Child Labor, Openness, Human Capital and Technology : A Panel Data Approach

I explore the link between openness to trade, investments in human capital and technology, credit market constraints, and child labor in a panel of 101 countries from 1980 to 2004. In a cross-country setting, Neumayer and De Soysa (2005) and Edmonds and Pavcnik (2006) find that countries that trade more and have a higher stock of FDI have less child labor. Here, I provide stronger support for their findings and extend their analysis by addressing the simultaneity issue that their cross-country approach might have. I employ a Fixed Effects Instrumental Variables (FE-IV) panel data model that is robust to the correlation between unobserved heterogeneity amp;#956;(sub i) and regressors X(sub it). I also account for the endogeneity of child labor and individual-country-specific effects by using IVs when necessary. I find that trade openness, investments in human capital and technology, and financial development are associated with a reduction of child labor. For instance, a one-standard-deviation increase in trade volume is associated with a reduction of child labor of 0.62 to 0.91 percentage points. A one-standard-deviation increase in public spending on education is associated with a reduction of child labor of 0.31 to 0.56 percentage points. Finally, a one-standard-deviation improvement in the credit constraints measure is associated with a reduction of child labor of 0.63 to 0.94 percentage points