Comprehensive evidence implies a higher social cost of CO2

Abstract

The social cost of carbon dioxide (SC-CO₂) measures the monetized value of the damages to society caused by an incremental metric tonne of CO₂ emissions and is a key metric informing climate policy. Used by governments and other decision-makers in benefit-cost analysis for over a decade, SC-CO₂ estimates draw on climate science, economics, demography and other disciplines. However, a 2017 report by the US National Academies of Sciences, Engineering, and Medicine¹ (NASEM) highlighted that current SC-CO₂ estimates no longer reflect the latest research. The report provided a series of recommendations for improving the scientific basis, transparency and uncertainty characterization of SC-CO₂ estimates. Here we show that improved probabilistic socioeconomic projections, climate models, damage functions, and discounting methods that collectively reflect theoretically consistent valuation of risk, substantially increase estimates of the SC-CO₂. Our preferred mean SC-CO₂ estimate is $185 per tonne of CO₂ ($44-$413 per tCO₂: 5%-95% range, 2020 US dollars) at a near-term risk-free discount rate of 2%, a value 3.6 times higher than the US government's current value of $51 per tCO₂. Our estimates incorporate updated scientific understanding throughout all components of SC-CO₂ estimation in the new open-source Greenhouse Gas Impact Value Estimator (GIVE) model, in a manner fully responsive to the near-term NASEM recommendations. Our higher SC-CO₂ values, compared with estimates currently used in policy evaluation, substantially increase the estimated benefits of greenhouse gas mitigation and thereby increase the expected net benefits of more stringent climate policies.

Fig. 1. RFF-SP socioeconomic scenarios and the resulting climate system projections.

a–c, Probabilistic socioeconomic projections for global population (a), per capita GDP growth rates (b), and carbon dioxide emission levels (c) from the RFF-SP scenarios. d–f, Corresponding climate system projections that account for parametric uncertainty in FaIR and BRICK for atmospheric carbon dioxide concentrations (d), global surface temperature changes relative to the 1850–1900 mean (e), and global mean sea-level changes relative to 1900 (f). In all panels, solid centre lines depict the median outcome, with darker shading spanning the 25%–75% quantile range and lighter shading spanning the 5%–95% quantile range.

Fig. 2. SC-CO₂ distributions vary with the choice of near-term discount rates.

Distributions of the SC-CO₂ based on RFF-SP scenario samples, a stochastic, growth-linked discounting framework, uncertainty in the FaIR climate and BRICK sea-level models, and uncertainty in climate damage parameters. Colours correspond to near-term average discount rates of 3.0% (blue), 2.5% (orange), 2.0% (red, our preferred specification) and 1.5% (teal). Dashed vertical lines highlight mean SC-CO₂ values. Box and whisker plots along the bottom of the figure depict the median of each SC-CO₂ distribution (centre white line), 25%–75% quantile range (box width), and 5%–95% quantile range (coloured horizontal lines) values. All SC-CO₂ values are expressed in 2020 US dollars per metric tonne of CO₂.

Fig. 3. Partial SC-CO₂ estimates and uncertainty levels strongly differ across the four climate damage sectors.

Box and whisker plots for the climate damage sectors included in the GIVE model, based on partial SC-CO₂ estimates for each sector. The figure depicts the median (centre white line), 25%–75% quantile range (box width), and 5%–95% quantile range (coloured horizontal lines) partial SC-CO₂ values. Black diamonds highlight each sector’s mean partial SC-CO₂, with the numeric value written directly above. All SC-CO₂ values are expressed in 2020 US dollars per metric tonne of CO₂.

Extended Data Fig. 1. SC-CO₂ distributions are robust to different damage function specifications ($ per tCO₂).

Distributions of the SC-CO₂ using the damage functions from GIVE (orange, our preferred specification), DICE-2016R (blue), and Howard & Sterner (red) for near-term discount rates of 1.5%, 2.0%, 2.5% and 3.0%. All results use the RFF-SP scenarios, a stochastic growth-linked discounting framework, and sample uncertain climate, sea-level and damage function parameters, including for DICE-2016R and Howard & Sterner damage functions. The DICE-2016R damage function is based on ref.  (see page 2 of that work’s supporting information). The Howard & Sterner damage function is based on the base coefficient in their table 2, specification (8). All SC-CO₂ values are expressed in 2020 US dollars per metric tonne of CO₂.

Extended Data Fig. 2. Discounted marginal damages by year, preferred 2% near-term discount rate case.

Solid line represents mean discounted marginal damages for a one-tonne CO₂ emissions pulse in 2020, dotted line represents the median, with darker shading spanning the 25%–75% quantile range and lighter shading spanning the 5%–95% quantile range. All SC-CO₂ values are expressed in 2020 US dollars per metric tonne of CO₂.