household income at the year of 2002 to account for the potential buffering effect of family resources on individual labour market experiences. Although firm size is an important factor in the Korean labour market, it was not included in this analysis due to data limitations and our focus on individual-level trajectories across various employment types, including self-employment and unemployment. The occupational class was analysed based on the most frequent occupational class during this period.
ANALYSIS AND RESULTS
PWI and trajectories of precarious work
In our study on the trajectories of labour market experience measured by PWI, we conducted GBTM and determined a five-group solution to be the most suitable based on its strong classification accuracy. The result was indicated by an average posterior probability of assignment (APPA) of 0.94 and high entropy (0.91). While the BIC continued to improve with the addition of more groups, the six-group model presented no practically distinct features and showed a slight decrease in classification precision. Moreover, the six-group model resulted in diminished entropy (0.87) and an APPA of 0.91. Consequently, we determined that the five groups offered the most parsimonious and interpretable model, striking an optimal balance between the statistical indicators and practical significance of the groups.
As illustrated in Figure 1, our analysis identified five distinct groups, each characterised by unique trajectories of labour market experience over a 20-year period. These trajectories shed light on the changing dynamics of precarious work.
The increasing precariousness (IP) group, accounting for 9.7% of the sample, starts with a low level of precariousness but experiences a gradual and consistent increase in precarity over time. This group represents a concerning trend of growing labour market instability for a notable portion of the workforce.
In contrast, the steady decline (SD) and persistently low precarious (PLP) groups, which we consider relatively non-precarious, collectively comprise 55.9% of the sample. The SD group, showing a consistent downward trend in precariousness over the study period, represents 17.6% of the sample. The PLP group, the largest among all groups at 38.3% of the sample, maintains a consistently low level of precariousness throughout the study duration.
These findings highlight the diverse experiences of precarity within the labour market, ranging from persistent high precarity to steady improvements. The substantial proportion of individuals experiencing persistent precarity (34.4% combined for PHP and PP) underscores the significant challenges faced by a large segment of the workforce. Conversely, the majority (55.9%) experiencing low or declining precarity suggests the potential for positive trajectories in labour market experiences for some workers.
Table 3 presents the annual average PWI scores for the five trajectory groups compared to the overall sample average, offering insights into the varying degrees of precarity experienced by different workforce segments over two decades.
| Year | Total | PHP | PP | IP | SD | PLP |
|---|---|---|---|---|---|---|
| 2002 | 1.22 (1.01) | 2.04 (0.97) | 1.60 (0.88) | 0.61 (0.80) | 1.76 (0.96) | 0.78 (0.88) |
| 2003 | 1.06 (0.98) | 2.05 (0.98) | 1.52 (0.89) | 0.56 (0.72) | 1.52 (0.93) | 0.58 (0.77) |
| 2004 | 1.04 (1.01) | 2.56 (0.69) | 1.54 (0.86) | 0.50 (0.71) | 1.51 (0.94) | 0.42 (0.67) |
| 2005 | 1.01 (1.02) | 2.50 (0.77) | 1.59 (0.81) | 0.35 (0.62) | 1.37 (0.98) | 0.36 (0.67) |
| 2006 | 0.95 (0.99) | 2.59 (0.60) | 1.49 (0.90) | 0.37 (0.60) | 1.20 (0.87) | 0.35 (0.61) |
| 2007 | 0.86 (0.97) | 2.73 (0.51) | 1.39 (0.82) | 0.41 (0.66) | 1.14 (0.91) | 0.24 (0.51) |
| 2008 | 0.81 (0.96) | 2.77 (0.46) | 1.38 (0.79) | 0.39 (0.60) | 0.94 (0.95) | 0.17 (0.40) |
| 2009 | 0.78 (0.98) | 2.66 (0.62) | 1.39 (0.83) | 0.54 (0.79) | 0.89 (0.90) | 0.08 (0.34) |
| 2010 | 0.78 (0.95) | 2.66 (0.62) | 1.47 (0.74) | 0.41 (0.57) | 0.76 (0.80) | 0.08 (0.31) |
| 2011 | 0.73 (0.91) | 2.72 (0.54) | 1.33 (0.70) | 0.58 (0.72) | 0.66 (0.74) | 0.05 (0.22) |
| 2012 | 0.81 (0.97) | 2.64 (0.66) | 1.49 (0.68) | 0.65 (0.83) | 0.68 (0.80) | 0.04 (0.25) |
| 2013 | 0.75 (0.95) | 2.61 (0.58) | 1.46 (0.76) | 0.56 (0.71) | 0.46 (0.68) | 0.03 (0.18) |
| 2014 | 0.76 (0.97) | 2.51 (0.90) | 1.54 (0.72) | 0.59 (0.72) | 0.46 (0.76) | 0.03 (0.20) |
| 2015 | 0.78 (0.98) | 2.41 (0.71) | 1.55 (0.77) | 0.85 (0.83) | 0.42 (0.74) | 0.03 (0.21) |
| 2016 | 0.73 (0.94) | 2.51 (0.76) | 1.44 (0.73) | 0.79 (0.73) | 0.22 (0.47) | 0.05 (0.23) |
| 2017 | 0.69 (0.91) | 2.44 (0.69) | 1.35 (0.71) | 0.97 (0.72) | 0.14 (0.39) | 0.07 (0.35) |
| 2018 | 0.68 (0.93) | 2.34 (0.76) | 1.32 (0.78) | 1.02 (0.89) | 0.11 (0.33) | 0.07 (0.36) |
| 2019 | 0.75 (0.97) | 2.21 (1.02) | 1.41 (0.79) | 1.32 (0.88) | 0.15 (0.45) | 0.06 (0.32) |
| 2020 | 0.78 (0.99) | 2.25 (0.96) | 1.42 (0.85) | 1.35 (0.85) | 0.15 (0.48) | 0.09 (0.37) |
| 2021 | 0.68 (0.93) | 2.17 (1.02) | 1.29 (0.78) | 1.17 (1.01) | 0.11 (0.34) | 0.04 (0.21) |
- Note: PWI scores in bold are higher than the annual average. Standard deviation statistics reported in parentheses.
The overall trend shows a decline in average PWI scores from 1.22 in 2002 to 0.68 in 2021, suggesting a general reduction in precarity levels. This decline likely reflects the expansion of social insurance coverage in Korea during this period. However, this trend masks significant variations among different groups, highlighting the complexity of precarious work experiences.
The PHP and PP groups, which we identify as the ‘Precarious groups’, consistently score above the average throughout the study period. The PHP group maintains scores more than double the average in most years (e.g., 2.77 in 2008 compared to the average of 0.81), underscoring the severity and persistence of their precarious conditions. The PP group, while less extreme, also shows consistently above-average scores (e.g., 1.60 in 2002, 1.29 in 2021). Notably, while the overall sample's PWI scores decrease over time, these two groups either maintain or increase their scores, indicating a widening gap in labour market experiences.
In contrast, the PLP group consistently scores well below the average, with their PWI score decreasing from 0.78 in 2002 to 0.04 in 2021, indicating increasingly stable and secure labour market experiences. This stark difference between the precarious groups and the PLP group points to growing polarisation and inequality in the labour market.
The IP group presents a unique trajectory, starting below average (0.61 in 2002) but surpassing it around 2015 (0.85) and reaching 1.17 by 2021, indicating a gradual deterioration of their labour market conditions. Conversely, the SD group begins above average (1.76 in 2002) but drops below it by 2009 (0.89), ending at 0.11 in 2021, suggesting significant improvement over time.
These diverse trajectories challenge Standing's (2011) concept of a uniform precariat class characterised by shared precarious experiences and collective identity. Our results reveal a spectrum of precarity levels, suggesting that the precariat may be more heterogeneous than previously conceptualised. The distinct trajectories of the PHP and PP groups, in particular, indicate that they may represent two separate classes rather than a single, homogeneous precariat.
Table 4 provides critical insights into the characteristics of the trajectory groups, illuminating the complex nature of precarious work experiences. The occupational distribution provides further nuance. Mixed service functionaries (MSFs) constitute a significant portion across all groups but are particularly prevalent in the PHP (41.0%) and PP (46.6%) groups. Low service functionaries (LSFs) are disproportionately represented in the PHP group (29.9%), indicating the vulnerability of lower-skilled service occupations to persistent precarity. CAs have minimal representation in the PHP group (4.0%) but are more prevalent in the PLP group (21.6%).
| Total (N = 1135) | 1. PHP (N = 62) | 2. PP (N = 328) | 3. IP (N = 110) | 4. SD (N = 200) | 5. PLP (N = 435) | ||
|---|---|---|---|---|---|---|---|
| Precarious work index score, mean (SD) | 2002 | 1.2 (1.0) | 2.0 (1.0) | 1.6 (0.9) | 0.6 (0.8) | 1.8 (1.0) | 0.8 (0.9) |
| 2021 | 0.7 (0.9) | 2.2 (1.0) | 1.3 (0.8) | 1.2 (1.0) | 0.1 (0.3) | 0.04 (0.2) | |
| Employment instability | 2002 | 0.4 (0.5) | 0.5 (0.5) | 0.5 (0.5) | 0.2 (0.4) | 0.5 (0.5) | 0.3 (0.4) |
| 2021 | 0.2 (0.4) | 0.7 (0.4) | 0.4 (0.4) | 0.4 (0.4) | 0.02 (0.1) | 0.02 (0.1) | |
| Income inadequacy | 2002 | 0.3 (0.5) | 0.7 (0.6) | 0.4 (0.6) | 0.1 (0.4) | 0.5 (0.6) | 0.2 (0.5) |
| 2021 | 0.1 (0.3) | 0.6 (0.5) | 0.2 (0.4) | 0.2 (0.5) | 0.01 (0.1) | 0.001 (0.03) | |
| Lack of social protection | 2002 | 0.5 (0.4) | 0.8 (0.2) | 0.7 (0.3) | 0.3 (0.4) | 0.7 (0.3) | 0.3 (0.4) |
| 2021 | 0.4 (0.5) | 0.9 (0.4) | 0.7 (0.5) | 0.6 (0.5) | 0.1 (0.3) | 0.03 (0.2) | |
| Most frequent occupation N (%) | |||||||
| Capital accumulator | 172 (15.2) | 2 (4.0) | 36 (11.0) | 14 (12.3) | 27 (13.4) | 93 (21.6) | |
| Mixed service functionary | 464 (41.1) | 25 (41.0) | 153 (46.6) | 45 (41.3) | 61 (30.6) | 180 (41.6) | |
| Blue collar workers | 273 (24.1) | 14 (24.0) | 77 (23.5) | 33 (30.2) | 60 (29.8) | 89 (20.6) | |
| Socio-cultural professional | 120 (10.6) | 1 (1.2) | 27 (8.3) | 11 (10.2) | 35 (17.9) | 46 (10.6) | |
| Low service functionary | 101 (8.9) | 19 (29.9) | 35 (10.7) | 7 (6.0) | 16 (8.2) | 24 (5.6) | |
| Gender (2002), N (%) | Male | 748 (66.0) | 21 (34.0) | 207 (63.4) | 89 (80.6) | 124 (62.4) | 307 (70.7) |
| Female | 385 (34.0) | 41 (66.0) | 120 (36.6) | 21 (19.4) | 75 (37.6) | 128 (29.3) | |
| Age, mean (SD) | 2002 | 28.1 (4.3) | 29.5 (4.0) | 29.1 (4.1) | 29.0 (3.9) | 27.3 (4.6) | 27.4 (4.2) |
| Household income (yearly, Korean 10,000 won), mean (SD) | 2002 | 2816 (1794) | 2208 (1488) | 2629 (1553) | 2813 (1993) | 2518 (1524) | 3178 (1987) |
| 2021 | 7386 (5350) | 4922 (2388) | 6163 (4755) | 6837 (3.407) | 6.925 (4221) | 9296 (6432) | |
| Education (year, 2002), mean (SD) | 13.4 (2.0) | 11.9 (2.5) | 12.8 (2.0) | 13.5 (2.0) | 13.4 (1.8) | 14.2 (1.8) | |
- Note: Occupation profiles were measured as the most frequent occupations held by each individual during the study period.
The gender composition across groups reveals significant patterns. The PHP group is predominantly female (66.0%), highlighting the gendered nature of severe precarity. In contrast, the PLP group has a higher proportion of males (70.7%), suggesting a gender-based disparity in labour market stability.
Socioeconomic factors play a crucial role in shaping precarity trajectories. The PHP group shows the lowest household income and least years of education. Conversely, the PLP group demonstrates the highest levels of both income and education. This pattern suggests that higher socioeconomic status and educational attainment serve as protective factors against precarious work. These findings paint a complex picture of precarious work. The interplay of gender, education, income, and occupation in shaping precarity trajectories highlights the multifaceted nature of labour market vulnerabilities.
Determinants of precarious work trajectories
Next, a multinomial logistic regression analysis was conducted, using the CA group as the reference category. Using the multinomial logistic regression result (Number of obs. 1121, LR chi2(32) = 280.69, p < 0.000, pseudo R2 = 0.09), Table 5 reports the average marginal effects of explanatory variables in terms of probability. The coefficients in the table can be interpreted as the probability of being in each group when the independent variable changes by one unit.
| PHP | PP | IP | SD | PLP | |
|---|---|---|---|---|---|
| Mixed service functionary (ref. CA) | 0.04+ (1.79) | 0.08+ (1.76) | 0.01 (0.34) | −0.05 (−1.36) | −0.07+ (−1.77) |
| Blue collar workers (ref. CA) | 0.02 (0.96) | −0.05 (−1.28) | 0.02 (0.62) | 0.02 (0.53) | −0.003 (−0.07) |
| Socio-cultural professional (ref. CA) | −0.02 (−0.85) | −0.02 (−0.29) | 0.05 (1.04) | 0.07 (1.35) | −0.09 (−1.53) |
| Low service functionary (ref. CA) | 0.06* (2.21) | 0.02 (0.32) | 0.01 (0.23) | −0.01 (−0.16) | −0.08 (−1.33) |
| Gender (ref. men) | 0.07*** (4.88) | 0.04 (1.33) | −0.07** (−2.98) | 0.03 (1.08) | −0.06+ (−1.92) |
| Age | 0.003** (2.15) | 0.01*** (4.50) | 0.003+ (1.76) | −0.004+ (−1.92) | −0.02*** (−5.00) |
| Education | −0.01*** (−3.76) | −0.05*** (−7.26) | −0.002 (−0.51) | 0.003 (0.44) | 0.06*** (8.27) |
| Log equivalised household income | −0.03*** (−2.84) | −0.02 (−1.11) | 0.01 (0.43) | −0.03+ (−1.87) | 0.08** (3.27) |
- Note: Z statistics reported in parentheses.
- + p < 0.1;
- * p < 0.05;
- ** p < 0.01;
- *** p < 0.001.
The analysis reveals a complex relationship between occupational categories and precarious work trajectories, challenging traditional class theory in modern labour markets. Notably, BCWs did not show a statistically significant likelihood of belonging to precarious trajectories (PHP and PP groups), contradicting the conventional assumption that working-class occupations are inherently precarious.
Instead, LSFs and MSFs emerge as most susceptible to precarious work paths. MSFs show a 4% increase in likelihood of placement within the PHP group, while LSFs are associated with a 6% rise. Importantly, LSFs have a stronger and more significant impact on PHP compared to PP at the 0.05 confidence level, suggesting their particular vulnerability to persistently high levels of precarity. This shift from traditional BCW to service-oriented occupations aligns with Standing's (2011) conceptualisation of the precariat as distinct from the traditional proletariat, reflecting the changing nature of labour markets with their flexible work arrangements and limited protections.
Gender emerges as a significant determinant, with women exhibiting a 7% higher likelihood of belonging to the PHP group and a 7% lower chance of being in the IP group. This extends Standing's argument about women constituting the precariat, suggesting that while women are more likely to experience persistent high precarity, those who initially enter stable employment are less likely to transition into precarious groups later.
Age emerges as a strong factor influencing precarity across different trajectory groups. According to Table 5, for each year increase in age, there is a 0.3% higher probability of belonging to the PHP group and a 1% higher probability of being in the PP group. Conversely, age is associated with a 2% lower probability of being in the PLP group. This pattern likely reflects the unique characteristics of the Korean labour market, particularly early retirement and subsequent re-entry into less secure employment. It offers a more complex view of age-related precarity than Standing's assertion that young people are a primary precariat demographic, suggesting instead that precarity can increase with age in certain contexts.
Education and household income prove to be significant protective factors against precarity. Each additional year of schooling decreases the probability of being in the PHP group by 1% and the PP group by 5%. An increase in log household income correlates with a 3% reduction in PHP probability and an 8% higher likelihood of PLP categorisation.
In conclusion, our analysis reveals that precarity is not uniform across the workforce, as illustrated by the distinction between PHP and PP groups. This supports Wright's (2016a, 2016b) call for a more nuanced understanding of class categories within the precariat framework. While our findings support Standing's identification of service sector workers as key components of the precariat, they also suggest a more heterogeneous precariat than initially proposed, with distinct sub-groups experiencing different levels and types of precarity. The complex interplay of occupational categories, demographic factors, and socioeconomic variables indicates that precarity operates on a spectrum rather than as a binary condition, challenging traditional notions of class and precarity.
DISCUSSION AND CONCLUSION
Discussion
This study investigated the trajectories and determinants of precarious work in Korea over a 20-year period, employing a two-pronged analytical strategy that yields significant insights into the nature of precarious work and the concept of the precariat. By integrating dimensions of employment instability, income inadequacy, and lack of social protection into the PWI, we expanded the understanding of precarious labour's complexity. Our use of GBTM to analyse longitudinal changes in PWI revealed distinct trajectories of precarious work experiences, offering a nuanced view of labour market segmentation and challenging existing theories.
Our findings both support and refine Standing's concept of the precariat. While we identify substantial groups experiencing persistent precarity (PHP and PP), aligning with the idea of a precariat class, the diversity of trajectories suggests that precarity experiences are more varied than a single class designation might imply. The identification of five distinct trajectories—PHP, PP, IP, SD, and PLP—illustrates the dynamic nature of labour market experiences and the complex stratification of precarious work, moving beyond static dualism.
Crucially, our research reveals that the precariat is not a monolithic entity as initially proposed by Standing, but rather a spectrum of precarious experiences shaped by various factors. The PHP and PP groups, in particular, cut across traditional occupational boundaries, representing a new form of stratification in the labour market. This finding challenges the argument of dual labour market segmentation and a uniform precariat class, indicating instead the existence of multiple precariat sub-classes with varying degrees of precarity.
Our multinomial logistic regression analysis provided further insights into the occupational and demographic characteristics associated with precarious work trajectories. Contrary to assumptions about blue-collar roles, we found that LSFs and MSFs more frequently navigate precarious pathways. This shift supports Standing's distinctions on the precariat while necessitating a re-evaluation of traditional class theories, including Wright's (2016a, 2016b) commentary that the precariat and the proletariat are essentially the same.
The analysis also revealed significant gender disparities within precarious work trajectories, particularly within the PHP group. Women's disproportionate representation in the most precarious segments underscores the need for gender-sensitive labour market policies. Additionally, our findings indicate that higher education and income act as buffers against precarity, highlighting the role of socioeconomic factors in shaping labour market outcomes.
In the context of the Korean labour market, known for its high proportion of non-standard employment relationships, our findings highlight the persistence of precarity even as workers progress through their careers. This suggests that labour market segmentation in Korea may be more entrenched than previously thought, with implications for policy interventions aimed at reducing inequality.
These findings collectively point to the emergence of a new, heterogeneous class of precarious workers that transcends traditional occupational and class boundaries. This emerging class formation challenges existing theoretical frameworks and calls for a renewed approach to class analysis in labour market studies. Our study contributes significantly to this theoretical landscape by providing empirical evidence for the complex, multifaceted nature of precarious work experiences.
Moving forward, there is a clear need for policy interventions that address the persistent precarity faced by vulnerable populations, including women and older workers. These interventions should be tailored to the diverse experiences of precarity revealed by our study, recognising that a one-size-fits-all approach may be insufficient to address the complex realities of the modern labour market.
Limitations
Our sample selection process, particularly the exclusion of cases with more than 5 years of missing PWI values, likely leads to an underestimation of precarity among women in our sample. This is because women are more likely to experience extended periods of economic inactivity, which are coded as missing values in our PWI construction. If we were to include these periods of pure economic inactivity, the level of precarity among women might be even higher than what our current analysis suggests. This limitation should be considered when interpreting our results, especially regarding gender differences in precarious work experiences. Our analysis, employing the PWI, provides a quantitative overview of precarious work yet may not capture the full scope of precarious experiences. This index's emphasis on quantifiable aspects risks neglecting important qualitative factors such as job satisfaction, which are integral to a thorough understanding of precarity (Kalleberg, 2009). The occupational classification used in this study warrants further scrutiny, as the correlation between occupational groups and precarious work trajectories appears less definitive than expected, indicating the need for additional research.
Additionally, our study acknowledges the gaps in addressing the subjective dimensions that contribute to the discourse on the precariat as a class (Wright, 2016a, 2016b). The notion of the precariat extends beyond material conditions, encompassing factors such as political identification, class consciousness (‘class-for-itself’), labour union involvement and mobilisation potential. However, our study only provides longitudinal empirical insights into the material basis of class through the lens of precarious work conditions and identifies different groups of precarious work.
Finally, the concept of the precariat within the Korean labour market context is another aspect that requires clarification. The interplay of occupation, gender, age and education within Korean society in relation to precarious work needs to be defined more explicitly through comparative research in the future.
Conclusion: Precarious classes in the making
Our study offers a renewed perspective on precarious work, advocating for a multidimensional framework that underscores the importance of income stability and social protection alongside the traditional facets of employment security. Examining South Korea's labour market through this lens reveals a variegated and evolving landscape of work that disrupts the perceived uniformity of the precariat, which was assumed by prior studies. This narrative is substantiated by our identification of distinct precarious work trajectories, each delineating a unique labour market journey, thus enriching the discourse on labour market segmentation.
Crucially, our findings suggest that a new class of precarious workers is in the making, one that transcends traditional occupational boundaries and challenges existing class theories. This emerging class is characterised by its heterogeneity, with subgroups experiencing varying degrees and forms of precarity. Our research underscores the need to bring class discussion back into labour market inequality analysis, but with a new theoretical framework that synthesises traditional class theory with contemporary precariat discourse.
This study contributes significantly to this emerging theoretical landscape. By identifying distinct trajectories of precarious work and their determinants, we provide empirical evidence for the formation of a complex, multifaceted precarious class. This new class formation is not a monolithic entity as initially proposed by Standing, but rather a spectrum of precarious experiences shaped by factors such as occupation, gender, age, and education.
Our research has pivotal implications for the current understanding of occupational classes and their relationship with precarious work. It suggests a more intricate nexus between occupational roles and employment precarity, compared with Standing's singular concept of the precariat (2011). Recognising the stratified and fluid nature of the labour market is crucial for understanding the complex realities of class in contemporary society.
These findings call for a critical reassessment of the precariat concept, considering the varied experiences that constitute precarious work and the policies that cater to this diversity. The insights from this study underscore the need for comprehensive policy approaches that reflect the multiplicity of the experiences of precarity, thus fostering a more nuanced understanding of class and inequality in the modern workforce.
While our study provides a robust empirical foundation for understanding the material basis of this emerging class, we acknowledge that class formation extends beyond economic conditions. Future research should incorporate subjective elements such as political identification and class consciousness, which are critical in transforming the precariat from a ‘class-in-itself’ to a ‘class-for-itself’. The role of labour unions in this process also warrants further investigation. This approach would allow for a more complete picture of the precariat and its implications for social and political change.
In conclusion, our findings contribute significantly to a broader understanding of precarious work and class formation in contemporary society. By providing empirical evidence for the emergence of a new, heterogeneous precarious class, we lay the groundwork for a renewed class analysis that can better capture the complexities of modern labour markets. This study serves as a stepping stone for future research that can further explore the subjective dimensions of class formation and the potential for collective action among precarious workers.
No comments:
Post a Comment