Skip to content
Stanford Doerr School of Sustainability

Earth System Science Seminar Series: Dr. Ken Caldeira, “Using Climate Models to Test Econometric Methods for Projecting Climate Damage”

ESE Seminar - Matthias Ihme: Probing Thermodynamic Processes at the Molecular Scale: Phase Transition and Structural Relaxation in Supercritical Fluids

Event Details

EARTH SYSTEM SCIENCE

SEMINAR SERIES

November 13, 2025

12:00 - 1:00 PM

Y2E2, Room 299

Ken Caldeira, Ph.D.

Visiting Scholar, Sustainable Solutions Lab,

Earth System Science Department, Stanford University

(also Senior Scientist at Gates Ventures)

“Using Climate Models to Test Econometric Methods for Projecting Climate Damage”

 

Abstract

Economists often project the economic impacts of climate change by statistically relating historical variation in economic output to historical variation in weather, and then extrapolating these relationships into the future. These econometric approaches underpin models such as Nordhaus’s DICE and inform estimates of the social cost of carbon by the U.S. EPA. However, because projections rest on counterfactual future scenarios that are not empirically determinable, such methods may not be directly falsifiable.

We evaluate the predictive skill of these econometric approaches using climate models in two ways. First, we apply standard econometric methods to simulated land-surface responses to historical climate variability, then test whether these methods can predict land-surface responses to future simulated climate change. Second, we construct a stylized economic model, akin to the core of DICE, embed known climate-response functions, and drive the model with climate simulations across grid points and scenarios. We then apply econometric methods to the historical simulations and test their ability to project the changes simulated for future scenarios. In both cases, initial results show limited predictive skill for the econometric methods considered.

These findings suggest that while econometric methods remain valuable for identifying current sensitivities to climate variability, they may be less reliable for long-term projection. A productive path forward may be to use econometric insights to inform development of mechanistic models, and then use these mechanistic models to project the economic consequences of future climate change.

Biography

Visiting Scholar, Sustainable Solutions Lab, Earth System Science Department, Stanford University (also Senior Scientist at Gates Ventures).

https://sustainablesolutions.stanford.edu/people/ken-caldeira

https://sustainablesolutions.stanford.edu/conceptual-investigations-unit

Topics & Tags

Explore More Events