Abstract
Gig economy, characterized by its flexible and fragmented work patterns, has profoundly transformed traditional labor market by introducing novel forms of work and reshaping employment structures and ways of job fulfillment. Using data from the China General Social Survey (2015–2021), this study employs a Probit model to analyze the impact of gig economy development on individual employment. The results indicate that the gig economy development has a significant negative impact on the employment of working-age individuals, while medium education and substantial work experience can mitigate this negative effect. Furthermore, the study identifies the pivotal mechanism through which the gig economy influences the employment of working-age individuals, namely, the decrease in workers’ bargaining power. In contrast, the gig economy development positively impacts the employment of elderly workers, primarily due to their labor characteristics and accumulated work experience, which align well with the demand of gig market. Moreover, the study reveals that the value of formal and traditional education diminishes, while work experience becomes a more critical determinant of labor market success in gig economy. These findings offer important insights into the dynamics of the gig economy development and its implications for employment, serving as a valuable reference for corporate strategy and public policy formulation.
Introduction
Gig economy, with its unique flexibility and fragmented work model, is widely reshaping the contemporary labor market. The term “traditional gig economy” refers to an economic model centered on skilled, independent workers engaging in project-based work with a focus on outcomes (Zheng and Yang, 2019). Under the influence of digitalization, the integration of internet technologies has transformed and expanded the gig economy. This study defines it as the “new gig economy,” characterized by the economic activities of flexible workers in the digital era. These activities primarily occur through online labor markets and app-based on-demand work, where tasks resemble virtual production lines (Mo and Li, 2022). Gig platforms have significantly improved job-matching efficiency while generating new employment forms (Seamans and Zhu, 2014). Flexibility and autonomy are its defining features (Meijerink and Keegan, 2019; Muldoon and Raekstad, 2023). Hallmarks of this economy include flexibility in work hours, locations, tasks, contracts, and roles (Davidescu et al., 2020; Dunn et al., 2023; Lehdonvirta, 2018), and autonomy in methods, decision-making, and self-management (Kossek et al., 2021). However, with the continued advancement of digital technologies, the nature of work is undergoing constant transformation(Vertesi et al., 2020). Despite its benefits, it also poses notable social and economic challenges.
Previous studies have extensively explored the driving factors behind the development of the gig economy and its subsequent economic and social impacts (Acquier et al., 2019; Demailly and Novel, 2014; John, 2013; Palgan et al., 2017). For instance, technological innovations in areas such as data analytics and cloud computing have significantly enhanced work efficiency and market transparency, laying a solid foundation for the expansion of the gig economy (Gansky, 2010; Yeganeh, 2021). Digitalization has profoundly reshaped the consumer market, where rising demands for speed, convenience, and user-friendliness in products and services have accelerated the rise of the gig economy model (Pilkington, 2016). It has profoundly influenced the economy and society by reshaping labor relations, increasing income volatility, constraining career development, and challenging organizational efficiency and human resource management (Acemoglu and Autor, 2011; Friedman, 2014; Kuhn et al., 2021; Muldoon and Raekstad, 2023; Parente et al., 2018). Through the intervention of internet platforms, it has completely changed traditional employment relationships, making work more flexible and autonomous (Meijerink and Keegan, 2019; Muldoon and Raekstad, 2023).
Based on the transformation of labor relations brought by the gig economy, some scholars hold a positive view of its development, believing it can provide more employment and entrepreneurial opportunities for workers (Ahsan, 2020; Mo and Li, 2022; Roy and Shrivastava, 2020), increase sources of income (Greenwood et al., 2017; Mulcahy, 2017), and offer opportunities to develop diverse skills (Fox et al., 2018), as well as better job matching (De Stefano, 2016), which is conducive to individual employment. The gig economy’s features—fragmentation, flexibility, remote work, and impersonalization—have reshaped business tasks and management practices, altering costs, efficiency, and risks (Meijerink and Keegan, 2019). These shifts offer substantial operational advantages, such as lower operating costs and enhanced risk diversification (Gandhi and Sucahyo, 2021; Malin and Chandler, 2017; Mulcahy, 2017; Zwick, 2018).
However, some scholars argue that the gig economy undermines workers’ bargaining power. On the one hand, it relies on digital platforms, emphasizing flexibility and autonomy. However, some argue that algorithmic control and digital surveillance compel workers to strictly adhere to platform demands (Feng and Zhan, 2019; Kuhn and Maleki, 2017). On the other hand, gig workers are often classified as independent contractors. As a result, workers may lose various labor rights and legal protections, including wage guarantees, working hours restrictions, and working conditions (De Stefano, 2015; Healy et al., 2017; Steinbaum, 2019; Stewart and Stanford, 2017). Moreover, traditional workers enhance their bargaining power through union-based collective action (Ashenfelter and Johnson, 1969), while gig workers may lose the ability to engage in collective bargaining (Johnston et al., 2024), making them more vulnerable to bargaining pressures and weakening their market leverage. Corporations also face increased labor disputes (Friedman, 2014; Tassinari and Maccarrone, 2020). These concerns confirm that gig economy development has negative implications on individual employment.
Although scholarly interest in the gig economy has increased significantly in recent years, there remains substantial disagreement regarding its overall impact. Such divergence may be rooted in contextual variations, including economic conditions, individual characteristics, and cultural settings. While some studies underscore the gig economy’s potential to create employment opportunities, this study argues that it functions primarily as a substitution mechanism—shifting labor from formal to informal employment rather than expanding aggregate labor demand in certain contexts. The substitution effect is driven by the combination of factors, including technological advancement, structural transfer of labor from traditional manufacturing to platform-based sectors, employers’ cost-minimization strategies, and ongoing erosion of workers’ bargaining power.
In the traditional labor market, education and work experience are seen as the most direct indicators of human capital. However, in context of new gig economy, this view is contested. For instance, some authors argue that higher education does not always confer a competitive advantage (Ghaffarzadegan et al., 2017; Herrmann et al., 2023). Traditionally, educational level is positively related with wage returns, with higher education leading to higher income, a presumption in classical labor economics (Layard and Psacharopoulos, 1974). Given the ongoing debate in literature, the study may enrich existing research by examining the relationship between gig economy development and individual employment behavior using micro-level data from China. It integrates key dimensions of human capital, namely education and work experience, to assess their moderating effects through theoretical and empirical analysis. Additionally, the study explores the mediating role of wage bargaining power in gig economy and employment among working-age individuals, which contrasts its effects on elderly workers. It offers a fresh perspective on gig economy dynamics.
This study may make several key contributions. First, the diversity of gig economy, high workforce mobility, unstable labor relations, and lack of authoritative data pose major challenges to impact assessment. Drawing on Johnes (2019) and data from the China General Social Survey, this study constructs regional gig economy development indices to analyze their relationship with individual employment decisions, offering a micro-level perspective on employment effects. Second, by incorporating education and work experience into the analytical framework, the study highlights the crucial role of individual qualifications and labor market competitiveness, enriching our understanding of how human capital shapes the relationship between gig economy growth and employment behavior. Third, it examines the mediating role of wage bargaining power, offering new insights into the mechanisms through which the gig economy affects individual employment choices. Fourth, the study explores the gig economy’s impact on older workers relative to the working-age population, shedding light on its heterogeneous effects across labor groups. Finally, considering the distinct labor market, policy environment, and cultural context as an emerging economy in China, the findings provide valuable comparative insights and contribute meaningfully to global gig economy research.
Theory and research hypothesis
Theoretical basis
Bargaining power
It first proposed by Schelling, has become one of the classic theories in the field of negotiation and bargaining (Schelling, 1956). The main content covers several key aspects. First, it points out existence of asymmetric bargaining power between parties in negotiations. This means that one party may possess a stronger position, thus being more likely to achieve its negotiation objectives (Binmore et al., 1986). Asymmetry can stem from a variety of factors such as resource allocation, market position, and information advantages (Lippman and Rumelt, 2003; Qing et al., 2017; Rivard et al., 2006), shaping different power dynamics in negotiation. Secondly, negotiating parties can employ a variety of strategies to enhance their bargaining power. These include threats, concessions, mastering information, and protecting interests, among other methods (Wagner, 1988). By strategically employing these techniques, negotiators can shape the tone and direction of negotiation to advance their interests and achieve desired outcomes. Lastly, the theory emphasizes that negotiation outcomes reflect the relative bargaining power of parties, rather than purely objective facts and logic (Eisenberg, 1975). Success thus depends not only on information and reasoning, but also on the strategic skills exercised during the bargaining process. These theoretical perspectives offer a vital framework for understanding wage bargaining between gig workers and employers, enabling deeper analysis of the mechanisms through which the gig economy influences individual employment.
Information asymmetry
Akerlof (1970) systematically studied the situation of asymmetric information in transactions through an in-depth analysis of the used car market phenomenon (Akerlof, 1970). The research revealed presence of information asymmetry in market transactions, especially in used car market, where sellers usually have more information about the quality of the vehicles than buyers. However, this observation extends beyond the used car market, prompting scholars to consider the broader implications of information asymmetry. Subsequent research identifies two primary issues: adverse selection before transactions and moral hazard afterward (Hansen, 1987; Pauly, 1978; Prescott and Townsend, 1984). Adverse selection occurs in situations of information asymmetry, leading to a higher likelihood of inferior products or services in the market, as buyers find it difficult to accurately assess the true value of products (Prescott and Townsend, 1984). Moral hazard involves unethical actions that the party with information advantage might take after the transaction, causing harm to the party at an information disadvantage (Pauly, 1978; Smith, 1987). These two issues form core content of this theory, highlighting the adverse effects of information inequality on market transactions. In labor market transactions, especially in employer-employee relationships, the application of information asymmetry theory is particularly significant (Grossman, 1979). Employers typically possess more information-such as details on working conditions, salary structures, and company prospects-while employees often face information deficits (Graham et al., 2017; Rosenblat and Stark, 2016). The asymmetry leads to employers being more likely to have an advantageous position in negotiations, thereby maximizing their benefits, while employees may suffer losses due to insufficient information (Epstein, 2013). Therefore, information asymmetry is not only prevalent in market transactions but also critically shapes labor market dynamics. It offers a key theoretical lens for understanding employer-employee relations in the gig economy, particularly regarding bargaining power and transaction efficiency.
Gig economy development and individuals’ employment among the working-age population
While some studies view gig economy as having a “job creation effect” (Mo and Li, 2022)—offering flexible employment opportunities to marginalized groups through platform-based models—its actual impact on net employment may be limited in certain contexts. Specifically, when the overall size of the labor market cannot expand significantly in the short term, the absorption of workers by gig platforms may represent not genuine job creation, but rather a structural substitution of traditional employment.
In China, overall employment has remained relatively stable in recent years, indicating that aggregate labor demand has also been largely steady. However, significant structural shifts have occurred, with labor demand increasingly moving from traditional manufacturing to services and high-tech sectors (Li, 2018; Su et al., 2022). Workers who would otherwise participate in formal labor market may be crowded out by gig economy’s flexible employment model and cost advantages, leading to weaker labor protections, reduced job stability, and the erosion of formal employment opportunities (De Stefano, 2015; Healy et al., 2017; Steinbaum, 2019; Stewart and Stanford, 2017).
From the demand side of labor market, employers often prefer gig workers because they are typically not entitled to traditional benefits such as health insurance or pensions. This allows firms to significantly reduce labor costs(Duggan et al., 2020; Schor et al., 2020) and maximize profits (Howcroft and Bergvall-Kåreborn, 2019; Van Doorn, 2017; Wu et al., 2019). For reasons of cost efficiency, employers may reduce full-time formal positions and instead turn to gig labor (Zervas et al., 2017). Thus, rise of gig platforms has not only reshaped labor organization through technology but also fundamentally altered hiring practices. Employers increasingly rely on gig workers to cut costs and enhance flexibility. The shift has led to outsourcing of some formal jobs into nonstandard arrangements, thereby shrinking share of formal employment in labor market.
From supply side, gig work attracts not only older workers-valued for their experience-but also individuals seeking greater autonomy and those with lower expectations for wages and fixed workplaces (Spreitzer et al., 2017), as well as workers transitioning from traditional industries. For elderly individuals, regardless of skill level, the gig economy offers a way to supplement income while maintaining a sense of social engagement and purpose (Andersen and Sundstrup, 2019).
In contrast, younger workers tend to prefer stable jobs with clear career paths and social recognition, making them less willing to enter gig work—especially those with higher education. As supply of formal jobs shrinks but preferences remain unchanged, the mismatch may lead to structural unemployment or prolonged job search periods. Studies consistently show that youth unemployment exceeds the general rate, particularly in urban areas and among the highly educated (Demissie et al., 2021; Feng et al., 2024). According to official reports, the unemployment rate among young people aged 16–24 has reached a relatively high level in recent years, reflecting the shortage of formal jobs and structural imbalances in labor market that jointly contribute to the persistence of youth unemployment.
Technological advancement has accelerated labor market shifts. Since the mid-2010s, manufacturing sector has undergone capacity reduction, automation, and offshoring in China. When automation becomes more cost-effective than human labor, workers are replaced by machines, reducing overall labor demand and exerting downward pressure on employment and wages (Acemoglu and Restrepo, 2020; Autor et al., 2003). As a result, many middle-aged, low-skilled blue-collar workers have lost stable manufacturing jobs and moved into urban services—particularly the gig economy, which offers low entry barriers and quick onboarding. While this shift may appear to reflect job creation by platform economies, it in fact signals a transition from formal to informal employment. This transition often comes at the cost of weaker labor protections, less income stability, and narrower paths for career development.
In gig economy, weakened worker bargaining power and growing market inequities have reduced the willingness of working-age individuals to participate in the labor market. Theory of bargaining power suggests that level and quality of employment for workers are influenced by their bargaining power in labor market (Wood et al., 2019). In gig economy, classification of workers as independent contractors rather than formal employees increases their reliance on platforms, undermines their bargaining power, and hampers their ability to obtain fair wages and working conditions. Sense of powerlessness and frustration may lead to a decrease in enthusiasm for work and commitment to their jobs (Birnbaum and De Wispelaere, 2021; Grossman and Oberfield, 2022). As this employment relationship falls outside traditional legal definitions of “employee” or “independent contractor,” workers risk losing key labor rights and protections, including wage security, limits on working hours, and safeguards for working conditions (De Stefano, 2015; Healy et al., 2017; Steinbaum, 2019; Stewart and Stanford, 2017). Traditional workers can engage in collective bargaining through unions to improve wages and working conditions (Ashenfelter and Johnson, 1969). However, gig workers have difficulty forming or joining unions, thus losing the right and ability to collectively bargain (Johnston et al., 2024). This makes them more susceptible to individual employer negotiation pressures, weakening market bargaining power.
Advancements in automation further erode workers’ bargaining power by increasing risk of replacement and diminishing their leverage in wage and benefit negotiations (Lobel, 2017). Simultaneously, automation weakens the link between wages and labor productivity, intensifying labor market instability (Duggan et al., 2020). In addition, information asymmetry exacerbates market inequities. On digital labor platforms, operators possess superior knowledge of market conditions and pricing, while individual gig workers often lack access to such information, limiting their task selection (Graham et al., 2017; Rosenblat and Stark, 2016) and exposing them to potential exploitation (Epstein, 2013). Collectively, these factors undermine both the motivation and capacity of workers to engage in labor market.
H1: The development of gig economy is negatively related to working-age individuals’ employment.
Moderating effects: education and work experience
The relationship of the gig economy and employment in working-age individuals shows significant individual variability. We explores the moderating effects of educational level and work experience on the relationship between the development of the gig economy and the employment of working-age individuals, from the perspectives of individual professional qualifications and labor market competitiveness. Therefore, within gig economy, an individual’s level of education and work experience are external “signals” that employers need to pay attention to, and are important measures of their bargaining power.
Education
In labor economics, education is traditionally viewed as positively correlated with wage returns, as higher educational attainment enhances worker attractiveness by signaling greater skills and reducing adverse selection risks. However, this relationship is increasingly challenged in gig economy, where employers prioritize task outcomes over formal qualifications or workplace context, thereby diminishing relevance of education as a determinant of earnings (Herrmann et al., 2023).
Research on the impact of education on gig workers remains underdeveloped. Wheelahan and Moodie (2022) argue for a stronger connection between education and job content (Wheelahan and Moodie, 2022). Based on existing literature, we indicate that effects of gig economy differ depending on an individual’s educational level. For highly educated individuals, the mismatch between theoretical knowledge and practical skills (Kuzminov et al., 2019), the gap between high expectations and actual remuneration(Ghaffarzadegan et al., 2017) and the decreasing alignment between industry demands and educational qualifications (Graham et al., 2017) may be the cause and reduce their employment. They usually possess rich theoretical knowledge and analytical abilities. However, the gig economy places more emphasis on practical skills, making it difficult for them to immediately translate their education into concrete work efficacy (Kuzminov et al., 2019). They may be more likely to pursue stable full-time jobs with high career prospects, but the nature of the gig economy is short-term and informal, with limited remuneration levels and career advancement opportunities (Graham et al., 2017). Herrmann et al. (2023) selected gig workers in 14 Western countries for research and concluded that higher education levels do not increase the hourly wages of online gig workers. As the gig economy evolves, most industries may focus more on practical skills and experience, while a high level of education may not always provide a direct competitive advantage (Ghaffarzadegan et al., 2017; Herrmann et al., 2023). These factors lead to highly educated individuals no longer being as attractive in the gig economy as they are in traditional industries, thereby affecting their employment.
For individuals with lower education levels, the substitutability of automation technology (Duggan et al., 2020), insufficient skill matching (Cappelli, 2015), and limited career development opportunities (Al-Asfour et al., 2017) are important factors affecting their employment. Firstly, individuals with lower education often engage in relatively simple, repetitive tasks that are more easily replaced by automation with technological advancements (Rincón and Martínez, 2020; Vrontis et al., 2022). Secondly, these individuals typically possess only basic, low-skill work abilities, for which market demand is limited. With insufficient skill matching, low-educated workers often struggle to find suitable jobs (Palmer, 2017). Even if they do find work, skill mismatches may lead to low work efficiency or even loss of job opportunities. Lastly, due to nature of their work in low-skill, entry-level fields and lack of relevant training, these individuals face limited career progression, making it difficult to secure job opportunities (Al-Asfour et al., 2017). These limiting factors affect not only their current employment status but also exacerbate the uncertainty of their future job prospects.
In gig economy, individuals with medium education levels are often seen as at a balance point. They do not have overly high expectations like highly educated individuals, nor do they face challenges in certain positions due to being “overqualified” (Herrmann et al., 2023). Their education level does not exceed what is practically needed. At the same time, compared to lower-educated individuals, those with medium education have more knowledge and skills, making them more competitive and adaptable in gig market. These individuals can flexibly adjust their expectations and mindset to accommodate the fluctuations and uncertainties of gig market (Berg, 2015; Johnston et al., 2024). Their expectations align more closely with reality, avoiding the psychological gap and disappointment that highly educated individuals may experience (Berg, 2015). Due to relatively lower expectations, individuals with medium education might be more content with basic positions, thereby increasing their employment opportunities (Thompson, 2018). They are not as lacking in confidence and skills as those with lower education and possess stable learning abilities, thus holding certain advantages in gig economy.
H2: Medium education moderates the relationship between gig economy development and employment, such that the negative relationship is weaker for individuals with medium education.
Work experience
Empirical research indicates that individuals with longer working tenures tend to possess greater practical experience and domain-specific skills (Deal et al., 2010; Kuijpers et al., 2006; Quińones et al., 1995). In flexible context of gig economy, on-demand employment, such competencies enhance adaptability to varied tasks and shifting job requirements. The adaptability not only increases their chances of securing suitable opportunities but also buffers the adverse effects of gig economy expansion on individual employment.
The gig economy is considered a form of informal employment by the International Labor Organization (Günther and Launov, 2012). Value of labor capital, especially work experience, is one of the key factors determining income (Mincer, 1958). Rich work experience significantly impacts individual income (Luan, 2003; Ren et al., 2015). In gig employment sector, such experience may endow individuals with stronger competitiveness and adaptability, thus enhancing their bargaining power and wage returns. Therefore, the premium effect of work experience is also significant in gig economy (Healy et al., 2017; Mas and Pallais, 2020).
The gig economy features low entry barriers and diverse participation formats, allowing most individuals to access opportunities more readily than in traditional employment (Mo and Li, 2022). Workers with extensive experience often possess broader social networks and stronger interpersonal ties (Baron and Markman, 2000; Kalleberg, 2012), which serve as valuable resources for connecting with potential employers, clients, or collaborators. This social capital enhances their competitiveness, increases participation likelihood, and contributes to more favorable evaluations in gig market.
In gig economy platforms, evaluations and feedback are important aspects of employment relationships (Kellogg et al., 2020; Kuhn and Maleki, 2017). Digital labor platforms often rely on consumer feedback and rating systems to assess and monitor workers’ performance. Platform-based rating systems significantly shape workers’ online reputations, where negative feedback can harm future job prospects or result in deactivation (Johnston and Land-Kazlauskas, 2018). Individuals with substantial formal work experience often possess established professional reputations and positive word-of-mouth, enhancing their credibility and likelihood of being hired in gig economy (Kokkodis and Ipeirotis, 2016). Research indicates that both workers and employers in the gig economy place significant emphasis on prior experience and online ratings (Herrmann et al., 2023). Consequently, individuals with extensive experience possess greater bargaining power and enhanced competitiveness compared to their less experienced counterparts.
H3: Work experience moderates the relationship between gig economy development and employment, such that the negative relationship is weaker for individuals with extensive work experience.
Methods
Sample and data
This study utilizes data from the Chinese General Social Survey (CGSS), which is the earliest nationwide, comprehensive, and continuous academic survey project in China. The CGSS has released data for the years 2003, 2006, 2008, 2010, 2011, 2012, 2013, 2015, 2017, 2018, and 2021, and continues to update its dataset to the present. Core module questions, in use since 2010, mainly cover social demographic attributes, housing, health, migration, lifestyle, social attitudes, and related topics. The structure and question types of the CGSS questionnaire are diverse, providing comprehensive insights into gender, educational level, work experience, occupation, and occupational income, thereby meeting the data needs of this study.
We specifically selected data from CGSS2015, CGSS2017, CGSS2018, and CGSS2021, focusing on the post-2015 period, which aligns with the rapid expansion of the gig economy in China. Around 2014, platforms such as Didi, Baidu Delivery, and Meituan Delivery catalyzed the growth of gig-based labor, with rideshare drivers and food delivery couriers emerging as the primary workforce. Food delivery, in particular, expanded rapidly due to urbanization and the rise of a rich culinary culture in China. Compared to rideshare services, food delivery attracted workers across a wider range of age groups because of its lower entry barriers (e.g., no vehicle purchase required).
Recruitment data from major job platforms, including Zhaopin, 51Job, and Liepin, indicate that the earliest job postings containing the “food delivery” tag appeared in 2014, highlighting the sector’s influence on employment choices. To ensure data relevance and minimize bias from earlier periods, the study restricts the sample to respondents from 2015 onward. The final dataset includes women aged 18–55 and men aged 18–60, excluding records with substantial missing data, yielding a valid sample size of 17,605.
Model setting
In order to test the theoretical hypothesis of the study, we set the Probit model as follows:
(1)
EMP is the dependent variable in model (1), representing whether the i-th respondent is engaged in employment. GIG is the explanatory variable, indicating the development level of the gig economy in the respondent’s region. Moderators include medium level of education (MEDU) and work experience (WEX). K represents all control variables. Additionally, the model also controls for dummy variables for year and region.
Measures
Dependent variable
We refer to the study by Li and He (2023) and use the question, “What is your work experience and status?”, from the survey to measure individual employment status (Li and He, 2023). When an individual responds with “Currently engaged in non-agricultural work”, it is coded as 1 and 0 otherwise. According to this definition, the proportions of individual employment in the years 2015, 2017, 2018, and 2021 were 66.88, 68.81, 69.08, and 61.86%, respectively.
Independent variable
We follow the research method of Johnes (2019) and use the sample data from the CGSS database to measure the scale of the gig economy. Specifically, the study calculates proportion of gig, casual, or freelance workers in selected sectors relative to the total employment of each province or city (Johnes, 2019). We used formula (2) to measure the development level of the gig economy in each province or city. The calculation formula is as follows:
(2)
NPCS represents the number of part-time, casual, and self-employed workers in specific sectors within a given province or city. TNPCS represents the total number of workers in these same sectors, including full-time, part-time, casual, and self-employed workers, within that province or city. These specific sectors are chosen because they typically represent gig economy, including construction and engineering, design and creativity, financial and business services, information technology and communications, education and health services, sales and services, crafts and technical work, transportation and delivery, and simple labor tasks. Therefore, the proportion of gig, casual, or freelance workers in these typical gig employment sectors provides an effective proxy variable for measuring and comparing the scale of the gig economy in different regions.
Moderating variables
Medium education is defined as a binary variable that equals 1 if respondent i completed high school, vocational high school, technical secondary school, or vocational school, and 0 otherwise. In China, compulsory education comprises nine years of primary and junior high school. The subsequent educational stage—encompassing the aforementioned institutions—is classified as medium education, as it follows compulsory education but does not constitute higher education (e.g., college level or above). For robustness, the analysis also employs years of schooling and its squared term as alternative measures. Following prior research (Hu and Lu, 2011), this study uses accumulated years of non-agricultural employment as a proxy for work experience. Specifically, it is measured as the total duration from an individual’s first to last non-agricultural job, regardless of current employment status. For individuals with no such experience, the value is set to zero. In robustness checks, this variable is also coded as a binary indicator, equal to 1 if individual i’s work experience meets or exceeds the average for their province (or city) in that year, and 0 otherwise.
Control variables
Following practices of existing literature (Harding and Rosenthal, 2017; Li and He, 2023; Simoes et al., 2016; Viitanen, 2005; Zhang et al., 2008), this study selects the following control variables: (1) Basic demographic characteristics. Age, gender, race, and religious beliefs are important factors influencing an individual’s employment choices and job opportunities (Simoes et al., 2016). (2) Human and political capital. Education level and health status are key components of human capital. Political capital, in a broad sense, may provide individuals with advantages in job access and career development (Li and He, 2023). (3) Family and economic background factors. Family structure (such as marital status) and childcare responsibilities significantly influence an individual’s employment choices (Viitanen, 2005; Zhang et al., 2008). Unmarried or single individuals may participate more actively in the labor market due to the need to maintain their livelihood. In contrast, married or partnered individuals may reduce their work participation because of family dependencies, although they may also need to work due to life pressures. The household registration system, as a basic institutional arrangement for social management in China, is a key social factor affecting individual employment (Li and Gu, 2011). Meanwhile, differences in urban and rural living environments may affect individual employment opportunities and types of work. Additionally, an individual’s economic conditions, such as home ownership, are also considered, as economic security and asset status may influence an individual’s employment motivation and choices (Harding and Rosenthal, 2017). Considering that the data spans multiple years, we controlled for time variables to eliminate the impact of time trends and cyclical factors. Furthermore, considering differences in infrastructure construction and economic development levels across different regions of China, we categorized the provinces of residence of individuals into Eastern, Central, and Western regions to control for the potential impact of regional differences on employment. Table 1 provided the definitions of main variables used in this study.
By Nature – https://www.nature.com/articles/s41599-025-05970-x
