Childcare Costs and Migrant and Local Mothers' Labor Force Participation in Urban China

ABSTRACT

This study empirically analyzes the impact of childcare costs on the labor force participation (LFP) and childcare utilization of migrant and local mothers of preschool-age children in urban China, using data from the 2010 National Dynamic Monitoring Survey of Floating Populations. The estimates show that childcare costs have a strong negative effect on the LFP and childcare utilization of migrant and local mothers. Compared to local mothers, the LFP and childcare utilization rate of migrant mothers are, respectively, more and less sensitive to changes in childcare costs. The analysis indicates that the LFP and childcare choices of migrant mothers are more constrained by the lack of access to affordable childcare than are local mothers in China.

KEYWORDS:

• Childcare
• migrant mothers
• labor force participation
• China

JEL Codes:

• J16
• J13
• R23

ACKNOWLEDGMENTS

This work was carried out with the aid of a grant from the International Development Research Center of Canada (Project no. 107579) and a grant from the National Social Science Foundation of China (Grant no. 15BRK036). We also benefited greatly from the comments and suggestions of Rachel Connelly and the anonymous reviewers.

Notes

1 See Du and Dong (2013) for a review of the evolution of childcare policy in China during the economic transition.

2 In this contribution, the term “childcare services” includes the services provided by both nurseries and kindergartens. The former refers to center-based childcare for children ages 0–2, and the latter refers to center-based childcare, including preschool programs, for children ages 3–6. Center-based childcare is termed “formal childcare.”

3 The five cities covered by the survey, namely Beijing, Chengdu, Zhengzhou, Zhongshan, and Suzhou, all have populations of over five million. We are unable to derive an average childcare cost at the community level because the survey does not provide information that can be used to identify neighborhood communities. We do not apply the approach of Connelly (1991, 1992), which relates childcare costs to a mother's paid work hours, because in our sample, many mothers not working in the labor market also enrolled their children in childcare programs for preschool education.

4 This model assumes that childcare services are available if the price is sufficiently high. This assumption is valid for urban China in the 2010s, where the private childcare market has grown rapidly. As in other market economies, the essence of childcare accessibility in urban China is the lack of access to affordable childcare among families of low socioeconomic status.

5 Less than 10 percent of migrant mothers and less than 5 percent of local mothers have more than one preschool-age child in the sample.

6 The husband's earnings are assumed to be exogenous to a mother's LFP decision. This assumption is plausible, given the influence of traditional gender norms in Chinese society. Despite women's emancipation movements during Mao's era, the husband is still deemed the primary earner and the wife the secondary earner in the post-reform period.

7 Arguably, some of the household composition variables, such as the presence of a child ages 0–2 and the presence of an elderly parent, may be endogenous if a mother's LFP decision is made jointly with the family's fertility and co-residence decisions. Due to data limitations, we are unable to address the potential endogeneity bias of these variables. Nevertheless, as Rachel Connelly et al. (2006) point out, the endogeneity of fertility is a matter of degree. Biologically speaking, fertility cannot be planned exactly and therefore always contains exogenous elements. China's birth control policies generate additional exogenous variations in fertility, thereby weakening its simultaneous association with maternal LFP decisions. With respect to the presence of elderly parents, the variable is statistically insignificant in most cases, and its exclusion does not cause noticeable changes in the estimates.

8 Constrained by the lack of information on the features of childcare a mother used, we do not control for the quality of childcare.

9 By measuring childcare costs as expenditures on childcare per hour the mother worked in the labor market, Connelly (1992) and Powell (1997) corrected for selection bias in the childcare cost equation by estimating a bivariate probit model that takes into account two decisions: the use of paid care decision and the LFP decision. Our selection equation limits our attention to the decision to use paid childcare, as our measure of childcare costs takes into account the decisions of mothers who did not participate in the labor market but did enroll their children in childcare for preschool education.

10 The estimation procedure is inspired by Jackline Wahba (2015).

Additional information

Notes on contributors

Yueping Song

Yueping Song is Associate Professor at the Institute of Population and Development at Renmin University of China. Her research interests focus on migration, gender issues in the labor market, and health economics.

Xiao-yuan Dong

Xiao-yuan Dong is Professor in the Department of Economics at the University of Winnipeg, Canada, and Adjunct Professor at the National School of Development, Peking University, China. She has served on the board of directors of the International Association of Feminist Economics and is an Associate Editor of Feminist Economics. Her research focuses on China's economic transition and development, with an emphasis on labor and gender issues.