Welcome to follow

The Jounal of Chinese Sociology

2025年7月1日，The Journal of Chinese Sociology（《中國社會學學刊》）上線文章Trends in women’s education and fertility in Japan and China: a comparative analysis（《中國和日本女性的受教育程度和生育率：一項比較分析》）。

作者簡介

張佳潔，東京大學博士研究生。主要研究方向：社會分層、家庭社會學。

Keywords

Fertility behavior; Educational attainment; Cohort changes; Japan and China

Abstract

East Asian countries have undergone significant demographic shifts characterized by declining fertility rates and the expansion of higher education opportunities for women. Adopting a comparative perspective, this study examines cohort-based changes in the relationship between women’s educational attainment and fertility behaviors in China and Japan via discrete-time event history models and difference-in-differences analysis. The results reveal divergent trends: Japan shows a weakening negative association between higher education and fertility over time, whereas China demonstrates a persistent trend influenced by historical population control policies. These findings underscore the crucial roles played by sociocultural contexts and policy environments in shaping women’s fertility behaviors. Although China can draw lessons from Japan, the unique features of the Chinese socioeconomic landscape call for more targeted policies to address low fertility. This study further illustrates that ultralow fertility can emerge even in societies where traditional family values remain strong, contributing to broader discussions on the second demographic transition in East Asia.

Introduction

The proliferation of higher education opportunities has altered the educational composition of the female population of childbearing age, resulting in a significant increase in the number of highly educated women. Despite this shift, the debate surrounding the persistent negative association between women’s education and fertility rates is ongoing. Researchers continue to explore the complex interplay of factors that contribute to this relationship, seeking to understand the underlying mechanisms and to identify potential solutions.

This relationship demonstrates distinct mechanisms across fertility stages, historical periods, and nations. For example, over the past few decades, Western countries have experienced a second demographic transition alongside the expansion of higher education opportunities for women. In this context, several studies have reported a negative relationship between women’s educational attainment and completed fertility (Weinberger 1987; Jejeebhoy 1995; Martin 1995; Cleland 2002; Howe 2016). However, other studies have shown that welfare and social policies can mitigate these negative effects (Breierova and Duflo 2004; Monstad et al. 2008; McCrary and Royer 2011; Geruso and Royer 2018; Kan and Lee 2018; Kramarz et al. 2023) or have highlighted a positive association between women's educational attainment and fertility behaviors (Braakmann 2011; Fort et al. 2016). Given the diverse findings across various contexts, examining heterogeneity across populations, societies, and cohorts is crucial to obtain an in-depth understanding of the relationship between women’s educational attainment and fertility behaviors. However, there is limited research on how this relationship has evolved over time (Lazzari et al. 2021).

Among the regions experiencing rapid demographic changes, Asia has shown the fastest decline in fertility rates. Many Asian countries have witnessed a significant drop in birth rates over the past few decades, often accompanied by an increase in women’s educational attainment. This trend has led to growing concerns about the consequences of population aging and the sustainability of economic growth in these regions.

China and Japan are two noteworthy cases within Asia. Both countries have experienced dramatic declines in fertility rates, with Japan being one of the first countries to enter the “ultralow fertility” category and China recently following suit. These two countries have also experienced rapid expansion of higher education, with women’s tertiary enrollment rates surpassing 50% (Gender Equality Bureau 2020; China Ministry of Education 2021). Moreover, China and Japan represent two distinct paths of economic development and social change. Japan experienced rapid economic growth in the postwar decades, followed by gradual deceleration, culminating in extended stagnation after the economic bubble in the 1990 s collapsed. In contrast, China has undergone a more compressed and accelerated transformation, achieving industrialization and urbanization, particularly following the economic reforms of the late 1970s.

The comparison of China and Japan provides valuable insights into fertility behaviors in East Asia because of their shared cultural foundations and similar demographic challenges. Both countries exhibit a strong link between marriage and fertility, with minimal nonmarital childbearing (Raymo et al. 2015), and have witnessed significant increases in women’s educational attainment and labor force participation in recent decades (Raymo et al. 2015; Yu and Xie 2015). Rapid economic development has brought about substantial changes in social norms and family structures in both societies (Kumagai 2010; Qian and Qian 2015). Furthermore, both countries are grappling with the challenges of aging populations and the associated socioeconomic implications (Muramatsu and Akiyama 2011; F. Wang and Wang 2021). Notably, both China and Japan have maintained relatively low levels of immigration, with foreign-born individuals constituting a small proportion of the population (Liang 2016; Igarashi et al. 2022).

Despite these similarities, the timing and pace of demographic challenges differ between Japan and China. Japan has been addressing low fertility and increasing female labor force participation since the 1990 s, thus offering a 30-year precedent for studying long-term effects and policy responses (Atoh 2017). In contrast, China’s more recent economic development, coupled with decades of strict population control policies, has led to more severe demographic challenges (Zhao and Chen 2011; Feng et al. 2013). An examination of Japan’s experiences can provide insights into potential demographic trends in China and identify effective policy interventions. This comparative approach can shed light on how cultural similarities interact with economic development, policy measures, and social changes to shape fertility decisions in East Asia.

This study examines how educational attainment shapes fertility patterns across different cohorts in China and Japan. Using cohort analysis and cross-national comparison, we analyze the evolving education?fertility relationship while testing the impact of China’s one-child policy abolition through difference-in-differences methodology. This comparative approach reveals distinct mechanisms operating in each country’s unique socioeconomic context. Our contribution is twofold: first, we illuminate how education?fertility dynamics have shifted over time in two East Asian societies with contrasting policy approaches to low fertility, thus potentially refining existing theoretical frameworks; and second, we provide evidence on how changing gender roles and family structures create challenges in parity progression across educational strata.

Literature review

Education and fertility

in China and Japan

Gary Becker’s (1960) economic theory of fertility posits that the higher opportunity costs of childbearing for more educated women lead to lower fertility rates. This relationship is moderated by policies supporting work?family balance, such as childcare services and parental leave (Kravdal and Rindfuss 2008; Del Boca et al. 2009). In countries with comprehensive support systems, such as Nordic nations, the negative association between women’s educational attainment and fertility behaviors is weaker (Kravdal and Rindfuss 2008; Andersson et al. 2009). However, increased competition among highly educated women in the labor market may strengthen this negative correlation, especially in countries lacking adequate social policy support.

China and Japan exemplify contrasting policy approaches to low fertility rates. Japan has implemented various work?family balance policies, including expanded childcare services and the promotion of paternal involvement (Suzuki 2013). In contrast, China has focused on abolishing population control policies without providing sufficient social welfare benefits for childbearing ( Chen 2021). This disparity in policy focus may lead to differing impacts on the relationship between women’s education and fertility behaviors in these countries.

Both China and Japan have experienced substantial expansions in women’s higher education opportunities over the past several decades, albeit under different socioeconomic conditions and policy regimes. In China, higher education began transitioning from an elitist system to a mass and universal system in the late 1970 s, following the Cultural Revolution and the reinstatement of the national college entrance examination in 1977 (Yan and Xue 2019). Enrollment rates increased steadily—from 2.8% in 1985 to 8.4% in 1999—and then accelerated after 1999, when universities adopted tuition-based, market-oriented admission policies. By 2019, the enrollment rate exceeded 50%. Over the same period, China implemented a one-child policy (1980–2015), which limited most couples to a single child and contributed to increased educational attainment among women in single-child families (Lu et al. 2019). This policy was phased out in 2015, initially allowing all couples to have two children and later permitting up to three. However, declining fertility persists, reflected in the decrease to 9.56 million total births in 2022 and a total fertility rate (TFR) falling below 1.3, with second- and third- or higher-order births accounting for 38.9% and 15.0% of all births, respectively (National Health Commission of the People’s Republic of China 2023). Policy reforms such as allowing couples to have more than one child have not reversed the downward trajectory, partly because childcare responsibilities still fall overwhelmingly on families and, in many cases, on mothers (Ji et al. 2017; Jing Zhang and Emery 2023). Urban reforms have dismantled the work-unit (" danwei (單位)") system that once subsidized childcare, leaving private enterprises and individuals to bear the costs (Du and Dong 2013). Traditional patriarchal norms, persistent labor discrimination, and insufficient formal childcare resources have further exacerbated the opportunity costs of childbearing for educated Chinese women (Connelly et al. 2018; Wang and Zhao 2022).

Japan has also undergone multiple waves of higher education expansion, initially from the postwar period until the mid-1970s and then more rapidly from the 1990 s onward, once regulations on establishing new universities were relaxed (Toyonaga 2020). By 2005, the enrollment rate surpassed 50%, indicating universal access to higher education (Utagawa 2022). Although this expansion has enabled more women to join the labor force, Japan’s fertility dynamics differ from those of China. Rather than strict population control, Japan has grappled with declining birth rates since the 1970 s, with its TFR decreasing below replacement levels and remaining under 1.5 children per woman since the early 1990 s (Atoh 2017). Although the government promoted measures such as up to 2 years of parental leave and childcare allowances, the birth rate in Japan continues to decrease due to increasing nonmarriage. In 2020, the lifetime unmarried rate reached 28.3% for men and 17.8% for women (National Institute of Population and Social Security Research 2022). The economic bubble of the 1990 s led to employment instability, which reduced men’s income prospects and complicated families’ long-term financial plans (Tsutsui 2015). Moreover, Japanese women’s increasing educational attainment fosters greater economic independence and higher expectations for partnership, which, in turn, can delay or deter marriage and childbearing (Mugiyama 2017). Although government policies offering economic support for childrearing and work–life balance have moderated the impact of having children on women’s employment, many are still reluctant to marry or have larger families, citing insufficient policy effectiveness and rigid workplace norms (Atoh et al. 2011).

Despite these different historical legacies, China and Japan share the challenges posed by ultralow fertility, increasing female educational attainment, and the strain of balancing work and family. Both countries’ policy interventions focus on encouraging childbirth during marriage—either by loosening restrictions (as in China) or by providing economic support (as in Japan). Nevertheless, the structural and cultural realities faced by each country appear to be outpacing current policy responses. As fertility continues to decline, policy-makers in both countries must address evolving gender roles, increasing labor discrimination against mothers, and the high opportunity costs of childbearing, all of which profoundly shape the fertility decisions of highly educated women (Zhang 2023).

Hypotheses

The relationship between women’s educational attainment and fertility behaviors operates through several interconnected mechanisms that vary across sociocultural contexts. By examining these mechanisms in China and Japan, we can develop empirically testable hypotheses about education–fertility patterns in each country.

Becker’s (1960) economic theory posits that higher education increases women’s earning potential, thereby increasing the opportunity costs of childbearing. This mechanism predicts that highly educated women will delay or forego childbearing due to career considerations. However, the strength of this effect depends critically on institutional support systems. In contexts with comprehensive work?family policies, the negative relationship between education and fertility may weaken (Kravdal and Rindfuss 2008; Del Boca et al. 2009). In China, we observe limited development of childcare infrastructure following market reforms (Du and Dong 2013), with persistent workplace discrimination against mothers that disproportionately affects educated women (Ji et al. 2017). This observation leads us to propose the following:

Hypothesis 1a : In China, the negative relationship between women’s educational attainment and the timing of the first birth persists across cohorts, reflecting the enduring influence of traditional gender roles and the limited development of childrearing support measures.

In contrast to the opportunity cost perspective, higher education can increase the amount of resources available for childrearing. Educated women may have a greater financial capacity to support larger families and better access to information about reproductive health. In China, Shuang Chen (2022)reported a positive effect of education on fertility in low-fertility regions, suggesting that resource advantages may outweigh opportunity costs in certain contexts. Following the relaxation of China’s population control policies, some evidence indicates that educated women have a greater propensity for second births ( Chen and Duan 2019; Chen and Gu 2020). This research leads us to propose the following:

Hypothesis 1b: In China, the relationship between women’s educational attainment and the likelihood of having a second child differs from that of first births, with a potentially weaker negative association or even a positive association for highly educated women in recent cohorts.

The institutional context shapes how education influences fertility by determining the compatibility of motherhood with employment. Japan has implemented various work?family balance policies, including extended parental leave and childcare expansion (Suzuki 2013), which may reduce the opportunity costs of childbearing for educated women. Raymo et al. (2015) reported increasing acceptance of maternal employment in Japanese society, whereas Setsuya Fukuda et al. (2020) found that highly educated women now show greater marriage propensity than in previous decades. These institutional shifts lead us to hypothesize the following:

Hypothesis 2a : In Japan, the negative relationship between women’s educational attainment and the timing of the first birth has weakened in recent cohorts, reflecting changing societal norms and policy environments.

The education?fertility relationship evolves across cohorts as societal conditions change. In Japan, Cyrus Ghaznavi et al. (2022) and Kazuyasu Sakamoto (2023) observed a weakening negative association between education and fertility in recent cohorts, particularly for second births. Yuko Nozaki (2017) further reported that work–family balance policies have had stronger positive effects on fertility intentions among educated women. As institutional adaptations progress and gender norms evolve, we expect the following:

Hypothesis 2b: In Japan, the negative relationship between women’s educational attainment and the likelihood of having a second child has weakened in recent cohorts, suggesting a convergence in fertility behaviors across educational groups.

By testing these hypotheses through cohort analysis and cross-national comparison, we can better understand how socioeconomic contexts and policy environments mediate the education?fertility relationship in East Asian societies experiencing ultralow fertility.

Data and methods

Analytical framework

This study employed an empirical approach to investigate the factors influencing women’s fertility behaviors. The analytical framework was structured as follows: First, a descriptive analysis was performed to examine the temporal progression from marriage to first birth and then first birth to second birth, stratified by women’s education levels, thus providing an overview of childbearing patterns across educational groups. Second, a discrete-time event history model was used to analyze the relationship between women’s educational attainment and childbearing events (‘marriage to first birth’ and ‘first to second birth’) by birth cohort, assessing the impact of education on parity transitions while accounting for cohort effects. Third, a DID approach was used to estimate the impact of the abolition of China’s population control policy on fertility, comparing the outcomes of affected (treatment) and unaffected (control) women before and after the policy shift and identifying its causal effect while controlling for confounding factors. Finally, by synthesizing the results comparatively, the study elucidates the similarities and differences in fertility behaviors between China and Japan, considering their distinct socioeconomic, cultural, and policy contexts, and provides insights into the factors shaping childbearing decisions in East Asian societies.

Data sources

and key measures

Data

For China, data from waves 1 to 6 of the China Family Panel Studies (CFPS) (from 2010 to 2020) were used to analyze the Chinese societal context. The CFPS is a longitudinal survey that has been conducted biennially by Peking University since 2010, with samples drawn through stratified multistage random sampling in 25 provinces. The number of recovered samples was 42,590, with a household response rate of 81.3% and an individual response rate of 84.1% (Xie and Lu 2015).

This study analyzed both the transition to marriage and the transition to first and second births. For the marriage analysis, the sample included unmarried women aged 16 years and older. The analysis began by determining the time from the age of 16 years to marriage, creating person–year data until the event of marriage or censoring the data if marriage did not occur by 2020.

For the fertility analysis, this study focused on first-married couples, excluding those who had children before marriage or with missing childbearing age values, resulting in 14,992 analyzed couples. The risk starting point for the first birth was the time of marriage, and observations continued until the first birth, creating person–year data, which were censored if childbirth did not occur by 2020 or if the woman had reached the age of 50 years. The period from first to second births was similarly analyzed and treated. A total of 14,446 couples had one child, and 9,057 had two children.

With respect to Japan, data from 29 surveys of the Japan Panel Survey of Consumers (JPSC) conducted by Keio University from 1993 to 2021 were used to analyze the Japanese societal context. When the JPSC started in 1993, it targeted 1500 women aged 24 to 34 years, with younger respondents added approximately every five years to avoid age gaps. By 2018, the target age group had expanded to include women aged 29 to 59 years. Sampling was conducted nationally via a stratified two-stage random sampling method. The average recovery rate for each survey throughout the study period was 95%, thus maintaining a high level of data reliability and consistency (Keio University Panel Data Research Center 2022).

This study analyzed both the transition to marriage and the transition to first and second births. For the marriage analysis, the sample included unmarried women aged 16 years and older. The analysis began by determining the time from the age of 16 years to marriage, creating person–year data until the event of marriage or censoring the data if marriage did not occur by 2021. This study selected age 16 as the earliest point to capture both early legal marriages in Japan and consensual unions in China. In Japan, individuals can legally marry at age 16 with parental consent. In China, 17% of the sample entered consensual unions between the ages of 16 and 19 years, before the legal marriage age of 20 years. By capturing data from age 16 onward, the study offers insights into early marriage patterns that might otherwise be overlooked in analyses starting at later ages and thus provides a comprehensive view of marriage patterns.

For the fertility analysis, this study focused on first-married couples, excluding those who had children before marriage and those with missing childbearing age values, resulting in 2,455 analyzed couples. The main analysis targeted married women, treating the year a single woman was married as the year of marriage. The period from marriage to first birth was analyzed, creating person–year data, which were censored if childbirth did not occur by 2021 or if the wife had reached the age of 50 years. The period from first to second births was similarly analyzed and treated. A total of 2,122 couples had one child, and 1,602 had two children. This approach allows for a more direct comparison between the two countries and aligns with their dominant cultural norms of family formation. According to the CFPS data, 6.87% of births occurred before marriage, whereas according to the JPSC data, this figure was 2.1%. These relatively small proportions align with national statistics in China and Japan (Li 2022; World Population Review 2024), further justifying the focus on marital fertility. Owing to limitations in the datasets regarding the fertility histories of remarried couples, the analysis concentrated on first-married couples to ensure consistency and accuracy.

The data were converted into person–year format to analyze the interval from a woman’s marriage to the birth of her first and second child (or age 50) via discrete-time logit models. The person–year format covers three periods: from age 16 to marriage, from marriage to first birth, and from first birth to second birth. This approach estimates the hazard of childbirth events without requiring an exact alignment of survey years between China (2010–2020) and Japan (1993–2021). Although this restructuring partially mitigates the issue of differing survey years, it does not fully control for time-varying variables with cross-sectional information, an inherent limitation of both the data and the chosen method. Despite these limitations, this approach offers valuable insights into the relationship between educational attainment and fertility behaviors in China and Japan.

Key measures

The dependent variables in the analysis were the occurrences of the following events: “from the age of 16 years to marriage,” “from marriage to first birth,” and “from first birth to second birth.” The key independent variable was women’s educational attainment, which was categorized into five groups in the Chinese data: “primary school,” “lower secondary,” “upper secondary,” “college (3 years),” and “university graduate.” In the Japanese data, the categories were “lower secondary,” “upper secondary,” “college (2-year degree),” and “university graduate,” as all participants had completed more than primary school.

This study analyzed temporal changes in the impact of women’s educational attainment on fertility behaviors across birth cohorts. The Chinese birth cohorts were divided into “before the 1960 s,” “1960s,” “1970s,” and “1980s and later.” The Japanese birth cohorts were divided into “1960s,” “1970s,” and “1980s and later.” In the Chinese data, the pre-1960s cohort was used to understand the impact of educational attainment on fertility behaviors before the one-child policy; however, the Japanese analysis focused on cohorts born in the 1960 s and later due to data scarcity. Other control variables include the number of houses owned by the family, urban or rural status, wife’s age at marriage, sex of the first child, wife’s age at first birth, and year of marriage.

Analytical method

This study also employs a multiperiod difference-in-differences (DID) framework (Beck et al. 2010) to analyze the staggered implementation of China’s partial and universal two-child policies from 2011 to 2015. This study begins by identifying women who, as of 2010, resided in areas enforcing either the one-child policy or a conditional two-child policy (which included exceptions for ethnic minorities or rural households with a firstborn daughter) and had not yet had a second child by the relevant policy start date. Three distinct treatment events are defined, corresponding to the dates when new groups of women became eligible to have a second child.

For the 2011 "shuangdu (雙獨)" policy, which allowed couples to have two children if both parents had been born under the only child policy, this study considers women treated if, by 2011, they were subject to the one-child policy or the conditional two-child policy in 2010, had not had a second child before 2011, and both spouses were only children, making them eligible under the Shuangdu policy. For the 2013 "dandu (單獨)" policy, which allowed couples to have two children if either of the parents was an only child, women are regarded as treated if, by 2013, they were subject to the one-child policy or conditional two-child policy in 2010 and had not had a second child before 2013 and at least one spouse was an only child, thus qualifying the parents under the "dandu(單獨)" policy. For the 2015 universal two-child policy, women are deemed treated if, by 2015, they were subject to the one-child policy or conditional two-child policy in 2010 and had not had a second child by 2013 and both spouses had siblings, which meant that the couple were newly eligible under the comprehensive two-child policy.

Results

Descriptive analysis

Tables 1 and 2 present the descriptive statistics of the variables used in the analysis for China and Japan, respectively. These tables provide an overview of the sample characteristics, including the distributions of women’s educational attainment, birth cohorts, and other relevant variables. Table 3 presents a crosstabulation of women’s educational attainment and birth cohorts at the time of marriage for both countries. This table shows the changes in educational attainment across cohorts. Notably, for cohorts born after 1980, both in China and Japan, the proportion of women with a college education increased significantly, reflecting the expansion of higher education in these societies.

Figures 1a–c present Kaplan?Meier survival curves for life events among Chinese women, whereas Fig. 1d–f show the corresponding results for Japanese women.

Figure 1a shows Kaplan?Meier survival curves from the age of 16 years to marriage in China. Chinese women with higher education levels tended to marry later, but most eventually married. Despite the trend toward later marriages, the low lifelong unmarried rate suggests that Chinese society remains a universal marriage society (Zhai et al. 2022).

Figure 1b shows that most couples in China progressed to the birth of a first child after marriage. log-rank and Wilcoxon tests (p?< 0.001) indicate significant differences in survival functions (duration until first birth) among groups with different educational levels, with more highly educated women having longer durations from marriage to first birth than their less educated counterparts did.

Figure 1c depicts Kaplan?Meier survival curves from the first birth to the second birth in China. The wife's educational attainment influenced the likelihood of having a second child, with higher education levels correlating with a lower probability of having a second child. The log-rank and Wilcoxon tests (p?< 0.001) revealed statistically significant differences among the educational groups. A higher level of education correlated with a lower probability of progressing to a second birth, whereas a lower level of education indicated a greater tendency toward higher parity.

Figure 1d shows the Kaplan?Meier survival curve for Japanese women from the age of 16 years to the year of marriage, illustrating the relationship between educational attainment and the timing of marriage. Like Chinese women, Japanese women with higher education levels tended to delay marriage. However, Japan has higher unmarried rates than China does, with approximately 25% of women with lower secondary education remaining unmarried throughout their lives.

Figure 1e reveals that married women with higher educational attainment—except for those with a bachelor’s degree—showed no substantial difference in the progression to first birth compared with those with lower educational attainment, despite the statistically significant results of the log-rank and Wilcoxon tests (p?< 0.001). These tests revealed a statistically significant difference in survival function (time until first birth) among groups with different educational levels. However, the actual differences in the survival curves were relatively small, suggesting that the effect of educational attainment on the timing of the first birth is less pronounced in Japan than in China.

Figure 1f displays the Kaplan?Meier survival curve for the interval between the first and second births among Japanese women. The log-rank and Wilcoxon tests (p?< 0.001) suggest statistically significant differences in the likelihood of having a second child based on the wife’s educational background. However, disparities in the progression to a second birth among Japanese women were less pronounced than those among Chinese women, as shown in Fig. 1c. This finding indicates that while educational attainment influences the likelihood of having a second child in both countries, the effect is weaker in Japan than in China.

Temporal variation in

the education–fertility link

Discrete-time model analysis results

for Chinese and Japanese women

Table 4 presents the discrete-time event history analysis results for the transitions to marriage, the first birth, and the second birth across the four birth cohorts in China. Figure 2a–c were drawn based on Table 4, which shows the changes in the predicted probabilities. Table 5 presents the discrete-time event history model results for the transitions to marriage, the first birth, and the second birth across the three birth cohorts in Japan. Figure 2d–f was drawn based on Table 5 and shows the predicted probabilities of the influence of women’s education levels on the results.

Figure 2a shows that higher educational attainment was associated with delayed entry into marriage across all cohorts. This finding reflects the prevalent notion of “establishing a career before starting a family” in Chinese society, where individuals consider marriage after completing their education and securing stable employment and income. With the extension of years spent in education, the age at marriage continues to increase for highly educated individuals.

Figure 2b indicates that for cohorts born before and during the 1960 s, women’s educational attainment did not significantly influence the timing of the first birth. During this period, China had a planned economy, and compulsory education had not yet been widely implemented, resulting in low educational attainment for most people. In this social context, fertility behaviors were influenced primarily by traditional beliefs and family planning policies, with women’s education levels having a limited impact on the timing of childbirth. In the cohort born in the 1970 s, women began to delay the birth of their first child due to higher educational attainment, but no significant difference was observed between college and university graduates. This period marked the beginning of China’s reform and opening-up, with a modest expansion of higher education that was still limited to a relatively small population. Influenced by traditional beliefs, many women chose to marry and have children soon after completing their education, and the postponement of childbearing among highly educated women was not widespread. The 1980 s cohort showed a persistent negative association between women’s educational attainment and age at the first birth, with university graduates exhibiting a greater propensity to delay childbearing. This finding reflects the comprehensive implementation of compulsory education and the massification of higher education in China during the 1980 s, which led to a more equitable distribution of educational resources, enabling women to benefit from more equal access to higher education. Highly educated women are more deeply influenced by modern ideas about marriage and childbearing, and they commonly postpone childbirth. Moreover, as educational attainment increases, women face greater pressure to develop their careers, making it difficult to balance work and family, which also contributes to delays in childbearing.

Figure 2c presents the predicted probabilities for the timing of the second birth on the basis of women’s educational attainment. For the cohorts born before 1960, in the 1960 s, and in the 1970 s, higher educational attainment was associated with delayed progression to the second birth. However, the cohort born in the 1980 s exhibited a distinct pattern, with a generally greater probability of having a second child than earlier cohorts did. This pattern can be attributed to the implementation of the “universal two-child” policy in 2016, which coincided with the childbearing years of the cohort born in the 1980s. Interestingly, women with university, high school, primary school, and junior high school educations were more likely to have a second child, whereas those with college degrees had the lowest probability. This difference suggests that in the era of higher education expansion, college-educated women may face greater challenges in balancing their career and family responsibilities, leading to a lower likelihood of having a second child. The effectiveness of population control policies, particularly the second-child policy, in shaping fertility behaviors will be examined further in Sect. 4.2.2.

Figure 2d illustrates that Japanese women across all cohorts tended to delay marriage as their educational attainment increased, mirroring the trend observed in China. Importantly, this analysis includes all women, even those who never marry. When analyzing only married women, as in Fig. 2e, this approach excludes unmarried women, most of whom are highly educated.

Figure 2e, which focuses on the timing of the first birth among married women, reveals that those born in the 1980 s demonstrate the fastest and earliest transition to the first birth. In the 1960 s cohort, women with upper secondary education comprised the highest proportion (refer to Table 4) and exhibited the earliest timing of the first birth. For the 1970 s cohort, college-educated women represented the largest group and progressed most rapidly to the first birth. Interestingly, among the 1980 s cohort—despite a significant increase in university-educated women—college graduates still had the earliest timing of the first birth. The gap in birth timing between women with lower and upper secondary educations has narrowed over time. Additionally, the tendency for university graduates to delay the first birth relative to college graduates has become less pronounced in recent cohorts. These patterns suggest two key trends. First, the expansion of higher education opportunities in Japan has reduced the likelihood of highly educated women entering marriage. Second, among women who marry, the effect of educational differences on birth timing has diminished. This nuanced relationship underscores the importance of considering both marriage and fertility decisions when examining the impact of education on family formation patterns in Japan.

Figure 2f shows that women born in the 1980 s, with the exception of those with lower secondary education, had the fastest timing of the second birth after having their first child. For the cohort born in the 1960 s, the association between higher educational attainment and the earlier timing of the second birth was less evident. However, for the cohort born in the 1970 s, higher educational attainment was associated with faster progression to the second birth, with university-educated women having the highest probability of having a second child. This finding indicates that women with lower educational levels tend to delay progression to higher parities. Among women born in the 1980 s and later, there was no significant difference in the timing of the second birth among high school, college, and university graduates. This finding suggests that the universalization of higher education in Japan has narrowed the time gap between the first and second births for university-educated women.

Multiperiod DID analysis

of the abolition of the one-child policy

This study employs a multiperiod DID (difference-in-differences) framework to analyze China’s staggered implementation of partial and universal two-child policies from 2011 to 2015. Table 6 presents the results of the multiperiod DID analysis, which examines the impact of China’s gradually relaxed fertility policies, culminating in the universal two-child policy. Figures 3 and 4 provide visual representations of these findings, offering insights into the policy’s effects across educational levels and over time.

The positive and statistically significant coefficient for the interaction term (“Treat#After”) in Table 6 indicates that the progressive easing of fertility restrictions significantly increased couples’ likelihood of having a second child. This finding underscores the overall effectiveness of the policy in stimulating fertility. However, the magnitude of this effect varies considerably across levels of women’s educational attainment, revealing a nuanced pattern of responses to the policy changes.

Figure 3 illustrates the predicted probability of having a second child by women’s education levels. Notably, women with lower secondary education levels demonstrate the strongest positive response to the policy changes. This group’s heightened responsiveness may be attributed to a combination of factors, including reduced opportunity costs of childbearing and potentially stronger adherence to traditional family values that favor larger families. In contrast, women with upper secondary and three-year college degrees exhibit a more muted response. This difference could be due to competing priorities between career advancement and family expansion or perhaps a more cautious approach to family planning despite the policy relaxation. The relatively subdued response from this group suggests that policy changes alone may not be sufficient to overcome other socioeconomic factors influencing fertility decisions for women at these educational levels. Interestingly, university-educated women show a slightly stronger response than do those with upper secondary and college degrees, albeit the response is still less pronounced than that of women with lower secondary education. This pattern could be explained by their greater financial resources and potentially more flexible work arrangements, which might facilitate the decision to have a second child. However, the fact that their response is not as strong as that of women with lower secondary education indicates that higher education continues to exert a moderating influence on fertility decisions, even under more permissive policies.

The robustness check shown in Fig. 4 was performed to examine parallel trends, revealing the dynamic effects of the policy changes. The sharp rise in second-birth rates shortly after each policy relaxation suggests an immediate positive response from couples who were previously constrained by the restrictions. The gradual decrease in this effect suggest two possible explanations: the policy’s impact has reached its maximum potential, or additional factors beyond policy measures have emerged to constrain fertility rates.

In conclusion, while China’s shift toward a universal two-child policy has successfully encouraged second births overall, its impact is far from uniform across educational strata. The policy changes have effectively reduced the opportunity costs of childbearing for less educated women while allowing highly educated women to leverage their stronger financial positions to expand their families, albeit to a lesser extent than their less educated counterparts did.

Comparison between

China and Japan

In China, the negative association between higher educational attainment in women and childlessness has emerged in recent cohorts, partially supporting Hypothesis 1a. For cohorts born before and during the 1960 s, there was no clear negative correlation between women’s educational attainment and the timing of the first birth. However, for women born in the 1970 s and 1980 s, the negative correlation became more pronounced, with highly educated women exhibiting a lower propensity for having a child.

The decision to have a second child varies among families, with higher educational attainment consistently associated with delayed second birth across most cohorts. However, the negative correlation between women’s educational attainment and second birth behavior diminished in the 1980 s, which was likely influenced by the abolition of the one-child policy in 2015. Some highly educated women in this cohort benefited from the policy change and had relatively more family resources and thus chose to have a second child. Nevertheless, the positive effects of the policy gradually weakened, partially supporting Hypothesis 1b.

The growing proportion of highly educated women in Chinese society has led to a decreased birth rate, aligning with Becker’s theory of fertility opportunity costs, which suggests that increased educational attainment leads to higher opportunity costs of childbearing and, consequently, lower fertility. As women’s educational attainment continues to increase, the negative association between women’s educational attainment and fertility behaviors is expected to persist, posing challenges for China’s future population growth and age structure.

In Japan, the relationship between higher educational attainment in women and a lower propensity for childbearing has weakened due to the increasing proportion of highly educated women, which supports Hypothesis 2a. In the 1960 s, higher educational attainment hindered the progression to the first birth, but this relationship reversed in the 1970 s and 1980s. The cohort born in the 1970 s showed a positive association between women’s educational attainment and the timing of to first birth, particularly among college graduates, who showed the earliest timing. This relationship persisted in the cohort born in the 1980 s, but the differences between lower and upper secondary graduates and between college and university graduates diminished.

Hypothesis 2b was also supported, as the negative association between higher educational attainment in women and progression to the second birth was attenuated. In the 1970 s, higher educational attainment among women was associated with earlier progression to the second birth. However, among women born in the 1980 s, no discernible difference in the timing of the second birth was observed between upper secondary, college, and university graduates. The weakening association between women's educational attainment and fertility behaviors in Japan suggests that Becker’s theory of fertility opportunity costs may not fully explain the observed patterns. The expansion of higher education and the implementation of policies supporting work?family balance have likely contributed to the attenuation of the negative relationship between women’s educational attainment and fertility behaviors in Japan.

Discussion

Low fertility in China

China has entered an era of persistently low fertility, although it has followed a path distinct from that of Japan. While the unmarried rate is gradually increasing, marriage remains nearly universal for older cohorts (Zhai et al. 2022). The findings reported in this study indicate that women’s higher educational attainment coincides with delayed first births and a reduced propensity for second births, largely due to higher opportunity costs. Among women born before the 1970s , those with university degrees enjoyed comparatively better labor market opportunities, but subsequent cohorts have faced greater discrimination and career obstacles, leading to additional delays in marriage and childbearing.

The abolition of the one-child policy in 2015 eased the formal restrictions on second births but generated unequal effects across different educational groups. Women at both the lower and higher ends of the educational continuum displayed stronger fertility responses than those with intermediate (college-level) education did. This pattern suggests that, among less educated women, decreased policy barriers lowered the financial and logistical hurdles of having a second child, whereas highly educated women, despite facing substantial opportunity costs, were able to leverage greater economic resources or workplace advantages to accommodate an additional birth. Nonetheless, China’s policy measures remain largely confined to lifting fertility restrictions rather than establishing robust family support systems such as affordable childcare or comprehensive parental leave.

Low fertility in Japan

In Japan, low fertility is attributable primarily to rising rates of nonmarriage. Among those who marry, the traditionally negative relationship between female education and fertility has weakened (Raymo et al. 2015). Two interrelated processes help explain this change. Japan has introduced limited policy and corporate measures—such as parental leave provisions and daycare subsidies—that have begun to reduce the immediate costs of childbearing for working mothers. In addition, marriage itself has become selective: Women who find traditional gender roles unappealing may opt to remain single, thereby influencing Japan’s overall fertility rate via nonmarriage rather than within-marriage fertility declines. Despite the expansion of these moderate support programs, persistent workplace gender inequality and societal expectations of intensive motherhood continue to impede a significant recovery in Japan’s fertility rate, given that many women remain discouraged by the incompatibility of full-time career aspirations with traditional family demands.

Explaining China–Japan

divergence and overlaps

A key finding from this comparative perspective is that while both countries are experiencing low fertility in the context of increasing female education, the dominant pathways differ. China carries the legacy of the one-child policy, which both established strong norms against large families and entrenched institutional structures limiting second births. Moreover, a comprehensive support system for working parents is lacking, thus increasing the opportunity costs of having additional children. Japan, in contrast, has undergone a more extended transition to low fertility and has implemented partial, although imperfect, measures to facilitate childbearing. The emphasis on marriage-based family policies, however, overlooks the growing prevalence of singlehood. Although China’s current marriage rates remain high, the rapid pace of socioeconomic change could lead it to follow Japan’s trajectory of long-term low fertility and delayed or foregone marriage, albeit at a faster pace if no robust family policies are adopted.

The findings of this study suggest that removing formal birth restrictions in China does not guarantee a rebound in the fertility rate, particularly among women who face significant obstacles in balancing career and family obligations. In Japan, a gradual but limited array of support for married mothers has partially mitigated the opportunity costs of childbearing but has been insufficient to counterbalance the demographic effects of widespread nonmarriage. Both countries could benefit from more systematic policies that integrate gender-equitable labor practices, expanded parental leave, and improved childcare options. China might benefit from imitating Japan’s incremental policy attempts while taking care to avoid postponing broader reforms until the norm of single-child families becomes even more deeply entrenched.

A cross-national view reveals that, in both China and Japan, female educational expansion has reshaped fertility behavior, but the precise mechanisms differ owing to distinct policy histories and social structures. In China, fertility decline is still rooted predominantly in marital childbearing, and increasing education has contributed to delays in both marriage and birth timing. In Japan, a protracted economic slowdown and insufficiently progressive family policies have contributed to increasing rates of nonmarriage, making it the central driver of low fertility. Although China has taken steps to loosen restrictive birth policies, it lacks the coordinated family support networks necessary to encourage higher-order births among the newly eligible population, especially as more women obtain higher education. Japan’s experience with piecemeal reforms aimed primarily at married couples shows that without broader strategies that address gender norms and support the growing number of unmarried adults, fertility rates may remain well below replacement. Future research with more detailed causal identification and additional comparative policy analyses could provide further insight into how specific policy interventions, workplace practices, and cultural expectations shape fertility trajectories in these two low-fertility contexts.

Conclusion

This study highlights the complex interplay between educational expansion, policy environments, and fertility behaviors in China and Japan, offering critical insights into the demographic challenges faced by East Asia. At the core, both countries share cultural foundations that emphasize marriage-based fertility, yet their divergent policy histories and socioeconomic trajectories have produced distinct fertility patterns. In China, the enduring influence of population control policies, coupled with inadequate family support systems, continues to increase the opportunity costs of childbearing for highly educated women. In contrast, Japan has implemented incremental policy measures that slightly alleviate the burden of childrearing within marriage, yet its fertility rate remains constrained by widespread nonmarriage and entrenched gender norms.

The findings suggest that lifting fertility restrictions in China alone is insufficient to reverse demographic decline without addressing the broader structural challenges of balancing career and family obligations. Similarly, Japan’s limited support measures have not fully mitigated the demographic impact of rising singlehood. Both countries must adopt comprehensive policies that integrate gender-equitable labor practices, expanded childcare options, and societal reforms to address the deeper roots of low fertility.

By situating these trends within the broader context of East Asia’s second demographic transition, this study emphasizes the importance of cross-national perspectives in understanding how sociocultural and policy factors intersect to shape fertility behaviors. Future research should further explore the causal pathways and long-term policy implications, thus offering a foundation for developing more effective interventions in similar sociocultural contexts.

免費閱讀并下載全文

引用本文

Zhang, J. Trends in women’s education and fertility in Japan and China: a comparative analysis. J. Chin. Sociol. 12, 10 (2025). https://doi.org/10.1186/s40711-025-00236-0

https://journalofchinesesociology.springeropen.com/articles/10.1186/s40711-025-00236-0

文章僅為作者觀點，不代表本刊立場

更多相關文章

以上就是本期JCS Weekly的內容啦！

定期查收講座、征文信息/趣文推薦/熱點追蹤/主題漫談

學術路上

JCS陪你一起成長！

關于JCS

《中國社會學學刊》(The Journal of Chinese Sociology)于2014年10月由中國社會科學院社會學研究所創辦。作為中國大陸第一本英文社會學學術期刊，JCS致力于為中國社會學者與國外同行的學術交流和合作打造國際一流的學術平臺。JCS由全球最大科技期刊出版集團施普林格·自然(Springer Nature)出版發行，由國內外頂尖社會學家組成強大編委會隊伍，采用雙向匿名評審方式和“開放獲取”(open access)出版模式。JCS已于2021年5月被ESCI收錄。2022年，JCS的CiteScore分值為2.0（Q2），在社科類別的262種期刊中排名第94位，位列同類期刊前36%。2023年，JCS在科睿唯安發布的2023年度《期刊引證報告》（JCR）中首次獲得影響因子并達到1.5（Q3）。2025年JCS最新影響因子1.3，位列社會學領域期刊全球前53%（Q3）。

歡迎向《中國社會學學刊》投稿！

Please consider submitting to The Journal of Chinese Sociology!

官方網站：

https://journalofchinesesociology.springeropen.com