Mental Health Impacts of Child Labour: Evidence from Vietnam and India

Abstract

A number of recent studies have investigated the relationship between child labour and physical health. However, there has been little empirical evidence that child labour affects children’s emotional and behavioural development. This study departs from existing literature by examining the mental health impacts of child labour in Vietnam and India, as measured by the Strengths and Difficulties Questionnaire. The potential endogeneity of child labour is addressed by using rainfall as the instrument. The findings show that children engaged in child labour suffer from mental health issues as measured by peer problems and reduced prosocial behaviour in both countries. There is a significant gender difference in the impact of child labour in India. Finally, doing household chores, an accepted social and cultural work in developing countries is found to be associated with the better mental health of children in Vietnam.

Acknowledgements

I would like to thank two anonymous referees for their constructive comments and suggestions that significantly improved the paper, and to this article’s responsible editors, for their kind and generous patience during the submission of the paper. I would also like to thank Simon Feeny and participants at the 41st Australian Health Economics Society Conference for their helpful comments on earlier drafts of this paper. The data used in this publication come from Young Lives, a 15-year study of the changing nature of childhood poverty (www.younglives.org.uk). Young Lives are funded by UK aid from the Department for International Development (DFID). The views expressed here are those of the author. They are not necessarily those of Young Lives, the University of Oxford, DFID or other funders. All errors are my own. Interested scholars can email the author for data and code.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary Materials

Supplementary Materials are available for this article which can be accessed via the online version of this journal available at https://doi.org/10.1080/00220388.2020.1746280

Notes

1. In this study, we apply the ILO’s definition of child labour (International Labour Organization, 2015), which refers to the exploitation of children through any form of work that deprives children of their childhood, interferes with their ability to attend regular school, and is mentally, physically, socially or morally harmful. According to ILO minimum age convention (C138) of 1973, child labour refers to any work performed by children under the age of 12, non-light work done by children aged 12–14, and hazardous work done by children aged 15–17. In this study, we define child labour as those who perform any work given that they aged between seven and nine in the first round of Young Lives.

2. For more details about the sampling approach in India and Vietnam, see Kumra (2008) and Nguyen (2008).

3. Informed consent is obtained from everyone involved in the Young Lives Project. Further, anonymity and confidentiality of participants in the survey are protected. More information about the ethical process can be found at: https://www.younglives.org.uk/content/research-ethics.

4. For a recent review of the impact of climate change on agriculture, see Dell, Jones, and Olken (2014).

5. The results are consistent with our main findings (see Table A11 in the Supplementary Online Materials).

6. Note that a higher score of prosocial scale means better prosocial behaviour.

7. Unfortunately, we are not able to directly test the impact of rainfall deviation on household income as this information is not available in the first round of Young Lives (Briones, 2018).

8. Using two-stage least squares approach provides a consistent result. We examine the potential problem of weak instruments using the critical values in Stock and Yogo (2002). We test for weak instruments using the first stage F-statistic and the Kleibergen–Paap Wald F-statistic. Under the null hypothesis, we assume that the instrument is weak. The critical value of 16.38 (Stock & Yogo, 2002) implies a rejection of the null hypothesis in both countries.

9. It should be noted that the relationship between parental education and child mental health is not causal. In fact, differential impacts of parental schooling on child outcomes have been found in the literature (e.g. Behrman & Rosenzweig, 2002; Black, Devereux, & Salvanes, 2005).

10. We acknowledge that child BMI might be endogenous in this context. For example, poor mental health may cause weight gain which is a part of BMI’s calculation, or other parental socio-economic status may affect both child BMI and mental health. The literature has shown that overcoming the endogeneity issue is challenging given that weight is not randomly assigned and correlated with a number of unobservable factors. Willage (2018) proposes an ideal instrument for BMI using genetic information derived from a biological sample and laboratory analysis. This information, however, is not available in the context of our study. Still, we attempt to find an instrument for BMI by using an area-based measure. Specifically, we follow Morris (2006) and employ the mean BMI across children living in the same commune (district in India) as the instrument. We argue that area BMI is a good predictor of individual level BMI while controlling for other covariates, and at the same time is less likely to directly affect mental health indicators. The results are presented in Tables A12a and A12b (Supplementary Online Materials) for Vietnam and India, respectively. Firstly, we find consistent impacts of child labour on mental health in both countries. Secondly, we observe a negative correlation between child BMI and mental health, and the relationship is statistically significant. In addition, there is a strong association between child BMI and our instrument (mean area BMI), as illustrated in the first stage of our estimation (see Panel B).

11. The findings should be interpreted with caution given that the amount of support as well as type of support is not available in the Young Lives Project.