No.72-Parental Migration Decisions and Child Health Outcomes: Evidence from China
Lin, Carl; van der Meulen Rodgers, Yana
Published: 2018/8/13 22:00:00    Updated time: 2018/8/15 11:02:55
Abstract: This study uses migrant household survey data from 2008 and 2009 to examine how parental migration decisions are associated with the nutritional status of children in rural and urban China. Results from instrumental variables regressions show a substantial adverse effect of children’s exposure to parental migration on height-for-age Z-scores of left-behind children relative to children who migrate with their parents. Additional results from a standard Blinder-Oaxaca decomposition, a quantile decomposition, and a counterfactual distribution analysis all confirm that children who are left behind in rural villages – usually because of the oppressive hukou system – have poorer nutritional status than children who migrate with their parents, and the gaps are biggest at lower portions of the distribution.
Keywords: migration; China; children; health; nutrition

( forthcoming in Research in Labor Economics 2018)

Authors:

Carl Lin, Bucknell University, CIID, and IZA

Yana van der Meulen Rodgers, Rutgers University

 

I. Introduction

       Insufficient food consumption and the lack of a healthy diet for children can result in unwanted weight loss, fatigue, headaches, poor mental health, and frequent illness. Childhood health in turn serves as an important determinant of an individual’s health status in adulthood and of his or her likelihood of developing costly and debilitating health conditions. In addition, children’s nutritional status and health are associated with performance in school and years of educational attainment, both of which serve as important predictors of future labor market outcomes, especially wages and occupational attainment. The literature provides strong evidence that relates childhood health and nutritional status to cognitive development, school performance, and future success in the labor market.ii See Alderman et al. (2006) and Currie (2009) for reviews of the literature on the long-term effects of children’s health. For example, Victora et al. (2008) conducted an extensive meta-analysis as well as their own analysis of data for five developing countries and found that that low height-for-age and weight-for-age at two years of age are associated with long-term impairment in educational attainment, school performance, adult height, productivity, and earnings.  The authors concluded that nutritional deprivation among children is an important mechanism that can undermine the health outcomes of successive generations.

Children’s nutritional status is affected by a number of factors that include environmental exposure, food intake, illnesses, and other external determinants that are influenced by socioeconomic status (Puffer and Serrano 1973).  One of these determinants is parental employment, and a large body of work indicates that one of the most important channels through which parental employment affects child health and nutritional status is through the income that they earn.iiii For evidence on the relationship between household socioeconomic status and child health, see, for example, Bhattacharya et al. (2004), Chowa et al. (2010), Currie and Lin (2007), and Rodgers (2011).    Yet parents’ participation in the labor market can entail a fundamental tradeoff. The income that parents earn contributes to the household’s ability to purchase goods and services that improve children’s health and nutritional status. However, parents’ market-based work could reduce the quantity or quality of time spent caring for children, with potentially adverse effects on child well-being. Just like household income, time spent with children also affects the degree to which parents can engage in care practices that influence child nutrition and health (Tracey and Polachek 2018). 

This tradeoff between income from market-based work and time spent away from children can be heightened for parents who have migrated to urban areas and left their children behind in rural villages in the care of others. Parental remittances may improve children’s nutritional status through the purchase of more nutritious foods and through housing improvements that are conducive to children’s health. However, migrant parents are apart from their children and the quality of care from substitute care-providers may be inferior.  Migrant parents who bring their children with them also face this tradeoff given the pressure that migrants face to work long hours in paid employment in order to stave off the risk of economic hardship that comes with rural-to-urban migration. 

This study, to the best of our knowledge, is the first to examine how parental migration decisions relate to the nutritional status of both children left behind and children who migrate with their parents. In 2010, approximately 61 million children ages 0-17 were left behind in rural villages, accounting for 38 percent and 22 percent of all rural children and all children nationally, respectively (China Women’s Federation Research Team 2013).iiiiii The study defines left-behind children as cases in which both parents or just one (father or mother) have migrated from rural to urban areas while leaving the child behind in the rural village to which the household registration (hukou) belongs. Among the 61 million left-behind children in 2010, 47 percent have had both parents migrate to cities; 36 percent have had just fathers migrate; and 17 percent have had just mothers migrate. Our paper uses the same definition. The number of left-behind children has been increasing rapidly, with, for example, an increase of 2.42 million children from 2005 to 2010.  Note that this study defines left-behind children as cases in which both parents or just one (father or mother) have migrated from rural to urban areas while leaving the child behind in the rural village to which the household registration (hukou) belongs.iviv  As a robustness check, we created four separate categories of left behind children: all left behind children, both parents work in city and leave child behind; mother works in city and child is left-behind with father; and father works in city and child is left behind with mother.  We then created a set of kernel density graphs showing difference between migrating children and children in each of these categories. These graphs (available upon request) show that the differences in HAZ and WAZ scores between the four categories of left behind children are negligible. Among the 61 million left-behind children in 2010, 47 percent have had both parents migrate to cities; 36 percent have had just fathers migrate; and 17 percent have had just mothers migrate.

China constitutes an important case study not only because it is the world’s most populous country that is experiencing the biggest internal migration flow ever (estimated at 245 million people in 2016 by National Bureau of Statistics of China), but also because it has tried to manage an enormous flow of rural to urban migration with an institutionalized system of household registration known as the hukou that may have unintended consequences for children’s well-being. In particular, the hukou is a household registration system based on either a rural or urban classification that depends mostly on birthplace of the household head and is very difficult to change.  Moreover, many public services in urban areas are restricted to individuals with urban hukou only, thus excluding rural-to-urban migrants who still have their rural hukou and denying them access to public healthcare, schooling, and social services. 

This analysis utilizes data from the Longitudinal Survey on Rural Urban Migration in China (RUMiC), a rich dataset on migrant workers and their households that has detailed information on human capital indicators, socioeconomic status, food expenditures, and health. The data are used to examine the determinants of children’s nutritional status, as measured by weight-for-age Z-scores (WAZ scores) and height-for-age Z-scores (HAZ scores), with a focus on how children’s cumulative exposure to parental migration affects the health outcomes of migrating children relative to left-behind children. Multiple approaches are used to clearly identify these effects, including ordinary least squares regressions, instrumental variables, a standard Blinder-Oaxaca decomposition, and quantile decompositions based on re-centered influence function (RIF) regressions (Firpo et al. 2009; Fortin et al. 2011).  Lastly, we estimate a set of counterfactual quantile treatment effects (Chernozhukov et al. 2013) by constructing a counterfactual scenario which captures what the WAZ or HAZ distribution would be if left-behind children were to live with their parents in cities.

 

________________________
i See Alderman et al. (2006) and Currie (2009) for reviews of the literature on the long-term effects of children’s health.
ii For evidence on the relationship between household socioeconomic status and child health, see, for example, Bhattacharya et al. (2004), Chowa et al. (2010), Currie and Lin (2007), and Rodgers (2011).
iii The study defines left-behind children as cases in which both parents or just one (father or mother) have migrated from rural to urban areas while leaving the child behind in the rural village to which the household registration (hukou) belongs. Among the 61 million left-behind children in 2010, 47 percent have had both parents migrate to cities; 36 percent have had just fathers migrate; and 17 percent have had just mothers migrate. Our paper uses the same definition.
iv As a robustness check, we created four separate categories of left behind children: all left behind children, both parents work in city and leave child behind; mother works in city and child is left-behind with father; and father works in city and child is left behind with mother. We then created a set of kernel density graphs showing difference between migrating children and children in each of these categories. These graphs (available upon request) show that the differences in HAZ and WAZ scores between the four categories of left behind children are negligible.

 

[生成二维码分享]
Share to