Children's cognitive ability, schooling and work: Evidence from Ethiopia

• The paper uses two rounds of cohort data from the Ethiopia Young Lives project.
• It analyzes the association between children's cognitive ability (as measured by test score) and their school/work status.
• The most common reasons for school non-enrollment are ‘financial problems’ and ‘to work for household or for pay’.
• Higher ability children are more likely to be in school and less likely to work in the market, and work fewer hours.
• The results continue to hold when ability is measured as a residual (net of schooling) and when using data on sibling pairs.

I investigate the relationship between children's cognitive ability and parental investment using a rich dataset on a cohort of children from Ethiopia. The data come from Young Lives, a long-term international study of childhood poverty in four countries. Ability is measured by scores on a cognitive test. A child's enrollment in school, participation in work and work hours are employed as measures of parental investment in human capital. The results provide strong evidence of reinforcing parental investment – higher ability children are more likely to be enrolled in school and less likely to work and, conditional on participation, also work fewer hours. These results are mostly robust to addressing potential feedback effects between schooling and test scores and household heterogeneities. On the policy front, the results suggest that the seeds of adulthood inequality in human capital and earnings capability may be sown quite early in childhood, and thereby underscore the importance of interventions that, among others, attempt to improve prenatal and early life health and nutrition, which are often cited as the sources of deficiencies in children's cognitive ability.