Incorporating risk in the economic analysis of agronomic trials: fertilizer use on barley in Syria

In the drier areas of Syria yields of barley, the principal crop, are low. Due to the variability in rainfall, fertilizer use is perceived as risky. Barley-fertilizer trials have been conducted on farmer's fields over a period of four years to investigate whether the large yield response to fertilizer observed on research sites could be achieved under farmers' conditions. Data were pooled across years and sites and response functions estimated. Simple optimization analysis shows that economic optimum fertilizer rates vary considerably with rainfall and relative prices. Historical rainfall data are combined with the estimated response functions, and stochastic dominance analysis is used to compare the riskiness of fertilizer treatments in terms of net benefits and benefit-cost ratios. Results show that, given the estimated expected rainfall in barley producing areas, fertilizer use, especially at low levels, may not be as risky as has been believed. Extensions of the methodology to include other environmental variables, or to target recommendations, are discussed.