Impacts of boarding on primary school students’ mental health outcomes – Instrumental-Variable evidence from rural northwestern China

Highlights

• Using students’ home-to-school distance as an instrumental variable to estimate impacts of boarding on primary school students’ mental health.

• Boarding has a significantly negative effect (0.455 SDs) on their mental health status.

• The negative effects of boarding on mental health are larger for students with relatively advantageous backgrounds.

Abstract

This paper estimates the impacts of boarding on primary school students’ health outcomes, using data on 7606 students from rural areas of two northwestern provinces (Qinghai and Ningxia) of China. Exogenous variations in students’ home-to-school distance are exploited to address potential endogeneity in their boarding status. Instrumental variable estimates suggest that while boarding has little impact on students’ physical health (measured by height-for-age and BMI-for-age z-scores and hemoglobin concentration levels), it has a significantly detrimental effect on their mental health status, amounting to 0.455 standard deviations (SDs) of the distribution of scores on a Mental Health Test (a modified version of the Children’s Manifest Anxiety Scale). The effect of boarding is more pronounced for students with relatively advantageous backgrounds. For example, boarding boys scored 0.544 SDs higher on the Mental Health Test (suggesting more anxiety problems) than nonboarding boys, and boarders from relatively wealthier families scored 0.754 SDs higher than wealthier nonboarders. ‘

Introduction

China’s population control policies, along with its rapid economic development, have greatly reduced Chinese parents’ demand for children. As a consequence, the total fertility rate in rural China has declined by more than 60 % since the late 1970s.¹ The number of children under age 17 also dropped rapidly from 410 million in 1982 to 345 million in 2000 (National Bureau of Statistics of China (NBSC, 1982, 2000). The shrinking population of school-age children since the early 1990s has led to substantial reductions in school enrollment in rural China (Chen et al., 2015). Small-scale enrollments and the resulting high teacher-to-student ratios have, in turn, caused substantial waste of the already-scarce educational resources in rural China (Ministry of Education, 2010).

In reaction to this problem, the Chinese government launched a national school merger program in 2001 (State Council, 2001; Liu et al., 2010). The key part of this program involved plans to close small primary schools in remote villages, especially non-complete primary schools and teaching points that had only 3 grades, and merge them with larger, complete primary schools (with all 6 grades) (Ministry of Education, 2010). Approximately 300,000 rural schools were then merged to bigger and better-equipped schools during the first decade of the new Millennium (Wang, 2017). While also accommodating nonboarding students, boarding schools have become the major platform of education provision and management in rural China since 2004: by 2015, the numbers of boarding students in rural primary schools had reached 9,550,000, accounting for 14.4 % of the total enrollments of rural primary schools (Ministry of Education, 2016).² The 19th Congress of China, held in October 2017, further stressed the key role boarding schools ought to play in the development of China’s rural education system in the foreseeable future.³ Yet to better serve their mission, boarding schools need to first overcome a series of problems plaguing China’s rural education system, the most pressing one being poor child health (Yue et al., 2018).

While it is relatively easy to monitor students’ physical health status through more frequent and regular checkups, it is much more difficult to detect their mental health problems. Worse yet, if not discovered in time, mental health problems can lead to more serious consequences. A series of devastating suicide incidents stemming from students’ mental health issues occurring the early 2000s have since induced the Chinese government to devote considerable efforts to reducing mental health problems at all levels of schooling (Hesketh et al., 2002; Zhang et al., 2013). A question naturally arising in this context is: How does boarding, the new platform for education provision and management in rural China that may last for decades, affect students’ mental well-being?

The answer to this question is theoretically ambiguous. On the positive side, the collective living environment supervised by designated student-care staff at school may help ensure boarding students’ health and safety (Wang et al., 2017). School rules and teachers’ instruction may also help boarders develop more living skills and form healthier living habits (Liu, 2005). For students with disadvantaged backgrounds, living away from their disadvantaged home environment may improve their subjective well-being (Child, 2016; Martin et al., 2014). On the negative side, however, boarding implies a lack of close parental care, without which boarding students’ personalized emotional need may not be easily met. Homesickness, schooling-related stress, and lack of constant and continuous support from parents may also incur psychological costs (e.g. Fisher et al., 1986; Cookson, 2009). Since the relative strengths of these boarding-related effects are unclear in theory, how boarding affects students’ mental health is ultimately an empirical question.

Unfortunately, existing empirical evidence from China also fails to provide a conclusive answer. While some studies found significantly negative associations between boarding and students’ mental health outcomes (e.g. Du et al., 2010; Wang et al., 2016), others found virtually no association between them (Zhou et al., 2018). As recently pointed out by Wang et al. (2016), the lack of a consensus in existing findings may be due to two methodological drawbacks: relatively small sample size and lack of exogenous variations to identify boarding effects.

To overcome these drawbacks, the present study employs a large dataset on 7606 students collected from 74 rural primary schools in two northwestern provinces of China (Ningxia and Qinghai) to estimate the causal effect of boarding on students’ mental health status, including subdimensions such as study anxiety, social anxiety, and loneliness, etc. (− effects of boarding on students’ physical health outcomes, including height, weight and hemoglobin concentration levels, are also examined for comparison and interpretation purposes). To address potential endogeneity in students’ boarding status, we exploit exogenous variations in their home-to-school distance to create an instrumental variable (IV) for it. Admittedly, this IV is not free of concerns, a major one being that home-to-school distance may capture the level of prosperity of one’s home village relative to his/her school location, thereby confounding the effect of boarding. For instance, students who grew up in well-developed villages may be physically and mentally healthier than those who grew up in remote and less-developed villages, regardless of their boarding status. We address this concern by (1) controlling for school fixed effects, and (2) restricting our analysis to the sample of students with a relatively short home-to-school distance (≤ 3 km). The rationale is that, conditional on the school being attended, identification of boarding effects comes from within-school variations in students’ home-to-school distance, which are unlikely to capture the influence of home-village prosperity when the focus is placed on students with a short home-to-school distance. Our IV estimates suggest that while boarding has a negligible effect on students’ physical health, it has a significantly negative effect on their mental health status.

Our study makes two contributions to the literature. First, to the best of our knowledge, our study is the first to identify causal effects of boarding on students’ mental health outcomes in the context of (rural) China. To measure students’ mental health status, we administered a mental health test, adapted from the widely-used Children’s Manifest Anxiety Scale originally developed by Reynolds (1980), to all sampled students, assessing their mental well-being in eight subdimensions including study anxiety, social anxiety, loneliness, self-punishment, physical anxiety symptoms, sensitivity, fear and impulsiveness. This test helps to depict a comprehensive profile of students’ mental health status in rural northwestern China and assess how boarding affects it. Second, our analysis provides new evidence on the education-health nexus from an often-overlooked angle. The majority of previous studies on this nexus have mainly focused on estimating the impact of the quantity of education (e.g. years of schooling or the highest degree earned) on one’s health behavior and outcomes (e.g. Arendt, 2005; De Walque, 2007;Groot and Maassen van den Brink, 2007; Albouy and Lequien, 2009; Braakmann, 2011; Zhong, 2016; Parinduri, 2017).⁴ Our study, in contrast, examines how boarding, as a particular form for education provision and management, affects one’s (mental) health. This line of inquiry also contributes to the emerging literature on the role boarding schools play in forming students’ human capital (Curto and Fryer, 2014; Behaghel et al., 2015; Gregg, 2018).

The remainder of this paper is structured as follows. The next section describes our data. Section 3 develops a framework underlying our empirical analysis. Section 4 reports and discusses our main findings. The final section draws conclusions and suggests several directions for future research.