The project will evaluate the impact of a program that teaches farmers how to conduct field trials on their own plots. We evaluate the value of own-farm experimentation, measure the returns to new inputs, and evaluate the program's impact on the efficiency of input adoption.
Intervention
The EASE project aims to train small-holder maize farmers in Eastern Uganda to engage in small-scale experimentation on their own farm to identify which agricultural technologies are most profitable for them.
Experimental design
Villages and farmers are randomly selected and assigned to treatment or control. Treated farmers benefit from training to experiment with agricultural inputs, while control participants receive inputs but no training in how to conduct experiments do not receive this component. Data are collected through surveys and field observations.
Design details
Randomization method
Randomization conducted in office by computer, reproducible (Stata).
Randomization unit
Villages are first randomly assigned to treatment villages or control villages. Within each selected treatment village, eligible maize-growing farmers are listed, and a subset of 8 is randomly assigned to participate in the experiment (treated farmers).
Clusters
120 villages
Observations
8 farmers per village (960 units of observations).
Arms
75 villages treated (600 treated farmers) and 45 villages control (360 control farmers).
Primary outcomes
1. We study the value of own-farm experimentation (in terms of the profitability of the generated input recommendations), versus a range of alternative sources of input recommendations.
2. We study the impact of the EASE program on
a. Farmer’s adoption of improved seeds and chemical fertilizers.
b. Farmer’s adoption of experimental methods in the subsequent season.
c. Projected profit gains due to choice of input adoption.
Secondary outcomes
- We provide estimates of average returns to different inputs, complementarities between these inputs, and heterogeneity, including dependence on soil characteristics.
- We measure how farmers' beliefs about input returns respond to own-farm experimentation.
Cited by
No citation data available for this registration yet.