Interannual Climate Variation, Climate Prediction, and Agricultural Trade: the Costs of Surprise versus Variability

Agriculture remains sensitive to variation in rainfall and temperature. Fortunately, our ability to predict the lower frequency variation in the earth's atmosphere is increasing rapidly. While information cannot affect the underlying source of variability, it allows unexpected shocks to be anticipated and acted upon. This paper develops and analyzes an intertemporal Ricardian trade model with Bayesian beliefs. Results in this paper show that improved climate prediction reduces expected prices, but increases price variability. In addition, trade is crucial to realizing the potential benefits of climate prediction. The value of climate prediction is highest with both storage and trade, and falls by over 300% when a country is in autarky.