The rise in child labour and the negative effect of it on child schooling outcomes is an important policy issue in developing countries. However, despite almost universal agreement that child labour is undesirable, there is wide disagreement on how to tackle the problem. The formulation of policies that are effective in curbing child labour requires a clear understanding of the key determinants of child employment. This article contributes to the debate by providing an analysis of the key determinants of child labour and schooling in an upper middle-income country, Botswana. The study used the Labour Force Survey (LFS) 2005/06 data from the Central Statistics Office (CSO) and the multinomial logit model for analytical work. The results suggest that the probability of children working while schooling is negatively and significantly influenced by the age of the child, being from a female headed household and employment status of the household head. However the probability of child labour and schooling is positively and significantly influenced by child education level, the number of children in the household, and the household head being engaged mainly in the agricultural sector.
Key Words: Child Labour, Multinomial Logit Model, Botswana,
JEL: J2, C25, C3