Crop Failures from Temperature and Precipitation Shocks: Implications for US Crop Insurance

The effects of global warming on crop yield risk are critically important to US agriculture, particularly to crop insurance programs. We introduce a nonparametric model, using a copula density approach, to construct flexible conditional yield distributions given temperature and precipitation. Our approach has two advantages over the traditional approaches. First, our nonparametric copula approach allows us to estimate complex, flexible interaction effects of temperature and precipitation. Second, because we know the full distribution, we can coherently examine the effects on not only mean yields as in regression analyses, but also the effects on the probability of disastrous outcomes, variance, skewness, and other risk measures. This approach facilitates probabilistic predictions of quantities such as the probability of crop disasters and large crop insurance payouts in response to temperature and precipitation shocks.