Child labor, agricultural shocks and labor sharing in rural Ethiopia

The author studies the effect of an agricultural shock and a labor sharing arrangement (informal social network) on child labor. Albeit bad parental preference to child labor (as the strand of literature claims), poor households face compelling situations to send their child to work. This is, especially, true when they are hit by an income shock and face a binding adult labor constraint. The author used panel data from the ERHS and employed a fixed effects model to pin down causal relation between shocks, membership in a labor sharing arrangement and child labor. It was found that child labor is, indeed, a buffer stock. Though a labor sharing arrangement doesn't affect child labor at normal times, it helps households to lessen the pressure to rely on it when hit by idiosyncratic shocks. While almost the whole effect of these shocks is offset by participation in a labor sharing arrangement, the covariate shock is not. Even if this may well affect a child's academic performance, school attendance doesn't decrease. This differential effect of shocks on child labor in participant households might be because of the extra adult labor made available or due to mutual support that comes with these social networks. This paper is indicative of the importance of considering social networks in smoothing out consumption. Further, it highlights the difficulty to cope up with covariate shocks and hence, calls for development interventions that are particularly meant to address their impact.