Parents’ Migration and Adolescents’ Transition to High School in Rural China: The Role of Parental Divorce

Abstract

Using both quantitative and qualitative data collected in a migrant-sending county from 2012 to 2013, this article examines the mechanisms through which parental migration could shape adolescents’ transition to high school in rural China. Though parental migration improves children’s educational outcomes via social remittance of education value, it also leads to a decline in children’s educational achievements by increasing the odds of parental divorce. The likelihood of divorce rises with the migration of mother or both parents, and this significantly increases the risks of discontinuing schooling and transitioning to vocational high schools, relative to attending academic high schools. In contrast to the conventional explanations of economic resources and psychological health, this article emphasizes the significant role of marital instability in the link between parental migration and children’s educational outcomes.

Keywords

parental migration divorce adolescent educational outcome transition to high school rural China left-behind children

Introduction

Labor out-migration has led to major transformations in family structure and dynamics in many developing societies (All China Women’s Federation, 2013; Bryant, 2005). Due to the costs and uncertainties of migration and resettlement and the exclusion of migrants and their families from social welfare and services in destination areas, most parents prefer to leave their children with the other parent or with relatives when they relocate to pursue work. As a result, an increasing number of children of migrants are growing up in the absence of one or even two parents. In 2010, 61 million rural children were left behind in China alone when one or both parents migrated, most of whom moved to urban areas in search of better economic opportunities (All China Women’s Federation, 2013).

News reports and ethnographic studies in rural China often suggest that left-behind children tend to suffer academically, emotionally, and socially (Xiang, 2007; Ye & Pan, 2011). However, some recent comparative studies based on nationally representative samples found that left-behind children fare as well as their non-left-behind counterparts, when children’s demographic characteristics, parents’ socioeconomic status, and neighborhood characteristics are controlled for (Ren & Treiman, 2013; Xu & Xie, 2015; Yeung & Gu, 2016). Another study performing meta-analysis on data from 27 surveys—which are conducted by a single research team using a common sampling strategy and a standardized data collection instrument between 2009 and 2013—found that overall both left-behind children and non-left-behind children performed poorly on most of the examined indicators of health, nutrition, and education, even if the former did as well as or better than the latter (Zhou et al., 2015). These studies, however, tend to focus on the total effect of parental migration on child development and do not differentiate the multiple ways—both positive and negative—through which parental migration can affect children’s lives.

This study conceptualizes parental migration as a dynamic and multifaceted phenomenon, and simultaneously investigate multiple mechanisms through which parental migration affects children’s well-being. In addition to the conventional explanations of economic resources and psychological well-being, I highlight the significant role of marital instability in the link between parental migration and children’s educational outcomes. Labor out-migration of adult members of a Chinese family often involves labor division between genders and generations (Fan, 2003, 2008). Spouses may become separated from each other for long periods of time due to out-migration of one partner or migration of both partners to different destinations (Li & Zhang, 2016). Long-term spousal separation is likely to lead to marital instability or estrangement and dissolution. Given the numerous studies showing negative effects of parental divorce on children’s lives (Amato, 2000; Amato & Cheadle, 2005; Amato & Keith, 1991; Kim, 2011; MacLanahan & Sandefur, 1994), it is worth exploring whether parental migration affects educational outcomes through parental divorce.

This study on parental migration and children’s educational outcomes is situated in the context of China, where, as a result of the country’s rapid economic transformation, the largest human movement in history is taking place. Specifically, I examine the impact of parental migration on left-behind adolescents’ transition from middle school to high school in rural China. Transition to high school is a crucial landmark in rural youth’s lives, given the importance of educational attainment in securing a better job and lifelong economic prospects. Since the college expansion policy in the late 1990s, the opportunity of higher education has increased considerably for young Chinese, though benefiting different social groups unequally (Yao, Wu, Su, & Wang, 2010; Yeung, 2013). Compulsory education in China covers only 6 years of primary school and 3 years of middle school. Middle school graduates have to compete for limited slots in high school by taking the High School Entrance Exam. Based on the results of this exam, adolescents are placed in two different tracks of high school: academic versus vocational; and not surprisingly, an academic high school provides much better prospects of higher education after graduation. Researchers have shown that the transition from middle school to high school remains a bottleneck for rural youth in educational advancement (Hannum, An, & Cherng, 2011; C. Liu et al., 2009).

How does parents’ migration shape adolescents’ transition from middle school to high school in rural China? I investigate this central question using both detailed survey and in-depth interview data on final-year adolescents in middle schools located in a migrant-sending county in central China. I assess how the effect of parents’ migration varies by parents’ gender. Throughout the analysis, I focus on the social and economic mediating mechanisms underlying the effect. Generalized structural equation modeling (GSEM) is used to delineate the direct and indirect effects of parental migration on educational outcomes.

Parents’ Migration and Children’s Education

Research on how parental migration affects children’s educational outcomes in developing countries has produced mixed findings. Parental migration can generate income and relieve household financial constraints, promoting children’s education by providing resources for food, health care, and education (Lu & Treiman, 2011; Yang, 2008). But, the lengthy absences of migrant parents can jeopardize children’s psychosocial well-being by reducing the care, supervision, and academic assistance that children receive at home, thereby unfavorably affecting their educational outcomes (Dreby, 2007; Lahaie, Hayes, Piper, & Heymann, 2009; Xiang, 2007). Parental migration may encourage children of migrants to pursue migrant labor, like their parents, and thus discourage them from pursuing higher educational aspirations (Kandel & Kao, 2001). Again, a number of studies have found neutral impact of parental migration on the education of left-behind children (Arguillas & Williams, 2010; Jordan & Graham, 2012; Lu, 2012; Xu & Xie, 2015; Zhou et al., 2015). Moreover, the relationship between parental migration and children’s education can vary across different contexts (Lu, 2014). These findings highlight the importance of identifying the underlying economic and social mechanisms through which parental migration affects children’s educational outcomes.

This article develops a synthesized framework of parental migration and examines multiple mechanisms linking parental migration to adolescents’ educational outcomes (see Figure 1). The research framework clearly demonstrates that parental migration has different implications for distinct forms of family and parental resources that are important to the development and well-being of children. The overall effect of parental labor migration on children’s educational outcomes depends on the beneficial pathways countervailing the adverse ones. It is thus imperative to examine not only the total effect of parental migration on children’s well-being but also the various underlying mechanisms. It is also important to distinguish the migration of the father from that of the mother. Previous studies have revealed that the association between parental migration and child outcome varies by the gender of migrant parent (Lu, 2012).

Figure 1.

A research framework of parental migration and children’s educational outcomes.

The Economic Resources Mechanism

The most studied mediating channel linking parental migration and child development is the economic resources mechanism. Increased financial resources, enabled by parental labor migration, ease access to adequate food, shelter, other material goods, stimulating learning materials, and educational investment. This may positively affect children’s nutrition, health, cognitive development, and educational attainment. A number of empirical studies have shown positive effects of parental labor migration and remittance on left-behind children’s educational outcomes (Acosta, 2006; Edwards & Ureta, 2003; Lu & Treiman, 2011; Morooka & Liang, 2009; Yang, 2008). In this study, I use the number of books available at home to capture the educational resources parents invest in children. Migrant parents may have more resources at their disposal to buy books for their children. Previous research has shown that growing up in homes with many books has a substantial positive effect on children’s educational attainment (Evans, Kelley, Sikora, & Treiman, 2010).

The Psychological Well-Being Mechanism

Nonmaterial family resources such as attention, care, and supervision are also important to child development (Coleman, 1988). In particular, attachment theory posits that developing relationships with caregivers plays a major role in the child’s social and emotional development (Ainsworth, 1989; Bowlby, 1988). Parents are usually the primary caregivers who monitor the child’s activities, respond to the child’s needs, and provide the child with care, security, protection, and emotional support. A secure attachment to parents forms a basis from which children can learn and develop new skills and interests in life. The physical absence of one or both parents may decrease parental attention and care, disrupt attachment behavior, and increase risks of emotional and psychological stress for the child. In rural China, left-behind children are usually separated from migrant parents for extended spells of time. Studies suggest that left-behind children tend to feel lonely, afraid, abandoned, depressed, and become self-enclosed or even lose interest in life (Ye & Pan, 2011). Using survey data on students in rural Anhui, Chongqing, and Guizhou, Liu et al. (Z. Liu, Li, & Ge, 2009) found that children separated from parents at a younger age, and children left behind by mothers or both parents, displayed more symptoms of anxiety and depression.

On the other hand, researchers have documented a significant negative association between levels of social and psychological stress and academic performance among adolescents (Fröjd et al., 2008; Owens, Stevenson, Hadwin, & Norgate, 2012). Depressed adolescents may concentrate on depressive thoughts and lose interest and initiative in learning, both of which reduce the cognitive and noncognitive resources for learning and lead to poor school performance.

Given the claimed association between parental migration and psychological well-being, and the link between psychological well-being and academic performance, this research tests the role of psychological well-being as a mediating channel between parental migration and educational outcomes.

The Social Remittance Perspective

The social remittance perspective focuses on the ideas, behaviors, norms, and values transmitted by migrant workers from their work destinations to their origin communities (Levitt, 1998). Migrant parents are likely to adopt new knowledge, values, and practices, from destination cities, that are often perceived as more advanced both socially and economically than their home communities. They are likely to transmit their newly acquired values, attitudes, and aspirations, to their children. This study develops and focuses on one aspect of the social remittance associated with parental labor migration: the values about education. Many migrant parents learn from their own experiences, or by observing the urban industrial world around them, that education is the key to white-collar jobs, and is perhaps the only way for their children to achieve a different life from theirs. Despite the physical distance, migrant parents try to maintain frequent communications with their children and monitor their health status and school performance (Hu, 2017). They tend to tell their own stories as counterexamples to emphasize the importance of educational qualification to their children. Moreover, the remaining parent and grandparents often use the hardships of migrant life and the self-sacrifices the migrant parent makes for the child’s future to motivate the child to value schooling more and study harder. Children who are socialized to perceive education as crucial to their future development are more willing to devote time and effort to study and are more likely to do well in school and move on to the next level of the school system.

The Parental Divorce Mechanism

Another unexplored potential mechanism linking parental migration to children’s educational outcomes is parental divorce. A theoretical foundation for hypotheses about the relationship between migration and marital instability is drawn from studies of social integration and divorce (Booth, Edwards, & Johnson, 1991; Breault & Kposowa, 1987; Glenn & Shelton, 1985; Glenn & Supancic, 1984; Trovato, 1986). The basic argument made in this literature is that the level of social integration is negatively related to divorce rates. Where social integration is high, married persons are embedded in a tight-knit network of friends, acquaintances, and extended family members. This helps enforce compliance with social norms emphasizing marital commitment and family cohesion. The framework has received empirical support from studies both at the aggregate level (Breault & Kposowa, 1987; Fenelon, 1971; Finnäs, 1997; Glenn & Shelton, 1985; Shelton, 1987; Trovato, 1986) and at the individual level (Frank & Wildsmith, 2005; Landale & Ogena, 1995).

The second framework linking mobility and marital dissolution can be found in the micro-level analysis of benefits and costs (Becker, Landes, & Michael, 1977; Levinger, 1976). This framework posits that individuals make marital decisions based on the relative benefits or costs of remaining in the same marriage vis-à-vis getting divorced. Individuals will choose to remain married only if the benefits are greater than the expected benefits of marital dissolution.

The third framework is concerned with shifting normative values about marriage and family life (Thornton, 2001; Thornton & Lin, 1994). Over the past few decades, the norms against divorce, premarital sex, cohabitation, and childbearing outside of marriage have weakened substantially and the divorce laws have been liberalized in northwestern Europe and other developed countries. Educational expansion, urbanization, and mass media have greatly facilitated the access to and adoption of these new ideas in other parts of the world.

A positive association between individual labor migration and marital dissolution can be expected by applying these three frameworks in the context of China’s internal labor migration. First, for a substantial amount of time during their married lives, husbands and wives involved in internal labor migration live apart from each other (Li & Zhang, 2016). The long-term absence of migrant spouse from the home combined with the anonymity of migrant life in industrial towns and cities may increase the chances of transgression for both parties. A national survey conducted in 2000 showed that the proportion of respondents who had experienced extramarital sex is about 23% to 24% for migrant workers, higher than that among rural (9.1%) and urban (15.7% to 19.1%) residents (Pan, Parish, Wang, & Laumann, 2004). Second, the rewards associated with marriage such as sharing living costs, receiving physical help, company, care, daily intimacy, and sexual life have drastically decreased in the case of spousal separation due to labor migration. This may gradually lead to estrangement between the couple and their declining commitment to the union. Third, work and life in the cities may expose migrant workers to different values and attitudes about spousal roles and marital relations. For example, migrant workers may start to have different expectations for their spouses as a result of their observation of the supposedly more egalitarian marital relationship in urban areas (Fan, 2008).

Scholars have suggested a positive correlation between labor migration and marital dissolution in China (Du, 2010; Shi, 2006). This association, however, is likely to be gendered. Traditional gender norms have remained strong in rural China and the age-old inside–outside dichotomy in gender roles defines the woman’s place to be inside the family and the man’s sphere to be outside. The gendered division of labor helps explain why fathers are more likely to be migrant workers than mothers, and rural children are more likely to stay behind with mothers than with fathers (All China Women’s Federation, 2013). This arrangement of father being a migrant worker and mother staying at home is consistent with the traditional gender roles and expectations, and thus is expected to be least disruptive to the marital relationship. On the contrary, mothers’ labor out-migration not only leads to their absence from home but also increases their economic independence and autonomy. The rising economic status of female migrants relative to their nonmigrant spouses is likely to rewrite the power balance between husband and wife, which may threaten the masculine identity of the husband as the primary breadwinner, and strain the marital relationship (Fan, 2008). This study hypothesizes that the out-migration of mothers but not fathers significantly increases the risk of parental divorce.

In family structure and child development literature, children in divorced or single-parent families, compared with those who stay with both parents, are generally found to be disadvantaged in terms of educational outcomes and other aspects of well-being (Amato & Cheadle, 2005; Fomby & Cherlin, 2007; MacLanahan & Sandefur, 1994). Their disadvantages are likely due to fewer economic resources, less parental attention and care, and psychological challenges associated with adjusting to parental conflict and new family arrangements (Amato, 2000; Amato & Keith, 1991; Astone & McLanahan, 1991; Carlson, 2006). Given the theoretical and empirical evidence suggesting that labor migration is associated with higher risks of divorce, this study attempts to explore the role of parental divorce as a potential channel linking parental labor migration and child outcome.

This study endeavors to contribute a more comprehensive understanding of the consequences of parental labor migration on adolescents’ educational outcomes, by combining and simultaneously examining the different pathways. In particular, this study highlights the more nuanced pathway through parental divorce in China and how this relationship affects the well-being of rural adolescents. As transition to high school has long-reaching implications for rural youths’ educational attainment and future prospects; focusing on the educational outcome of final-year middle school adolescents offers some insight into the long-term impacts of parental migration on children’s well-being.

Method

Data Collection

The data for this study were collected between September 2012 and October 2013 from a migrant-sending county located in Hubei Province of Central China. Hubei, the ninth most populous province according to the 2010 census, is one of the largest providers of interprovincial labor migrants in China. The percentage of rural children who are left behind by migrant parents in Hubei has exceeded 40% (All China Women’s Federation, 2013). The GDP per capita for Hubei in 2012 is about 38,572 RMB, equivalent to around 6,100 USD, close to the national average (Hubei Bureau of Statistics, 2013). Hubei, being a middling province in China, both socially and economically, has the potential to showcase illustrative insights into the influence of parental migration on left-behind children.

The target population is final-year adolescents in middle school. Gaining access to middle schools has been crucial to this study, as final-year students spend most of their time on campus studying and preparing for the High School Entrance Exam. This research was carried out in Tongcheng County—a valuable setting with a high percentage of left-behind adolescents and an average economic and demographic profile in Hubei. The county consists of two townships and nine towns, which vary in population size, share of agricultural population, income level, and out-migration rate. On average, towns tend to have lower share of agricultural population and higher population density than townships, and the county seat—the town where county government is based—is usually the most economically and socially developed town of the whole county. There are 18 middle schools distributed across this county: 6 are located in the county seat, 9 in the other eight towns, and 3 in the two townships. I selected one school located in a township, one school located in a town, and one school from the county seat to conduct this study. All final-year adolescents with a total number of 452 from these three schools, their caregivers, and teachers have been recruited for this study.

I used mixed methods with a concurrent triangulation design to collect information from adolescents, parents or other caregivers, teachers, and schools. The purpose is to better understand the research problem by obtaining different but complimentary data and to validate quantitative results with qualitative findings. Adolescent questionnaire and caregiver questionnaire have collected information on adolescents’ school life and educational aspirations, attitudes toward education, psychological well-being, parental migration status, parent–child relationship, and other family background information. For the qualitative part, I selected a subset of the survey sample (17 boys and 21 girls) for in-depth interview with a quota sampling method based on gender, school, and parental migration status. The interview questions centered on adolescents’ daily lives, childhood experiences, relations with parents and other caregivers, and relations with peers and teachers. Some caregivers and teachers of these adolescents have also been interviewed to get more information on their home life, campus life, and family situation and relations (see a list of main interview questions in Section A in the supplementary materials [All supplementary materials are available in online version of the article.]).

The information collected from the 452 adolescents in the whole sample reveals that 16 adolescents have lost one or both parents. These adolescents have been excluded from data analysis to reduce the complication of examining how children from migrant families fare, compared with those from nonmigrant families. However, 13 more adolescents have also been excluded from data analysis, as their final transitioning outcome was incomplete on account of missing information. The final sample is of 423 adolescents, about 94% of the total number of adolescents recruited. A comparison of the characteristics of those with and those without missing value on transitioning outcome has not shown any significant difference in parental migration status, parental marital status, academic performance, depressive symptom scores, number of books at home, value toward education, gender, age, number of siblings, parental education, and family economic status. The nonresponse rates of variables used in the final analytic sample are below 7%.

Variables

The dependent variable, the outcome of transition from middle school to high school, has been coded into three categories: leaving school, going to vocational high school, and going to academic high school. The key independent variable is parents’ migration status; it is measured by four categories: nonmigrant, father-migrant, mother-migrant, and parents-migrant.

I include four variables to examine the mediating mechanisms linking parental migration to children’s transition to high school. Parental divorce is a dichotomous variable that indicates if a divorce has occurred between parents by the time of the survey in October 2012. Social remittance (value toward education), educational investment (number of books at home), and children’s psychological well-being (depressive symptom scores) are the other mediating variables that have been tested in this study. A scale to measure adolescents’ value about education has been created. Adolescents have indicated on a 1 (completely disagree) to 4 (completely agree) scale on how much they agree with the following statements: (a) college education is necessary for me to do what I want to do in the future, (b) I need to get good scores in school in order to get a good job when I grow up, and (c) performing well in school is the best way to future success for me. The Cronbach’s α coefficient for the education value scale is .83.

To capture the educational investment of parents on adolescents, I asked adolescents to report the number of books they had at home by choosing from the following categories: none; less than 10; 10 to 21; 21 to 50; 51 to 100; 101 to 500, or more than 500. The answers have been recoded and a dummy variable has been created—“having more than a few books at home”: 0 if they have chosen none or less than 10, and 1 if they have marked the other categories.

The survey adopted the six-item Center for Epidemiologic Studies Depression scale of depressive symptoms from the 2010 Child Questionnaire of Chinese Family Panel Studies conducted by Beijing University (Luo & Wu, 2014; Xie, Xu, Zhang, & Xu, 2012), for psychological well-being. For a discussion of the application of Center for Epidemiologic Studies Depression scale to the context of China, please refer Zhang et al. (2010). The list includes (a) feeling depressed, unexcited about anything; (b) feeling anxious; (c) feeling uneasy, restless, finding it difficult to remain calm; (d) feeling the future is hopeless; (e) finding it difficult to do anything; (f) feeling life is meaningless. Adolescents have indicated their frequency of feeling in specific ways in the past month on a 1 to 5 scale in which: 1 = never, 2 = sometimes, 3 = half of the time, 4 = often, and 5 = almost every day. Next, an index of depressive symptoms has been constructed, by taking the mean score over the six items. The Cronbach’s α coefficient for the depressive symptoms scale is .85.

Chinese and Math scores in a countywide comprehensive examination, held by the county bureau of education in November 2012, have been used to measure academic performance. All final-year adolescents in middle schools, in the county, took part in this uniform and compulsory exam, at the same time and on the same dates. I obtained these scores from the schools. The theoretical range of both Chinese and Math test scores is from 0 to 120. In the multivariate analyses, the standardized total scores of Chinese and Math with a mean of 0 and a standard deviation of 1 are controlled for.

A series of demographic and socioeconomic status variables are controlled for in the model, which may intervene in the association between parental migration and educational attainment. For gender, female is coded as 1 and male is coded as 0. Age and the number of siblings are continuous variables. Father’s and mother’s educational attainment are grouped into three categories: primary school or below, middle school, and high school or above. Family computer ownership, the best available piece of information on family economic situation in the dataset, has been used to capture the wealth status of the family. It is a dummy variable that indicates whether an adolescent’s family owns a computer at home.

Data Analyses

The survey data used in this study has an inherent multilevel structure: students nested within classes and classes nested within schools. Because there are only three groups (schools) in the data, multilevel modeling, which generally requires a large number of groups for accuracy of estimates and high power of tests (Hox, 2010), is not an appropriate method to use in this research. An alternative analytical strategy for hierarchical data is to use a fixed effects model by including a set of dummy variables for groups. This study implements a school fixed effects model, with school dummies acting as a control for all unobserved factors related to location of school/neighborhood. This facilitates the examination of the effect of parental labor migration on outcome variables.

To start with, simple descriptive statistics has been shown and one-way analysis of variance and chi-square test has been conducted, to examine if the four groups of adolescents differ in any of the mediating and outcome variables. GSEM, or more specifically generalized path analysis, has then been used for a full-scale analysis of the direct and indirect effects of parental migration on transitioning outcome of their children. I use path analysis to examine mediation effects, because it allows researchers to analyze more complicated models with multiple independent and dependent variables (Cheung, 2007; Stage, Carter, & Nora, 2004).

In-depth interview data are also used to provide contextual understanding of children’s experiences of parental migration, and to illustrate the nuance or underlying mechanism linking parental migration to educational outcomes.

Results

Descriptive Statistics

Table 1 presents the descriptive statistics of the final analytical sample. As shown in Table 1, the majority of the adolescents have one or two parents who migrated to work. About 39.4% of adolescents are from nonmigrant families, 22.6% from father-migrant families, 14.0% from mother-migrant families, and 24.0% from both-parents-migrant families.

Table 1.

Educational Outcomes and Sociodemographic Characteristics of Adolescents by Parental Migration Status.

	N	Total (421)	Nonmigrant (166)	Father-migrant (95)	Mother-migrant (59)	Parents-migrant (101)	Bivariate test
Transitioning outcome	421						Pearson’s chi-square (df = 6) = 14.58*
% Leaving school		10.45	9.04	6.32	11.86	15.84
% Going to vocational high school		21.38	15.66	28.42	30.51	18.81
% Going to academic high school		68.17	75.30	65.26	57.63	65.35
Chinese scores (0-120)	421	73.90 (14.21)	76.24 (12.62)	71.49 (15.67)	72.15 (15.09)	73.97 (13.73)	One-way ANOVA F statistic = 2.77*
Math scores (0-120)	421	66.05 (25.81)	66.70 (26.43)	62.77 (24.33)	64.25 (25.83)	69.87 (25.72)	One-way ANOVA F statistic = 1.38
Parental divorce (%)	410	10.73	4.88	7.61	20.69	17.71	Pearson’s chi-square (df = 3) = 17.68**
Values toward education (1-4)	409	2.81 (0.73)	2.78 (0.70)	2.94 (0.71)	2.63 (0.74)	2.85 (0.77)	One-way ANOVA F statistic = 2.32^†
Having more than a few books at home (%)	418	67.22	73.94	67.02	61.02	60.00	Pearson’s chi-square (df = 3) = 6.78^†
Depressive symptoms score (1-5)	395	1.98 (0.76)	1.98 (0.73)	1.93 (0.70)	2.06 (0.82)	2.00 (0.83)	One-way ANOVA F statistic = 0.36
Age	421	14.91 (0.71)	14.87 (0.67)	14.94 (0.71)	14.99 (0.62)	14.89 (0.68)	One-way ANOVA F statistic = 0.57
Female (%)	421	50.36	51.81	47.37	47.46	52.48	Pearson’s chi-square (df = 3) = 0.86
Number of siblings	400	0.95 (0.62)	1.00 (0.58)	0.88 (0.51)	1.00 (0.72)	0.90 (0.72)	One-way ANOVA F statistic = 1.01
Father’s education	412						Pearson’s chi-square (df = 6) = 7.69
% Primary school and below		20.15	14.11	22.58	22.41	26.53
% Middle school		49.27	52.15	45.16	51.72	46.94
% High school and above		30.58	33.74	32.26	25.86	26.53
Mother’s education	407						Pearson’s chi-square (df = 6) = 3.64
% Primary school and below		31.70	30.67	27.66	29.63	38.54
% Middle school		52.83	52.76	58.51	53.70	46.88
% High school and above		15.48	16.56	13.83	16.67	14.58
Family computer ownership (%)	414	45.89	50.31	42.55	41.38	44.55	Pearson’s chi-square (df = 3) = 2.24
Location of school	421						Pearson’s chi-square (df = 6) = 30.08***
% Township		16.63	16.27	12.63	10.17	24.75
% County seat		58.91	69.88	61.05	49.15	44.55
% Town		24.47	13.86	26.32	40.68	30.69

Note. ANOVA = analysis of variance; df = degrees of freedom.

Standard deviations in parentheses

†

p < .1. *p < .05. **p < .01. ***p < .001.

The cross-tabulation shows an association between the transitioning outcome and the type of parental migration. While 75.3% of adolescents from nonmigrant families have transitioned to academic high school, only 65.3% of those from father-migrant families, 57.6% of those from mother-migrant families, and 65.4% of those from parents-migrant families have managed to do the same. Relatively more adolescents from mother-migrant families (11.9%) or parents-migrant families (15.8%) have discontinued schooling than did their counterparts from nonmigrant families (9.0%) and father-migrant families (6.3%). Bivariate test shows that adolescents’ transitioning outcome differs significantly by their parents’ migration status (χ = 14.58, degrees of freedom [df] = 6, p < .05). The differences in Chinese test scores by parental migration status are statistically significant, with non-left-behind adolescents at the top, followed by adolescents staying with neither parent, only mother, and only father. Parental migration status does not seem to make a difference in their children’s Math test scores.

Overall, about 10.7% of the adolescents are from divorced families. As expected, children from migrant families are more likely to have parents who had divorced by the time of the survey. Less than 5% of children from nonmigrant families have parents who had divorced, compared with about 7.6% of children from father-migrant families, nearly 21% of children from mother-migrant families and about 17.7% of children from parents-migrant families. Bivariate analysis confirms that there is a significant association between parental divorce and parental labor migration (χ = 17.68, df = 3, p < .01). As to the other potential mediating factors, parental migration status is associated with the scores on values toward education and book availability at home with marginal statistical significance, but has no effect on depressive symptoms score.

In Table 1, I further compare the demographic characteristics and socioeconomic background of adolescents by the migration status of their parents. The mean age of adolescents is 14.9. Half of them are girls. The average number of siblings adolescents have is about 1. Among the fathers, 20.2% had primary education or less, 49.3% had middle school education, and 30.6% had at least high school education. The fathers of children who are left behind by mother or both parents are less educated than the fathers of non-left-behind children and children left behind by fathers. The mothers, on average, have attained lower education, with 31.7% having only primary education or less and only 15.5% having high school education or above. The share of high school and above graduates is the highest among the mothers of children left behind by mothers (16.7%) and the lowest among the mothers of children left behind by fathers (13.8%). However, these differences in the educational attainment of fathers and mothers by parental labor migration status are not statistically significant. About 46% of adolescents reported that their families own a computer at home. To summarize, the four groups of adolescents do not differ significantly from each other in terms of demographic characteristics and family socioeconomic status.

With respect to the location of school, there are clear variations by parents’ migration status. The township school has the highest share of adolescents left behind by parents (24.8%), the county seat school non-left-behind adolescents (69.9%), and the town school adolescents left behind by mothers (40.7%). In all multiple analyses, school heterogeneities are controlled for by including school fixed effects.

The Effects of Parental Migration on Adolescents’ Transition to High School

To test the effects of parental migration on the outcome of adolescents’ transition to high school and the potential mediating effects of book availability at home, depressive symptoms, value toward education, and parental divorce, I conducted analyses using the GSEM of the Stata program (see Figures 2a and 2b). Adolescents’ demographic characteristics and family background including gender, age, number of siblings, father’s education, mother’s education, family wealth status, and school fixed effects are controlled for. For the full results, please refer to Appendix A. Note that GSEM does not produce standardized parameter estimates, nor does it provide goodness-of-fit statistics such as RMSEA or R². For robustness check, I run the same model using the “lavaan” package of R and reported the results in the supplementary materials, which confirm the main results reported below.

Figure 2a.

Path diagram of the effects of parental migration on transitioning outcome, full model.

Figure 2b.

Path diagram of the effects of parental migration on transitioning outcome, alternative model.

Not surprisingly, total test scores are strongly negatively associated with the likelihood of discontinuing schooling or going to vocational high school, relative to going to academic high school.

The Economic Resources Mechanism: Book Availability at Home

Home environment conducive to learning might have an accumulative effect in terms of cultivating adolescents’ interest in pursuing education over the long run. Moreover, investing more resources to buy books also signals parents’ commitment to children’s education. The results in Figure 2a show that having more than a few books at home is negatively associated with the likelihood of leaving school relative to going to academic high school. But book availability at home does not seem to matter with regard to test scores or which type of high school adolescents go to. However, there is no evidence suggesting that book availability at home helps explain the gross effects of parental migration on transitioning outcome. This is because parental migration has no significant direct effect on book availability. Migrant parents or caregivers may not have the time to go to a bookstore to buy books for their children because the bookstores are located far from their villages. Another explanation could be that many migrant parents may not be aware of the educational value of having books at home because most of them grew up in bookless homes and did not even go to high school.

Although potential greater economic resources associated with parental labor migration do not seem to be translated into greater book availability at home for left-behind children, it does not mean that there are no economic benefits associated with parental labor migration. About 80% of caregivers indicated that parental labor migration has improved the family’s economic situation. More specifically, both adolescents and caregivers have frequently mentioned the beneficial effects of remittance, manifested as improvements in living conditions, payment for education costs, better study environment and education opportunities, and more financial resources for college education.

The Psychological Well-Being Mechanism: Depressive Symptoms

Similarly, psychological well-being (measured by depressive symptoms) has a direct and independent effect on transitioning outcome when it comes to whether to pursue further education, but do not have an effect on test scores or the type of high school adolescents attend. However, parental migration has no effect on the depressive symptoms of adolescents. Therefore, the mediating role of depressive symptoms linking parental migration to educational outcomes is not supported.

The lack of effect of parental migration on psychological well-being is consistent with previous findings (Xu & Xie, 2015). Potential explanations include the coping strategies adopted by migrant parents and their left-behind children, and the coresidence of grandparents or other relatives as caregivers. My interview data reveals that migrant parents and left-behind children are actively exercising their agency to deal with the challenge of parental absence. Migrant parents make use of short messages, phone calls, and Internet chatting to care for and supervise their children at home, which helps alleviate the supposedly negative impact of parental absence on adolescents’ emotional well-being. Migrant parents also enlist help from other extended family members, enabled by the prevalence of multigenerational coresidence and social support networks among extended kin. In this sample, about four in five adolescents left behind by both parents, have been taken care of by grandparents, and the rest by uncles, aunties, or other relatives.

The Social Remittance Mechanism: Value Toward Education

Researchers have argued that one of the main channels through which labor migration transforms origin communities is social remittance, or transmission of ideas, knowledge, behaviors, and values from destination cities to rural areas through migrants. This research focuses on the transmission of education value. Migrant workers’ working experiences and urban living may strengthen their perception of the importance of education for better job prospects and economic security, and they may, in turn, pass this strong, positive evaluation of education to their children.

The results in Figure 2a show a marginally significant positive effect of father’s migration on educational opportunities through enhanced value toward education. Father’s migration decreases the relative likelihood of going to vocational high school instead of academic one via strengthening education value. Moreover, father’s migration is shown to decrease the likelihood of leaving school or going to a vocational high school, relative to going to an academic one, through influencing education value that in turn affects test scores. These results suggest that the transmission of education value generates some protective effects of keeping adolescents in school.

The positive effect of father’s migration on educational outcomes through education value is best exemplified by the story of Ye and his father. Having completed only 3 years of primary school education and having worked as a migrant worker for more than 15 years, Ye’s father wants more education for his son. He calls Ye once every 2 to 3 days and contacts Ye’s class teacher, Mr. Yang, two to three times every semester to talk about how Ye is doing in school. When I asked him how he encourages Ye to study, he shared the following story:

I usually tell him, we [I and Ye’s mother] don’t have much schooling, and so we can only do manual labor. It’s hard. I ask him to study hard and be like his aunty. [She] sits in an office. [There is] air-conditioning in the summer and heating in the winter. It’s easy, and the salary is high. Two years ago, he [Ye] said he does not want to go to school anymore. I took him to Guangzhou to see [my factory life]. I asked him if he could endure the hardship of doing migrant work. I asked him if he wants to do migrant work or to study. He said he wants to go home and study. I took him to my factory [to see around], his aunt took him to her factory [to see around], [and then] he said he wants to study. (Ye’s father)

Ye’s father was notable for his involvement in the child’s education, and Mr. Yang named him as one of the few parents who particularly value education. Not surprisingly, Ye was hardworking and making steady progress in school. Below is a quote from Mr. Yang commenting on Ye’s academic dedication and achievement:

Ye has a very good attitude towards study. He is very focused during class, and pays great attention to assignments. Although his academic performance is not excellent, he has surpassed X [name of another student], who greatly outperformed him in Grades 7 and 8.

Ye’s performance in the high school entrance exam earned him a place in an academic high school and he continued his schooling after graduating from middle school.

The Parental Divorce Mechanism

In addition to the social remittance channel, another potential channel, neglected in the literature, is the link between parents’ migration and divorce. Figure 2a shows that the migration of mother or both parents is positively and significantly associated with the risk of parents’ divorce, when father’s education, mother’s education, adolescent’s gender, and school fixed effects are controlled for.

This pattern is corroborated by the in-depth interviews. Among the subsample selected for in-depth interview, there have been four adolescents with divorced parents and the accounts of three of them connect their parents’ migration with parental divorce. For example, in the case of Tian, her mother started working in a coastal city in Eastern China when Tian was less than 1-year old. When Tian was about 2 or 3 years old, her mother met someone else and her parents got divorced and she has been living with her father ever since. When asked about her relationship with her father, the reason for her parents’ divorce emerged in her narrative:

But I don’t go to [maternal] grandma’s home often, because the reason my parents got divorced is that, my mom, um, she found an uncle [new partner] outside [the county]. So I know every time I visit my [maternal] grandma, my dad actually minds, though he said he does not mind.

It is worthy of note that parental migration is not significantly associated with parental marriage dissolution for those from father-migrant families. The arrangement of the father doing migrant work, while the mother staying at home, is in agreement with the traditional gendered norms regarding labor division, and thus, may pose the least challenge to family structure and functioning among the three types of parental migration. In contrast, adolescents from mother-migrant and parents-migrant families are much more likely to live in divorced or stepfamilies.

It is possible that parental divorce might lead to parental migration rather than the other way around. For example, after the marital relationship falls apart, one party or both parties may turn to labor out-migration as a solution to the difficult situation, or a way to start a new life. Taking this possibility of the reverse causal relationship into consideration, I checked the relative timing of parental migration and divorce. Among those cases with available information, the onset of parental migration has been prior to parental divorce for a great majority. Excluding the cases in which parental migration occurred after the parental divorce, the patterns remains unchanged. Again, including only those cases in which the onset of parental migration has been prior to parental divorce in the analysis, similar results have been obtained. These results have been reported in Appendix B.

As shown in Figure 2a, parental divorce has a significant direct positive effect on leaving school and going to vocational high school, relative to going to academic high school. Parental divorce has no significant effect on test scores. Therefore, it is unlikely that parental migration and parental divorce would influence adolescents’ transition to high school through affecting their test scores.

Combining these results, as Table 2 illustrates, there is a significant negative effect of mother’s migration and parents’ migration on adolescents’ transitioning outcome through increased risk of parental divorce. In other words, when parental migration is associated with parental divorce, children seem to be most disadvantaged in terms of educational opportunity. The significant and substantial effects of parental divorce on children’s transitioning outcome presented here are also validated by the qualitative data.

Table 2.

Indirect Effects, Direct Effects, and Total Effects of Parental Migration on Transitioning Outcome (Unstandardized Coefficient; Corresponding to the Model Presented in Figure 2a).

Parental migration (ref = nonmigrant)	Through mediating pathway	Transitioning outcome (ref = academic high school)	Estimate	Standard error	Two-tailed p value
Father-migrant	Parental divorce	Leaving school	1.09	1.24	0.380
Mother-migrant			2.92	1.77	0.098
Parents-migrant			3.08	1.79	0.086
Father-migrant		Vocational high school	0.66	0.74	0.376
Mother-migrant			1.77	1.02	0.083
Parents-migrant			1.86	0.99	0.060
Father-migrant	Parental divorce—test scores	Leaving school	−0.26	0.30	0.389
Mother-migrant			−0.69	0.46	0.135
Parents-migrant			−0.73	0.47	0.120
Father-migrant		Vocational high school	−0.18	0.21	0.388
Mother-migrant			−0.50	0.34	0.139
Parents-migrant			−0.52	0.34	0.124
Father-migrant	Education value	Leaving school	−0.03	0.08	0.742
Mother-migrant			0.01	0.03	0.761
Parents-migrant			−0.02	0.05	0.748
Father-migrant		Vocational high school	−0.14	0.09	0.133
Mother-migrant			0.04	0.07	0.531
Parents-migrant			−0.09	0.08	0.256
Father-migrant	Education value—test scores	Leaving school	−0.10	0.05	0.057
Mother-migrant			0.03	0.05	0.533
Parents-migrant			−0.06	0.05	0.220
Father-migrant		Vocational high school	−0.07	0.04	0.057
Mother-migrant			0.02	0.04	0.539
Parents-migrant			−0.04	0.04	0.215
Father-migrant	More books	Leaving school	0.30	0.41	0.469
Mother-migrant			0.43	0.52	0.406
Parents-migrant			0.70	0.49	0.157
Father-migrant		Vocational high school	0.05	0.11	0.614
Mother-migrant			0.08	0.15	0.597
Parents-migrant			0.12	0.20	0.541
Father-migrant	More books—test scores	Leaving school	0.01	0.05	0.808
Mother-migrant			0.02	0.07	0.799
Parents-migrant			0.03	0.11	0.789
Father-migrant		Vocational high school	0.01	0.04	0.808
Mother-migrant			0.01	0.05	0.799
Parents-migrant			0.02	0.08	0.788
Father-migrant	Depressive symptoms	Leaving school	−0.04	0.09	0.697
Mother-migrant			0.04	0.12	0.723
Parents-migrant			−0.01	0.10	0.941
Father-migrant		Vocational high school	−0.02	0.04	0.706
Mother-migrant			0.02	0.05	0.723
Parents-migrant			−0.00	0.04	0.941
Father-migrant	Depressive symptoms—test scores	Leaving school	−0.00	0.02	0.698
Mother-migrant			0.01	0.02	0.729
Parents-migrant			−0.00	0.02	0.941
Father-migrant		Vocational high school	−0.00	0.01	0.699
Mother-migrant			0.01	0.02	0.731
Parents-migrant			−0.00	0.01	0.941
Father-migrant	Direct effects	Leaving school	−0.93	0.73	0.203
Mother-migrant			−0.84	0.87	0.335
Parents-migrant			−0.19	0.67	0.778
Father-migrant		Vocational high school	0.63	0.48	0.190
Mother-migrant			0.17	0.52	0.746
Parents-migrant			−0.37	0.55	0.503
Father-migrant	Total effects	Leaving school	0.04	1.24	0.977
Mother-migrant			1.94	1.86	0.297
Parents-migrant			2.81	1.81	0.121
Father-migrant		Vocational high school	0.93	0.72	0.197
Mother-migrant			1.62	1.01	0.110
Parents-migrant			0.98	0.98	0.316

Note: Bolded values and texts indicate that the estimated coefficients of the pathway are marginally statistically significant (p < 0.1).

Parental divorce may lead to children’s estrangement from biological father or mother and feelings of resentment and anger, which may weaken their interest in schooling and commitment to pursuing further education. Taking Tian as an example, as noted earlier, Tian’s parents divorced when she was still very young and her father remarried when she was in primary school. She feels resentful toward her biological mother and does not get along well with her stepmother, who she claims favors her half-siblings over her. Tian is not close to her father either:

My dad and I, [choking her tears back], we had nothing to say. I go home every Sunday, although sometimes I was alone with my dad, but most of these moments were filled with silence. . . . When I was young, I felt envious of my classmates so much, I felt envious of my classmates so much, because [choking her tears back] when it rained, it got cold, and it’s their mom or dad who brought umbrella for them, but me, every time, it’s my grandma. When I was in Grade 6, in the class beside mine, there was a boy with the surname Chen; he seemed to know my situation. He told schoolmates that my parents were divorced. For this, I got into a fight with him. He was short, and in the same grade as me. I beat him and blood came out from his nose.

According to her class teacher, Tian was not doing very well in class and became weary of studying. Her class teacher made the following remarks about her family circumstances and disinterest in school:

Her [biological] mom is outside [of the county]. Her step-mother has a tendency to abuse her by saying things like “you will be just like your mom, a prostitute.” So she feels very resentful and rebellious. She sometimes can’t help crying while talking [to me]. . . . Her academic performance is about average or below. She has now grown weary of studying. She said she can’t wait to go out to do migrant work, to earn money herself.

Tian has dropped out of school at the beginning of her last semester in middle school to start her life as a migrant worker. Similar to Tian, two others boys among the adolescents I interviewed, Zhou and Ling, whose parents were migrant workers and then got divorced, have lost interest in studying and have dropped out of school in their last semester of middle school. Unlike Tian and Ling who were relatively young when their parents got divorced, Zhou was already in Grade 8 at the time of his parents’ divorce. Zhou used to feel close to his mother, but his relationship with her deteriorated since the divorce. When asked how he feels or thinks about his parents’ divorce, he said,

Sometimes I become very angry with my mom. She doesn’t call me. Sometimes when I call her, she does not pick up the phone. . . . That time, it was a Sunday, and we also had a Friday off, I went to my maternal grandma’s home, she was there. I did not talk to her. She does not care for me as she used to. I didn’t want to talk to her. She did not talk much with me either.

Shortly after the Chinese New Year when he was in Grade 9, Zhou dropped out of middle school and left home for a coastal city, despite his father’s objection.

Ling has always felt close to his mother, who he said would tell him stories and reason with him if he did something wrong and would encourage him if he did something right. Both of his parents remarried and had a new family. He wished to leave his father’s family and join his mother’s family when his father beat him. He still feels sad about his parents’ divorce, which he has been enduring. Ling lamented:

Sad, [I was] very sad. By that time, I knew [his parents’ divorce]. I saw that others had both dad and mom taking care of them, but my mom was not at home, I had only dad. So I asked him, and he told me [the divorce]. . . . In primary school, they [his parents] were strict with my study. Before they got divorced, I was doing very well in primary school. My mom was very strict with my study at that time. I was a bit afraid of her. And then they got divorced, and stopped supervising my study closely. Gradually I lost interest in studying, and I stopped putting effort into study.

His father’s remarriage has not improved parental supervision he could receive and his disengagement from study has continued.

I was angry, my dad found me a new mom. She sometimes finds fault with me and tells my dad everything [about me], and then my dad will beat me, and I will get angry.

Parental marital dissolution may also put economic stress on the family and the limited economic resources may force children to stay out of school. One factor related to increased financial pressure is the larger sibling size as a result of parental remarriage. Larger sibling size means fewer resources available for each individual child’s educational investment. For example, Tian’s father has remarried and has two more children with his current wife. He has told Tian that if she does not manage to make it to the best academic high school in the county after the High School Entrance Examination, there will be no further schooling for her. Financial concern is, thus, part of the reason that Tian has discontinued her schooling.

The in-depth interviews seem to suggest that insufficient economic resources may affect girls more than boys. Since sons and not daughters carry on the bloodline of the paternal family, it is possible that daughters may receive less educational investment than sons, from divorced parents. In the case of Tian, had she been a boy, she may have felt differently about her father’s commitment to financially supporting her education.

Because neither the economic resources mechanism nor the psychological well-being mechanism is supported by the results shown in Figure 2a, I removed these two mechanisms and presented a more parsimonious model in Figure 2b. A comparison of the Akaike information criterion and Bayesian information criterion of these two models favors the more parsimonious model. The results about the education value pathway and the parental divorce pathway from both models are essentially the same.

In addition to those indirect effects mediated through parental divorce and education value, Table 2 further reveals that parental migration does not exert a significant direct effect on adolescents’ transition to high school. Overall, the total effects of parental migration on transitioning outcome tend to be negative (estimated coefficients are positive because the reference category is the highest level) in the case of mother-migrant and parents-migrant families, although they are not statistically significant.

Discussion

I examined the relations between parental migration and rural Chinese adolescents’ educational outcomes in their transition from middle school to high school. Studies on internal labor migration in China have revealed that migrant workers and their remaining family members are trying to maximize their economic and social security through flexible household strategies, such as division of labor and collaboration between genders, generations, and households, and circular movements between their places of work and the home community (Fan, 2009; Fan & Wang, 2008). Under the structural and social constraints of China’s urban–rural bifurcation, many migrant parents, in their efforts to improve their children’s future prospects, are physically absent from their children’s daily lives for years.

Contrary to what migrant parents may have hoped, the results suggest no overall beneficial effects of parental migration on adolescents’ transition from middle school to high school. Specifically, though the effect of parental migration on transition to high school through social remittance pathway (value toward education) is positive, it is only marginally significant for adolescents in father-migrant families. Different from the theoretical predictions, parental migration is not associated with greater book availability at home. This might be because migrant parents may not have enough knowledge to translate their greater economic resources into more books at home by visiting bookstores and purchasing books. Also, parental migration does not negatively affect educational outcome via worsening adolescents’ psychological well-being, because left-behind adolescents do not differ significantly from their non-left-behind counterparts as measured by depressive symptoms in this study. This could be due to the role of extended family members in caring and supervising adolescents in the absence of parents, and the fact that middle school adolescents in rural China spend a lot of time on campus studying, socializing, living with peers, and being closely supervised by teachers (Hu, 2017).

Most importantly, the results show that mother’s migration and parents’ migration increase the likelihood of adolescents’ leaving school or going to vocational high school, relative to going to academic high school via increased chances of parental divorce. On one hand, adolescents from mother-migrant and parents-migrant families are more likely to experience parental divorce, compared with adolescents whose parents are not migrants. On the other hand, parental divorce greatly increases the risk of leaving school or going to vocational high school, relative to enrolling in academic high school.

Adolescents in father-migrant families do not differ from their counterparts in nonmigrant families in the likelihood of experiencing a parental divorce. This is probably because the arrangement of father being a migrant worker and mother staying at home is consistent with the traditional gendered division of labor and, thus, least disruptive among the three types of parental migration, with regard to marital relationship dynamic and family functioning.

Analysis of the qualitative data reveals that parental divorce has a long-term negative impact on adolescents’ educational outcomes. First, adolescents of divorced parents are more likely to be psychosocially disturbed and to lose interest in schoolwork and further education. Second, parental divorce may be associated with decreased economic resources available for adolescents’ educational investment. Parental remarriage often leads to an increase in siblings for the adolescent and, thus, more competition for family resources, including those that support education. The evidence also suggests that the effect of parental divorce on educational outcomes may differ by gender with daughters receiving less educational investment than sons do in divorced families.

However, the mediating effects of parental divorce between parental labor migration and adolescents’ transition to high school, are likely overestimated to some extent, due to omitted variable bias. I acknowledge the possibility that in certain cases, the preexisting marital problems or other unobserved characteristics may lead to both divorce and labor out-migration. Fortunately, the rich and nuanced interview data have lent credence to the results presented by the statistical model, suggesting a strong association between parents’ migration and divorce, as well as a substantial negative impact of parental divorce on children’s educational outcome. These findings reveal the need to pay greater attention to parental marital instability as a potential pathway mediating the impact of parental migration on children’s outcomes. In future research, it will also be important to investigate the underlying mechanisms between individual labor migration and marital instability in rural China.

This research has several limitations. The results are based on a relatively small, nonrepresentative sample limiting their generalizability. However, by delving deeply into the mediating mechanisms through which parental migration affects children’s well-being, the findings shed light on understanding the dynamics and consequences of parental migration in many other similar migrant-sending communities across rural China. Another problem is that the time-span of the data collection is limited. An ideal situation would be to have longitudinal data covering a longer period of these adolescents’ lives, to better deal with the selectivity issue in linking parental migration with parental divorce. Moreover, the potential economic benefits have not been better examined, due to missing information.

These limitations are offset partially by the richness and triangulation of the data. The data has been collected using both survey and interview methods with multiple sources, including adolescents, parents, and other key figures such as caregivers and teachers. Overall, the data have been valuable and rich for advancing the theoretical understanding concerning the channels of how parental labor migration affects adolescents’ educational achievement in a migrant-sending community of central China. The disadvantages of adolescents from divorced families in terms of educational outcomes have profound implications for the life chances of rural children and their families, and social inequality in China as a whole. Social researchers and policy makers should pay attention to the social costs of massive labor migration and discriminatory urban development borne by rural families.

Supplemental Material

supplementary_materials – Supplemental material for Parents’ Migration and Adolescents’ Transition to High School in Rural China: The Role of Parental Divorce

Supplemental material, supplementary_materials for Parents’ Migration and Adolescents’ Transition to High School in Rural China: The Role of Parental Divorce by Shu Hu in Journal of Family Issues

Footnotes

Appendix A

Estimated Path Coefficients of Generalized Structural Equation Modeling on Transitioning Outcome.

	Coefficient	Robust standard error
Transitioning outcome = leaving school
Parental migration status (ref = nonmigrant)
Father-migrant	−0.93	0.73
Mother-migrant	−0.84	0.87
Parents-migrant	−0.19	0.67
Standardized total test scores	−2.06***	0.35
Parental divorce	2.08**	0.91
Values toward education	−0.12	0.37
Having more than a few books at home	−1.29*	0.58
Depressive symptoms score	0.94**	0.31
Female	−1.33*	0.61
Age	−0.05	0.35
Number of siblings	1.05**	0.47
Father’s education (ref = primary or below)
Junior high	−1.16^†	0.63
Senior high and above	−1.69^†	0.88
Mother’s education (ref = primary or below)
Junior high	0.11	0.61
Senior high and above	0.69	1.00
Family computer ownership	−0.28	0.60
Location of school (ref = county seat town school)
Township school	0.76	0.68
Town school	2.00**	0.63
Constant	−3.22	5.66
Transitioning outcome = vocational high school
Parental migration status (ref = nonmigrant)
Father-migrant	0.63	0.48
Mother-migrant	0.17	0.51
Parents-migrant	−0.37	0.55
Standardized total test scores	−1.48***	0.30
Parental divorce	1.26*	0.50
Values toward education	−0.65*	0.31
Having more than a few books at home	−0.23	0.36
Depressive symptoms score	0.40^†	0.25
Female	0.40	0.38
Age	−0.09	0.25
Number of siblings	0.39	0.27
Father’s education (ref = primary or below)
Junior high	−0.44	0.50
Senior high and above	−0.48	0.57
Mother’s education (ref = primary or below)
Junior high	−0.54	0.44
Senior high and above	0.21	0.55
Family computer ownership	0.34	0.40
Location of school (ref = county seat town school)
Township school	1.69**	0.45
Town school	2.38***	0.48
Constant	−0.45	4.11
Standardized total test scores
Parental divorce	0.24^†	0.12
Values toward education	0.23***	0.07
Having more than a few books at home	0.03	0.10
Depressive symptoms score	−0.08	0.06
Female	0.17	0.10
Age	−0.13^†	0.07
Number of siblings	−0.13	0.08
Father’s education (ref = primary or below)
Junior high	0.21	0.14
Senior high and above	0.43**	0.16
Mother’s education (ref = primary or below)
Junior high	−0.08	0.11
Senior high and above	−0.43*	0.16
Family computer ownership	−0.03	0.11
Location of school (ref = county seat town school)
Township school	−0.07	0.15
Town school	−0.48***	0.12
Constant	1.60	1.18
Parental divorce
Parental migration status (ref = nonmigrant)
Father-migrant	0.52	0.55
Mother-migrant	1.40**	0.53
Parents-migrant	1.48**	0.46
Female	0.37	0.33
Father’s education (ref = primary or below)
Junior high	−0.35	0.46
Senior high and above	−1.31*	0.65
Mother’s education (ref = primary or below)
Junior high	0.35	0.49
Senior high and above	1.71**	0.69
Location of school (ref = county seat town school)
Township school	−0.10	0.46
Town school	−0.46	0.44
Constant	−3.07***	0.51
Values toward education
Parental migration status (ref = nonmigrant)
Father-migrant	0.22*	0.09
Mother-migrant	−0.07	0.10
Parents-migrant	0.13	0.10
Female	0.21**	0.08
Age	0.02	0.05
Number of siblings	0.06	0.06
Father’s education (ref = primary or below)
Junior high	0.20*	0.10
Senior high and above	0.29*	0.11
Mother’s education (ref = primary or below)
Junior high	−0.10	0.09
Senior high and above	−0.15	0.12
Family computer ownership	−0.02	0.07
Location of school (ref = county seat town school)
Township school	−0.21*	0.10
Town school	−0.17^†	0.09
Constant	2.20**	0.77
Having more than a few books at home
Parental migration status (ref = nonmigrant)
Father-migrant	−0.23	0.31
Mother-migrant	−0.34	0.38
Parents-migrant	−0.54^†	0.30
Female	−0.41	0.25
Age	−0.01	0.17
Number of siblings	−0.09	0.20
Father’s education (ref = primary or below)
Junior high	0.11	0.31
Senior high and above	0.80*	0.38
Mother’s education (ref = primary or below)
Junior high	−0.20	0.28
Senior high and above	−0.57	0.46
Family computer ownership	0.84***	0.25
Location of school (ref = county seat town school)
Township school	−0.69*	0.31
Town school	−0.50^†	0.29
Constant	1.30	2.63
Depressive symptoms score
Parental migration status (ref = nonmigrant)
Father-migrant	−0.04	0.10
Mother-migrant	0.05	0.13
Parents-migrant	−0.01	0.11
Female	0.22**	0.09
Age	0.06	0.06
Number of siblings	0.06	0.07
Father’s education (ref = primary or below)
Junior high	−0.09	0.12
Senior high and above	−0.10	0.14
Mother’s education (ref = primary or below)
Junior high	−0.01	0.10
Senior high and above	−0.02	0.15
Family computer ownership	0.04	0.09
Location of school (ref = county seat town school)
Township school	0.17	0.11
Town school	0.14	0.09
Constant	0.95	0.93
Variance (standardized total test scores)	0.78***	0.05
Variance (values toward education)	0.47***	0.03
Variance (depressive symptoms score)	0.55***	0.07
N	404

†

p < .1. *p < .05. **p < .01. ***p < .001.

Appendix B

Odds Ratios of Logistic Regressions on Parental Divorce, Robustness Check.

	Excluding the cases of migration after the divorce occurs	Including only the cases of migration before the divorce occurs
Parental migration status
Nonmigrant (ref.)
Father-migrant	1.41 (0.81)	4.29^† (3.74)
Mother-migrant	4.04** (2.11)	8.06* (7.15)
Parents-migrant	3.71** (1.75)	8.65** (6.71)
Female	1.71 (0.60)	3.84* (2.24)
Father’s education
Primary or below (ref.)
Junior high	0.64 (0.29)	0.71 (0.49)
Senior high and above	0.27^† (0.18)	0.35 (0.32)
Mother’s education
Primary or below (ref.)
Junior high	1.37 (0.67)	0.80 (0.53)
Senior high and above	4.56* (3.30)	3.55 (3.28)
Location of school
County seat school (ref.)
Township school	0.90 (0.45)	1.60 (1.00)
Town school	0.70 (0.31)	0.67 (0.43)
Log pseudolikelihood	−111.96	−63.29
Pseudo R²	0.088	0.159
N	395	377

Note. Robust standard errors in parentheses.

†

p < .1. *p < .05. **p < .01. ***p < .001.

Acknowledgements

I am grateful to the adolescents, caregivers, and teachers who participated in this study. I thank Jean Yeung and all the anonymous reviewers for helpful comments on earlier versions of this article.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The fieldwork for this study was funded by the Graduate Research Support Scheme of the Faculty of Arts and Social Sciences at National University of Singapore.

Supplemental Material

The online supplementary material for this article is available online.

ORCID iD

Shu HU

References

Acosta

(2006). Labor supply, school attendance, and remittances from international migration: The case of El Salvador (World Bank Policy Research Working Paper No. 3903). Retrieved from http://documents.worldbank.org/curated/en/103531468234850236/pdf/wps3903.pdf

Ainsworth

M. D. S.

(1989). Attachments beyond infancy. American Psychologist, 44, 709-716. doi:10.1037/0003-066X.44.4.709

All China Women’s Federation. (2013). A research report on rural left-behind children and urban migrant children in China. Beijing, China: Author.

Amato

P. R.

(2000). The consequences of divorce for adults and children. Journal of Marriage and the Family, 62, 1269-1287.

Amato

P. R.

Cheadle

(2005). The long reach of divorce: Divorce and child well-being across three generations. Journal of Marriage and Family, 67, 191-206.

Amato

P. R.

Keith

(1991). Parental divorce and the well-being of children: A meta-analysis. Psychological Bulletin, 110(1), 26-46.

Arguillas

M. J. B.

Williams

(2010). The impact of parents’ overseas employment on educational outcomes of Filipino children. International Migration Review, 44, 300-319.

Astone

N. M.

McLanahan

S. S.

(1991). Family structure, parental practices and high school completion. American Sociological Review, 56, 309-320.

Becker

G. S.

Landes

E. M.

Michael

R. T.

(1977). An economic analysis of marital instability. Journal of Political Economy, 85, 1141-1187.

10.

Booth

Edwards

J. N.

Johnson

D. R.

(1991). Social integration and divorce. Social Forces, 70, 207-224. doi:10.1093/sf/70.1.207

11.

Bowlby

(1988). A secure base: Parent-child attachment and healthy human development. New York, NY: Basic Books.

12.

Breault

K. D.

Kposowa

A. J.

(1987). Explaining divorce in the United States: A study of 3,111 counties, 1980. Journal of Marriage and the Family, 49, 549-558. doi:10.2307/352200

13.

Bryant

(2005). Children of international migrants in Indonesia, Thailand and the Philippines: A review of evidence and policies (UNICEF Innocenti Working Paper 2005-05). New York City, NY: UNICEF.

14.

Carlson

M. J.

(2006). Family structure, father involvement, and adolescent behavioral outcomes. Journal of Marriage and Family, 68, 137-154.

15.

Cheung

M. W.

(2007). Comparison of approaches to constructing confidence intervals for mediating effects using structural equation models. Structural Equation Modeling: A Multidisciplinary Journal, 14, 227-246.

16.

Coleman

J. S.

(1988). Social Capital in the Creation of Human Capital. The American Journal of Sociology, 94, S95-S120.

17.

Dreby

(2007). Children and power in Mexican transnational families. Journal of Marriage and Family, 69, 1050-1064.

18.

(2010). The impact of China’s rural-urban labor migration on marriage stability. Comparison of Economic and Social Institutions, 5, 105-112.

19.

Edwards

A. C.

Ureta

(2003). International migration, remittances, and schooling: Evidence from El Salvador. Journal of Development Economics, 72, 429-461.

20.

Evans

M. D.

Kelley

Sikora

Treiman

D. J.

(2010). Family scholarly culture and educational success: Books and schooling in 27 nations. Research in Social Stratification and Mobility, 28, 171-197.

21.

Fan

C. C.

(2003). Rural-urban migration and gender division of labor in transitional China. International Journal of Urban and Regional Research, 27, 24-47.

22.

Fan

C. C.

(2008). China on the move: Migration, the state, and the household. New York, NY: Routledge.

23.

Fan

C. C.

(2009). Flexible work, flexible household: Labor migration and rural families in China. In Keister

L. A.

(Ed.), Work and organizations in China after thirty years of transition (pp. 381-412). Bingley, England: Emerald.

24.

Fan

C. C.

Wang

W. W.

(2008). The household as security: Strategies of rural-urban migrants in China. In Nielsen

Smyth

(Eds.), Migration and social protection in China (pp. 205-243). London, England: World Scientific.

25.

Fenelon

(1971). State variations in United States divorce rates. Journal of Marriage and the Family, 33, 321-327. doi:10.2307/349418

26.

Finnäs

(1997). Social integration, heterogeneity, and divorce: The case of the Swedish-speaking population in Finland. Acta Sociologica, 40, 263-277. doi:10.1177/000169939704000303

27.

Fomby

Cherlin

A. J.

(2007). Family instability and child well-being. American Sociological Review, 72, 181-204.

28.

Frank

Wildsmith

(2005). The Grass Widows of Mexico: Migration and union dissolution in a binational context. Social Forces, 83, 919-947. doi:10.1353/sof.2005.0031

29.

Fröjd

S. A.

Nissinen

E. S.

Pelkonen

M. U.

Marttunen

M. J.

Koivisto

A.-M.

Kaltiala-Heino

(2008). Depression and school performance in middle adolescent boys and girls. Journal of Adolescence, 31, 485-498.

30.

Glenn

N. D.

Shelton

B. A.

(1985). Regional differences in divorce in the United States. Journal of Marriage and the Family, 47, 641-652. doi:10.2307/352265

31.

Glenn

N. D.

Supancic

(1984). The social and demographic correlates of divorce and separation in the United States: An update and reconsideration. Journal of Marriage and the Family, 46, 563-575. doi:10.2307/352598

32.

Hannum

Cherng

H. Y. S.

(2011). Examinations and educational opportunity in China: Mobility and bottlenecks for the rural poor. Oxford Review of Education, 37, 267-305. doi:10.1080/03054985.2011.559387

33.

Hox

J. J.

(2010). Multilevel analysis: Techniques and applications. New York, NY: Routeledge.

34.

Hubei Bureau of Statistics. (2013). Hubei Statistical Yearbook 2013. Beijing, China: China Statistics Press.

35.

(2017). “It’s for Our Education”: perception of Parental Migration and Resilience among Left-Behind Children in Rural China. Social Indicators Research, 1-21. doi: 10.1007/s11205-017-1725-y.

36.

Jordan

L. P.

Graham

(2012). Resilience and well-being among children of migrant parents in South-East Asia. Child Development, 83, 1672-1688. doi:10.1111/j.1467-8624.2012.01810.x

37.

Kandel

Kao

(2001). The impact of temporary labor migration on Mexican children’s educational aspirations and performance. International Migration Review, 35, 1205-1231.

38.

Kim

H. S.

(2011). Consequences of parental divorce for child development. American Sociological Review, 76, 487-511. doi:10.1177/0003122411407748

39.

Lahaie

Hayes

J. A.

Piper

T. M.

Heymann

(2009). Work and family divided across borders: The impact of parental migration on Mexican children in transnational families. Community, Work & Family, 12, 299-312.

40.

Landale

N. S.

Ogena

N. B.

(1995). Migration and union dissolution among Puerto Rican women. International Migration Review, 29, 671-692. doi:10.2307/2547499

41.

Levinger

(1976). A social psychological perspective on marital dissolution. Journal of Social Issues, 32, 21-47. doi:10.1111/j.1540-4560.1976.tb02478.x

42.

Levitt

(1998). Social remittances: Migration driven local-level forms of cultural diffusion. International Migration Review, 32, 926-948.

43.

Zhang

(2016). Migrating together or staying behind? An empirical research on household arrangement during rural-urban migration. Sociological Studies, 5, 139-163.

44.

Liu

Zhang

Luo

Rozelle

Sharbono

Shi

(2009). Development challenges, tuition barriers, and high school education in China. Asia Pacific Journal of Education, 29, 503-520.

45.

Liu

(2009). Left too early: The effects of age at separation from parents on Chinese rural children’s symptoms of anxiety and depression. American Journal of Public Health, 99, 2049-2054.

46.

(2012). Education of children left behind in rural China. Journal of Marriage and Family, 74, 328-341.

47.

(2014). Parental migration and education of left-behind children: A comparison of two settings. Journal of Marriage and Family, 76, 1082-1098. doi:10.1111/jomf.12139

48.

Treiman

D. J.

(2011). Migration, remittances and educational stratification among blacks in apartheid and post-apartheid South Africa. Social Forces, 89, 1119-1143. doi:10.1093/sf/89.4.1119

49.

Luo

(2014). The measurement of psychological health in China Family Panel Studies 2012 (China Family Panel Studies Technical Report No. 26). Bejing, China: China Family Panel Studies.

50.

MacLanahan

Sandefur

(1994). Growing up with a single parent: What hurts, what helps. Cambridge, MA: Harvard University Press.

51.

Morooka

Liang

(2009). International migration and the education of left-behind children in Fujian, China. Asian and Pacific Migration Journal, 18, 345-370.

52.

Owens

Stevenson

Hadwin

J. A.

Norgate

(2012). Anxiety and depression in academic performance: An exploration of the mediating factors of worry and working memory. School Psychology International, 33, 433-449. doi:10.1177/0143034311427433

53.

Pan

S.-M.

Parish

Wang

A.-L.

Laumann

(2004). Sexual behavior and relation in contemporary China. Beijing, China: Social Sciences Documentation Publishing House.

54.

Ren

Treiman

D. J.

(2013). The consequence of parental labor migration in China for children’s emotional well-being (China Family Panel Studies Working Paper Series No. WP13-001). Bejing, China: China Family Panel Studies.

55.

Shelton

B. A.

(1987). Variations in divorce rates by community size: A test of the social integration explanation. Journal of Marriage and the Family, 49, 827-832.

56.

Shi

(2006). The impact of youth’s out-migration on marriage in rural China. Population Journal, 155(1), 32-36.

57.

Stage

F. K.

Carter

H. C.

Nora

(2004). Path analysis: An introduction and analysis of a decade of research. Journal of Educational Research, 98(1), 5-13.

58.

Thornton

(2001). The developmental paradigm, reading history sideways, and family change. Demography, 38, 449-465.

59.

Thornton

Lin

H.-S.

(1994). Theoretical mechanisms of family change. In Thornton

Lin

H.-S.

(Eds.), Social change and the family in Taiwan (pp. 88-115). Chicago, IL: University of Chicago Press.

60.

Trovato

(1986). The relationship between migration and the provincial divorce rate in Canada, 1971 and 1978: A reassessment. Journal of Marriage and the Family, 48, 207-216. doi:10.2307/352245

61.

Xiang

(2007). How far are the left behind left behind? A preliminary study in rural China. Population, Space and Place, 13, 179-191.

62.

Xie

Zhang

(2012). The composite variables of China Family Panel Studies 2010 (Part Two): Educational attainment and depression scale (China Family Panel Studies Tecahnical Report No. 12). Bejing, China: China Family Panel Studies.

63.

Xie

(2015). The causal effects of rural-to-urban migration on children’s well-being in China. European Sociological Review, 31, 502-519. doi:10.1093/esr/jcv009

64.

Yang

(2008). International migration, remittances and household investment: Evidence from Philippine migrants’ exchange rate shocks. Economic Journal, 118, 591-630.

65.

Yao

Wang

(2010). The impact of higher education expansion on social justice in China: A spatial and inter temporal analysis. Journal of Contemporary China, 19, 837-854.

66.

Pan

(2011). Differentiated childhoods: Impacts of rural labor migration on left-behind children in China. Journal of Peasant Studies, 38, 355-377.

67.

Yeung

W.-J. J.

(2013). Higher education expansion and social stratification in China. Chinese Sociological Review, 45(4), 54-80.

68.

Yeung

W.-J. J.

(2016). Left behind by parents in China: Internal migration and adolescents’ well-being. Marriage & Family Review, 52, 127-161. doi:10.1080/01494929.2015.1111284

69.

Zhang

Z. Y.

Fang

Han

B. X.

Chen

Z. Y.

(2010). Development of the Chinese age norms of CES-D in urban area. Chinese Mental Health Journal, 24, 139-143.

70.

Zhou

Sylvia

Zhang

Luo

Liu

. . . Rozelle

(2015). China’s left-behind children: Impact of parental migration on health, nutrition, and educational outcomes. Health Affairs, 34, 1964-1971. doi:10.1377/hlthaff.2015.0150

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.37 MB