Essays on Child Labour and Schooling in Ghana

This thesis consists of three papers on child labour and schooling in Ghana. The first paper examines the correlates of child labour and schooling, as well as the trade-off between work and schooling of children aged 5-17 years with the 2013 Ghana Living Standard Survey data. A bivariate probit model is used since the decisions to participate in schooling and in the labour market are interdependent. The results show that there is a gender gap both in child work and schooling. In particular, boys are less likely to work (and more likely to be enrolled in schools) relative to girls. Whereas parent education, household wealth and income of the family are negatively correlated with child work, these factors influence schooling positively. In addition, parents' employment status, ownership of livestock, distance to school, child wage and schooling expenditure increase the probability of child labour and reduce the likelihood of school enrolment. In terms of the relationship between child labour and schooling, the results show that an additional hour of child labour is associated with 0.15 hour (9 minutes) reduction in daily hours of school attendance; and the effect is bigger for girls relative to boys. Also, one more hour of child labour is associated with an increase in the probability of a child falling behind in grade progression by 1.4 percentage points. The second paper estimates the impact of Ghana's Livelihood Empowerment Against Poverty (LEAP) cash transfer programme on schooling outcomes (enrolment, attendance hours, repetition and test scores) and child labour in farming and non-farm enterprises. Using longitudinal data, the paper employs three different quasi-experimental methods (propensity score matching, difference-in-difference, and difference-in-difference combined with matching). Overall, the results show that the LEAP programme had no effect on school enrolment and test scores, but it increased the weekly hours of class attendance by 5.2 hours and reduced repetition rate by 11 percentage points for children in households that benefited from the programme. In addition, there was heterogeneity in these impacts, with boys benefiting more relative to girls. In terms of child labour, the results show that the programme had no effect on the extensive margin of child labour in farming and non-farm enterprises. However, the LEAP programme reduced the intensity of farm work done by children by as much as 2.6 hours per day. The largest impact of the programme, in terms of reduction in the intensity of child labour in farming, occurred in female-headed and extremely poor households. The last paper investigates the impact of mothers' autonomy or bargaining power in the household on their children's schooling and child labour in Ghana. The paper uses a noneconomic measure of women's autonomy, which is an index constructed from five questions on power relations between men and women. The paper employs both an Ordinary Least Square (OLS) and an Instrumental Variable (IV) approach. Overall, the results suggest that ignoring the endogeneity of mothers' autonomy underestimates its true impact on schooling and child labour. They also show that an increase in mothers' autonomy increases school enrolment and hours of class attendance, with girls benefiting more than boys. The paper finds a negative relationship between mothers' autonomy and both the extensive and intensive margin of child labour. In addition, it demonstrates that improvement in women's autonomy has bigger impacts on rural children's welfare relative to urban children.