摘要：The present study sets up a thought experiment calibrated to represent risks of a high-risk production activity (farming), and investigating whether the structure of risk aversion (i.e., the changes in absolute or relative risk aversion associated with changes in wealth) can be estimated with reasonable precision. Findings strongly suggest that typical production data are unlikely to allow identification of the structure of risk aversion. A flexible utility parameterization is found to worsen technology parameter estimates. Findings also indicate that even under a restricted utility specification, the quality of utility parameters estimated from small samples is very poor.
