Abstract
This paper evaluates the potential for global reallocation between agricultural and nonagricultural production to contribute to climate change adaptation. Empirical estimates using a global sample of firms suggest that rising temperatures reduce productivity less in nonagriculture than angriculture, implying large potential gains if hot countries could increase food imports and shift labor toward manufacturing. However, model counterfactuals show that subsistence consumption needs and high trade barriers combine to create a “food problem” in which climate change instead intensifies agricultural specialization in especially vulnerable regions. Simulations suggest that reducing trade barriers can significantly reduce climate damages, especially in poor countries.