摘要：Peasant farmers in Sahelian West Africa adjust to rainfall uncertainties in the agricultural season by making decisions sequentially as a function of the evolving rainfall patterns. Understanding such flexibilities in farmer decision-making is central to technology introduction. This paper determines how sequential decision-making under weather uncertainty affects the adoption and farm-level effects of cereal technologies in Niger. The study also draws policy implications for a price floor to arrest the substantial fall in cereal prices in good rainfall years when farmers have more grains to sell. The methodology used is Discrete Stochastic Programming. This paper shows that the ability of peasant farmers to adapt cropping and resource-management strategies to the rainfall patterns is the basis for their survival in this highrisk environment. Model results show that by (a) carrying a portfolio mix of varieties of varying maturities, and (b) making sequential decisions based upon rainfall expectations, farmers can adapt to the production uncertainties. Breeding programs should therefore be diversified to develop not only early-maturing cultivars, but also improved intermediate and long-season varieties.
