The contrasting consequences of hazardous and nonhazardous child labor on cognitive and psychosocial outcomes in Afghanistan and Bangladesh

• In Event: Education in Conflict Situations: Inequality, Interruption, and Survival

Proposal

1. INTRODUCTION

The diverse nature of child labor demands a more nuanced analysis than has been traditionally offered. Not all work performed by children carries the same weight of consequences. While hazardous child labor—characterized by long hours, dangerous conditions, and risk to life and limb—is universally recognized as detrimental, less harmful forms, such as household chores or contributing to family businesses, may have ambiguous or even beneficial effects on a child’s development (Maconachie et al., 2022). Historically, researchers have often been forced to combine these categories due to data limitations, leading to an overstatement of the negative effects of non-hazardous work and an understatement of the true harm caused by hazardous labor (Posso, 2018).

This quantitative study examines the distinct effects of hazardous and non-hazardous child labor on a child’s cognitive and psychosocial outcomes in Afghanistan and Bangladesh. The topic is directly relevant to the field of comparative and international education, which has long explored the relationship between child labor, schooling, and learning outcomes (Post, 2002). This research also complements existing qualitative and quantitative efforts to disentangle the effects of different types of child labor (Lieten, 2011; Posso, 2018).

Grounded in human capital theory, our conceptual framework posits that hazardous child labor is a significant obstacle to a child's human capital formation, as measured by literacy, numeracy, and psychosocial well-being. The effects of non-hazardous child labor, however, are considered ambiguous, potentially having both positive and negative influences on a child's development.

The comparative analysis of Afghanistan and Bangladesh is particularly insightful. Although both are South Asian, Muslim-majority countries, they represent dramatically different contexts of human development, educational advancement, and institutional strength. These disparities are crucial because the prevalence and nature of child labor are often shaped by a country's institutional and social environment. We hypothesize that Bangladesh's stronger institutions may mitigate the most severe impacts of child labor, even in its hazardous forms, compared to Afghanistan, where institutional instability may exacerbate its negative consequences.

2. DATA AND METHODS

This study leverages detailed data from the UNICEF Multiple Indicator Cluster Surveys (MICS), a globally recognized household survey program that provides internationally comparable data on the well-being of children and women. The MICS Round 6 data, collected in 2019, are employed for this analysis. Our focus is on the Questionnaire for Children Age 7-14, which contains data on child labor, foundational learning skills, and psychosocial well-being. These data are merged with household-level information on parental education and household composition.

The outcome variables are:

- Literacy skills: A child's ability to read and comprehend a given text, measured by the number of words read correctly.

- Numeracy skills: A child's ability to perform basic mathematical operations and compare numbers.

- Psychosocial well-being: A composite index constructed from a child's self-reported difficulties related to communication, learning, making friends, and emotional states such as anxiety or sadness.

The key explanatory variable is the weekly hours of child labor. We identify hazardous child labor using MICS questions that detail the work environment and tasks, such as carrying heavy loads, working with dangerous tools, or exposure to hazardous materials like dust, fumes, or chemicals.

The analysis will use multivariate regression models to estimate the relationship between child labor and the three outcome variables. We will analyze the effects of total weekly work hours and then introduce an interaction term between weekly hours and a dummy variable for hazardous child labor to isolate the distinct effects of dangerous work. Ongoing research will explore more advanced econometric methods, such as dose-response models, to better understand how the number of hazardous work activities influences outcomes (Cerulli, 2015).

3. FINDINGS AND CONTRIBUTIONS

Preliminary descriptive statistics reveal high rates of child labor in both countries: 85.9% in Afghanistan and 85.6% in Bangladesh. A notable difference emerges in the prevalence of hazardous work, which affects 73.9% of children in Afghanistan versus 44.0% in Bangladesh. This disparity aligns with our initial hypothesis about institutional and human development differences.

The regression results demonstrate a critical point: the consequences of child labor are highly contextual and outcome-specific.

Afghanistan: No statistically significant association was found between child labor and literacy or numeracy scores. However, the interaction between weekly hours and hazardous child labor was a large and significant predictor of psychosocial outcomes. Specifically, each hour of hazardous child labor was associated with a 0.8-point decrease in psychosocial scores. This suggests that in contexts with weak institutions, the greatest harm from dangerous work is on a child's mental and emotional well-being.

Bangladesh: Child labor hours were negatively and statistically significantly associated with numeracy and psychosocial outcomes, but not literacy. Each hour of work was associated with a 0.5-point decrease in numeracy and a 0.1-point decrease in psychosocial scores. The interaction term for hazardous child labor, however, was statistically insignificant across all outcomes, indicating that, on average, the additional harm of hazardous work is not as pronounced as it is in Afghanistan. This may be because Bangladesh’s stronger institutional and legal frameworks mitigate the most extreme forms of hazardous work.

In summary, this study provides novel quantitative evidence of the contrasting effects of child labor. It underscores that the impact of child labor is not uniform but depends on the specific context, the type of child labor (hazardous vs. non-hazardous), and the particular outcome being measured (cognitive vs. psychosocial). Our findings contribute to a more nuanced understanding of child labor by moving beyond a single, aggregated category and providing comparative evidence from two distinct developmental contexts.

4. REFERENCES

Cerulli, Giovanni. (2015). Ctreatreg: Command for fitting dose-response models under exogenous and endogenous treatment. The Stata Journal 15(1), 3-20.

Lieten, G.K. (2011). Hazardous Child Labour in Latin America. Springer.

Maconachie, Roy, Neil Howard, and Rosilin Bock. (2021). Re-Thinking ‘Harm’ in Relation to Children’s Work: A ‘Situated,’ Multi-Disciplinary Perspective. Oxford Development Studies 50 (3), 259–71.

Posso, Alberto (2018). The health consequences on hazardous and non-hazardous child labor. Review of Development Economics 23, 619-639.

Post, David. (2002). Children’s Work, Schooling, and Welfare in Latin America. Westview.

Author

• M. Najeeb Shafiq, University of Pittsburgh