Farmer Education and Technology Adoption: The Choice of Education Measures

This paper investigates how we should measure farmer education in empirical specifications. Many existing studies find positive effects of education on the adoption of agricultural innovations by farmers. However, the empirical specification of education differs between studies, and most of the studies do not justify the measure of education that is used. In this context, the two relevant issues relate to the level of education and which member of the farm household's education matters. In this paper, I use data on rural Bangladesh to estimate and compare the effects of 14 education measures on the probability of adopting newly disseminated crops. The empirical results suggest that the average and minimum years of schooling and the presence of one literate member in the household have positive effects on technology adoption. By contrast, years of schooling of the most educated person and the household head have no significant effects. These results cast doubt on the arbitrary choices of education measures in existing studies