2000年-世界发展银行全球_Revisiting_the_Link_between_Poverty_and_Child_Labor___The_Ghanaian_Experience_26页_1mb

Summary of "Revisiting the Link Between Poverty and Child Labor: The Ghanaian Experience"

Core Content

This working paper by Niels-Hugo Blunch and Dorte Verner examines the relationship between poverty and child labor in Ghana, focusing on harmful child labor, which is defined as labor that directly conflicts with the accumulation of human capital. The study challenges recent claims that poverty is not a main determinant of child labor, and instead reinstates the positive link between poverty and harmful child labor.

The authors analyze a new dataset from the 1997 Core Welfare Indicators Questionnaire (CWIQ) in Ghana, which covers 14,514 households and 60,686 individuals. The paper explores the determinants of child labor and schooling, with a particular emphasis on the interplay between poverty, gender, and location.

Main Findings

1. Poverty and Child Labor

• Positive relationship: The study reaffirms that children from poor households are more likely to be engaged in harmful child labor.
• Robustness: This relationship is statistically significant across all sub-samples, including urban and rural areas, and across gender groups, except for urban girls.
• Age effect: The likelihood of harmful child labor increases with age, particularly for girls.

2. Gender Gap in Child Labor

• Girls more vulnerable: Girls are consistently more likely to engage in harmful child labor than boys, both at the aggregate level and across sub-samples.
• Cultural norms: The gender gap may reflect cultural norms and expectations, though further research is needed to explore this.
• Education and work: The study suggests that girls' work may subsidize the education of their brothers, highlighting a complex interplay between gender and labor.

3. Location and Child Labor

• Rural vs. urban: Children in rural areas are more likely to engage in harmful child labor than those in urban areas, possibly due to the reliance on child labor in agricultural activities.
• Structural differences: There are structural differences in the processes underlying child labor between rural and urban locations.

4. Determinants of Child Labor

• Household characteristics:
  • Children of the household head are less likely to engage in harmful child labor.
  • Ownership of land, sheep, and cattle is associated with an increased likelihood of child labor.
• Socioeconomic status: Children of self-employed workers (both in agriculture and non-agriculture) are more likely to engage in harmful child labor.
• Supply factors:
  • Distance to the nearest primary and secondary school positively affects the likelihood of child labor.
  • This suggests that access to education is a key constraint in the decision to engage in child labor.

5. Schooling and Work Decision

• Schooling is almost universal: The majority of children in Ghana attend school, with school attendance rates above 95% for most age groups.
• Work decision focus: Given the near universality of schooling, the authors focus on the work/no-work decision in their econometric analysis.
• Model used: A univariate probit model is employed to estimate the likelihood of child labor, with the dependent variable being a binary indicator of whether the child works.

Key Information

• The paper is part of the World Bank's broader effort to understand the processes underlying child labor.
• It challenges the notion that poverty is not a main determinant of child labor, based on recent literature.
• The study emphasizes the importance of structural differences across gender, location, and poverty levels in shaping child labor outcomes.
• It highlights the need for further research into the role of cultural norms and the impact of disability on child labor decisions.

Policy Implications

• Policymakers should focus on identifying the most vulnerable groups, particularly girls and children from poor households.
• Interventions should address both the supply and demand factors influencing child labor, such as school accessibility and household income.
• The findings suggest that improving access to education and reducing the economic pressures on poor households may help reduce harmful child labor.

Methodology

• The paper uses a probit model to analyze the determinants of child labor.
• It includes individual, household, and community variables such as age, poverty quintile, location, and socioeconomic status.
• The model is estimated for full samples and sub-samples across gender, location, and poverty levels to capture structural differences.

Conclusion

The study concludes that poverty and gender play significant roles in determining the likelihood of harmful child labor in Ghana. It underscores the importance of understanding the underlying mechanisms and structural differences that influence child labor decisions, and calls for targeted policy interventions to address these issues.