Reducing child labor with micro-insurance

Micro-insurance is insurance for the poorest against what sometimes seem trivial risks to us. Yet, for the poorest of this world small shocks can be devastating. It is also believed that such shocks are often the reason why parents have to send their children to work even though they know it is bad for their future. It is thus natural to study whether the introduction of micro-insurance reduces the incidence of child labor.

Andreas Landmann and Markus Frölich benefit from a randomized experiment in rural Pakistan, where a micro-credit enterprise is also offering micro-insurance to members that can be voluntarily extended to additional family members. We are talking about microscopic insurance here, like health insurance for a dollar a year. It turns out that in villages where the extension is offered the incidence of child labor is lower by up to a fourth, when combined with help filling claims. The authors interpret their results not to come from the mitigation of adverse shocks, but rather from the feeling of security.

As with all randomized experiments, the question on whether these results can generalize is open. One would need many more (costly) experiments to find assurance that micro-insurance is effective in reducing child labor. In addition, this kind of study cannot measure the aggregate impact of interventions. Imagine micro-insurance were available to everyone and child labor incidence is reduced. This is bound to increase adult wages, which may reinforce the decrease in child labor as even fewer parents need to send their kids to work. But for this to happen, we need more than a few villages.