Fertility and Child Occupation: Theory and Evidence from Senegal

This paper analyzes household fertility and child occupation decisions in a risky environment. Fertility decisions are made first, when only the distribution of shocks is known. When shocks are realized and fertility is fixed, parents adapt by allocating children's occupations, i.e. school, paid work and domestic chores. Fertility is decreasing with the shock probability and increasing with parental permanent income. Households facing an adverse shock make more use of child labor and send fewer children to school, unless the total number of children is small. These predictions are tested with data from the Senegalese SEHW (2003) following this two-step methodology. A Poisson model estimates the number of children with classical instruments and household-level information on shock distribution, confirming the theory's predictions on fertility. A multivariate Tobit model estimates the determinants of children occupations, including the occurrence of shocks and accounting for the endogeneity of fertility. The number of children increases (decreases) the probability of child specialization (multiple activities). Shock-related variables have an adverse effect on schooling.