Agricultural Productivity and Anticipated Climate Change in Sub-Saharan Africa: A Spatial Sample Selection Model

A cereal yield response function is estimated conditional upon environmental and topographical features to detect the effects of spatial heterogeneity and spatial dependence in explaining agricultural productivity across Sub-Saharan Africa. Controlling for direct and localized spillover effects, we then estimate the effect that projected changes in temperature and precipitation as a result of global climate change will have on agricultural production. We find that the estimated declines found in the climatological literature may overestimate actual declines, and factors such as spatial heterogeneity (i.e., country fixed effects) are profoundly more important to agricultural production.