Policies, projections, and the social cost of carbon: Results from the DICE-2023 model 

What happens when Nobel-winning climate models meet the latest science?

Lint Barrage and William Nordhaus propose updates to the famous DICE integrated assessment model in « Policies, Projections, and the Social Cost of Carbon: Results from the DICE-2023 Model » (2024).

They update this classic climate-economy model with the latest data to see how optimal climate strategies and the Social Cost of Carbon (SCC) have shifted since its previous iteration:

• DICE-2023 introduces better risk accounting, a revised carbon cycle, updated damage estimates, inclusion of non-CO2 greenhouse gases, and new discount rate assumptions.
• The revised model yields a lower optimal temperature path at 2.6°C by 2100 and finds it economically justified to limit warming more stringently than the 3.6°C suggested by policy baselines.
• The projected cost of achieving net-zero emissions falls from about 11% of output in 2020 to just 2.7% of output by 2100, making ambitious climate goals more economically attainable.
• The estimated social cost of carbon in DICE-2023 is roughly $66 per ton of CO2 in 2020, a large increase from Nordhaus’ prior DICE estimates in the $30–$40 range a decade ago.
• Damages are now almost twice as large for a given temperature rise compared to DICE-2016: at 3 °C warming, DICE-2023 expects about a 3.1% hit to global GDP (7% at 4.5 °C).

Governments using DICE-2023 would aim for tighter emissions limits and possibly earlier carbon neutrality, since the model now finds avoiding higher temperatures is cost-beneficial.

Notably, the social cost of carbon rises to ~$66 in this update, which, while lower than some estimates, still represents a substantial increase compared to today’s market value.

Lower clean technology costs and manageable mitigation expense suggest that transitioning to a low-carbon economy may impose less drag on growth than feared.

As a limitation, DICE-2023 still uses a highly aggregated damage function that treats global temperature increase as the sole driver of damages, which omits dynamic effects and overlooks regional vulnerabilities.