CRED Estimating Climate Risks and Economic Damages
Estimating Climate Risks and Economic Damages

Estimating Climate Risks and Economic Damages

GW’s Climate and Health Institute is creating data and tools to estimate climate risks and economic damages (CRED)

Climate change impacts human health and societal systems in different ways. From heat waves to wildfires, hurricanes, and sea level rise, climate change can worsen environmental pollution, accelerate its impacts, disrupt healthcare, and reduce economic productivity. As the evidence linking climate change with societal impacts continues to strengthen, the opportunity to comprehensively evaluate these damages and inform solutions has never been greater.

The GW Climate and Health Institute at George Washington University is advancing data and tools to quantify and monetize the many pathways by which climate change impacts health and welfare. We are developing new science that links climate drivers with health impacts and other societal outcomes and integrating this new knowledge into analytical tools that have been applied in some of the most consequential greenhouse gas policies and regulations in the U.S., making the tools more comprehensive. We are also reconstructing tools to make them more broadly accessible in a variety of contexts, including national and state-level legislation and legal proceedings.

Engagement is central. The CHI is coordinating with a range of stakeholders to help shape our approaches for characterizing impacts and economic damages from climate change and ensure they reflect real‑world needs. 

CRED Overview

CRED projects the impacts of climate change on health and the economy in the United States, by state, sector and year. It builds on EPA's Framework for Evaluating Damages and Impacts (FrEDI), which was used for EPA regulatory analyses in 2023 and 2024. EPA removed the FrEDI webpage in December 2025 (EPA, 2026). At CHI, we are working to keep the tool current, adding missing health impacts, and making it easier to use.

What is CRED?

CRED is a climate impacts estimation tool written in R. The FrEDI core of CRED was originally developed as part of EPA's Climate Change Impacts and Risk Analysis (CIRA) project, under the guidance of Marcus Sarofim and Rona Birnbaum from the CHI team while they were at EPA. Projections of global temperature and sea level, and US population and income serve as inputs, and CRED returns annual impact estimates for the 48 contiguous states and the District of Columbia. These include physical impacts, such as premature deaths and new cases of childhood asthma, and their economic value.

CRED uses a temperature binning approach (Sarofim et al., 2021): reduced form impacts functions are estimated based on sectoral studies that estimate impacts at each degree of warming. These damage functions depend on the amount of warming rather than on a specific emissions scenario. The same functions can evaluate an entire future scenario or the impacts of emissions changes due to a single regulation. A century-long analysis is run on a laptop, and results can be extended to 2300 (Hartin et al., 2023).

Most of the projected damages come from health effects (Hartin et al., 2023). The health sectors cover:

• temperature-related deaths

• deaths and illness from climate-driven changes in ozone and fine particle pollution

• deaths and illness from wildfire smoke

• deaths and illness from dust in the southwest, Valley fever, Lyme disease and vibriosis

A separate module estimates deaths and asthma from methane-derived ozone, based on McDuffie et al. (2023). The remaining sectors cover infrastructure, labor, agriculture, learning loss, recreation and natural resources, for 25 impact sectors in total.

EPA's documentation of FrEDI went through public comment and external peer review in 2024 (EPA, 2024a), in addition to FrEDI's use in multiple peer-reviewed journal publications.

Where has CRED been used?

Federal analysis

• EPA used FrEDI in the regulatory impact analyses for several rules in 2023 and 2024, including standards for vehicles, oil and gas methane, and power plants (e.g., EPA, 2024b). For the Waste Emissions Charge on methane, EPA ran FrEDI to estimate the avoided impacts in the U.S. resulting from the rule's emission changes and reported the results by sector, region and socially vulnerable group (EPA, 2024c, 2024d).

• EPA's 2023 Report on the Social Cost of Greenhouse Gases, NHTSA's 2024 fuel economy standards, and OMB's 2022 assessment of federal climate risk also used FrEDI (EPA, 2023; OMB, 2022).

Research

•  Sarofim et al. (2021) introduced the temperature binning approach.

• Hartin et al. (2023) described FrEDI and ran it for 10,000 scenarios. Mean damages reached $2.9 trillion a year by 2090 (2020 dollars), mostly from health effects.

• Kopits et al. (2025), an EPA working paper, compared U.S. damages per ton of carbon dioxide from three models. For 2030 emissions, FrEDI gave $33 per ton, compared with $22 from GIVE and $14 from DSCIM (2020 dollars, 2 percent near-term discount rate).

• Yuan et al. (2026) linked FrEDI to an economy-wide model and found that economy-wide damages were about 20 percent larger than the sum of the sector damages in 2050, and about 50 percent larger in 2100.

Public comments

• In November 2026, the attorneys general of 22 states and the District of Columbia, joined by Chicago, Denver, New York City and the Bay Area Air Quality Management District, commented on EPA's proposal to rescind its greenhouse gas findings for power plants (EPA-HQ-OAR-2025-0124-3584). In Appendix 2, Marcus Sarofim used FrEDI to estimate that EPA's chosen scenario, cutting power plant carbon dioxide emissions in half through 2100, would avoid 1,606 premature deaths a year in 2100, about 50,000 deaths between 2025 and 2100, and about 1,800 new cases of childhood asthma a year by 2100. Avoided damages through 2300 total $0.44 trillion at a 3 percent discount rate and $3.08 trillion at a 2 percent Ramsey rate (2020 dollars).

Where is CRED going?

The goal is a tool that U.S. city and state leaders, health systems, insurers and community groups can use to see what climate hazards will cost them, where the burden falls hardest, and how much a given action would save. An open, rigorously documented tool, maintained by a university and shaped with the people who use it, gives them information they can trust and apply to their own circumstances. As federal or congressional needs arise, the tool will be in an improved and ready-state for application. 

Easier to use

We are planning to create a web interface for a version of CRED to serve users who don't program in R. For users who do, we will also provide examples of how to drive CRED with standard models such as the FaIR climate model (Leach et al., 2021) and the BRICK sea level model (Wong et al., 2017), using emissions scenarios from van Vuuren et al. (2026) and socioeconomic projections from Resources for the Future (Rennert et al., 2022).

More impacts

CRED is not a comprehensive model. We are looking for existing impact studies that can be adapted into damage functions within the CRED framework, particularly for health impacts. Examples include hurricane impacts on mortality (Young and Hsiang, 2024) and power outages (Guikema et al., 2025), heat-related emergency department visits (Lay et al., 2018; Young et al., 2026), high-tide flooding and elder mortality (Sheahan et al., 2025), and effects on anxiety and depression (Belova et al., 2026).

Uncertainty

CRED can be driven by ensembles that sample uncertainty in climate (using FaIR and BRICK) and in socioeconomics (using the RFF projections). Pattern uncertainty is included within CRED: for many sectors, damage functions are fit to several global climate models. However, CRED's standard results use the central impact values from each underlying study, though some of those studies assessed uncertainties in the relationship between temperature change and impacts. A few sectors do include alternatives, such as low and high variants for temperature-related deaths and multiple exposure-response functions for wildfire smoke. CHI researchers and colleagues are developing a global analysis of wildfire smoke and health (Kingdon et al., in preparation), which could provide an alternative function for the wildfire sector. We would work to carry more of these uncertainties into CRED and to report them in its results.

Applications

CRED could support a wider range of decisions. For policy analysis, it can estimate the avoided health and economic impacts of federal and state policies that change emissions, either by running a policy's emission changes directly or by applying damages per ton. For adaptation planning, results by state, sector and year show where impacts are projected to be largest, and several sectors, such as roads, rail and coastal property, compare impacts with and without adaptation. States can run CRED with state population projections and use the results in climate and health planning. We would like to work with state agencies, health departments and other users to tailor CRED to create targeted analyses and provide credible and useful results to the decisions they face.

Key references

• Belova, A., et al. (2026). Projecting and valuing climate change impacts on anxiety and depression in the contiguous United States. Lancet Planetary Health 10(2), 101426. https://doi.org/10.1016/j.lanplh.2025.101426

• EPA (2023). Report on the Social Cost of Greenhouse Gases: Estimates Incorporating Recent Scientific Advances. Docket EPA-HQ-OAR-2021-0317.

• EPA (2024a). Technical Documentation for the Framework for Evaluating Damages and Impacts (FrEDI). EPA 430-R-24-001.

• EPA (2024b). Regulatory Impact Analysis: Multi-Pollutant Emissions Standards for Model Years 2027 and Later Light-Duty and Medium-Duty Vehicles. EPA-420-R-24-004, section 9.

• EPA (2024c). Regulatory Impact Analysis of the Proposed Waste Emissions Charge. EPA-430/R-23-005, January 2024, Appendix B.

• EPA (2024d). Regulatory Impact Analysis of the Waste Emissions Charge. EPA-430/R-24-007, November 2024, section 9 and Appendix B.

• EPA (2026). Response to Request for Correction 25001, April 23, 2026. https://www.epa.gov/system/files/documents/2026-04/rfc-25001-technical-document-on-fredi_final-response-letter_signed.pdf

• Guikema, S., Pagan-Cajigas, Z., Fant, C., Boehlert, B., Maier, C.X., Emanuel, K., Hartin, C., and Sarofim, M.C. (2025). Climate change impacts on tropical cyclone-induced power outage risk: Sociodemographic differences in outage burdens. PNAS 122(43), e2502266122. https://doi.org/10.1073/pnas.2502266122

• Hartin, C., et al. (2023). Advancing the estimation of future climate impacts within the United States. Earth System Dynamics 14, 1015–1037. https://doi.org/10.5194/esd-14-1015-2023

• Kingdon, C., Kerr, G.H., Anenberg, S.C., and Cromar, K.R. (in preparation).

• Kopits, E., et al. (2025). Economic Damages from Climate Change to U.S. Populations: Integrating Evidence from Recent Studies. EPA National Center for Environmental Economics Working Paper 25-01.

• Lay, C.R., et al. (2018). Emergency department visits and ambient temperature: Evaluating the connection and projecting future outcomes. GeoHealth 2(6), 182–194. https://doi.org/10.1002/2018GH000129

• Leach, N.J., et al. (2021). FaIRv2.0.0: a generalized impulse response model for climate uncertainty and future scenario exploration. Geoscientific Model Development 14, 3007–3036. https://doi.org/10.5194/gmd-14-3007-2021

• McDuffie, E.E., Sarofim, M.C., et al. (2023). The social cost of ozone-related mortality impacts from methane emissions. Earth's Future 11(9). https://doi.org/10.1029/2023EF003853

• OMB (2022). Climate Risk Exposure: An Assessment of the Federal Government's Financial Risks to Climate Change. Office of Management and Budget white paper, April 2022.

• Rennert, K., et al. (2022). The social cost of carbon: Advances in long-term probabilistic projections of population, GDP, emissions, and discount rates. Brookings Papers on Economic Activity, Fall 2021.

• Sarofim, M.C., et al. (2021). A temperature binning approach for multi-sector climate impact analysis. Climatic Change 165, 22. https://doi.org/10.1007/s10584-021-03048-6

• Sheahan, M.B., Hauer, M., Mueller, V., Sheriff, G., Fant, C., Holland, I., Sarofim, M.C., Hartin, C., and Neumann, J.E. (2025). Projections of future mortality risk in older adults from high-tide flooding in coastal areas of the USA: an economic modelling study. Lancet Planetary Health, 101382. https://doi.org/10.1016/j.lanplh.2025.101382

• van Vuuren, D.P., et al. (2026). The Scenario Model Intercomparison Project for CMIP7 (ScenarioMIP-CMIP7). Geoscientific Model Development 19, 2627–2656. https://doi.org/10.5194/gmd-19-2627-2026

• Wong, T.E., et al. (2017). BRICK v0.2, a simple, accessible, and transparent model framework for climate and regional sea-level projections. Geoscientific Model Development 10, 2741–2760. https://doi.org/10.5194/gmd-10-2741-2017

• Young, M., Dahl, K.A., and Saha, S. (2026). Climate change increases emergency department visits for heat-related illness in the United States. Climatic Change 179, 190. https://doi.org/10.1007/s10584-026-04247-9

• Young, R., and S. Hsiang (2024). Mortality caused by tropical cyclones in the United States. Nature 635, 121–128. https://doi.org/10.1038/s41586-024-07945-5

• Yuan, M., et al. (2026). General equilibrium effects from partial equilibrium-driven climate impacts: Implications from FrEDI-USREP. Climate Change Economics 17(4), 2650007. https://doi.org/10.1142/S2010007826500077