Using a Farmer's Beta for Improved Estimation of Expected Yields

Abstract

Effects of sampling error in estimation of farmers' mean yields for crop insurance purposes and their implications for actuarial soundness are explored using farm-level corn yield data in Iowa. Results indicate that sampling error, combined with nonlinearities in the indemnity function, leads to empirically estimated insurance rates that exceed actuarially fair values. The difference depends on the coverage level, the number of observations used, and the participation strategy followed by farmers. A new estimator for mean yields based on the decomposition of farm yields into systemic and idiosyncratic components is proposed, which could lead to improved rate-making and reduce adverse selection.

Keywords

• actual production history
• adverse selection
• crop insurance
• mean yields estimation
• sampling error