Subjective versus objective yield distributions as measures of production risk

Subjective probabilities of farmers with nonlinear utility are estimated from distributions elicited by applying a linear scoring rule with monetary payoffs. Estimated subjective distributions are compared to objective estimates of yield distributions in a risky rainfed rice-growing area in the Philippines. The results indicate that farmers are knowledgeable about mean nitrogen response, but seriously underestimate year-to-year variability in yield response to nitrogen. The implication is that accurately estimated objective distributions can be used to estimate upper limits of the effects of risk aversion on allocative decisions.