Miracle Seeds: Biased Expectations, Complementary Input Use, and the Dynamics of Smallholder Technology Adoption

To fully benefit from new agricultural technologies such as improved seed varieties, significant investment in complementary inputs such as fertilizers and pesticides and practices such as systematic planting, irrigation, and weeding are also required. Farmers may fail to recognize the importance of these complements, leading to disappointing crop yields and outputs and, eventually, dis-adoption of the improved variety. Using a field experiment, we test an information intervention among smallholder maize farmers in eastern Uganda that points out these complementarities. We find that farmers adopt less after they have been sensitized about the need to use complementary inputs to unlock the adoption premium. We rationalize this finding with a simple theoretical model where farmers have misspecified mental models of the technology production function and conclude that most farmers in our sample do indeed believe in miracle seeds.