THE MOST RECENT ONE
Mobile-phone based agricultural advice for smallholder farmers: evidence and implications
Customized agricultural advice, delivered by mobile phones, holds the promise of delivering actionable, relevant information to small-holder farmers at a very low cost. We present evidence from a series of experiments testing the efficacy of mobile-phone based agricultural advice, conducted in Africa and South Asia. We find some positive evidence: in a trial in India advice changed practices, increasing yields in cumin (26%) and cotton (8.6%, for a sub-group receiving reminders). In Kenya, SMS encouragement to adopt agricultural lime increased lime purchases by 10-15% among maize farmers. A series of experiments with One Acre Fund in Kenya and Rwanda show that lime encouragement SMS campaigns increased lime purchases by similar magnitudes. We also find settings in which such services are not effective, or optimized: For instance, SMS advice by a Kenyan government research group referencing a specific soil chemistry condition had no effect on farmer behavior; government-designed soil health cards in India were incomprehensible to most farmers without visual and audio aids; and a simple tweak to the design of a mobile advisory service in another African country significantly increased service usage. Together, these underscore the value of iterating on the design, and testing the efficacy of mobile agricultural extension services through randomized experiments. We also discuss ongoing work to test behavioral and social learning theories in the context of agricultural decision-making.