The Effect of Improvements in Agricultural Productivity on Child Labor: Theory and Empirical Evidence from the Green Revolution

• Applicants: Dominik Naeher, Ph.D.; Professor Dr. Matthias Schündeln
• Subject Area: Economic Policy, Applied Economics
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 561142299

Child labor remains a pressing issue, with nearly one in ten children worldwide working according to recent estimates. This has both direct effects on children’s welfare and indirect effects on their education. While public attention often focuses on child labor in export industries, most child labor occurs in the agricultural sector. The economic literature mainly views child labor as driven by subsistence poverty, rooted in the `luxury axiom' – the idea that parents will only send their children to work if the household falls below a minimum consumption threshold. Based on this view, policymakers often emphasize the need to increase agricultural productivity and incomes to reduce child labor. However, the empirical evidence on the relationship between productivity and child labor is mixed: in some cases, increased productivity has led to more child labor. The uncertainty surrounding this apparent puzzle contributes to the difficulties policymakers face in this context. This project seeks to clarify the link between agricultural productivity, child labor, and education through a synthesis of theoretical and empirical work. A preliminary model suggests that child labor decisions may be determined by two opposing effects. The first effect stems from the luxury axiom and implies that higher productivity contributes towards reducing child labor. On the other hand, higher productivity may also increase the demand for labor on farms. If labor markets are imperfect and family labor is not easily replaced by hired labor, then households may find it optimal to increase child labor when productivity rises. These countervailing effects predict an inverse-U shaped relation between agricultural productivity and child labor. Initially, as farm productivity raises, poor households increase child labor, but wealthier households reduce it. This could reconcile the seemingly contradictory empirical findings in the literature, reflecting differences in household characteristics across studied populations. Our project will develop a model that helps to clarify the link between agricultural productivity and child labor at a general level, keeping assumptions about the utility and production functions flexible. Then we will test the implications of the model empirically, using a unique dataset constructed from different sources and an identification strategy based on a difference-in-differences and shift-share (Bartik) instrumental variable approach. We will exploit variation in the timing and diffusion of high-yielding seed varieties for different crops during the Green Revolution, providing quasi-experimental evidence on the effects of agricultural productivity changes on child labor during this period and beyond. The empirical work will focus on India but also provide cross-country evidence. Finally, we will apply the obtained insights to discuss implications of changes in productivity caused by climate change, trade liberalization, and technological change.

• DFG Programme: Research Grants