International trade literature tends to focus heavily on the production side of general equilibrium, leaving us with a number of empirical puzzles. There is, for example, considerably less world trade than predicted by Heckscher-Ohlin-Vanek (HOV) models. Trade among rich countries is higher and trade between rich and poor countries lower than suggested by HOV and other supply-driven theories, and trade-to-GDP ratios are higher in rich countries. Our approach focuses on the relationship between characteristics of goods and services in production and characteristics of preferences. In particular, we find a strong and significant positive correlation of more than 45% between a good's skilledlabor intensity and its income elasticity, even when accounting for trade costs and cross-country price differences. Exploring the implications of this correlation for empirical trade puzzles, we find that it can reduce HOV's overprediction of the variance of the net factor content of trade relative to that in the data by about 60%. Since rich countries are relatively skilled-labor abundant, they are relatively specialized in consuming the same goods and services that they are specialized in producing, and so trade more with one another than with poor countries. We also find a positive sector-level correlation between income elasticity and a sector's tradability, which helps explain the higher trade-to-GDP ratios in high-income relative to low-income countries.

While climate change impacts on crop yields has been extensively studied, estimating the impact of water shortages on irrigated crop yields is challenging because the water resources management system is complex. To investigate this issue, we integrate a crop yield reduction module and a water resources model into the MIT Integrated Global System Modeling framework, an integrated assessment model linking a global economic model to an Earth system model. We assess the effects of climate and socioeconomic changes on water availability for irrigation in the U.S. as well as subsequent impacts on crop yields by 2050, while accounting for climate change projection uncertainty. We find that climate and socioeconomic changes will increase water shortages and strongly reduce irrigated yields for specific crops (i.e., cotton and forage), or in specific regions (i.e., the Southwest) where irrigation is not sustainable. Crop modeling studies that do not represent changes in irrigation availability can thus be misleading. Yet, since the most water-stressed basins represent a relatively small share of U.S. irrigated areas, the overall reduction in U.S. crop yields is small. The response of crop yields to climate change and water stress also suggests that some level of adaptation will be feasible, like relocating croplands to regions with sustainable irrigation or switching to less irrigation intensive crops. Finally, additional simulations show that greenhouse gas (GHG) mitigation can alleviate the effect of water stress on irrigated crop yields, enough to offset the reduced CO₂ fertilization effect compared to an unconstrained GHG emission scenario.

While climate change impacts on crop yields has been extensively studied, estimating the impact of water shortages on irrigated crop yields is challenging because the water resources management system is complex. To investigate this issue, we integrate a crop yield reduction module and a water resources model into the MIT Integrated Global System Modeling framework, an integrated assessment model linking a global economic model to an Earth system model. We assess the effects of climate and socioeconomic changes on water availability for irrigation in the U.S. as well as subsequent impacts on crop yields by 2050, while accounting for climate change projection uncertainty. We find that climate and socioeconomic changes will increase water shortages and strongly reduce irrigated yields for specific crops (i.e., cotton and forage), or in specific regions (i.e., the Southwest) where irrigation is not sustainable. Crop modeling studies that do not represent changes in irrigation availability can thus be misleading. Yet, since the most water-stressed basins represent a relatively small share of U.S. irrigated areas, the overall reduction in U.S. crop yields is small. The response of crop yields to climate change and water stress also suggests that some level of adaptation will be feasible, like relocating croplands to regions with sustainable irrigation or switching to less irrigation intensive crops. Finally, additional simulations show that greenhouse gas (GHG) mitigation can alleviate the effect of water stress on irrigated crop yields, enough to offset the reduced CO2 fertilization effect compared to an unconstrained GHG emission scenario.