Revisiting the role of education for agricultural productivity

While the majority of micro studies finds that rural education increases agricultural productivity, various recent cross-country regressions analysing the determinants of agricultural productivity were only able to detect insignificant or even surprising negative effects of schooling. In this paper, we argue and show that this failure to find a positive impact of education in the international context is rather a data problem related to the use of enrolment and literacy indicators. Using a panel of 95 developing and middle-income countries from 1961 to 2002 together with the newest version of the Barro-Lee educational attainment dataset, we show that education indeed has a highly significant, positive effect on agricultural productivity which is robust to changes in the control variables and in the econometric methods applied. Distinguishing between different levels of education further reveals that only primary and secondary schooling have significant positive impacts while tertiary education remains insignificant. Finally, the effect of education is estimated separately for countries with different income levels. Results indicate that the coefficient of the education variable remains insignificant for countries from the poorest three income quintiles, while it is positive and highly significant for the richest two quintiles. This finding can be interpreted as support for the prominent argument claiming that education leads to higher agricultural productivity only in the presence of rapid technical change where education will help farmers to adjust more readily to the new opportunities.