Non-Parental Childcare Services and Time Allocation of Mothers with Young Children in China

Abstract

This study explores the impact of access to and affordability of paid and unpaid childcare services on the time allocation of mothers with children ages 0–6. The study employs a fixed-effect seemingly unrelated regression model on longitudinal data from the China Health and Nutrition Survey for 2004–11, when women’s employment in China was declining rapidly. The study finds that holding constant other determinants, doubling daily wages of nannies and tuition fees of childcare centers tends to reduce a mother’s market work time by 5.2 and 1.3 h per week and increases her time spent on housework by 1.7 and 0.5 h per week, respectively. Mothers who live with an older, woman relative spend 5.5 h fewer per week on childcare. Access to local childcare centers reduces mothers’ time spent on childcare by 13.3 h per week, and these mothers’ wage rates have no effect on their time allocation.

HIGHLIGHTS

• In China, rising parenting costs contribute to declines in women’s labor participation rate and the fertility rate.

• Increasing prices of childcare services reduce mother’s time on paid work and increase time on housework.

• Access to childcare has no impact on mothers’ time on activities beyond childcare.

• The government should subsidize early childhood education as it subsidizes elementary education.

• Childcare leave and flexible work arrangements may alleviate mothers’ time burdens.

KEYWORDS:

• Non-parental child service
• maternal time allocation
• fixed-effect seemly unrelated regression estimator
• mothers with young children
• China

JEL:

• Codes: J16
• J22
• D13

ACKNOWLEDGMENTS

The authors thank the associate editor and anonymous reviewers for their valuable comments and suggestions. This research uses data from China Health and Nutrition Survey (CHNS). We thank the National Institute of Nutrition and Food Safety, China Center for Disease Control and Prevention, Carolina Population Center, the University of North Carolina at Chapel Hill, the NIH (R01-HD30880, DK056350, and R01-HD38700) and the Fogarty International Center, NIH for financial support for the CHNS data collection and analysis files from 1989 to 2006 and both parties plus the China–Japan Friendship Hospital, Ministry of Health for support for CHNS 2009 and future surveys.

Notes

1 Dr. Yanyan Xiong is the corresponding author of this article.

2 In China, the average age of a college graduate is 22 years, and the average retirement age for women is 55 years. Women’s participation rates are calculated using Chinese Urban Household Survey data.

3 The real per capita tuition fees are deflated by consumer price index and based on 2017 prices. In comparison, the real tuition fees for primary and junior high education (in 2017 prices) increased fewer than 6 times, from 72 and 124 yuan per capita per year in 1996 to 410 and 844 yuan per capita per year in 2017, respectively.

4 Authors’ calculation according to the statistics from China Labour Statistical Yearbook (National Bureau of Statistics and Ministry of Human Resources and Social Security 2018).

5 The real per capita tuition fees are deflated by consumer price index and based on 2017 prices.

6 Du and Dong (2013) elaborate on the changes in China’s childcare policy and reforms. For a more detailed discussion of childcare in China, please refer to their excellent work.

7 Extensive evidence shows that sleep time is also affected by demographic and social factors, thereby correlating with time allocation in other activities (Biddle and Hamermesh 1990; Basner et al. 2007).

8 If t*_j<0, then t_j = 0 in our sample.

9 Daily wages of nannies have been reported by the interviewees. The following question was asked in the questionnaires of the CHNS: What is the daily wage for this occupation (nanny hired by households) in the village/neighborhood?

10 This information was also provided by interviewees. Questions were asked about the availability and price of childcare centers in the questionnaires: Is a childcare center present for this age group in this village/neighborhood? What is the average monthly fee for this age group in this center?

11 D₁ equals one if there is an older, woman member living in the same household and zero otherwise. D₂ equals one if there is a childcare center for children younger than 2.

12 In our sample, on average, 26.5 percent of mothers with young children were not working or reported zero hours on market work. Results of the Heckman two-stage estimation are available upon request. The first stage is to estimate the selection equation with a binary dependent variable that is equal to 1 if the woman works (and thus, her wage level can be observed) and 0 otherwise. A woman’s age, age squared, years of schooling, whether she spends longer time on farm work relative to non-farm work, whether she has children ages 0 to 5, dummies of regions, and survey waves are controlled for in the first stage. The two independent variables, whether the mother spends longer time on farm work relative to non-farm work and whether she has children ages 0 to 5, are excluded from the second-stage outcome equation, and both variables have a substantial impact on the probability of selection. The first-stage estimates show that the variable of whether the mother spends longer time on farm work relative to non-farm work has a significant and positive effect on the probability of working, and whether she has children ages 0–5 has a significant and negative effect on the probability of working. Moreover, the two variables may not directly affect a woman’s wage rate, which may be primarily determined by her human capital and the region’s heteroscedasticity. The lambda from Heckman’s two-stage estimation is significant from zero at the 1 percent significance level, indicating that the two-step estimator produces the evidence of selection.

13 This study does not include the number of children. China had been implementing a one child policy for almost four decades until recently. In our sample, 95 percent of the families had only one child. Therefore, unlike studies in other countries, the number of children is not a key variable in determining the time use of women.

14 The detailed information on survey design can be found in the work of Barry M. Popkin et al. (2010), Du and Dong (2013), and B. Zhang et al. (2014).

15 A total of 340 households have been repeatedly interviewed in two waves; ninety-eight households have been surveyed in three waves, and three households stayed in four waves.

16 More specifically, leisure time in the current study may have been undercounted because the counted leisure activities listed in the questionnaires were not exhaustive. Physical activities include martial arts (for example, kung fu), gymnastics, dancing, acrobatics, track and field (running), swimming, soccer, basketball, tennis, badminton, volleyball, and others (ping pong, tai chi). Meanwhile, sedentary activities include watching TV, movies, and videos, playing video or computer games, surfing the Internet, reading (books, newspapers, and magazines), writing, and drawing.

17 Physical activities include martial arts (for example, kung fu), gymnastics, dancing, acrobatics, track and field (running), swimming, soccer, basketball, tennis, badminton, volleyball, and others (ping pong, tai chi). Meanwhile, sedentary activities include watching TV, movies, and videos, playing video or computer games, surfing the Internet, reading (books, newspapers, and magazines), writing, and drawing.

18 Additional information on the 2008 CTUS is available in the recent work of Dong and An (2015).

19 The number, 70.7 percent, is estimated using Compilation of Chinese Time Use Survey Data 2008.

20 The CTUS data are calculated by the authors using statistical data collected from Compilation of Chinese Time Use Survey Data 2008, released by National Bureau of Statistics (http://www.stats.gov.cn/ztjc/ztsj/2008sjly/).

21 The regional dummies are urban areas, suburbs, counties and villages, and sample province.

22 The Breusch–Pagan test of independence for pooled- and FE-SUREs was conducted, showing that the correlations among five-time equations exist and thus, SUR estimators are more efficient.

23 On average, 91.3 percent of women ages 16 and above worked at least forty hours per week in urban China in 2017 (National Bureau of Statistics and Ministry of Human Resources and Social Security 2018).

Additional information

Funding

This work was supported by the Fundamental Research Funds for the Central Universities; Social Science Foundation of Jiangsu Province 13ZHC007; National Natural Science Foundation of China Project 71573146 and Project 71673317; and The National Social Science Fund of China 18BJY051.

Notes on contributors

Jing Liu

Jing Liu is Professor in the School of Economics at the Minzu University of China. Her major is labor economics and her research focuses on the gender economics, children’s development, and human capital.

Liangshu Qi

Liangshu Qi is Associate Professor at Tsinghua University. She graduated from Peking University, China. Her research focuses on time poverty and intergenerational transfer.

Yanyan Xiong

Yanyan Xiong is ZJU100 Young Professor in School of Economics, and Research Fellow in Center of Social Welfare and Governance and in Research Center for Socialism with Chinese Characteristics, Zhejiang University. She is also Research Fellow (Honorary) at Lau Chor Tak Institute of Global Economics and Finance, The Chinese University of Hong Kong. Her research areas are human capital investment, education inequality, and China’s economy.