Assessing the Economic and Environmental Consequences of Climate Change

Emissions abatement alone cannot address the consequences of global warming for frequency of weather disasters. We model regional-level adaptation to mitigate disaster risks to capital stock. Optimal adaptation—a mix of private efforts and public spending funded by a tax on capital—depends on learning regarding the adverse consequences of global warming from disaster arrivals. We apply our model to country-level control of flooding from major tropical cyclones. Learning is needed to rationalize new findings on the response of asset prices to disaster arrivals. The value of adaptation with learning is much higher than under the counterfactual without learning. Finally, we quantify how learning and adaptation alters projections for the social cost of carbon over time.

As we begin deploying machine learning models in consequential settings like medical diagnostics or self-driving vehicles, we need ways of knowing when the model may make a consequential error (for example, that the car doesn’t hit a human). I’ll be discussing how to generate rigorous, finite-sample confidence intervals for any prediction task, any model, and any dataset, for little computational cost. This will be a chalk talk. I will primarily discuss conformal prediction and related methods, which work for a large class of prediction problems including those with high-dimensional, structured outputs (e.g. instance segmentation, multiclass or hierarchical classification, protein folding, and so on).

This paper evaluates the impact of extreme weather events for the social cost of climate change. To that end we develop a dynamic spatial model of the US economy divided in over 3,000 counties that features risk-averse individuals, forward-looking migration and investment, and idiosyncratic and aggregate climate risk. We achieve tractability by relying on recent methodological advances that leverage analytic perturbations of the Master Equation representation of the economy. We estimate the model on US counties by combining climatic county-level data for weather events such as heat waves, floods and hurricanes over the course of the 20th century with population, migration, investment and income data. We run event studies that trace out the impact of extreme weather events on productivity, amenities, mortality, and capital depreciation. Our findings are twofold. First, climate impacts on capital depreciation and mobility are the main source of climate damages. They substantially magnify the costs of climate change in the business-as-usual warming scenario. Second, adaptation through migration and investment are crucial mitigators of climate damages, without which welfare losses would be much more spatially concentrated.