Abstract
We apply a Structural Ricardian Model (SRM) to farm-level data from Ghana for seven principal food crops in order to model the factors which influence farmers' decisions about which food crops to grow and the revenue conditional on these choices. Our application of the SRM incorporates a flexible functional form to allow for the possibility that the effects of temperature and rainfall may not be linearly separable. We use this model as a basis for simulations of the effects of climate change. These simulations suggest that extreme climate change will lead to a significant reduction in the average net revenue per hectare from maize, which accounts for over half of current food production. Across a range of climate change scenarios, there is also substantial substitution of maize for heat-tolerant millet, and a reduction in the cultivation of other crops. Under most of the climate change scenarios that we consider, these results imply a substantial reduction in the aggregate value of agricultural production, since millet is the lowest-value crop.