Systemic Risk, Relative Subsidy Rates and Area Yield Insurance Choice

Xuche Gong, David A. Hennessy, Hongli Feng
September 2022  [22-WP 636] (Revised)

Download Full Text

Suggested citation:

Gong, X., D.A. Hennessy, and H. Feng. 2022. "Systemic Risk, Relative Subsidy Rates and Area Yield Insurance Choice." Working paper 22-WP 636. Center for Agricultural and Rural Development, Iowa State University.


We investigate the nature of crop yield systemic risk and its implications for farmers’ area yield insurance choices. A theory-grounded, normalized measure of systemic risk, R2, is developed and made amenable to empirical analysis through the logit link. The measure is estimated on a large-scale, unit-level corn yield dataset. We find that systemic risk explains less than half of total unit yield variability on average, suggesting that the risk management effectiveness of area yield insurance is low. By relating natural resource endowments with systemic risk, we find that more excessive heat and drought events lead to larger while more excessive precipitation events lead to smaller systemic risk. We then study whether current area yield insurance subsidy rates provide farmers with sufficient subsidy transfers to compensate for the uncovered risk exposure associated with area yield insurance. A new concept, the threshold relative area subsidy rate (TRASR), quantifies a lower bound on the ratio of area yield insurance subsidy rate over individual yield insurance subsidy rate below which risk-averse farmers should always prefer individual yield insurance to area yield insurance. The calibrated TRASR values indicate that current area yield insurance subsidy rates discourage farmers from choosing area yield insurance over individual yield insurance, especially at low area yield insurance coverage levels. We also find that TRASR is positively correlated with systemic risk at the unit-level of analysis, suggesting that farmers who like area yield insurance’s risk management features will dislike its transfer implications and vice versa.