Accounting for geographic basis risk in agricultural weather index insurance : how kriging prediction may improve protection against climate risk

Sammanfattning: Weather index insurance is a potential solution to a widely acknowledged problem of information asymmetries in agricultural crop insurance. Insurance where payouts depend on a weather index relies on accurate estimates of local weather, such as temperature and rainfall. Using a geostatistical kriging method and empirical weather and crop yield data from Illinois, I explore whether accounting for geographic approximation errors produces more desirable index insurance contracts. I find that switching to this so-called geographic basis risk-adjusted contract improves farmers’ utility for one of our two indices, but not for the other. Further, purchasing any index insurance contract only improves farmers’ utility during a particularly hot year. During a cooler year, purchasing WI insurance results in lower utility for risk neutral farmer, and constant utility for risk averse farmers.