LAUSR

Increasing agricultural productivity while protecting natural resources depends on proper understanding of farmers’ incentives to use intensification strategies, including fertilizer. Using a large-scale household dataset collected in rural Burkina Faso, we examined how the response of maize yield to fertilizer, and thus the economic incentives for its use, varied according to agro-ecological conditions. We employed a Control Function Approach with Correlated Random Effects in order to test and control for endogeneity of fertilizer use, measuring agro-ecological conditions at several scales. We investigated the profitability of fertilizer use with value-cost ratios. We found that productivity and marginal effects of fertilizer differ significantly according to agro-ecological conditions. Micro-variation appeared to be more critical than the definition of agro-ecological zone. Burkinabe soils are severely degraded and would benefit from greater application of fertilizer. However, at full market prices, fertilizer use was unprofitable. Though it was profitable with subsidized prices, transaction costs diminish the benefits of the subsidy. Profitability of fertilizer use with maize varied across agro-ecological conditions, even for field plots located in the same agro-ecological zone. Our results confirm that policy makers need to be cautious when generalizing across regions or drawing policy recommendations from a single agro-ecological zone because crop responses and economic incentives vary widely.