Information and behavior: Evidence from fertilizer quantity recommendations in Bangladesh

Abstract

We use a field experiment in Bangladesh to test if two types of variety-specific fertilizer recommendations – government provided, community-level recommendations and plot-specific recommendations based on individual soil tests – affect fertilizer use, yield, and profits. The treatments have a limited effect on quantity of most fertilizer types used by farmers. One exception is the case of TSP (phosphate fertilizer), which is recommended for all varieties but recommended quantities are significantly lower than the baseline average usage. A minority of treatment farmers over-react to the recommendations by stopping TSP use after the intervention. In the soil-testing treatment arm farmers also shift their seed choice to varieties for which their baseline fertilizer consumption aligns with the recommendation. Opting out of using TSP, an essential fertilizer, ultimately hurts productivity as farmers in the community-based recommendation arm experience a 4% yield reduction.