Child labor in Indonesia: Households response to income shocks

The discussion on child labor has shifted from a purely theoretical to a more applied approach as more sources of information have become available in developing countries, where the phenomenon is more prevalent. This paper focuses on how households in Indonesia resort to child labor in the aftermath of an income shock.

I use data from the latest round of the Indonesian Long Family Survey (IFLS). The ongoing survey tracks households and individuals across time which allows separation of the time of the shock and the response. Another advantage of the dataset is the inclusion of a specific module on child labor in the two most recent rounds (2000 and 2007). Child labor is measured as the number of hours worked by children under the age of 15: this includes market work, work on family business, work on non-family business and household work. I estimated the impact on child labor as a response to a variety of shocks among which are crop losses and the occurrence of a natural disaster (transitory shocks) and the death of an adult in the household (permanent shock). Endogeneity is a concern in this type of studies and part of the methods sections is devoted to explain to what extent these shocks can be considered exogenous.

I find little evidence that households use child labor as a buffer against a shock: a significant positive effect of a natural disaster. However there is evidence that holding some specific assets could protect against child labor, most notably land not used for farming and jewelry, which could be tied to the household borrowing possibilities.