Risk research in agricultural economics: opportunities and challenges for the next twenty-five years

Risk research has not convinced many of its importance due to a focus on problems where risk is likely less important. Data availability has influenced problem choice and explains excessive focus on short-run problems with temporal and spatial aggregation. Issues of long-term risk, which depend on serial correlation and farm failure, are likely more important. Standard assumptions of additive separability are absurd for such problems. Major self-imposed constraints on risk research are methodological and related to the heavy recent emphasis on duality and non-parametric approaches. Both movements have been counterproductive from the standpoint of risk research because standard applications have ignored risk aversion. Conditions of profit maximization are not tested but simply assumed. Some dual approaches can be generalized but little has been done to date and much of the convenience is lost under risk aversion. A comprehensive data base is crucial for uncovering behavior that goes beyond standard methodologies.