Abstract
Though experimentation by small farmers in low-income countries is fundamental to technology adoption, such experimentation often takes place mostly in private. The semi-private nature of this experimentation limits the positive externalities of the experimenters’ effort, slowing learning by others. We implement a village-level randomized controlled trial to evaluate the effects of three scalable strategies to promote farmer experimentation and learning about new hybrid seed varieties. Across 156 Kenyan villages and 4160 farmers, we vary the proportion of farmers to whom we offer an improved hybrid maize seed; we also divide the low concentration treatment study arm into an incentive and non-incentive group designed to compensate the participating farmers for their effort and the public goods that their experimenting provides. In the first treatment arm, 10% of villagers receive 0.5 kgs of maize seed; in the second, 10% of the farmers receive the maize plus financial compensation for their experimentation. In the third, 35% of the village farmers receive only the maize seed. All participant farmers receive signage for posting next to these experimental fields to promote awareness of their work. Our design allows us to test the effects of different levels of saturation within a given village against a model where farmers are compensated for their experimental efforts. We study the relative effects of on farmer experimentation, learning, yield expectations, and adoption.