Gender Income Inequality Within and Outside the State System in China, 2003–2021: An Age–Period–Cohort Analysis

Abstract

Guided by Sustainable Development Goal (SDG) 5 on achieving gender equality and empowering all women and girls, our study examines the age, period, and cohort effects of gender income inequality across China’s public and private sector employment by utilizing hierarchical age–period–cohort cross-classification random-effects models (HAPC-CCREMs) and repeated cross-sectional data from the Chinese General Social Survey from 2003 to 2021 (N = 29,367). The results demonstrate the following: (1) Age effects of gender income inequality diverge between public and private sector employment. In public sector employment, inequality undergoes a progressive decline over individuals’ career spans, as age is institutionalized as a sector-specific capital and compresses inequality through seniority-based accumulation. In private sector employment, inequality follows an inverted U-shaped trend as age is marketized as a proxy for labor productivity, producing steeper inequality in individuals’ early careers and sharp declines thereafter. (2) Period effects of gender income inequality manifest significant developing differences across public and private sector employment between 2003 and 2021. In public sector employment, the state redistributive mechanism maintains inequality at a consistently low and stable level. In private sector employment, inequality fluctuates with China’s post-transition economic restructuring, expanding during rapid market growth (2003–2008), contracting amid structural upgrading (2010–2013), and rising again under deeper market integration (2015–2021). (3) Cohort effects are negligible, reflecting that mechanisms sustaining gender income inequality exhibit intergenerational continuity. These results demonstrate that institutional segmentation structures gendered income dynamics throughout the life course via distinct resource allocation mechanisms. Our study extends life course approaches to social inequality, emphasizing the role of gender-equality-oriented governance, lifecycle-spanning support mechanisms, and cross-sectoral coordination in mitigating gender disparities.

1. Introduction

The United Nations has designated gender equality as the fifth Sustainable Development Goal (SDG) and incorporated it into the 2030 Agenda for Sustainable Development, thereby emphasizing its significant role in achieving sustainable development [1,2,3]. However, an examination of global practices reveals a paradox. Marketization and globalization, in their pursuit of efficiency, may be eroding the gains made in gender equality. Specifically, under the dual pressures of efficiency-driven mechanisms and welfare cuts, the private sector is inadvertently reinforcing traditional gender divisions of labor, pushing women toward domestic roles or low-wage occupations [2,3,4]. Most concerning is the indication that this regression in equality is spreading into the public sector, which is traditionally regarded as offering stronger welfare protections [2,3,5]. In the context of social individualization and profound labor market liberalization, the examination and comparison of gender pay gaps within and between the public and private sectors has emerged as a practical imperative. This is essential for mitigating systemic risks and ensuring the realization of Sustainable Development Goal 5.

The worldwide influential SDG 5 has developed a distinctive trajectory in China. Gender equality is a fundamental institutional policy in China, viewed as an essential institution to strengthening social protection, advancing well-being, and improving quality of life [2,3,6]. Notably, since the major economic reforms of the 1980s, China has achieved remarkable socioeconomic development and made significant progress in enhancing gender equality. On the one hand, the gender gap in socioeconomic status has narrowed due to the improvements in women’s education levels and labor force participation rates. On the other hand, policies promoting gender equality, such as the slogan “women hold up half the sky”, have effectively curbed the cultural reproduction of China’s traditional patriarchal system, namely, the “men plough and women weave” division of labor [2,3,7,8,9]. These practical policies have also achieved equal pay for men and women during the planned economy era and the early stages of China’s emerging labor market.

Unlike Western labor markets, which are driven by competition and efficiency, China’s labor market reflects both the legacy of state-led redistribution and the institutionally managed transition to a market economy. These dual forces have created a segmented employment structure separating the public and private sectors [2,3,10,11]. However, China’s implementation of SDG 5 has recently encountered global challenges. The efficiency-driven management system characteristics of the private sector have had a considerable impact on the foundations of gender equality that were established during the period of the planned economy [2,3,12,13]. This has led to the resurgence of the conventional gendered division of labor at the family level, which is characterized by the notion that men should be employed outside the home while women should be employed inside the home, and gender discrimination, once concentrated in private sectors, has gradually spread into public sectors as well [2,3,14,15]. As one of the world’s largest developing economies, China’s institutional divide between the public and private sectors, coupled with the coexistence of redistribution and market efficiency mechanisms, offers a unique comparative context for exploring gender equality and sustainable development. The public sector, primarily composed of public service departments and state-owned monopolies or carbon-intensive enterprises led by state capital, contrasts with the private sector, which is dominated by high-tech industries and globally integrated private or foreign-owned firms [16,17]. Analyzing gender income inequality between these sectors within China’s distinctive context helps to understand the disparities in gender equality practices amidst the global energy transition and the growth of green jobs. It also reveals how these sectors may complement each other in achieving SDG 5, balancing institutional protections with efficiency priorities. This inquiry provides important theoretical and practical insights for advancing gender equality in segmented labor markets, ultimately contributing to the creation of more inclusive and sustainable development pathways.

The prevalence of gender inequality, both globally and within China, is intricately intertwined with the complex dynamics of market transition [2,3,18]. This transition is not a static outcome; rather, it is a dynamic process that continuously shapes individual life trajectories and social structures [2,3,19]. Therefore, to deeply understand the evolution of the gender income gap in both public and private sectors, it is essential to employ a dynamic analytical perspective. To address this, a life course perspective is adopted, along with the utilization of the age–period–cohort (APC) approach, which is achieved by establishing a correlation between age and occupational development, period and institutional change, and cohort and cultural shifts. This analytical approach enables the analysis of the dynamics of gender income inequality within China’s institutional segmented labor market structure. The central research question guiding this study is as follows: How does gender inequality evolve over time in the context of varying institutional mechanisms?

The remainder of this paper is structured as follows. To begin with, a literature review is provided on the primary theoretical perspective in order to explore gender income inequality in China, with critical emphasis placed on the neglect of labor market structure transfer dynamics within prior studies. Thereafter, hypotheses are formulated based on life course theory to interrogate the age, period, and cohort effects on gender income inequality within and outside China’s state system between 2003 and 2021. Subsequently, we present our longitudinal analysis, which is based on twelve waves of nationally representative data from the Chinese General Social Survey (CGSS) and employs Hierarchical Age–Period–Cohort Cross-Classified Random-Effects Models (HAPC-CCREMs). Ultimately, we conclude by discussing the broader theoretical contributions and practical implications of our findings.

2. Literature Review

Explanations for gender income inequality in labor markets traditionally cluster around three dominant theoretical paradigms. Individual differences theory inherits the analytical tradition of neoclassical economics. It attributes gender income inequality to measurable differences in human capital endowments and non-cognitive skills between men and women [2,3,20,21]. According to this paradigm, income disparities are understood as the cumulative outcome of differential investments in education, work experience, and occupational skills, which are subsequently priced through mechanisms of market competition and rational choice. In its extended form, this framework has sought to incorporate non-cognitive attributes—such as personality traits, individual dispositions, and emotional intelligence—to explain income differentials not captured by conventional human capital measures [2,3,22,23].

The institutional-structural perspective, which discards individual differences theory, is essential for understanding how inequality is woven into the very fabric of labor market stratification [2,3,24]. By introducing the labor market segmentation theory, it highlights the significant role of macro-institutional arrangements, such as the household registration system, regional hierarchies, and the urban–rural dual structure, in moderating gender income inequality. From this standpoint, gender disparities are not merely the cumulative outcome of individual choices but are ingrained in differentiated organizational logics and resource allocation mechanisms that determine who enters which sector, under what conditions, and with what prospects for advancement. At the meso level, the institutional–structural perspective also examines how segmented structures, such as industry, employment forms, and sectoral ownership regimes, produce systematic variations in gendered income outcomes [2,3,25,26,27].

The cultural-constructionist approach, with intellectual roots in gender role theory and a structural-functionalist narrative orientation, arises as an attempt to bridge the explanatory gap between individualistic and institutional–structural perspectives of gender inequality [2,3,28]. It emphasizes that enduring gender disparities are not merely the outcome of individual heterogeneity or structural stratification; rather, they are deeply embedded in symbolic orders, cultural expectations, and normative scripts that regulate behavior. Specifically, traditional gender norms and role expectations in China shape the conduct of men and women in ways that become self-reinforcing, rendering inequality both persistent and socially legitimate. In other words, these cultural mechanisms continue their reinforcement by naturalizing occupational segregation, valorizing gendered divisions of labor, and disciplining deviations from prescribed roles in China’s labor market [2,3,29,30,31].

Taken together, the aforementioned paradigms have advanced our understanding of gender income inequality while also revealing characteristic blind spots. Individual differences theory illuminates variation in human capital and skills but remains reductionist as it neglects the structural and cultural conditions under which such differences are produced and valued [2,3,32]. The institutional–structural perspective emphasizes the stratifying effect of segmented labor markets, yet it underestimates the micro-social and cultural mediation through which institutions are lived [2,3,33]. The cultural-constructionist approach aims to uncover how meanings, values, and norms are socially constructed and contested, yet it frequently marginalizes the institutional, historical, and material distributional logics through which power is structured and downplays how inequalities in resources, class, and institutions are both produced by and embedded in discursive formations [34]. Despite their complementarities, these accounts converge in one important limitation: they treat labor market structures as essentially static, thereby privileging synchronic explanations over diachronic processes. In essence, they have largely neglected the temporal dynamics of labor market restructuring over China’s market transitions [35].

Much of the existing scholarship proceeds on the assumption that patterns of China’s segmented labor market are relatively stable, treating gender income inequality at a specific point in time as an issue of functional alignment between individuals and structural arrangements [36]. Explanations rely on cross-sectional or fragmented research strategies, aiming to elucidate the structural characteristics and generation mechanisms of gender income inequality through static social facts [37,38,39]. Nevertheless, it is worth noting that a synchronic orientation obscures the diachronic specificity and institutional foundations of China’s gradual and path-dependent market transition. It further neglects the continuous interplay between individual life courses and macro-institutional orders over time, thereby failing to capture the cumulative consequences of structural change on gendered outcomes [40,41]. While some scholars have attempted to incorporate temporality by using instrumental variables to track macro-level indicators, employing propensity score matching to reconstruct career trajectories, or utilizing quasi-natural experiments to leverage regional variations, these strategies remain pseudo-diachronic [42,43]. They continue to count on cross-sectional data or limited time fragments, which constrains their ability to integrate the age, period, and cohort dimensions, and do not adequately account for the long-run dynamics of inequality across an individual’s life course.

In conclusion, despite their contributions, these paradigms continue to adhere to static conceptions of labor market order, thereby obscuring the cumulative dynamics through which inequality is produced and perpetuated. To overcome these limitations, our study concentrates on temporality, which is conceptualized through the interlinked dimensions of age, period, and cohort, and turns to life course theory as a framework for developing our theoretical model and hypotheses.

3. Theoretical Framework and Research Hypotheses

Life course theory, as developed comprehensively by Glen H. Elder, provides a theoretical framework that incorporates the social system of time into empirical social research [44]. Centered on process-oriented analysis, it addresses the limitations of purely synchronic studies, which often overlook the dynamic interplay between social change and individual development. The theory advocates tracing the temporal axis of individual lives and focusing on the causal relationships between specific events and personal trajectories. By considering multiple temporal dimensions—including age, period, and cohort—it explores the organic linkage between macro-level historical processes and micro-level life experiences [19]. Thus, hypotheses are grounded in this theoretical perspective.

3.1. Age Effects

Age effects manifest as a result of both the active processes of physical maturation, the accumulation of experience, and the evolution of social roles, as well as the structural regulation of life courses through role expectations, social norms, and institutional contexts [45]. Despite the paucity of studies that have directly examined the impact of age on gender income inequality in public sector employment, relevant research conducted across different dimensions—such as biological age, family life cycle, and career trajectory—provides significant insights for this study.

It has been demonstrated that female workers encounter gender discrimination and unequal treatment throughout their career development [46]. Some studies, which draw on human capital theory, concentrate on biological age. They utilize the age–income curve to attribute gender income inequality to the gendered accumulation of cognitive and non-cognitive skills throughout the life course [23,47,48]. Other research adopts a household division of labor perspective and examines the family life cycle. It emphasizes work–family conflict and demonstrates how motherhood impacts women’s career progression and earnings [49,50]. A further strand builds on status-attainment models and their revisions. These studies analyze the mechanisms, directions, and outcomes of occupational mobility to examine how institutional barriers relate to gender wage inequality [51,52]. Findings suggest that the interaction between unequal mobility opportunities and traditional household division of labor not only diminishes women’s likelihood of moving across sectors compared to men but also restricts the wage gains that mobility brings for women while producing larger gains for men. This pattern hampers women’s ability to capture institutional wage premiums or to benefit fully from market competition.

Nevertheless, these explanations have limitations within the context of China’s institutional labor market segmentation. Explanations that focus on biological age and the family life cycle tend to overemphasize individual accumulation processes while underestimating the structural reconfiguration inherent in public and private sector employment [53]. Specifically, the public sector generally provides stable career paths and comprehensive benefit systems, while the private sector is predominantly driven by efficiency imperatives. This renders women employed in the private sector more vulnerable to career interruptions due to familial responsibilities, thereby exacerbating the adverse impact on their income. However, it is important to note that protections within the public sector can also become an invisible driver of gender inequality. While the public sector generally provides more comprehensive maternity and parental leave systems, these provisions inadvertently reinforce gendered caregiving responsibilities, confining women to specific roles. This phenomenon is especially salient when women take parental leave or similar benefits, as prevailing societal and workplace perceptions about women’s roles—perceiving their primary duty as family care—remain entrenched [54]. This reinforcement of gender roles not only exacerbates the challenges women encounter in career advancement but also has the potential to reinforce expectations of their instability [55].

Therefore, the aforementioned explanations are inadequate in accounting for the reproduction and dynamic evolution of gender income inequality within the Chinese context of institutional labor market segmentation. Moreover, following the transition out of childbearing years, conventional household labor division and entrenched gender roles may impose certain constraints on women’s career advancement. However, the stable employment system in the public sector can, to a certain extent, provide women with the “courage” to pursue career development opportunities [56]. In contrast, the private sector exhibits a paucity of such institutional safeguards. As a result, women in this sector are more likely to experience career interruptions after childbearing years due to insufficient support [57], thereby exacerbating gender income inequality. Consequently, adopting a longitudinal perspective is imperative when assessing the age-related implications of gender income inequality between China’s public and private sectors. In accordance with these discussions, we hereby propose Hypothesis 1.

Hypothesis 1.

After controlling for period and cohort effects, disparities in gender income inequality manifest differently across public and private sector employment. Specifically, gender income inequality decreases with age in the public sector but increases with age in the private sector.

3.2. Period Effects

Period effects pertain to the impact of significant historical events and structural changes that simultaneously shape the experiences of individuals across all age groups [58]. Each period effect is indicative of the macro-social conditions that influenced all individuals of all age groups in a particular year. In China, the ongoing intensification of market reforms and the institutionalization of a market economy have fundamentally restructured employment relations, polarized the occupational structure of China’s labor markets, and altered the trajectory of gender income inequality across public and private sector employment [59,60].

Empirically, two contrasting processes shape gender income inequality in private sector employment, where resources are primarily allocated through market signals. First, skill-biased technological change, combined with the expansion of higher education, has enhanced women’s capacity to compete for high-skill positions and increased their earnings potential [61]. Second, service industries that rely on emotional labor and are structured around traditional gender norms tend to be female-dominated. This concentration channels women into mid- and low-level jobs with lower wages, limited promotion prospects, and narrow career paths, while men are more concentrated in labor- or technology-intensive industries, further reinforcing the gender income gap [62]. In contrast, compensation and promotion systems in public sector employment are supported and constrained by organizational structures. Standardized pay scales limit the growth of gendered wage discrimination, and relatively high job stability and welfare provision mitigate the effects of the motherhood penalty, helping to reduce gender income inequality [63,64].

Subsequently, scholars have scrutinized period effects of gender income inequality across public and private sector employment from various perspectives, including gender pay gap, occupational motherhood penalties, and traditional cultural gender discrimination [43,65,66]. Empirical studies utilizing data from 1989 to 2011 suggest that wage disparities across public and private sector employment are mainly due to traditional cultural gender discrimination. Compared to private sector employment, where gender income inequality has increased steadily over time, public sector employment shows lower wage gaps and less fluctuation over periods [67]. Research on the motherhood penalty from 1989 to 2015 further indicates that the wage penalty associated with motherhood is consistently more severe in private sector employment, and the inter-sectoral gap in this penalty has expanded over periods [65]. These findings offer direct evidence of a divergence in gender income inequality across public and private sector employment, indicating that the gender inequality between the public and private sectors showed a widening trend during the early stages of transition.

However, the academic understanding of gender income inequality across the public and private sector employment in China’s post-transition period (2000–2023) remains limited [68]. The accession of China to the World Trade Organization (WTO) in 2001 signified the nation’s transition into a reformed state, one that functions within the ambit of the global capitalist system. This transition precipitated a paradigm shift in labor relations and labor market flexibility, thereby ushering in a novel phase that is profoundly intertwined with the phenomenon of globalization [69]. Subsequent to this, China’s policy focus underwent a shift in the post-transition period, transitioning from a prioritization of economic growth to a focus on harmonious development. China has advanced towards a more institutionalized market order, with enhanced state redistributive capacity and policies promoting complementarities across ownership forms [70]. Against this backdrop, institutional practices supporting gender equality are expected to strengthen over time. Hypothesis 2 is as follows:

Hypothesis 2.

As historical periods progress, gender income inequality will decrease both in public and private sector employment, and the gap between them will converge.

3.3. Cohort Effects

Cohort effects illustrate how historical contexts influence life courses by shaping the accumulation of intergenerational resource disparities, resulting in enduring impacts on different cohorts [71]. It is essential to distinguish between period effects, which encapsulate the short-term impact of historical events or macro-social conditions affecting all age groups simultaneously, and cohort effects, which arise from the fact that individuals born into different socio-historical contexts experience distinct cumulative impacts of structural environments and opportunity regimes throughout their life course, thereby engendering durable generational imprints and intergenerational systematic disparities [72,73]. In China’s institutional segmented labor market, the dynamic interplay between institutional resilience and market forces constitutes a critical context for understanding cohort-based income inequality: the redistributive mechanisms facilitate the intergenerational transmission of institutional advantages, whereas marketization has continuously restructured the occupational opportunity across cohorts [74,75]. Despite the paucity of studies that have directly examined the cohort effects of gender income inequality across public and private sector employment, prior research consistently suggests that the stages of economic reform have produced distinct social configurations of gendered opportunities and constraints. These have yielded cohort-specific patterns in gender income inequality, and its cumulative impacts can be observed in the post-transition period.

During China’s planned economy era, a formal labor market had yet to develop, and public sector employment served as the primary career pathway for early cohorts born in the 1940s and 1950s. Given the relatively low level of socioeconomic development, both overall income levels and gender pay gaps remained modest under the state redistribution and gender equality policies. Nonetheless, persistent cultural norms favoring traditional gender roles concentrated women in lower-ranking positions, laying a social foundation for subsequent gender income inequality [9,76]. During the dual-track transition period from 1978 to 1991, cohorts born in the 1960s and 1970s generally experienced a divergence in employment pathways. Women employed in the public sector faced a “glass ceiling” in career advancement, yet the legacy of the “danwei” system and compensatory effects from cross-sectoral mobility restrained the expansion of gender income inequality in public sector employment [77,78]. In the context of private sector employment, the emergence of gendered wage rigidities was evident, driven by the dual-track pricing system, an influx of rural surplus labor, and the ongoing growth of labor-intensive industries [79]. This distribution pattern has led to an intensification of gender income differentiation in private sector employment, simultaneously resulting in a widening of the gender income inequality across public and private sector employment. From 1992 to 2007, during the market expansion period, cohorts born in the 1980s and 1990s—predominantly singletons—were marked by concentrated family resources, relaxed traditional gender norms, and swiftly increasing educational levels [80]. Women employed in the public sector reaped the benefits of dual protection from both institutional and familial domains, which alleviated the income penalties linked to motherhood. In contrast, private sector employment, gendered labor distribution, and heightened market competition intensified the penalties for motherhood and contributed to a widening inter-cohort gender income disparity [81,82]. Cohorts born in the 1990s and 2000s entered the labor market primarily during the period of deepening transition starting in 2008. Public sector employment, intensified entry competition, and a standardized pay and promotion system have tended to rigidly stabilize gender income gaps. In private sector employment, the polarization of employment structures has had a pronounced impact: while knowledge-intensive, high-end industries provide high-income opportunities for a subset of skilled women, most women occupy precarious, marginal positions with limited protections, resulting in an expanding gender income gap and pronounced within-group disparities among women [83,84].

As discussed above, gender income inequality in China is likely influenced by cohort effects and exhibits systematic variation across public and private sector employment. The path-dependent nature of redistributive mechanisms implies that cohort disparities in gender income inequality in public sector employment remain relatively stable and entrenched. By contrast, gender income inequality in private sector employment continues to expand across successive cohorts, driven by persistent cultural norms, increasingly gendered labor market competition, and insufficient employment guarantees. Accordingly, we propose Hypothesis 3 as follows:

Hypothesis 3.

Gender income inequality demonstrates significant cohort effects both in public and private sector employment. Specifically, it is expected to decrease gradually across successive cohorts in public sector employment, while progressively increasing in private sector employment.

4. Materials and Methods

4.1. Data

This study draws on twelve waves of the Chinese General Social Survey (CGSS) from 2003 to 2021. The CGSS employs a stratified sampling design, providing comprehensive, continuous, and nationally representative data. For the purposes of this study, the analytical sample was restricted to non-agricultural workers aged 18 to 60, who were of working age (many older adults lacked occupational and income data). After excluding cases with missing values on key variables, the final sample comprised 29,367 respondents, with a mean age of 38.72 years (SD = 9.92) and a balanced gender distribution.

4.2. Variables

4.2.1. Dependent Variable

The dependent variable in this study was income, measured by respondents’ self-reported “annual income in the previous year.” To account for inflation, reported incomes were deflated using the national Consumer Price Index (CPI) for urban residents, with 2003 as the base year. To mitigate the influence of extreme values and approximate a normal distribution, the deflated income values were log-transformed.

4.2.2. Independent Variables

The key independent variables included age, period, cohort, and gender. Age was measured as respondents’ actual age and treated as a continuous variable. Period corresponded to the survey year in which respondents were interviewed, covering twelve waves conducted in 2003, 2005, 2006, 2008, 2010, 2011, 2012, 2013, 2015, 2017, 2018, and 2021. Since reported income referred to the previous year, the actual period variable was coded as the survey year minus one. Cohort referred to individuals born within the same historical interval, which exposes them to similar formative social, economic, and institutional conditions. Based on respondents’ birth years (1943–2003), cohorts were defined in ten-year intervals, yielding seven successive birth cohorts, a standard convention widely adopted in the literature.

4.2.3. Conditional Variable

The conditional variable is the employment sector. Respondents were classified based on two survey items: “type of employing organization” and “employer.” Those employed in government agencies, public institutions, the military, social organizations, state- or collectively owned enterprises, and village committees were coded as public sector employment, whereas respondents working in private firms, foreign-invested enterprises, other business types, and self-employment were coded as private sector employment.

4.2.4. Control Variables

The control variables included educational attainment, political affiliation, marital status, household registration type, regional type, and class identity. The coding schemes and descriptive statistics for all variables are presented in

Table 1. Descriptive statistics of variables (N = 29,367).

Variables	Variables Description	Mean	Std.	Min	Max
Dependent Variable
Income	Continuous; log-transformed individual annual income	10.140	0.957	6.908	13.468
Independent Variable
Gender	0 = Female, 1 = male	0.565	0.496	0	1
APC Variables
Age	Continuous, 18–60	38.720	9.924	18	60
Period	Survey year	-	-	2003	2021
Cohort	Ten-year birth cohorts	3.845	1.102	1	7
Conditional Variable
Employment Sector	0 = Private sector employment, 1 = public sector employment	0.446	0.497	0	1
Control Variables
Education	0 = Primary or below, 1 = secondary, 2 = tertiary or above	1.264	0.608	0	2
Political Affiliation	0 = Non-Communist Party member, 1 = Communist Party member	0.157	0.364	0	1
Marital Status	0 = Not married, 1 = married	0.802	0.398	0	1
Hukou Type	0 = Rural, 1 = non-agricultural	0.692	0.462	0	1
Region Type	0 = Non-eastern region, 1 = eastern region	0.488	0.500	0	1
Class Identity	1–5, low to high	2.494	0.851	1	5

.

4.3. Analytical Approach

The age–period–cohort (APC) framework endeavors to identify the independent effects of three temporal dimensions. Specifically, age reflects stage-specific changes in individuals’ life courses; period captures historical fluctuations in the macro-social environment; and cohort originates from the shared experiences of distinct birth cohorts. In addressing the unidentifiability that arises from a fully linear relationship (Age + Cohort = Period), various methods have been proposed in the academic community [85]. These include Age–Period–Cohort-Constrained Generalized Linear Models (APC-CGLMs), Age–Period–Cohort Imputed Estimation (APC-IE), Age–Period–Cohort Interaction Models (APC-I), and Hierarchical Age–Period–Cohort Cross-Classified Random-Effects Models (HAPC-CCREMs).

However, the 12-wave repeated cross-sectional CGSS data from 2003 to 2021 utilized in this study encounters challenges in meeting the requirements of APC-EI or APC-CGLIMs for equidistant years and strict linearity constraints. That is, the repeated sampling data does not constitute a complete matrix encompassing age, period, and generation, rendering traditional linear APC models susceptible to unstable estimates. In contrast, HAPC-CCREMs do not rely on equidistant time periods or exogenous constraints, thereby enabling robust identification of APC effects within repeated cross-sectional data structures. This renders them more suitable for the data information structure in this study. The main advantages of these models are as follows: first, they stratify age, period, and cohort for estimation, with age as a fixed effect and period and cohort variation as random effects, solving the problem of linearity constraints on each other; second, there is no strict requirement for the year interval for multi-period cross-sectional survey data; and third, they can be directly used on microscopic survey data, greatly expanding the scope of use of the APC method. An HAPC-CCREM nests age in periods and cohorts; the three elements are not at the same level, and the time dimension at the higher order level is similar to the environmental variables, with regression coefficients and intercepts representing effects on the individual level [86,87]. The level 1 model takes the following form:

$$Y_{ijk} = {\beta }_{0jk}+{ \beta }_{1}AG{E}_{ijk} + {\beta }_{2}AG{E}_{ijk}^2 + {\beta }_3{GENDER}_{ijk} + {\beta }_{4}AG{E}_{ijk}{GENDER}_{ijk} +{ \beta }_5{Z}_{ijk} +{ \epsilon }_{ijk},$$

(1)

where Y_ijk denotes the income outcome for individual i in period j and cohort k and β denotes the regression coefficients. GENDER_ijk captures gender and includes an interaction term with age, AGE_ijkGENDER_ijk, to examine age-specific gender effects. Other control variables Z are incorporated into the first-level model, while ε_ijk ~ N(0, σ²) represents the individual-level random error.

The intercept effect in the level 2 model takes the following form:

$${\beta }_{0jk} = {\gamma }_0 + u_{0j} + {\nu }_{0k},$$

(2)

The effect of public sector employment in the level 2 model takes the following form:

$${\beta }_{3jk} = {\gamma }_3 + u_{3j} + {\nu }_{3k}\hspace{1em}{u}_{3j}\sim N\left(0,{\tau }_{3\mu }\right)\hspace{1em}{\nu }_{3k}\sim N\left(0,{\tau }_{3\vartheta }\right),$$

(3)

where u_0j and u_3j represent the cohort-specific effects on the intercept and the gender–income relationship, respectively, both assumed to follow a normal distribution. Similarly, v_0k and v_3k denote the period-specific effects on the intercept and the gender–income relationship, also assumed to be normally distributed. If the effect of gender on income varies across periods or cohorts, τ_3μ or τ_3ϑ can be assessed for statistical significance through variance testing. Substituting Equations (2) and (3) into Equation (1) yields the combined model:

$$Y_{ijk} = {\gamma }_0 +{ \beta }_{1}AG{E}_{ijk} + {\beta }_{2}AG{E}_{ijk}^2 + \left({\gamma }_3 +{ u}_{3j} + {\nu }_{3k}\right){GENDER}_{ijk} + {\beta }_{4}AG{E}_{ijk}{GENDER}_{ijk} + {\beta }_5{Z}_{ijk} + u_{0j} + {\nu }_{0k} + {\epsilon }_{ijk}$$

Model coefficients and fit statistics were estimated using the GLIMMIX procedure in SAS 9.4, and graphical representations and tables were generated with Excel.

5. Results

5.1. General Trajectories Analysis of Gender Income Inequality

Table 2. HAPC-CCREM benchmark regression results (N = 29,367).

Variables	Model 1	Model 2	Model 3	Model 4	Model 5
Intercept	7.8451 *** (0.1786)	7.8265 *** (0.1741)	7.8260 *** (0.1742)	7.9591 *** (0.1756)	7.8152 *** (0.1778)
Individual-level fixed effects
Age	0.0648 *** (0.0041)	0.0661 *** (0.0041)	0.0662 *** (0.0041)	0.0619 *** (0.0041)	0.0651 *** (0.0041)
Age-squared	−0.0008 *** (0.0000)	−0.0008 *** (0.0000)	−0.0008 *** (0.0001)	−0.0008 *** (0.0000)	−0.0008 *** (0.0000)
Male	0.3025 *** (0.0078)	0.2916 *** (0.0197)	0.2894 *** (0.0353)	0.3470 *** (0.0151)	0.3483 *** (0.0103)
Public sector employment	0.0214 ** (0.0108)	0.0219 ** (0.0108)	0.0334 ** (0.0108)	−0.3368 *** (0.0464)	0.0804 *** (0.0140)
Male × age			0.0002 *** (0.0008)
Public sector employment × age				0.0104 *** (0.0011)
Public sector employment × male				−0.1294 * (0.0516)	−0.1021 ** (0.0186)
Public sector employment × age × male				0.0004 ** (0.0012)
Control variables	Controlled	Controlled	Controlled	Controlled	Controlled
Random effect covariance parameters
Period effects
Intercept	0.2935 ** (0.1254)	0.2713 ** (0.1162)	0.2715 ** (0.1162)	0.2808 ** (0.1201)	0.2885 ** (0.1233)
Male		0.0027 * (0.0015)	0.0027 * (0.0015)	0.0013 * (0.0009)
Public sector employment × male					0.0012 * (0.0009)
Cohort effects
Intercept	0.0041 + (0.0032)	0.0050 (0.0039)	0.0047 + (0.0036)	0.0037 + (0.0029)	0.0045 + (0.0035)
Male		0.0004 (0.0007)
Public sector employment × male
Individuals	0.4213 *** (0.0035)	0.4206 *** (0.0035)	0.4207 *** (0.0035)	0.4180 *** (0.0035)	0.4204 *** (0.0035)
Fit statistics
AIC	58,152.00	58,130.95	58,142.37	57,961.35	58,109.45
BIC	58,153.46	58,120.95	58,134.37	57,953.35	58,101.45

Note: (1) *** p < 0.001, ** p < 0.01, * p < 0.05, and ⁺ p < 0.1. (2) Standard errors are reported in parentheses. (3) The control variables are the same as in Table 1.

presents the baseline age–period–cohort (APC) trends estimated using hierarchical models. Models 1 to 3 illustrate the age, period, and cohort variations in gender income inequality. Model 1 serves as the reference model, capturing the APC effects on income. According to the fixed effects in Model 1, age has a significant positive association with income (β = 0.0648, p < 0.001), while age-squared exhibits a significant negative association (β = −0.0008, p < 0.001). These findings are consistent with prior research, indicating that income among China’s labor force follows an inverted “U”-shaped curve in relation to age, with peak earnings occurring in midlife and subsequently decreasing. Furthermore, the significant random effects in Model 1 reveal a notable period difference in income (β = 0.2935, p = 0.0096), reflecting a consistent upward trend from 2003 to 2021 (figure omitted). However, the random effects in Model 1 indicate only marginally significant cohort effects (β = 0.0041, p = 0.0972), which display a nonlinear pattern: a decrease from the 1940s to the 1960s cohorts, an increase among the 1970s and 1980s cohorts, and a further decline for the 1990s and 2000s cohorts (figure omitted).

After controlling the random slopes for gender across periods and cohorts in Model 2, Model 3 introduces the interaction variable between age and gender to examine the moderating effect of gender on the age-related income patterns. The results of Model 3 indicate that the interaction between gender and age, as a fixed effect, has a significant influence on income (β = 0.0002, p < 0.001). This finding reflects a pronounced gender difference in the observed inverted “U”-shaped age–income profiles, independent of the period and cohort effects. The HAPC model enables the estimation and comparison of coefficients for age and age-squared in the age–income profiles of men and women, respectively. As illustrated in Figure 1, both the male and female age–income profiles follow inverted-U trajectories that peak around the mid-career age of 40. However, men exhibit higher peak incomes, higher overall income levels, and steeper increases and decreases compared to women. These patterns are consistent with the positive and significant interaction term between gender and age in Model 3. Moreover, as demonstrated by the derivative results and depicted in Figure 1, gender income inequality widens during the early career phase, reaches its maximum around the age of 38, and narrows during the mid-to-late career stage.

The random effects estimates from Model 2 indicate a significant gender variation in period effects (β = 0.0027, p < 0.05), but not in cohort effects (β = 0.0004, p > 0.1). It is important to clarify that the dependent variable in all models is the log of individual income; the gender coefficients reported here are period-specific random slopes that capture how the income difference between males (coded 1) and females (coded 0) varies across periods. Thus, the temporal pattern shown in Figure 2 does not represent a different outcome variable but rather the estimated period heterogeneity in the gender effect derived from the same HAPC-CCREM specification. These results reflect substantial period differences in gender income inequality, while generational differences remain statistically insignificant. As Figure 2 illustrates, gender income inequality expanded between 2003 and 2010 in a fluctuating yet accelerating fashion. Between 2011 and 2021, gender income inequality continued to grow but at a slower pace and with less volatility. Overall, the trajectory of gender income inequality is characterized by high levels and a decelerating rate of increase. Cohort effects on gender income inequality are statistically insignificant (p = 0.2587). Given this insignificance, Models 4 and 5 exclude random slopes for the interaction between gender and employment sector at the cohort level to avoid redundancy. Hypothesis 3 is not supported, whereas Hypotheses 1 and 2 are subject to further empirical examination.

5.2. Age Effects of Gender Income Inequality Across Public and Private Sector Employment

Model 4 incorporates a three-way interaction among gender, employment sector, and age to assess the variations in gender income inequality across public and private sector employment. As shown in Table 2, in Model 4, the interaction effect of public sector employment, gender, and age on income is statistically significant (β = 0.0004, p = 0.0073). This result indicates that the age effects on gender income inequality across China’s public and private sector employment differ significantly after controlling for period and cohort effects. The positive coefficient of the interaction term demonstrates that China’s institutional labor market segmentation amplifies gender income inequality through age-related dynamics.

To be more specific, an HAPC model was employed to estimate the coefficients of age and age-squared variables for income across gender and employment sectors. The derived results are presented in Figure 3, which illustrates the specific age effect trajectories for the four groups. The trajectories of public sector employment are relatively stable. Male earnings rise until approximately age 43.4 before declining, while female earnings peak later, around age 49.8. Males display steeper rates of both ascent and descent compared to females. In contrast, income from private sector employment follows a more pronounced inverted U-shape, with both males and females peaking near age 37 and experiencing rapid declines thereafter.

As illustrated in Figure 4, these sectoral contrasts result in distinct patterns of inequality. It should be clarified that the effects of age on gender income inequality discussed here also arise from the interaction between gender and age, whereas the dependent variable is individual log income throughout. Since gender is coded as 0 or 1, differences in predicted income trajectories between males and females reflect how gender inequality evolves over life courses within each sector. In public sector employment, gender income inequality gradually narrows across occupations and life courses, declining from 0.31 in the early stages to 0.07 towards the end. Conversely, the private sector employment disparities exhibit an inverted U-shaped pattern: they widen from ages 18 to 35, peak around age 35, and then begin to narrow, even reversing, in later career stages. Thus, Hypothesis 1 receives only partial support. Gender income inequality narrows progressively with age in public sector employment, but outside of it, inequality expands early and recedes later.

5.3. Period Effects of Gender Income Inequality in Public and Private Sector Employment

According to Table 2, Model 5 includes random slopes for the interaction between gender and sector based on Model 4, testing the moderating role of institutional segmented employment on the period and cohort effects of gender income inequality. In Model 5, the inclusion of random slopes for the interaction between gender and employment sectors reveals a significant period effect (p = 0.0423). A closer examination of the period effects shows marked divergence across public and private sector employment. In this context, the term “period effects of gender income inequality” signifies the fluctuations in the gender coefficient, which is represented by the income disparity between males and females, across different periods within each sector. As illustrated in Figure 5, gender income inequality in public sector employment remains relatively stable at a consistently low level across the observation period (2003–2021), with no clear upward or downward trajectory. By contrast, it exhibits pronounced fluctuations in private sector employment, with inequality rising between 2003 and 2008, declining between 2010 and 2013, and rising again from 2015 to 2021. The period effects of gender income inequality do not evolve in unison across public and private sector employment. Hypothesis 2 is not supported. A significant sectoral divergence is observed.

5.4. Robust Test

In order to verify the robustness of age, period, cohort effects, and key interaction effects, our study employed linear mixed-effects models to repeatedly estimate both the baseline model and the expanded interaction model. As illustrated in

Table 3. Robustness checks based on Models 1–5 (N = 29,367).

Variables	Model 1	Model 2	Model 3	Model 4	Model 5
Intercept	7.5873 *** (0.1395)	7.6023 *** (0.1398)	7.6022 *** (0.1398)	7.6941 *** (0.1412)	7.5564 *** (0.1402)
Individual-level fixed effects
Age	0.0554 *** (0.0065)	0.0552 *** (0.0065)	0.0551 *** (0.0065)	0.0567 *** (0.0065)	0.0573 *** (0.0065)
Age-squared	−0.0007 *** (0.0001)	−0.0007 *** (0.0001)	−0.0007 *** (0.0001)	−0.0008 *** (0.0001)	−0.0007 *** (0.0001)
Male	0.2929 *** (0.0081)	0.2692 *** (0.0470)	0.2692 *** (0.0470)	0.2970 *** (0.0516)	0.3209 *** (0.0146)
Public sector employment	0.2431 *** (0.0355)	0.2450 *** (0.0354)	0.2450 *** (0.0354)	−0.2621 *** (0.0693)	0.2475 *** (0.0422)
Male × age			0.0003 ** (0.0012)
Public sector employment × age				0.0115 *** (0.0014)
Public sector employment × male				−0.0894 * (0.0675)	−0.0930 * (0.0165)
Public sector employment × age × male				0.0002 * (0.0017)
Control variables	Controlled	Controlled	Controlled	Controlled	Controlled
Random effect covariance parameters
Cohort-level intercept SD	0.0000 (0.0002)	0.0000 (0.0001)	0.0000 (0.0002)	0.0000 (0.0002)	0.0485 (0.2258)
Cohort-level slope SD (gender)		0.0000 (0.0004)
Period-level intercept SD	0.4951 *** (0.0451)	0.4819 *** (0.0442)	0.4819 *** (0.0028)	0.4898 *** (0.0449)	0.4835 *** (0.0493)
Period-level slope SD (gender)		0.0723 *** (0.0133)	0.0723 *** (0.0133)	0.0633 *** (0.0130)	0.0654 *** (0.0130)
Period-level slope SD (public sector employment × male)					0.0002 * (0.0009)
Residual SD (individual level)	0.6743 *** (0.0028)	0.6735 *** (0.0028)	0.6717 *** (0.0028)	0.6713 *** (0.0028)	0.6714 *** (0.0028)
AIC	61,813.92	60,512.73	60,358.86	60,491.44	60,329.66
BIC	61,929.95	60,653.62	60,499.75	60,632.33	60,495.41
Individuals	0.4213 *** (0.0035)	0.4206 *** (0.0035)	0.4207 *** (0.0035)	0.4180 *** (0.0035)	0.4204 *** (0.0035)
AIC	58,152.00	58,130.95	58,142.37	57,961.35	58,109.45
BIC	58,153.46	58,120.95	58,134.37	57,953.35	58,101.45

Note: (1) *** p < 0.001, ** p < 0.01, * p < 0.05. (2) Standard errors are reported in parentheses. (3) The control variables are the same as in Table 1.

, the direction and significance of the fixed effects largely align with HAPC-CCREM analyses. Specifically, the effects of age and age-squared were found to be significant, while the main effects of male gender and formal employment were positive and significant. Notably, the interaction effect between gender and formal employment remained stable in direction. With respect to the random effects, the period intercept and gender random slope maintained statistical significance, while the cohort intercept variance exhibited a reduction in magnitude but demonstrated consistent trends. The residual variance demonstrated negligible change, suggesting a robust model fit. Overall, the robustness tests support the reliability of the HAPC-CCREM analysis conclusions, thereby confirming the high credibility of the preceding interpretations regarding age, period, cohort, and key interaction effects.

6. Discussion of Results

This study analyzes gender income inequality in China’s segmented labor market using a life-course perspective and the HAPC model, finding that, in the overall sample, gender income disparities persistently widen with age during the childbearing years (period β = 0.0027, p < 0.05), and this trend intensified between 2003 and 2021, showing intergenerational continuity (cohort β = 0.0004, p > 0.1). A second key finding is that the age effect differs significantly between the public and private sectors (β = 0.0004, p = 0.0073). In the public sector, the income curve peaks later, showing a gender convergence effect overall, while the private sector income curve is steeper, demonstrating a gender reversal effect. Lastly, the random effect of gender at the period level shows significant sectoral differences (β = 0.0012, p = 0.0423). When the twelve period-specific coefficients are plotted in chronological order, the gender effect remains stable in the public sector, while the gender effect follows a three-phase pattern in the private sector: an increase from 2003 to 2008, a decrease from 2010 to 2013, and a resurgence from 2015 to 2021. This pattern aligns with China’s economic restructuring [88], where market expansion and factor restructuring intensified gender inequality in the private sector, while the rise of knowledge-intensive industries was associated with a narrowing gender gap. However, structural reforms after 2015 deepened the occupational gender division. In contrast, the public sector’s standardized pay structure and stable redistribution system have maintained a relatively low gender gap. These findings suggest that institutionalized gender inequality is pervasive in Chinese society, influencing income returns across age, period, and cohort and constituting an important social foundation for the reproduction of inequality in contemporary China [89,90]. This underscores the importance of SDG 5, highlighting that economic growth alone, without targeted policy interventions, cannot address deeply entrenched gender inequality.

6.1. Theoretical Implications

Our study proposes a more process-oriented and structural analytical framework in contrast to existing research on SDG 5, which primarily conceptualizes gender equality as a performance indicator for sustainable development [91,92]. By defining the attainment of gender equality as a fundamental mechanism embedded in women’s life courses and incorporated into sustainable gender governance frameworks, our study contributes to three theoretical implications.

6.1.1. From Stability to Structuration: Transcending Gendered Labor Market Segmentation Theory from a Life Course Perspective

Traditionally, labor market segmentation has been studied as a relatively stable structural outcome, with a focus on the stable shaping of gender income inequality by institutional barriers and sectoral segmentation [93]. The requirements for “sustained people-centered development” as stated in SDG 5 are not met by this analytical approach. Utilizing the life course framework, this study illustrates that gender income inequality is not a static stock but rather a structuration process that is generated by the interaction between age accumulation, period shocks, and intergenerational disparities. The divergence between public and private sectors can be attributed to the dynamic interplay across multiple temporal dimensions between sector-specific characteristics and gendered allocation of employment opportunities, promotion rhythms, and social reproduction responsibilities. This has significant implications for women’s access to stable income, environmentally sustainable employment, and social security benefits. Consequently, this study transcends the static assumptions of segmentation theory, revealing the temporal mechanisms through which gender inequality is continuously reproduced and restructured within sustainable development and decarbonization transitions.

6.1.2. From Cross-Sectional Comparisons to Process-Oriented Tracing of Empowerment Trajectories

The majority of the current research on SDG 5 relies on single-point or short-term data, thereby constraining the comprehension of gender inequality to static comparisons of disparity levels while paying scant attention to the cumulative nature and path dependence of women’s empowerment throughout the life course [94]. Our findings demonstrate that different trends in age and period effects of gender income inequality across public and private sectors reflect the divergent trajectories of women’s economic empowerment unfolding within various institutional structures and opportunity environments. Furthermore, unequal access to technology has the potential to exacerbate already-existing disparities in empowerment, thereby solidifying gender inequality at both the procedural and structural levels, given changes in the sustainable development model and the emergence of new technologies as crucial elements in the decarbonization process. Gender income inequality should therefore be conceptualized not merely as an “outcome variable” but rather as a manifestation of empowerment pathways shaped by the interplay of institutional change, technological transformation, and life course dynamics. By shifting the focus from gap identification to the concept of empowerment trajectory governance, our study contributes to the advancement of research related to SDG 5 and provides a theoretical framework for the development of process-oriented, structured governance that is sensitive to gender dynamics.

6.1.3. Extending Ontological Commitment

At the ontological level, our study does not treat gender income inequality as a linear outcome of individual endowment differences or external institutional constraints. Instead, it is defined as a sustainable empowerment structure embedded within institutional arrangements, organizational practices, and life course trajectories. The following core implications are posited. First, gender inequality is embedded within economic models, industrial structures, and organizational governance mechanisms and affects women’s opportunities for participation in employment, leadership, resource access, and low-carbon transition. Secondly, the accumulation of economic, social, and technical capabilities across the life course is imperative for women’s empowerment and enables effective engagement in green employment and sustainable development. Thirdly, achieving gender equality is contingent upon institutional regulation and structural interventions throughout life courses to ensure women’s equal participation in political, economic, and public decision-making spheres. Consequently, our study proposes a sustainability framework for SDG 5′s gender governance practices, emphasizing the significance of gender equality as a governance practice. It involves the sustained temporal coordination between life trajectories and institutional structures and serves as the core mechanism for advancing women’s empowerment and decarbonization transformation.

6.2. Practical Implications

The practical implications lie in reinterpreting pathways for sustainable empowerment of female workers within the context of the decarbonization transition.

First of all, based on the divergent age effect trajectories across the public and private sectors, female workers undergoing decarbonization transitions must transcend the path dependency of regarding the public sector as a “safe harbor” for gender equality. It is imperative for them to transition toward establishing cross-sectoral economic resilience and the practice of subjectivity centered on life course dynamics instead. From a life course perspective, the public sector relies on seniority systems to establish a robust “age–income” relation structure, which not only leads to the convergence of gender income disparities in mid-to-late careers but also provides women with economic security expectations throughout life courses and enhanced bargaining power in household and societal division of labor [95]. However, this stable empowerment pathway is deeply intertwined with state redistribution mechanisms, rendering gender equality more dependent on external institutional structures than on workers’ transferable skills and sustained competencies [96]. In contrast, within the private sector, governed by effective and competitive market mechanisms, gender income inequality in early careers tends to widen due to market dynamics, skill preferences, and gendered cultural conditioning, whereas with the growth of green industries, the digital economy, and service-oriented roles, women’s educational capital and communication skills become increasingly transferable to sustainable employment opportunities in mid-to-late careers [97]. Consequently, women can develop cross-institutional and cross-sectoral mobility within their empowerment trajectories by planning career paths across sectors and accumulating skills aligning with low-carbon transition pathways.

Secondly, based on the sectoral sensitivity differences, the period effects of gender income inequality indicate that advancing gender equality requires “differentiated governance” to address the cyclical characteristics of segmented labor markets in transition economies. From a macro-period perspective, the public sector—embedded within the state’s redistribution system—consistently functions as a “stabilizer” across developmental stages [98]. The low-volatility income structure provides women with predictable occupational security and risk buffers. Furthermore, the institutional inertia of the planned economic era in question stabilizes women’s participation in public affairs and social governance [99]. This stability is not an inherent characteristic of the public sector; rather, it comes from the state’s strategic prioritization of stabilizing resource-intensive, high-carbon industries during economic reform and leads women’s macroeconomic security to be contingent upon state macro-control rather than market competitiveness or green technological capabilities [100]. Conversely, the gender income structure in the private sector exhibits heightened sensitivity to economic cycles and industrial restructuring: gender division of labor intensifies during the period of capital expansion, female workers gain advantages during the rise of knowledge-intensive industries, and the structural reform period has confined female workers to high-flexibility, low-security positions. Consequently, the advancement of SDG 5 is not contingent on the pursuit of “sheltered equality” within a specific sector. Instead, it is contingent on an enhancement in women’s cognitive capacity concerning labor market segmentation, industrial transition rhythms, and green transition windows [101]. This enables them to proactively identify risks, seize green employment opportunities, and navigate asymmetric fluctuations in the transitioning economy across different macroeconomic cycles.

Thirdly, the absence of cohort effects signifies that gender inequality exhibits structural inertia across generations [102]. Achieving this objective necessitates the establishment of sustainable institutional transmission structures across generations, thereby transforming the gender equality goals of SDG 5 from short-term policy adjustments into continuous governance practices that span the life course and generational transitions.

6.3. Limitations and Future Research

When interpreting our findings, it is also important to acknowledge certain theoretical and methodological limitations. Firstly, this research is grounded in China’s transition experience, where its sectoral segmentation structure, green transition pathways, and characteristics of state–market relations exhibit high scenario dependency. The extrapolation of these elements to transition economies with disparate political systems or stages of development may present a significant challenge. Secondly, this study’s focus on income dimensions may fail to encompass other dimensions of gender empowerment emphasized by SDG 5, such as access to economic resources, distribution of unpaid care responsibilities, and leadership in decision-making [103]. This risks underestimating the latent inequalities women face during decarbonization transitions, including risks to informal employment, marginalization during technological iteration, and path dependencies in household care responsibility allocation. Thirdly, while the HAPC-CCREM framework alleviates the constraints of traditional APC models in identifying age, period, and cohort, during periods of intensive decarbonization policy implementation, period effects often exhibit nonlinear, phased, or even discontinuous transitions. This renders the HAPC model susceptible to potential confounding risks between period and generation effects.

Therefore, future research should deepen along three pathways. First, future research should expand the institutional comparative perspective. Theoretically, it should explore pathways to sustainable gender equality under different institutional contexts by examining labor market segmentation, public–private sector disparities, and decarbonization pathways within the framework of institutional economics. Secondly, future research should introduce sequential analysis to dynamically map women’s development trajectories and empowerment pathways at the micro level within green employment systems, carbon-differentiated industries, and digitally intensive roles with high technological substitution potential. This will reveal how gendered social structures persistently shape gendered life opportunities during sustainable transitions. Thirdly, future research should incorporate macro-level institutional variables, including carbon pricing mechanisms, green fiscal subsidies, and gender support policy coverage, into a comprehensive APC model [104]. This approach will provide a more comprehensive perspective on the interactive mechanisms between decarbonization transitions and sustainable gender empowerment, offering theoretical underpinnings and empirical evidence for achieving sustained gender equality governance.

7. Conclusions and Policy Recommendations

This study introduces a life-course perspective into the discussion of SDG5 and, using the HAPC framework, systematically compares the age, period, and cohort effects on gender income inequality between China’s public and private sectors. Our findings demonstrate that gender income inequality does not stem from a mere “residual bias” in economic transition. Rather, it is a structural force that is perpetually reproduced across multiple temporal dimensions through the mechanisms inherent in sectoral institutions. The age effects reveal that the divergent institutional definitions of the “age–income” relationship in the public and private sectors give rise to distinct sectoral patterns of gender income inequality, operating under the “seniority–reward” and “productivity–reward” mechanisms. The period effects reveal that institutional inertia within the public sector has not translated into gender progress. In contrast, the private sector’s gender income structure undergoes continuous restructuring in response to economic cycles, industrial upgrading, and green economic expansion, exhibiting multidimensional evolution characterized by both opportunities and constraints. Finally, the cohort effects are statistically insignificant, indicating that gendered income structures exhibit intergenerational continuity. Overall, policymakers should prioritize gender income inequality as an institutional governance priority, rather than merely an outcome metric. Within the overarching institutional design of decarbonization transformation, a forward-looking, full-cycle gender governance system calibrated to the life course must be embedded. This will enable women to transition from marginal participants in the green transition to structural actors and key beneficiaries.

For transitional economies, it is imperative to establish a gender mainstreaming implementation framework for the “Fairness Transition,” confronting the inherent challenges of structural segmentation and sectoral heterogeneity within labor markets [105,106]. To ensure the deep integration and synergy of decarbonization processes with gender equality practices, the implementation of tiered governance is necessary, which involves the consolidation of equality foundations in the public sector, alongside the establishment of long-term empowerment mechanisms in the private sector. These measures are crucial for ensuring the participation rights and benefit entitlements of women throughout the green transition. For the public sector, efforts should focus on establishing synergistic mechanisms between the “green transition” and “gender equality practices” in high-carbon stock sectors [107]. State-owned enterprises in energy, industry, and transportation should integrate gender impact assessments, equitable employment standards, and female skill retraining programs into green technology upgrades and capacity enhancements. This transforms traditional closed-loop stability models (“seniority-based compensation”) into open-loop equality structures (“green skills–development opportunities–fair returns”), empowering women as active participants and beneficiaries of the green transition rather than passive bearers of externalized costs. For the private sector, it is essential to establish a “Full-Cycle Friendly” market ecosystem in green economy growth industries [108]. Addressing the structural feature of sharply widening private sector gender pay gaps in early careers, female empowerment mechanisms should be pre-established in emerging industries, such as renewable energy, circular economy, and digital carbon management. To mitigate the challenges faced by women in early careers, measures including pay transparency, oversight, anti-discrimination hiring practices, inclusive childcare support, and flexible work arrangements should be implemented. Furthermore, leveraging the public sector’s institutional wisdom of “delaying the peak,” lifelong learning, green skills certification, and career transition mechanisms should transform the “gender reversal” trend emerging in mid-to-late careers from a passive structural effect into active empowerment based on capability.