Abstract
We evaluate a technology upgrading intervention for smallholder dairy farmers in Uganda, showing that a simple version of the “contact farmer” extension model can meaningfully increase farmers’ revenues. In a program that provides no monetary incentives to local farmer trainers, backstopping by professional extension agent and advertising pro- social motivation reinforce impacts. We find no statistical evidence that program effects vary by farmers’ characteristics. Being exposed to a broad menu of 33 different feeding practices, farmer trainers re-adjust feeding practices to fit their labor and capital endowments; this selection according to comparative advantage may explain the program’s widespread and homogeneous benefits.