Women farmers and barriers to technology adoption; a randomized evaluation of BRAC's extension program in rural Uganda

This research project aims to shed light on the constraints that shape the adoption process of agriculture inputs and techniques by poor women in rural Uganda. We collaborate with the NGO BRAC to randomize the roll-out of their agriculture extension program and its different components across 166 villages between 2012 and 2014. Our evaluation is based on a two-level randomization design. First we randomize the allocation of treatment and the program components at the village level. In particular, 166 villages will be randomly allocated to 3 groups. The first group will receive the agriculture extension program in 2012; the second group will receive both the agriculture program and a microfinance program in 2012 and the third will receive neither until 2014. As microfinance is not available in these villages at baseline, comparing the effect of the agricultural extension program with and without the complementary credit program will allow us to evaluate the importance of credit constraints in the adoption process. Second, within each treated village we will randomly select community members who are chosen to serve as model farmers. This will allow us to identify which individuals are better suited for the task of promoting new technologies and the relevance of social networks for technology adoption. To do so, we will combine detailed information on social networks at baseline with the fact that some networks will be randomly treated to have one of their members selected to be a model farmer.