Farm risk management: past, present and prospects

The integration of ideas about utility and subjective probability into modern decision analysis is outlined. Despite being widely adopted, some aspects, are still contentious. The legitimacy and practicality of utility functions to reflect risk aversion has been challenged and there is far from general acceptance of subjective probabilities as a basis for risky choice. While subjective expected utility does not always explain behaviour under risk, it is arguably the best basis for prescriptive decision analysis. The focus on the limitations of utility theory may have diverted attention from the more important task of devising and applying ways to get good probabilities, especially in data-sparse cases. It is argued that more attention needs to be given to the important risks that farmers face. As these are often those for which frequency data are sparse or unavailable, the need for wider acceptance of well-considered subjective probabilities as descriptors of uncertainty is clear, requiring a difficult paradigm shift.