Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
I propose a new explanation for the large productivity gaps between agriculture and nonagriculture observed in developing countries, namely that the elasticity of substitution between skilled and unskilled workers is higher in agriculture than in nonagriculture. I estimate the elasticities in a cross-country data set, and use the results to simulate a simple two-sector model in which heterogeneous elasticities of substitution affect relative sectoral productivity levels through the allocation of human capital. Calibrated to match data from the United States, the model predicts sizable agricultural productivity gaps in countries where the share of highly educated workers in the labor force is low.