More than Mean Effects : Modeling the Effect of Climate on the Higher Order Moments of Crop Yields

The ability of farmers to manage crop production in the presence of severe weather uncertainty is an important factor for maintaining adequate agricultural production, and this uncertainty is expected to be exacerbated under most climate change scenarios. We present a generalized version of previous models linking weather to yield outcomes that is capable of identifying the casual implications of weather on the entire yield distribution. Importantly, this approach does not require an ex ante distributional assumption and is flexible enough to admit both positive and negative skewness. We consider the case of Arkansas, Mississippi, and Texas upland cotton to demonstrate how temperature and irrigation affect the shape of the yield distribution. Results suggest that high temperatures and lack of irrigation concentrate yield outcomes toward the lower tail of the distribution; thus having significant implications for price variability, risk management, crop insurance, and other commodity support programs