SLIDE 1 1
From Motherhood Premium to Motherhood Penalty? Heterogeneous Effects of Motherhood Stages on Women’s Economic Outcomes in Urban China
Menghan Zhao
Population Studies Center, University of Pennsylvania Abstract Women’s deteriorating position in the labor market after China’s economic reform has been documented in recent literature. However, few studies connect the relationship between the presence of children at different ages and women’s labor market outcomes with the local labor market development. Capitalizing on longitudinal data, this study uses person-fixed-effects model to investigate the relationship between different motherhood stages and married women’s economic outcomes in urban China, and considers how this relationship varies with the development of local labor market. We find that very young children inhibit mothers’ labor activity, whereas mothers’ income is positively correlated with the presence of children at school
- age. Our analysis further suggests that with the development of local labor market, the negative
association between very young children and women’s labor activity is exaggerated, while the positive relationship between children at school age and mothers’ economic outcomes is eroded. Findings also contribute to the literature that connects the labor market institution, gender-role ideology and women’s adjustment of economic activities to their childbearing and childrearing
Keywords Gender-role ideology; Economic reform; Urban Chinese women; Development of local labor market
SLIDE 2 2
From Motherhood Premium to Motherhood Penalty? Heterogeneous Effects
- f Motherhood Stages on Women’s Economic Outcomes in Urban China
Introduction Recently, the variation of the relationship between fertility and women’s economic activities across different socioeconomic contexts, related policy regimes and the gender equity levels have been documented (Brinton and Lee 2016; Esping-Andersen and Billari 2015; McDonald 2000; Rindfuss, Choe, and Brauner-Otto 2016). Typically, in an economically advanced society, which has more egalitarian division of labor within households and public policies that make it easier for women to combine the worker and mother roles, tends to have higher fertility rate. How the relationship between fertility and women’s labor force participation evolves with the economic development in China, however, has been less studied. Over the past three decades, China has witnessed an unprecedented pace of economic development and remarkable social changes. It has transitioned from a poor, centralized socialist economy to the worlds’ second largest economy and the ‘workshop of the world’ through promoting marketization. Meanwhile, growing income inequality during China’s rapid economic development in recent decades has been documented in many studies (Hauser and Xie 2005; Xie and Hannum 1996; Xie and Wu 2008; Xiang Zhou 2014; Xueguang Zhou 2000). Specifically, with the changes in the nature of employment, there are growing gender disparities in the labor market (Zhang and Hannum 2015). We suggest that China presents an interesting setting for the analysis of the interrelation
SLIDE 3 3
- f fertility, women’s economic activities, gender-role ideology and the changes of economic and
political institutions. First, at the height of the socialist period before the market transition, unemployment was an unknown phenomenon in China and the economic activity rate of women used to be at a high level that can hardly be surpassed. This, with the economic transition, leads to a decline of women’s economic activities, a trend that differs from most developed nations where female labor force participation rates and income increased rapidly during the process of industrialization (Goldin 2006). Second, though sharing common economic and political institutions prior to transition, the differences in trajectories of transitions across the former Soviet Bloc and other socialist societies have become more pronounced over time (Brainerd 2000; Heyns 2005). Instead of embracing the idea of capitalism and institutional arrangements that promised convergence with the West, China achieved its development by gradually experimenting with market mechanism. Also, the underlying fertility and family changes in China differs from the former socialist societies. China implemented strict birth control and experienced a rapid fertility transition from a fertility level higher than 6 in 1960s to below- replacement level in 1990s, comparing to a fertility level lower than 3 in 1960s in most Soviet Bloc societies (UNPD 2015) and even pronatalist policies during socialist era (Sobotka et al. 2008). Based on these reasons, we suggest that studying how the relationship between fertility and Chinese women’s economic activity changes with the evolving labor market may contribute to the extant literature that connects the socioeconomic institution to the relationship between fertility and women’s labor force participation.
SLIDE 4 4
In early studies, Chinese women were characterized by continuous employment because
- f the state intervention in promoting female labor force participation, whereas there is a growth
- f gender inequality in economic outcomes after China’s economic reform (Fincher 2016). This
is partly because the traditional gender-role ideologies and gender divisions within household still persists (Lu and Zhang 2016; Zuo and Bian 2001). Child care is one responsibility that women retain and women continue to shoulder responsibility for most domestic work. This traditional gendered labor division tends to lower women’s labor force participation during the plan-to-market economic style transition when the more gender-egalitarian state sectors have retreated (Hare 2016; Zhang and Hannum 2015). Women’s perceptions of discrimination also rise among those who enter the labor market after market reforms (Parish and Busse 2000). As estimated in previous literature, the statistical discrimination against female workers leads to two-thirds of the gender gap in payment (Xiu and Gunderson 2013), which accounts for much higher proportion for migrant workers (Min et al. 2016). Studies also suggested that, after China’s economic reform, urban women’s disadvantages in both the labor market and within the family mutually reinforce each other (Ji et al. 2017). However, previous studies on the relationship between women’s economic outcomes and children have rarely taken into account the heterogeneous relationship between different motherhood stages and women’s economic activities. Even fewer studies try to link this relationship to the development of the local labor market after 1990s, which is a period with profound socioeconomic and institutional changes in mainland China. Capitalizing on data from
SLIDE 5 5
a longitudinal survey, this study uses the person-fixed-effects model to examine how urban Chinese women’s economic outcomes are related with different motherhood stages in recent
- decades. Specifically, we try to link this relationship with the development of local labor market.
This study yields two significant findings. First, women’s labor force participation is closely related with different stages of motherhood. Young children exert an inhibiting effects on mothers’ economic activity, whereas the presence of children at school age is positively related with mothers’ income. Second, the negative correlation between young children and mothers’ involvement in the labor activity is exaggerated with the development of local labor market and the positive relationship between children at school age and women’s economic outcomes is
- eroded. Policy implications are also discussed.
Motherhood and Women’s Economic Activities Since the end of World War II, women’s educational attainment and labor force participation have increased globally (Charles 2011), which coincided with a decline in human
- fertility. Though the causal mechanisms linking fertility to women’s labor market remain elusive,
the association between fertility and women’s labor force activity reflects the challenges of balancing work and family that typifies industrialized societies. At individual level, the negative impacts of children on women’s economic outcomes have long been empirically observed and called ‘motherhood penalty’ (Angrist and Evans 1998; Budig and England 2001; Waldfogel 1997). Based on the comparative advantages of men and women, the economic model of within- household specialization posits a gendered labor division with the higher wage spouse (usually
SLIDE 6 6
husband) specializing in market work while another (usually wife) specializing in domestic sphere, and served as the main explanation for the motherhood penalty (Becker 1991). However, recent macro-level evidence suggests a positive relationship between fertility and development level (Myrskylä, Kohler, and Billari 2009), and even women’s labor force participation in advanced societies (Goldscheider, Bernhardt, and Lappegård 2015). This positive relationship between children and women’s economic outcomes challenges the dominant discourse about the negative relationship between children and women’s economic activities and urges for further explanations. Recently, variation of gender equity (Anderson and Kohler 2015; Esping-Andersen and Billari 2015; McDonald 2000), labor market institutions and related social policies (Brinton and Lee 2016; Rindfuss et al. 2016) have been introduced to explain the changing effects of children
- n women’s economic activities. Broadly speaking, the expansion of education (especially for
women), the birth of modern labor economics and the development of household labor-saving technology lead to a change of women’s economic role (Goldin 2006; Stevenson and Wolfers 2007). Women changed from secondary workers, who take the labor allocation within household as given, to active participants, who expect employment and make the labor force decisions jointly with other household members (Goldin 2006). These changes result in the increasing pressure for women to bear the brunt of conflicts between the demands of domestic work and labor market work, leading to the emergence of very low fertility in some European economically advanced countries in last century. However, with the increasing share of men’s
SLIDE 7
7
participation in household and childbearing task, higher level of within household gender equity can be achieved (Goldscheider et al. 2015). Specifically, the highly-educated groups are more likely to have a gender-egalitarian domestic-work-sharing relationships (Cherlin 2016), who are also those less likely to have motherhood penalty (Anderson, Binder, and Krause 2003; Budig and Hodges 2014). Further, Brewster and Rindfuss (2000) suggested a dynamic model that incorporates the institutional and normative arrangements should be used to analyze the relationship between fertility and labor force behavior. Empirical evidence from Europe also shows that countries with institutional arrangement and related policies that help women better balance work and family tend to see a recover of fertility (Rindfuss et al. 2016). Overall, the effects of fertility on women’s economic outcomes varies across different socioeconomic institutions, related policy regimes and the gender equity levels in both public and private sphere. Women’s Economic Activities in China during Transition During China’s economic reform, a period with changing market institutions and gender- role discourse, the relationship between children and women’s economic outcomes also changes. Since the founding of People’s Republic of China in 1949, Chinese women have been encouraged to join the labor market. It is widely accepted or even expected that women work full time after schooling. At the height of the socialist period, women in China had some of the highest labor force participation rates in the world (UNDP 1995). More recently, an erosion of women’s economic position relative to men’s has been highlighted by some research (Appleton,
SLIDE 8
8
Song, and Xia 2014; Berik, Dong, and Summerfield 2007; Chi, Li, and Yu 2011; MacPhail and Dong 2007), suggesting that women’s position in the labor market has deteriorated (Zhang et al. 2004; Wang 2005; Li and Li 2008). The worsening trend is concentrated among mothers (Zhang and Hannum 2015; Zhang, Hannum, and Wang 2008) and women’s labor force participation was increasingly discouraged by childbearing between 1970 and 2010 (Huang 2014). As explained below, women’s worsening labor market position in urban China during the market transition can be attributed to labor market changes during economic reform, the persistent gendered labor division within household and traditional gender-role ideologies. Like the other planned economies in the former Soviet Bloc, Chinese women were encouraged by the state to join the labor market before the economic reform (Croll 1983), in accordance with Marx and Engle’s doctrine that women’s emancipation is contingent on their participation in social production. Under state socialism, the work units (danwei) in urban China helped organize social production and built facilities, including dining halls, laundries, and childcare centers, which were either free or charged nominal fees to support families. Meanwhile, state propaganda was permeated with the image of the ‘Iron Girls’ and the slogan of ‘women can hold up half the sky’ (Honig 2000). Further, the birth control policies that have been implemented in China for more than three decades (Gu et al. 2007) have contributed to a fast fertility decline, reducing women’s time commitment to family obligations and increasing women’s economic activity (Wu, Ye, and He 2014). The high level of female labor force participation and the promoted egalitarian ideology leaded to the seeming elevation of women’s
SLIDE 9 9
status in the workplace. However, during China’s market transition, more gender-egalitarian state sectors have gradually retreated and stopped promoting gender equality via providing social services. First, the publicly-funded child care system that provided care to children from the earliest months until primary school has no longer covered children under three years old since 1989 (Du and Dong 2013). Thus, many publicly subsidized childcare centers were either shut down or started to provide commercial care services. Second, under mounting pressure for profits, the majority
- f Chinese urban enterprises ceased to provide subsidized childcare services to employees (Cook
and Dong 2011). These transitions have shifted the main burden of childbearing and childrearing back to women. Further, women were laid off disproportionally during the deepening reform of the state-owned enterprises after 1992 and the length of unemployment was longer for women than men (Du and Dong 2009). These changes during economic transition put women in an unfavorable position in the labor market, which had also been observed in some other transitional economies (Hunt 2002; UNICEF 1999)1. Recently, some studies drew attention to the intact gendered labor division and the traditional gendered roles within households under state socialism (Ji et al. 2017; Zuo 2012). In given historical junctures, Chinese women’s liberation appears as an integral part of nationalism, with the result that the gender equality was more obligation-oriented than individual rights-
- based. This gender obligation equality expected women to take almost the same work duties with
men for socialist production (Zuo 2012). Women’s labor force participation was supported by the
SLIDE 10 10
state through providing public services and taking over some of women’s responsibilities within
- household. Nevertheless, women still needed to do the rest of the unpaid household work, which
was treated as secondary to social production (Song 2012). Specifically, the patriarchy and traditional gendered labor division in families was retained by the state and the housework taken by women was recognized as contributions to social production. Thus working women, especially those with lower income and less help from other family members (Song 2012), still suffered from the double burden of paid work and family unpaid work, which was only alleviated by the work-unit system under state socialism (Ji et al. 2017). These legacies from state socialism of the unchallenged traditional gender ideals about family roles lead to the growing gender inequality in labor force participation in urban China during market transition. Further, during the market transition, the public discourse has also shifted from a state- dominated Marxist political discourse to a market-oriented discourse. This market discourse asserts that the gendered market outcomes result from distinct abilities deriving from essential gender differences, which is in a close alliance with the traditional gender order based on patriarchal norms, regulating the gender norms in urban China (Sun and Chen 2015). In post- socialist Vietnam, gender disparities in the household division of labor have also increased as a result of resurgent male-centered kin and family relations (Luong 2003). This revitalization of traditional gender values also contributes to the decline of women’s position in the labor market. In other developed East Asian countries which share the same patriarchal norms and Confucian ideology, the reduced labor force participation of women after marriage or childbirth has long
SLIDE 11
11
been observed. In Japan, women tend to show similar labor force participation with men when they have just left school; however, their labor force participation then decreases sharply after marriage or childbearing, and does not again recover to the earlier level (Brinton 1989). A similar situation is also observed in South Korea, where the labor force participation of married women is remarkably lower than unmarried women, even among college graduates (B. S. Lee, Jang, and Sarkar 2008). This is partly because, in East Asia, the conservative gender roles still persist and expect women to take the main responsibility for household chores and raising children. Further, persistent workplace masculinity in Japan that women exclusively concentrate in low-level management positions also hinders women’s employment (Nemoto 2013). Demand for long working hours and rare part-time employment made it hard for South Korean women to balance work and family (Ma 2014). Although the extant research has provided insightful discourse analyses of women’s position in the labor market in post-reform China, limited research has focused the relationship between different motherhood stages and their decisions to join the labor force and their income. As suggested in previous studies, women’s labor market decisions should be considered based on their life course events (Waite 1980). Typically, western empirical studies suggest that a small child has an inhibiting effect on mother’s work activity (Lu, Wang, and Han 2017; Maron and Meulders 2008; Waite 1976) and a woman who is pregnant or with preschool-age children is less likely to have voluntary job changes which usually increases wages and further ones’ career (Looze 2007). However, the negative effect of children decreases with the age of the youngest
SLIDE 12 12
child (Maron and Meulders 2008) and older children encourage women’s employment (Budig 2003), which contributes to the reverse of women’s labor force participation. In China, the conflict between work and family is also most intense when children are young and the presence
- f children at preschool age decreases Chinese women’s likelihood of participating in market
work (Hare 2016; Maurer-Fazio et al. 2011). Thus, we expect the relationship between different motherhood stages and women’s labor activities should vary. There have been several efforts in recent years to portray mothers’ increasing losses in the labor market over time. Capitalizing on data from China Health and Nutrition Survey from year 1991 to 2011, Hare (2016) suggested that having children under age seven has bigger inhibiting effects on women’s labor force participation after 2000 than before. By using the same survey data, Jia and Dong (2012) conducted person-fixed-effects models and found that urban women experienced substantial motherhood penalty between 1999 and 2005 rather than the previous period between 1990 and 1996. By interacting gender variable with year dummies, Zhang and Hannum (2015) also observed that mothers are increasing disadvantaged in wage earnings by the late 2000s. However, few empirical studies have directly examined how urban women’s labor market outcomes change with the development of local labor market. In this study, by adopting longitudinal survey data between 1991 and 2011, we investigate the heterogeneous relationship between different motherhood stages and urban women’s economic outcomes, including labor activity and income, and further connect this relationship with the local labor market development.
SLIDE 13 13
Data and Methods This study uses data from the China Health and Nutrition Survey (CHNS), a collaborative project between the Carolina Population Center and the China Academy of Preventive Medicine. This survey captured family changes from 1989 to 2011, a period with major events and socioeconomic development. The CHNS is a panel/longitudinal study of households initiated with eight provinces (Liaoning, Jiangsu, Shandong, Henan, Hubei, Hunan, Guangxi and Guizhou) in 1989. The survey then added a ninth (Heilongjiang) in 19972. This survey covers roughly half of China’s population in provinces that are geographically diverse (Jones-Smith and Popkin 2010). A multistate, random cluster design to select a stratified probability sample was used in the original survey. The initial primary sampling units consisted of 190 communities with substantial variations in level of economic development, including 31 urban neighborhoods, 31 suburban neighborhoods, 32 towns and 96 rural villages. Among the nine waves of surveys (1989, 1991, 1993, 1997, 2000, 2004, 2006, 2009 and 2011), 1989 wave data are not included in
- ur analysis because they only included a partial sample and had questionnaires substantially
different from those in the later waves (Gong, Xu, and Han 2015). Considering the focus of this study, we confine our analysis to women who are between 18 and 50 years old and exclude those who lived in rural villages. Only married women who stay in their first marriage are included, minimizing the impacts of selectivity into marriage. Because the data record the marriage and fertility histories of all women who are ever married under 52 years old, information about the children can be easily obtained. Childless women are also
SLIDE 14 14
included in our analysis. We stack the data into a women-wave structure for panel data analysis. As explained later in this article, we use person-fixed-effects models, which only use the within- individual variations, so we keep women with at least two-wave observations in our analytical
- sample. This exclusion does not substantially affect the distribution of the observations in terms
- f educational levels, stratums, labor force participation and different motherhood stages. To
better capture the impacts of children on urban women’s economic outcomes, rather than limit analytical sample to wage workers (Jia and Dong 2012; Yu and Xie 2014; Zhang and Hannum 2015), all women, either having worked or not, are included in analytic data. The final analytic sample has 6,370 person-wave observations for 1,931 women, the size
- f which is similar to the analytic sample size of previous studies using the same data source.
Among them, 117 women (with 305 observations) never reported working, 691 women (with 2,671 observations) experienced work interruption, 1,123 women (with 3,394 observations) worked for all the waves. As shown in Table 1, the last group is more educated than the first two groups. TABLE 1 ABOUT HERE Fixed-effects Model Person-fixed-effects models are conducted in this study to reveal how the within-person change (across waves) in labor force participation and income is associated with different stages
- f parenthood. By only using the information of variation within person, person-fixed-effects
models control for all unmeasured, unchanging characteristics of persons that contribute
SLIDE 15 15
additively to the estimation of their probability of working and income (Allison 2009). Results from random-effects models are also included for comparison. The conditional fixed-effects logistic regression model is used for the analysis of women’s labor activity (𝑞𝑗𝑢 is the probability of working for a woman 𝑗 at time 𝑢) : log ( 𝑞𝑗𝑢 1 − 𝑞𝑗𝑢 ) = 𝜈𝑢 + 𝛾𝑦𝑗𝑢 + 𝛿𝑨
𝑘𝑢 + 𝛽𝑗
Because the fixed-effects model uses the within-person variation, only women who have experienced work interruption are included. As a robustness test, we also use random-effects models with all women included for analysis (Appendix Table 1). For women who had jobs and reported income, the dependent variable is the logarithm of their annual income which is adjusted for CPI (inflated to 2009 yuan). The fixed-effects linear model is used: log (𝑏𝑜𝑜𝑣𝑏𝑚𝑗𝑜𝑑𝑝𝑛𝑓)𝑗𝑢 = 𝜈𝑢 + 𝛾𝑦𝑗𝑢 + 𝛿𝑨
𝑘𝑢 + 𝛽𝑗 + 𝜁𝑗𝑢
𝜈𝑢 allows for different constants at different waves. 𝑦𝑗𝑢 is a column vector of variables that vary both over individuals and over time. 𝑨
𝑘𝑢 represents community-level variables that vary over
community and over time.𝛽𝑗 represents all differences between individuals that are stable over time, which are a set of fixed constants that can be correlated with other measured predictors. 𝜁𝑗𝑢 is the idiosyncratic error term. For analysis of income, we also take into account the sample selection bias that income is
- bserved only for women choosing to participate in the labor force (Heckman 1977; Wu and Xie
2003) by using Heckman selection model as a robustness test. Results from the models that control for this selection bias are also shown in Appendix Table 2.
SLIDE 16 16
Variables Our main analytical outcome variables include women’s labor force participation and the logarithm of annual income. Women’s labor activity (working or not) depends on the reported working status4. By defining work broadly in terms of involvement in income-earning activities in general rather than wage employment, we capture the actual changing economic status of women more comprehensively by taking into account multiple moneymaking activities. Women’s annual income is calculated in the CHNS and built by adding each person’s income
- sources. Because the wage is imputed from adjacent waves if it is missing, the observations with
imputed wage are not included in our analysis, avoiding the bias in estimating the effects of children. The main time-varying variables that we are interested in are the different stages of parenthood: having children under three years old (very young children), having children between three and six years old (young children), having children between seven and fifteen years old (children at school age), and having children older than fifteen years old (children at working age). These measures capture different stages of parenthood by distinguishing the age of children, because, as explained in literature review section, women may hold different roles at different times in their life. However, most previous research on related topics lacked the life course perspective and only focused on the effects of motherhood or number of children on women’s wage in recent urban China (Jia and Dong 2012; Yu and Xie 2014). Our analysis focuses on the relationship between children and women’s economic outcomes across different
SLIDE 17 17
parenthood stages. Other time-varying variables are included in the model to control for women’s household
- contexts. The living status of mother or mother-in-law is included in the model to account for the
possible child care help from other household members. Five categories (living in the same household, living in the same neighborhood/village, living in the same city/county, living in
- ther city/county and not alive or unknown) represent different levels of help5: living in the same
household suggests that a woman can get the most child care help from mother or mother-in-law, both mother and mother-in-law are not alive or their living status are unknown represent the least child care help that a woman can get. Husband’s labor activity and the income of household
- ther than women’s income, which is a proxy for family economic resources, are included in the
model of women’s labor force participation. In random-effects models, we also control for women’s age, educational level and the sampling stratum of this survey. To capture the community labor market environments that change over time, we employ the community-level scale measure created by CHNS team: economic component score (Jones- Smith and Popkin 2010). An economic component score measures economic activity with a range between zero and ten, including the typical daily wage for ordinary male workers and the percentage of the population engaged in nonagricultural work. The information was obtained in the community survey from area administrators and official records (Monda et al. 2007). The increasing values of economic component scores indicate the development of community labor market environments (Table 2).
SLIDE 18 18
TABLE 2 ABOUT HERE Furthermore, to better control for the exogenous changes in the community, we also include quality of health score into the model, which measures health infrastructure, including the number and the type of health facilities in or nearby (≤ 12 kilometers) the community and the number of pharmacies in the community. Sanitation score and housing score for the community are also included. A sanitation score is a measure of the proportion of households with treated water and the prevalence of households without excreta present outside the house. A housing score measures the availability of electricity, indoor tap water, flush toilets and gas for cooking. Though social services score, which measures the provision of preschool for children under three years old and availability of different kinds of insurance, is also provided by the CHNS team, it is not included in our analysis because it only covers the waves after year 2000. Results Women’s Labor Activity Table 3 shows the descriptive statistics by year between 1991 and 2011. The labor activity of women in the sample dropped substantially from 1990s to 2000s, especially during the period between 2000 and 20046. The last four panels of Table 3 show the labor force participation of the subsamples for mothers at different stages of motherhood: children under age three, children between age three and six, children between age seven and fifteen, and children
- lder than age fifteen. As shown, except for women who have children older than age fifteen, the
mean ages of mothers for different motherhood stages between 1991 and 2011 do not differ
SLIDE 19
19
substantially, which ensures the comparison is less likely to be affected by age effect. The decline of female labor force participation is most striking for those who have very young children, the labor force participation of which dropped from almost 90% in 1991 to around 55% between 2004 and 2011. More than 90% of the women with young children joined the labor market before 2000, which also dropped greatly after 2000. However, the labor force participation is still higher than 60%. The labor force participation of mothers with children at school age remained higher than that of mothers with very young children. After a drop between 2000 and 2004, the female labor force participation for mothers with children at school age were around 70% between 2006 and 2011. For women who have children older than age fifteen, though the labor force participation rate dropped, it remained higher than the rate of women who have very young children. TABLE 3 ABOUT HERE Further, we conduct both conditional fixed-effects models and random-effects models of women’s labor activity, as shown in Model F1 and Model R1 of Table 4, respectively. In random- effects models, we include women’s age, sampling stratum and educational level which serves as a proxy for individual endowments. Both of these models suggest a negative relationship between having very young children and mothers’ labor market participation. Specifically, as suggested by conditional fixed-effects models, the odds of working are 60.74% (1 − e−0.935) lower if a woman gave birth during the past three years. The work interruption for those who gave birth after 1990s contradicts with some early findings that women in mainland China
SLIDE 20
20
continued their work across family life stages (Yi and Chien 2002). As explained in literature review part, the deepening reform of the state-owned enterprises after 1992 and the retreat of the publicly funded child care services contribute to the unfavorable labor market position of mothers with more childcare obligations (i.e. those with very young children) in recent decades. However, this negative relationship disappear in other parenthood stages. In Model R2 and Model F2, we include the interaction terms between different stages of motherhood and community-level economic component score of local labor market. As shown by the results, all of the four interactions are statistically significant. Specifically, the negative relationship between having very young children and mothers’ labor force is exaggerated with the development of local labor market. The positive relationship between mothers’ labor activity and having children at school age7 is eroded with the growth of the economic component score. As discussed, women’s labor market position might deteriorate with the deepening process of market transition and the development of local labor market. TABLE 4 ABOUT HERE In summary, the relationship between children and mothers’ labor activity varies across parenthood stages. That is, very young children have an inhibiting effect on mothers’ labor activity, while mothers’ labor force participation is less affected when children grow older. These relationships become more negative with the development of local labor market. Overall, mothers’ labor force participation declined after 1990s. As shown by the statistical models, the dummies for different years also have a clear trend of declining labor force participation rate of
SLIDE 21 21
women, which is consistent with the description about the decreasing percentage of women in the labor market in Table 3. Women’s Income Further analysis on women’s income is also conducted by using both fixed-effects model and random-effects model8. Besides, to account for the selection bias towards labor force participation, we also run robustness tests by including inverse Mills ratio into the model. The results in Appendix Table 2 also are consist with our findings. As shown in Table 5, overall, women’s income has been increasing after 1990s, because the dummies for waves show growing positive coefficients. Also, there is little evidence that women’s income is negatively related to having very young children because the coefficients of children under three years old are not statistically significant at 0.05 level. Nevertheless, both the fixed-effects model and random-effects model suggest a positive relationship between children at school age and mothers’ income. Specifically, mothers tend to have 11% higher income if they have children at school age than they do not, suggesting that, in response to the rising costs of children’s education or other needs as child grows up, mothers are motivated to work harder or resort to various moneymaking activities to increase income. This is also consistent with the findings from previous studies that an adaptive strategy is adopted by household members to support the education and wellbeing of children (Chen and Korinek 2010). The interactive terms are then included in both fixed-effects model and random-effects
- model. As shown, overall, the interactive terms are negative, suggesting a smaller positive
SLIDE 22 22
relationship between children at school age and mothers’ income with a higher local economic component score. Specifically, according to fixed-effect model, the income of women who have children at school age will drop by 3% for one unit increase in the economic component score of local labor market. TABLE 5 ABOUT HERE To sum up, there is no strong evidence that working mothers present with young children earn less, but rather, they have higher income when their children go to school. Nevertheless, the positive relationship between mothers’ income and children at school age is eroded when the local labor market develops. Conclusion and Discussion This paper contributes to recent Chinese research on the deteriorating position of urban women in the labor market by focusing on the heterogenous relationship between different stages
- f motherhood and women’s economic outcomes. By using the person-fixed-effects model on the
longitudinal survey data, we only use the information of within-person variation, controlling for unobserved factors that may be correlated with women’ s economic outcomes. This study also contributes to a comparative literature that connects the intuitional changes to the variation of female labor force participation by linking the relationship between children and women’s labor force participation with local labor market development. The results suggest that, overall, Chinese women’s labor activity declined after 1990. We also find that women’s economic activities respond to the demand of her family responsibility
SLIDE 23
23
that varies across different motherhood stages. That is, the probability of working is much lower for mothers who have very young children, while women with children at school age show higher income. This corresponds with the findings from western studies that the shifts in ages of family members and family expenses affect wives’ decisions about market activity (Waite 1980). Maternal labor supply is more likely to be affected by child costs and the age of children (Griffen, Nakamuro, and Inui 2015; G. H. Y. Lee and Lee 2014). When children grow and go to school, mothers take adaptive strategy, engaging in multiple moneymaking activities in response to the increasing expenses and resource demands. However, the historic trend of Chinese women’s labor force participation is different from the experience of other industrialized societies. In this sense, our research also complements studies that focus on gender equity and institutional changes by addressing how labor market changes might contribute to the gender inequality in the labor market as well as in the family. Before economic reform, women were encouraged by the state to participate in social production and their household work (especially childcare responsibilities) was alleviated by publicly funded services. However, during economic transition, both the more gender-egalitarian state sectors and the state-dominated Marxist political discourse have gradually retreated. Thus, in contrast to the growth of female labor force activity during the period of fast economic development in industrialized countries, Chinese women’s labor force participation has declined during the fast economic development period after 1990. Our analysis suggests that, for mothers at early stages of motherhood, the inhibiting effect of very young children is exaggerated with
SLIDE 24 24
the development of local labor market. For women who have children at school age, the positive association is eroded with the development of local labor market. This decline is partly because
- f the fact that the high level of female labor force participation that had been reached before
China’s economic reform has ‘regressed’ to the mean level of other market economies during its plan-to-market economic style transition. This resurgence of gender inequality in China’s labor market can also be attributed to the traditional gender-role ideologies within households that dominate persistently in East Asian regions remained intact under state socialism. Women are still expected to take the main responsibility of childrearing and household chores even when they have full-time jobs, whereas the convenient public funded child care system stopped during the economics transition. Recently, there is a heated discussion on how the 2015 loosened birth control policy, i.e. universal two-child policy, will affect the fertility level in China in the near future (Attané 2016b; Zhao 2015). The inadequate protection of Chinese women in the labor market and the lack of state policies supporting families in rearing children serve as the main reasons why the change in the birth control itself is not enough to expect a substantial rise in fertility level (Attané 2016a). According to McDonald’s description (McDonald 2000), very low fertility is likely to be
- bserved in postindustrial societies, where women’s labor market participation is normatively
accepted but a highly gendered division of labor remains within household. Our results suggest that the conflict between work and family for women gets severer in recent years. Further, under the universal two-child policy, there is a tendency of lower market position of women, because
SLIDE 25 25
employers will expect women to be less devoted to their work even for those who already have
- ne child and are not eligible to have a second child before the policy change. Thus we suggest
that family policies which support families in childrearing and promote gender equality within household should be also promoted aside from the loosening of birth control, which has been discussed more frequently in recent studies (Hu and Peng 2012; Zhao 2016; Zheng 2016). The impacts of various policies on the fertility trend have been observed in western countries that the policies helping women to combine the work and mother roles are more likely to see higher fertility rates (Brauner-Otto 2016; Rindfuss et al. 2016). Even in South Korea, the supportive policies are found to show positive effects on families’ fertility behavior (Yoon 2017). Overall, the long-term fertility trend in China will depend on the interactions between gender-role ideology, public policies and labor market intuitions in the future. Because our analyses focus on women in urban area, the results might not be generalized to rural area, where the economic structure is different. Previous studies suggest that, in rural China, economic development does not uniformly increase gender inequalities within household (Matthews and Nee 2000). The increasing rural-urban migration will also affects migrants’ family and childbearing behavior (Guo 2010; Xu 2016). Note
- 1. After reunification, East German employment rates were sharply lowered and the
unemployment was disproportionally high among women (Rosenfeld, Trappe, and Gornick,
SLIDE 26 26
2004). In Russia, the wage inequality between men and women increased across all percentiles of the wage distribution between year 1991 and 1994 (Brainerd, 1998).
- 2. Liaoning was unable to participate in the CHNS for 1997 wave but was added back in 2000.
- 3. First, Heckman model for maximum likelihood estimates is fitted to obtain the nonselection
hazard (inverse Mills’ ratio). This term is estimated from a probit model for working. In addition to the variables that have been included in the income model, husband’s labor force participation and the income of household other than women’s income are also used in the probit model to predict women’s labor force participation. Then, the nonselection hazard is included in the fixed-effects and random-effects linear model to account for the selection bias.
- 4. Robustness tests are conducted with women considered as working if she reported working
- r had positive income. The results of the robustness tests lead to the same conclusion.
- 5. Because we try to control for the possible child care help, thus the living status of mother
and mother-in-law are combined. One observation will be categorized into a higher level of child care help if either mother or mother-in-law can provide more help. For example, if a woman’s mother is living in the same city/county while her mother-in-law is living in the same household, the living status of mother or mother-in-law is living in the same household.
SLIDE 27 27
- 6. During this period, China joined the World Trade Organization (WTO) and launched a
drastic privatization in socialized services, shifting the family-related responsibilities back to women (Hare, 2016).
- 7. The main effect of having children between seven and fifteen years old is statistically
significant in random-effects model but not in fixed-effects model. This might result from the bigger standard error in fixed-effect model, because the effect will be significant at 0.05 level if we replace the standard error from fixed-effects model (0.164) with the standard error from random-effects model (0.132). Thus we suggest that the relationship between women’s labor activity and having children at school age is positive when the community economic component score is centered at 5. This finding is consistent with previous literature suggesting that Chinese women are likely to feel that it is their duty to work for the good of their children (Short et al. 2002).
- 8. We also include years of working (which is estimated by taking the differences between age
and the approximate age of obtaining the highest educational level) as a proxy for working experience in the random-effects models for a robustness check. The results are consistent.
SLIDE 28 28
References Allison, P. D. (2009). Fixed Effects Regression Models. SAGE Publications. Anderson, D. J., Binder, M., & Krause, K. (2003). The motherhood wage penalty revisited: experience, heterogeneity, work effort, and work-schedule flexibility. Industrial & Labor Relations Review, 56(2), 273–294. https://doi.org/10.1177/001979390305600204 Anderson, T., & Kohler, H.-P. (2015). Low fertility, socioeconomic development, and gender
- equity. Population and Development Review, 41(3), 381–407.
https://doi.org/10.1111/j.1728-4457.2015.00065.x Angrist, J. D., & Evans, W. N. (1998). Children and their parents’ labor supply: Evidence from exogenous variation in family size. The American Economic Review, 88(3), 450–477. Appleton, S., Song, L., & Xia, Q. (2014). Understanding urban wage inequality in China 1988– 2008: Evidence from quantile analysis. World Development, 62, 1–13. https://doi.org/10.1016/j.worlddev.2014.04.005 Attané, I. (2016a). Second child decisions in China. Population and Development Review, 42(3). https://doi.org/10.1111/j.1728-4457.2016.00151.x Attané, I. (2016b). The end of one child per family in China? Population & Societies. https://www.ined.fr/en/publications/population-and-societies/china-one-child-end/#tabs-2 Becker, G. S. (1991). A Treatise on the Family. Harvard University Press. Berik, G., Dong, X., & Summerfield, G. (2007). China’s transition and feminist economics. Feminist Economics, 13(3–4), 1–33. https://doi.org/10.1080/13545700701513954
SLIDE 29 29
Brainerd, E. (1998). Winners and losers in Russia’s economic transition. The American Economic Review, 88(5), 1094–1116. Brainerd, E. (2000). Women in transition: changes in gender wage differentials in Eastern Europe and the Former Soviet Union. ILR Review, 54(1), 138–162. https://doi.org/10.1177/001979390005400108 Brauner-Otto, S. R. (2016). Canadian fertility trends and policies: a story of regional variation. In
- R. R. Rindfuss & M. K. Choe (Eds.), Low Fertility, Institutions, and their Policies (pp.
99–130). Springer International Publishing. https://doi.org/10.1007/978-3-319-32997-0_5 Brewster, K. L., & Rindfuss, R. R. (2000). Fertility and women’s employment in industrialized
- nations. Annual Review of Sociology, 26, 271–296.
Brinton, M. C. (1989). Gender stratification in contemporary urban Japan. American Sociological Review, 54(4), 549–564. https://doi.org/10.2307/2095878 Brinton, M. C., & Lee, D.-J. (2016). Gender-role ideology, labor market institutions, and post- industrial fertility. Population and Development Review, 42(3), 405–433. https://doi.org/10.1111/padr.161 Budig, M. J. (2003). Are women’s employment and fertility histories interdependent? An examination of causal order using event history analysis. Social Science Research, 32(3), 376–401. https://doi.org/10.1016/S0049-089X(03)00012-7 Budig, M. J., & England, P. (2001). The wage penalty for motherhood. American Sociological Review, 66(2), 204–225. https://doi.org/10.2307/2657415
SLIDE 30
30
Budig, M. J., & Hodges, M. J. (2014). Statistical models and empirical evidence for differences in the motherhood penalty across the earnings distribution. American Sociological Review, 79(2), 358–364. https://doi.org/10.1177/0003122414523616 Charles, M. (2011). A world of difference: International trends in women’s economic status. Annual Review of Sociology, 37(1), 355–371. https://doi.org/10.1146/annurev.soc.012809.102548 Chen, F., & Korinek, K. (2010). Family life course transitions and rural Household economy during China’s market reform. Demography, 47(4), 963–987. https://doi.org/10.1007/BF03213735 Cherlin, A. J. (2016). A happy ending to a half-century of family change? Population and Development Review, 42(1), 121–129. https://doi.org/10.1111/j.1728-4457.2016.00111.x Chi, W., Li, B., & Yu, Q. (2011). Decomposition of the increase in earnings inequality in urban China: A distributional approach. China Economic Review, 22(3), 299–312. https://doi.org/10.1016/j.chieco.2011.03.002 Cook, S., & Dong, X. (2011). Harsh choices: Chinese women’s paid work and unpaid care responsibilities under economic reform. Development and Change, 42(4), 947–965. https://doi.org/10.1111/j.1467-7660.2011.01721.x Croll, E. (1983). Chinese women since Mao. http://www.popline.org/node/422940 Du, F., & Dong, X. (2009). Why do women have longer durations of unemployment than men in post-restructuring urban China? Cambridge Journal of Economics, 33(2), 233–252.
SLIDE 31 31
https://doi.org/10.1093/cje/ben034 Du, F., & Dong, X. (2013). Women’s employment and child care choices in urban China during the economic transition. Economic Development and Cultural Change, 62(1), 131–155. https://doi.org/10.1086/671714 Esping-Andersen, G., & Billari, F. C. (2015). Re-theorizing family demographics. Population and Development Review, 41(1), 1–31. https://doi.org/10.1111/j.1728-4457.2015.00024.x Fincher, L. H. (2016). Leftover Women: The Resurgence of Gender Inequality in China. Zed Books Ltd. Goldin, C. (2006). The quiet revolution that transformed women’s employment, education, and
- family. American Economic Review, 96(2), 1–21.
Goldscheider, F., Bernhardt, E., & Lappegård, T. (2015). The gender revolution: A framework for understanding changing family and demographic behavior. Population and Development Review, 41(2), 207–239. https://doi.org/10.1111/j.1728-4457.2015.00045.x Gong, X., Xu, D., & Han, W.-J. (2015). Household income and preschool attendance in China. Child Development, 86(1), 194–208. https://doi.org/10.1111/cdev.12294 Griffen, A. S., Nakamuro, M., & Inui, T. (2015). Fertility and maternal labor supply in Japan: Conflicting policy goals? Journal of the Japanese and International Economies, 38, 52–
- 72. https://doi.org/10.1016/j.jjie.2015.04.003
Gu, B., Wang, F., Guo, Z., & Zhang, E. (2007). China’s local and national fertility policies at the end of the Twentieth century. Population and Development Review, 33(1), 129–147.
SLIDE 32 32
Guo, Z. (2010). Impacts of floating population on current fertility in China. Population Research (in Chinese), 1, 19–29. Hare, D. (2016). What accounts for the decline in labor force participation among married women in urban China, 1991–2011? China Economic Review, 38, 251–266. https://doi.org/10.1016/j.chieco.2016.01.004 Hauser, S. M., & Xie, Y. (2005). Temporal and regional variation in earnings inequality: Urban China in transition between 1988 and 1995. Social Science Research, 34(1), 44–79. Heckman, J. J. (1977). Sample Selection Bias As a Specification Error (with an Application to the Estimation of Labor Supply Functions) (Working Paper No. 172). National Bureau of Economic Research. http://www.nber.org/papers/w0172 Heyns, B. (2005). Emerging inequalities in Central and Eastern Europe. Annual Review of Sociology, 31(1), 163–197. https://doi.org/10.1146/annurev.soc.30.012703.110637 Honig, E. (2000). Iron Girls Revisited: Gender and Politics of Work in Cultural Revolution 1966-1976. Redrawing Boundaries: Work, Households, and Gender in China. Barbara Entwisle and Gail E. Henderson (eds). University of California Press. Hu, Z., & Peng, X. (2012). China’s family policies in the context of family changes. Population Research (in Chinese), 2, 3–10. Huang, G. (2014). The buffering function of reproduction support in women’s career break: Based on the third issue of Chinese women’s social status investigation report. Collection
- f Women’s Studies (in Chinese), 4, 27–33.
SLIDE 33 33
Hunt, J. (2002). The transition in East Germany: When is a ten-point fall in the gender wage gap bad news? Journal of Labor Economics, 20(1), 148–169. https://doi.org/10.1086/323935 Ji, Y., Wu, X., Sun, S., & He, G. (2017). Unequal care, unequal work: Toward a more comprehensive understanding of gender inequality in post-reform urban China. Sex Roles, 1–14. https://doi.org/10.1007/s11199-017-0751-1 Jia, N., & Dong, X.-Y. (2012). Economic transition and the motherhood wage penalty in urban China: investigation using panel data. Cambridge Journal of Economics. https://doi.org/10.1093/cje/bes044 Jones-Smith, J., & Popkin, B. M. (2010). Understanding community context and adult health changes in China: Development of an urbanicity scale. Social Science & Medicine (1982), 71(8), 1436–1446. https://doi.org/10.1016/j.socscimed.2010.07.027 Lee, B. S., Jang, S., & Sarkar, J. (2008). Women’s labor force participation and marriage: The case of Korea. Journal of Asian Economics, 19(2), 138–154. https://doi.org/10.1016/j.asieco.2007.12.012 Lee, G. H. Y., & Lee, S. P. (2014). Childcare availability, fertility and female labor force participation in Japan. Journal of the Japanese and International Economies, 32, 71–85. https://doi.org/10.1016/j.jjie.2014.01.002 Li, C., & Li, S. (2008). Rising gender income gap and its dynamics in China: Market competition
- r sex discrimination ? Sociological Studies (in Chinese), 2, 94–116.
Looze, J. (2017). Why do(n’t) they leave?: Motherhood and women’s job mobility. Social
SLIDE 34 34
Science Research. https://doi.org/10.1016/j.ssresearch.2017.03.004 Lu, W., & Zhang, B. (2016). The effect of China’s family planning policy on women’s social
- status. Population Research (in Chinese), 40(4), 21–34.
Lu, Y., Wang, J. S.-H., & Han, W.-J. (2017). Women’s short-term employment trajectories following birth: Patterns, determinants, and variations by race/ethnicity and nativity. Demography, 1–26. https://doi.org/10.1007/s13524-016-0541-3 Luong, H. V. (2003). Gender relations: Ideologies, kinship practices, and political economy. Postwar Vietnam: Dynamics of a Transforming Society, 16, 201. Ma, L. (2014). Economic crisis and women’s labor force return after childbirth: Evidence from South Korea. Demographic Research, 31(18), 511–552. https://doi.org/10.4054/DemRes.2014.31.18 MacPhail, F., & Dong, X. (2007). Women’s market work and household status in rural China: Evidence from Jiangsu and Shandong in the late 1990s. Feminist Economics, 13(3–4), 93–124. https://doi.org/10.1080/13545700701439457 Maron, L., & Meulders, D. (2008). Having a child: A penalty or bonus for mother’s and father’s employment in Europe? Working Papers DULBEA, 8. http://dipot.ulb.ac.be/dspace/bitstream/2013/13588/1/dul-0072.pdf Matthews, R., & Nee, V. (2000). Gender inequality and economic growth in rural China. Social Science Research, 29(4), 606–632. https://doi.org/10.1006/ssre.2000.0684 Maurer-Fazio, M., Connelly, R., Chen, L., & Tang, L. (2011). Childcare, eldercare, and labor
SLIDE 35 35
force participation of married women in urban China, 1982–2000. Journal of Human Resources, 46(2), 261–294. https://doi.org/10.3368/jhr.46.2.261 McDonald, P. (2000). Gender equity in theories of fertility transition. Population and Development Review, 26(3), 427–439. Min, Q., Brown, J., Padmadas, S., Bohua, L., Jianan, Q., & Falkingham, J. (2016). Gender inequalities in employment and wage-earning among economic migrants in Chinese
- Cities. Demographic Research, 34(6), 175–202.
https://doi.org/10.4054/DemRes.2016.34.6 Monda, K. L., Gordon-Larsen, P., Stevens, J., & Popkin, B. M. (2007). China’s transition: The effect of rapid urbanization on adult occupational physical activity. Social Science & Medicine, 64(4), 858–870. https://doi.org/10.1016/j.socscimed.2006.10.019 Myrskylä, M., Kohler, H.-P., & Billari, F. C. (2009). Advances in development reverse fertility
- declines. Nature, 460(7256), 741–743. https://doi.org/10.1038/nature08230
Nemoto, K. (2013). When culture resists progress: masculine organizational culture and its impacts on the vertical segregation of women in Japanese companies. Work, Employment and Society, 27(1), 153–169. https://doi.org/10.1177/0950017012460324 Parish, W. L., & Busse, S. (2000). Gender and work. Chinese Urban Life under Reform, 209– 231. Rindfuss, R. R., Choe, M. K., & Brauner-Otto, S. R. (2016). The emergence of two distinct fertility regimes in economically advanced countries. Population Research and Policy
SLIDE 36 36
Review, 35(3), 287–304. https://doi.org/10.1007/s11113-016-9387-z Rosenfeld, R. A., Trappe, H., & Gornick, J. C. (2004). Gender and work in Germany: Before and after reunification. Annual Review of Sociology, 30(1), 103–124. https://doi.org/10.1146/annurev.soc.30.012703.110531 Short, S. E., Chen, F., Entwisle, B., & Fengying, Z. (2002). Maternal work and child care in China: A multi-method analysis. Population and Development Review, 28(1), 31–57. https://doi.org/10.1111/j.1728-4457.2002.00031.x Sobotka, T., Št’astná, A., Zeman, K., Hamplová, D., & Kantorová, V. (2008). Czech Republic: A rapid transformation of fertility and family behaviour after the collapse of state socialism. Demographic Research, S7(14), 403–454. https://doi.org/10.4054/DemRes.2008.19.14 Song, S. (2012). From visible to invisible: Housework in the collectivist period (1949-1966). Jiangsu Social Sciences (in Chinese), 1, 116–125. Stevenson, B., & Wolfers, J. (2007). Marriage and divorce: Changes and their driving forces. The Journal of Economic Perspectives, 21(2), 27–52. Sun, S., & Chen, F. (2015). Reprivatized womanhood: Changes in mainstream media’s framing
- f urban women’s issues in China, 1995–2012. Journal of Marriage and Family, 77(5),
1091–1107. https://doi.org/10.1111/jomf.12219
- UNDP. (1995). Human Development Report 1995. http://hdr.undp.org/en/content/human-
development-report-1995
- UNICEF. (1999). Women in Transition (MONEE Project, Regional Monitoring Report 6).
SLIDE 37 37
Florence: UNICEF Innocenti Research Centre.
- UNPD. (2015). World Population Prospects: The 2015 Revision. United Nations, Department of
Economic and Social Affairs, Population Division. Waite, L. J. (1976). Working wives: 1940-1960. American Sociological Review, 41(1), 65–80. https://doi.org/10.2307/2094373 Waite, L. J. (1980). Working wives and the family life cycle. American Journal of Sociology, 86(2), 272–294. Waldfogel, J. (1997). The effect of children on women’s wages. American Sociological Review, 62(2), 209–217. https://doi.org/10.2307/2657300 Wang, M. (2005). Gender wage differentials in China’s urban labor market. Economic Research Journal (in Chinese), 12, 35–44. Wu, X., & Xie, Y. (2003). Does the market pay off? Earnings returns to education in urban
- China. American Sociological Review, 68(3), 425–442. https://doi.org/10.2307/1519731
Wu, X., Ye, H., & He, G. G. (2014). Fertility decline and women’s status improvement in China. Chinese Sociological Review, 46(3), 3–25. https://doi.org/10.2753/CSA2162-0555460301 Xie, Y., & Hannum, E. (1996). Regional variation in earnings inequality in reform-era urban
- China. American Journal of Sociology, 101(4), 950–992.
Xie, Y., & Wu, X. (2008). Danwei profitability and earnings inequality in urban China. The China Quarterly, 195, 558–581. https://doi.org/10.1017/S0305741008000775 Xiu, L., & Gunderson, M. (2013). Gender earnings differences in China: base pay, performance
SLIDE 38
38
pay, and total pay. Contemporary Economic Policy, 31(1), 235–254. Xu, Q. (2016). Migration and its impact on rural migrants’ attitudes towards gender roles. Collection of Women’s Studies (in Chinese), 6, 54–65. Yi, C.-C., & Chien, W.-Y. (2002). The linkage between work and family: Female’s employment patterns in three Chinese societies. Journal of Comparative Family Studies, 33(3), 451– 474. Yoon, S.-Y. (2017). The influence of a supportive environment for families on women’s fertility intentions and behavior in South Korea. Demographic Research, 36(7), 227–254. https://doi.org/10.4054/DemRes.2017.36.7 Yu, J., & Xie, Y. (2014). The effect of fertility on women’s wages in China. Population Research (in Chinese), 1, 18–29. Zhang, L., Brauw, A. de, & Rozelle, S. (2004). China’s rural labor market development and its gender implications. China Economic Review, 15(2), 230–247. https://doi.org/10.1016/j.chieco.2004.03.003 Zhang, Y., & Hannum, E. (2015). Diverging fortunes: The evolution of gender wage gaps for singles, couples, and parents in China, 1989-2009. Chinese Journal of Sociology, 1(1), 15–55. https://doi.org/10.1177/2057150X14568769 Zhang, Yuping, Hannum, E., & Wang, M. (2008). Gender-based employment and income differences in urban China: Considering the contributions of marriage and parenthood. Social Forces, 86(4), 1529–1560. https://doi.org/10.1353/sof.0.0035
SLIDE 39 39
Zhao, M. (2016). Gender equity should be promoted by public policies under the two-child
- policy. Population Research (in Chinese), 40(6), 38–48.
Zhao, Z. (2015). Closing a sociodemographic chapter of Chinese history. Population and Development Review, 41(4), 681–686. https://doi.org/10.1111/j.1728-4457.2015.00090.x Zheng, Z. (2016). Review on birth-related public policies: from the perspective of women and
- family. Population and Family Planning, 4, 28–29.
Zhou, Xiang. (2014). Increasing returns to education, changing labor force structure, and the rise
- f earnings inequality in urban China, 1996–2010. Social Forces, 93(2), 429–455.
https://doi.org/10.1093/sf/sou073 Zhou, Xueguang. (2000). Economic transformation and income inequality in urban China: Evidence from panel data. American Journal of Sociology, 105(4), 1135–1174. Zuo, J. (2012). Women’s liberation and gender obligation equality in urban China: Work/Family experiences of married individuals in the 1950s. Science & Society, 77(1), 98–125. https://doi.org/10.1521/siso.2013.77.1.98 Zuo, J., & Bian, Y. (2001). Gendered resources, division of housework, and perceived fairness-a case in urban China. Journal of Marriage and Family, 63(4), 1122–1133. https://doi.org/10.1111/j.1741-3737.2001.01122.x
SLIDE 40
40
Tables Table 1 Descriptive statistics of married women at first wave of observation in CHNS Women who have never worked (n=117) Women who experienced work interruption (n=691) Women who have been working for all waves (n=1,123) Age at first wave/Std.Dev 34.58 /7.82 32.48/ (6.30) 34.82/6.81 Highest education level (%) Primary school or lower 14.53 22.43 23.33 Middle school 52.99 43.85 28.05 High school 27.35 23.88 19.59 College or above 5.13 9.84 29.03 Stratum (%) Cities 27.35 22.58 29.39 Suburban neighborhoods 41.03 41.68 36.06 Towns or county capital cities 31.62 35.75 34.55
SLIDE 41
41
Table 2 Value of economic component score of communities by year Wave 1991 1993 1997 2000 2004 2007 2009 2011 Mean 3.96 4.11 5.70 6.74 7.70 8.13 8.65 8.72 Std.Dev 1.63 1.42 3.14 2.92 3.05 2.75 2.64 2.60
SLIDE 42 42
Table 3 Descriptive statistics by year
Year 1991 1993 1997 2000 2004 2006 2009 2011 Number of observations 858 863 914 915 799 777 689 555 Age 34 36 37 38 39 40 40 41 Labor force participation
92.82% 86.54% 77.92% 62.83% 64.99% 66.04% 72.61% Income of working women 4077 5055 6518 8321 11002 12661 17425 18758 Logarithm of income 8.1 8.3 8.5 8.8 9.0 9.2 9.5 9.6 Children under age 3 17.48% 9.97% 10.61% 7.76% 8.01% 6.82% 5.95% 5.95% Children between age 3 and 6 33.80% 29.48% 15.43% 11.91% 11.64% 10.81% 12.77% 9.55% Children between age 7 and 15 55.24% 59.68% 57.11% 52.02% 42.43% 36.42% 35.27% 35.50% Children older than age 15 30.42% 32.44% 40.81% 46.78% 56.95% 59.07% 58.49% 60.54% Highest educational level Primary school or lower 34.15% 33.72% 29.43% 20.66% 15.27% 12.48% 9.58% 9.37% Middle school 33.45% 32.90% 32.49% 34.32% 36.92% 38.74% 38.17% 39.46% High school 22.03% 22.36% 21.66% 24.04% 24.03% 23.55% 20.75% 18.38% College or above 10.37% 11.01% 16.41% 20.98% 23.78% 25.23% 31.49% 32.79% Stratum City 25.06% 26.42% 27.57% 25.79% 23.40% 23.29% 22.06% 23.60% Suburban 38.81% 36.38% 40.44% 40.44% 43.55% 42.73% 41.80% 40.72% Town or county capital city 36.13% 37.20% 30.42% 33.77% 33.04% 33.98% 36.14% 35.68% Labor force participation of husband 98.83% 98.38% 96.17% 91.26% 80.48% 81.21% 81.86% 86.13% Logarithm of household income
9.0 9.0 9.2 9.3 9.6 9.5 9.9 10.1 Living status of mother or mother-in-law Living in the same household 27.86% 28.62% 25.71% 24.92% 24.28% 23.68% 24.96% 27.75% Living in the same neighborhood/village 32.87% 30.01% 27.13% 31.25% 33.67% 30.89% 27.00% 25.77% Living in the same city/county 23.19% 22.71% 27.90% 27.21% 25.41% 26.77% 32.80% 26.49% Living in other city/county 7.34% 7.18% 6.89% 6.23% 6.51% 6.95% 4.21% 6.31% Not alive or unknown 8.74% 11.47% 12.36% 10.38% 10.14% 11.71% 11.03% 13.69% Women who have children under age 3 Number of observations 150 86 97 71 64 53 41 33 Age 28 28 28 28 29 30 30 31 Labor force participation 88.67% 87.21% 88.66% 70.42% 53.13% 56.60% 56.10% 54.55% Income of working women 3483 3861 5374 7696 9994 13563 16933 17122 Logarithm of income 7.9 7.8 8.2 8.6 8.7 9.4 9.4 9.6 Women who have children between age 3 and 6 Number of observations 290 254 141 109 93 84 88 53 Age 31 32 31 30 32 32 33 33
SLIDE 43
43 Labor force participation 92.76% 90.94% 92.20% 86.24% 67.74% 60.71% 61.36% 75.47% Income of working women 3623 4716 5949 7843 10668 14944 18251 20187 Logarithm of income 8.0 8.2 8.4 8.7 8.9 9.4 9.5 9.7 Women who have children between age 7 and 15 Number of observations 474 515 522 476 339 283 283 197 Age 37 37 37 38 38 37 38 38 Labor force participation 94.73% 92.82% 88.89% 80.04% 64.61% 69.26% 71.19% 75.13% Income of working women 4232 5202 6755 8512 10692 12079 16813 18115 Logarithm of income 8.1 8.3 8.5 8.8 9.0 9.1 9.5 9.6 Women who have children older than age 15 Number of observations 261 280 373 428 455 459 403 336 Age 39 42 42 43 43 43 44 45 Labor force participation 90.04% 93.57% 81.77% 74.30% 57.80% 59.69% 60.05% 69.64% Income of working women 4367 5593 6725 7852 10623 11798 17259 18218 Logarithm of income 8.1 8.3 8.5 8.6 9.0 9.0 9.6 9.6
SLIDE 44 44
Table 4 Coefficients of logistic model predicting working (Reference group: not working)
Random-effects Model Conditional fixed-effects Model Model R1 Model R2 Model F1 Model F2 Coefficients S.E Coefficients S.E Coefficients S.E Coefficients S.E Children under 3 years old
0.193
0.197
0.278
0.267 *economic component score
0.057
0.070 Children between 3 and 6 years old
0.158 0.023 0.163
0.207
0.201 *economic component score
0.047
0.064 Children between 7 and 15 years old 0.353** 0.124 0.427** 0.132 0.197 0.175 0.267 0.164 *economic component score
0.039
0.056 Children older than 15 years old 0.094 0.146 0.136 0.151
0.195
0.199 *economic component score
0.041
0.061 Age
0.011
0.011 Highest education level (ref: Primary school or lower) Middle school 0.262* 0.132 0.268* 0.133 High school 0.442** 0.147 0.432** 0.148 College or above 0.592** 0.207 0.532* 0.208 Stratum (ref: city) Suburban 0.324* 0.150 0.341* 0.149 Town or county capital city
0.143 0.006 0.143 Working status of husband (ref: not working) 1.982*** 0.168 1.971*** 0.168 2.051*** 0.254 2.064*** 0.256 Income of household other than women’s income
0.035
0.035
0.044
0.045 Living status of mother or mother-in- law (ref: Living in the same household) Living in the same neighborhood/village
0.129
0.129
0.222
0.218 Living in the same city/county
0.143
0.143
0.234
0.226 Living in other city/county
0.220
0.219
0.368
0.369 Not alive or unknown
0.173
0.174
0.294
0.290 Community-level variables Economic component score
0.021 0.083+ 0.046
0.033 0.166* 0.068 Quality of health score 0.071** 0.024 0.068** 0.024 0.074** 0.025 0.069** 0.024 Sanitation score
0.030
0.030
0.060
0.059 Housing component score 0.046 0.035 0.048 0.035
0.082 0.000 0.081 Dummies for wave (ref: year 1991)
SLIDE 45 45
Year 1993
0.231
0.232
0.228
0.233 Year 1997
0.219
0.222
0.264
0.272 Year 2000
0.223
0.226
0.344
0.354 Year 2004
0.237
0.239
0.385
0.397 Year 2006
0.243
0.245
0.400
0.414 Year 2009
0.259
0.262
0.450
0.460 Year 2011
0.269
0.271
0.477
0.487
Note: Only women who experienced work interruption are included because of the settings of conditional fixed-effects logistic model. For a robustness check, random-effects logistic model is conducted (Appendix Table 1). Economic component score measures community-level economic activity with a range between 0 and 10, including typical daily wage for ordinary male workers and percentage of the population engaged in nonagricultural work. The variable is centered at 5.
+ p < 0.10, * p < 0.05, ** p < 0.01, *** p < 0.001
SLIDE 46 46
Table 5 Coefficients of linear model of women’s logged income
Random-effects Model Fixed-effects Model Model R1 Model R2 Model F1 Model F2 Coefficients S.E Coefficients S.E Coefficients S.E Coefficients S.E Children under 3 years old
0.045
0.046
0.054
0.058 *economic component score
0.013
0.014 Children between 3 and 6 years old 0.039 0.035 0.017 0.036 0.060 0.047 0.024 0.049 *economic component score
0.010
0.014 Children between 7 and 15 years old 0.112*** 0.030 0.119*** 0.031 0.108** 0.037 0.103* 0.040 *economic component score
0.009
0.012 Children older than 15 years old 0.026 0.035
0.035 0.017 0.045
0.049 *economic component score
0.010
0.012 Age 0.009*** 0.003 0.009*** 0.003 Highest education level (ref: Primary school
Middle school 0.084* 0.037 0.080* 0.038 High school 0.136** 0.042 0.130** 0.042 College or above 0.436*** 0.049 0.420*** 0.049 Stratum (ref: city) Suburban 0.077 0.070 0.076 0.070 Town or county capital city
0.066
0.066 Living status of mother or mother-in-law (ref: Living in the same household) Living in the same neighborhood/village 0.037 0.032 0.032 0.032 0.085 0.056 0.079 0.056 Living in the same city/county
0.034
0.034
0.061
0.061 Living in other city/county 0.022 0.050 0.018 0.050 0.120 0.095 0.115 0.095 Not alive or unknown
0.044
0.044
0.070
0.070 Community-level variables Economic component score
0.006 0.025* 0.011
0.010 0.013 0.014 Quality of health score
0.006
0.006
0.010
0.010 Sanitation score 0.027** 0.009 0.027** 0.009
0.019
0.018 Housing component score 0.039*** 0.010 0.044*** 0.010 0.005 0.026 0.012 0.026 Dummies for wave (ref: year 1991) Year 1993 0.119*** 0.035 0.118*** 0.035 0.161** 0.048 0.159** 0.048 Year 1997 0.274*** 0.041 0.273*** 0.041 0.465*** 0.080 0.462*** 0.079 Year 2000 0.461*** 0.046 0.462*** 0.046 0.691*** 0.113 0.687*** 0.112 Year 2004 0.654*** 0.053 0.649*** 0.053 0.951*** 0.133 0.937*** 0.133 Year 2006 0.781*** 0.056 0.774*** 0.056 1.139*** 0.125 1.123*** 0.125
SLIDE 47 47
Year 2009 1.156*** 0.060 1.144*** 0.060 1.608*** 0.134 1.578*** 0.135 Year 2011 1.251*** 0.063 1.240*** 0.063 1.770*** 0.141 1.734*** 0.142
Note: All models are not adjusted for employment selection
+ p < 0.10, * p < 0.05, ** p < 0.01, *** p < 0.001
SLIDE 48 48
Appendix Appendix Table 1 Coefficients of random-effects logistic model predicting working (Reference group: not working)
Model 1 Model 2 Coefficients S.E Coefficients S.E Children under 3 years old
0.196
0.204 *economic component score
0.055 Children between 3 and 6 years old
0.165
0.172 *economic component score
0.048 Children between 7 and 15 years old 0.050 0.132 0.152 0.139 *economic component score
0.039 Children older than 15 years old
0.154
0.162 *economic component score
0.042 Age 0.012 0.011 0.013 0.011 Highest education level (ref: Primary school or lower) Middle school 0.277+ 0.164 0.284+ 0.166 High school 0.696*** 0.186 0.690*** 0.188 College or above 2.751*** 0.244 2.708*** 0.245 Stratum (ref: city) Suburban 0.431 0.272 0.450+ 0.269 Town or county capital city
0.256 0.049 0.254 Working status of husband (ref: not working) 2.169*** 0.153 2.176*** 0.154 Income of household other than women’s income
0.036
0.036 Living status of mother or mother-in-law (ref: Living in the same household) Living in the same neighborhood/village
0.145
0.146 Living in the same city/county
0.154
0.155 Living in other city/county
0.233
0.234 Not alive or unknown
0.187
0.188 Community-level variables Economic component score
0.023 0.139** 0.047 Quality of health score 0.078** 0.025 0.072** 0.025 Sanitation score
0.039
0.039 Housing component score 0.011 0.046 0.025 0.046 Dummies for wave (ref: year 1991) Year 1993
0.215
0.216 Year 1997
0.222
0.224 Year 2000
0.236
0.238 Year 2004
0.256
0.258 Year 2006
0.269
0.270 Year 2009
0.288
0.290 Year 2011
0.302
0.303
Note: All women are included.
+ p < 0.10, * p < 0.05, ** p < 0.01, *** p < 0.001
SLIDE 49 49
Appendix Table 2 Coefficients of linear model of women’s logged income
Random-effects Model Fixed-effects Model Model R3 Model R4 Model F3 Model F4 Coefficients S.E Coefficients S.E Coefficients S.E Coefficients S.E Children under 3 years old
0.046
0.047
0.059
0.063 *economic component score
0.013
0.014 Children between 3 and 6 years old 0.036 0.036 0.014 0.036 0.060 0.047 0.024 0.049 *economic component score
0.010
0.014 Children between 7 and 15 years old 0.114*** 0.030 0.120*** 0.031 0.108** 0.037 0.102* 0.039 *economic component score
0.009
0.012 Children older than 15 years old 0.019 0.035
0.036 0.017 0.046
0.049 *economic component score
0.010
0.012 Age 0.010*** 0.003 0.009*** 0.003 Highest education level (ref: Primary school or lower) Middle school 0.083* 0.037 0.080* 0.038 High school 0.140*** 0.042 0.134** 0.042 College or above 0.471*** 0.055 0.450*** 0.056 Stratum (ref: city) Suburban 0.081 0.070 0.079 0.070 Town or county capital city
0.066
0.066 Living status of mother or mother-in- law (ref: Living in the same household) Living in the same neighborhood/village 0.027 0.033 0.024 0.033 0.085 0.057 0.082 0.057 Living in the same city/county
0.034
0.034
0.062
0.062 Living in other city/county 0.014 0.051 0.011 0.051 0.120 0.094 0.116 0.095 Not alive or unknown
0.046
0.046
0.070
0.070 Community-level variables Economic component score
0.006 0.024* 0.011
0.010 0.014 0.014 Quality of health score
0.006
0.006
0.010
0.010 Sanitation score 0.025** 0.009 0.025** 0.009
0.018
0.018 Housing component score 0.039*** 0.010 0.044*** 0.010 0.005 0.026 0.012 0.026 Dummies for wave (ref: year 1991)
SLIDE 50 50
Year 1993 0.119*** 0.035 0.118*** 0.035 0.161*** 0.048 0.159** 0.048 Year 1997 0.263*** 0.042 0.264*** 0.042 0.465*** 0.081 0.464*** 0.080 Year 2000 0.439*** 0.049 0.443*** 0.049 0.691*** 0.117 0.691*** 0.115 Year 2004 0.614*** 0.061 0.614*** 0.061 0.951*** 0.146 0.945*** 0.144 Year 2006 0.745*** 0.062 0.743*** 0.062 1.139*** 0.137 1.130*** 0.135 Year 2009 1.119*** 0.066 1.112*** 0.066 1.608*** 0.143 1.585*** 0.143 Year 2011 1.222*** 0.067 1.215*** 0.067 1.770*** 0.146 1.738*** 0.146
Note: All models are adjusted for employment selection (by including inverse Mills’ ratio estimated from Heckman two-step selection model clustering at individual level)
+ p < 0.10, * p < 0.05, ** p < 0.01, *** p < 0.001
