The social cost of carbon (SCC) can have a major impact on climate regulations established by the US government. While vitally important, estimating the value of the SCC also is a highly technical task. The federal government is currently implementing improvements to the process that will result in an updated value this January. Calculating the inputs for SCC estimates gets complicated; for example, very long-term economic growth, population, and greenhouse gas emissions involve substantial uncertainty on long time horizons. Research from Resources for the Future aims to account for these uncertainties and employ empirical data to make the best possible estimate of the SCC using the best available science.
The social cost of carbon (SCC) often has been referred to as the most important number you’ve never heard of. It’s the key economic measure of the benefits of mitigating climate change—an estimate, in dollars, of the economic cost (i.e., damages) that results from emitting each additional ton of carbon dioxide (CO₂) into the atmosphere. Conversely, the SCC represents the benefit to society of reducing CO₂ emissions by one ton—a number that can be used to inform policy decisions. Analogous metrics exist for the greenhouse gas pollutants methane and nitrous oxide.
The SCC has deep intellectual roots in economics. The Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel was awarded to William Nordhaus in 2018 (alongside Paul Romer) for his seminal work incorporating climate change into economic analysis, including the role of the SCC in informing policy. Textbooks often use carbon emissions and their role in causing climate change as the canonical example of an externality, which must be addressed to maximize human well-being. One specific method of addressing emissions and improving welfare comes from basic economic theory, which recommends that an optimal tax on CO₂ emissions (often called a carbon tax) should be set at a level where the SCC and the incremental cost of emissions control are equal.
But the relevance and application of the SCC go well beyond its potential role in environmental taxation. It’s been used by the US federal government for more than a decade in benefit-cost analyses that inform regulations like vehicle fuel economy standards and power plant emissions rules. The SCC also was the basis for valuing federal tax credits for carbon capture technologies enacted in 2018 and zero-emissions credits for nuclear power in New York State. The power grid operator for New York State currently is working to include the SCC as a cost “adder” on top of energy supply bids submitted by power plants, to reflect these social costs into market prices and power distribution. Many other states have used the SCC as the basis for various types of climate policies. Even more proposed applications include influencing federal procurement decisions, determining federal royalties on oil and gas leases on federal land, and the list goes on.
In other words, while political leaders and stakeholders continue to debate both the broad outlines and fine details of policies to mitigate carbon dioxide emissions, the SCC lies in the background as a remarkably important calculation that helps inform specific proposals.
Many of the factors underlying the calculation of the SCC are deeply uncertain. These factors include our understanding of the effect of climate change on economic outcomes, the time-sensitive consequences of today’s emissions projected in the future, and the science of Earth’s climate. The need for robust policy decisions implies that we should update the SCC over time to refine central estimates and their uncertainty as our scientific understanding progresses.
Credit: Martin Jozwiak / Unsplash
This article provides a high-level summary of important aspects of our efforts to update key determinants of the SCC, while reflecting the best available science, based on the recommendations of a landmark 2017 committee report by the National Academies of Sciences, Engineering, and Medicine (NASEM). The comprehensive details of our work will be published in a forthcoming article in the Brookings Papers on Economic Activity series.
This work is particularly relevant in light of the January 20, 2021, Executive Order 13990, which reestablished the Obama-era Interagency Working Group on the SCC and directed it to update the SCC in consideration of the NASEM report. The NASEM report provides extensive guidance to improve the scientific basis, provide more transparency, and better address uncertainties in the SCC. It also recommends establishing an institutional process for updating SCC estimates approximately every five years—an update cycle that would balance the benefit of incorporating the latest research with the need for a thorough process. The next updated estimates for the SCC are anticipated for January 2022.
What We Know
The federal government’s current interim value of $51 per ton of CO₂ reflects the expected value of the SCC, accounting for uncertainty in the climate’s warming response and five socioeconomic scenarios of economic growth, population, and emissions at a 3% constant discount rate. The NASEM report points out that prior SCC estimates by the US government—up to and including the current interim $51-per-ton SCC value—use somewhat dated and often overly simplistic approaches.
For example, the projections that inform SCC calculations apply across a very long time horizon—from decades to centuries—which makes predictions difficult. Complex uncertainties must be considered in those projections; for instance, the outlook for socioeconomic variables depends on factors such as new technologies, the mitigation policies implemented, and the shares of different sectors and regions in the global economy. And because the effects of climate change vary regionally, the socioeconomic projections ideally should provide enough detail to account for regional heterogeneity in climate impacts.
Credit: Señor Salme
An assessment of damages from future climate change is strongly influenced by underlying projections of socioeconomic variables such as population, economic growth, and emissions, and estimates of the SCC have been shown to exhibit significant sensitivity to the projections for these variables. But the five socioeconomic scenarios in routine use all have been treated as equally likely in prior estimates of the SCC by the US government, even though those five scenarios were not developed with any formal probabilities attached. The discounting approach also has used a constant discount rate, rather than treating the discount rate as contingent on each scenario—a distinction that becomes increasingly important as the relevant timescale shifts further into the future. To address such shortcomings, the NASEM report issued a series of recommendations, and Executive Order 13990 specifically directs the Interagency Working Group to take those NASEM recommendations into consideration when updating the SCC.
Specifically, the NASEM report proposes four “modules,” each corresponding to a step in SCC estimation, along with an overall framework that integrates the modules and considers their various interdependencies (Figure 1). Each of the modules listed below can characterize inherent uncertainty, resulting in a distribution of estimates rather than a single value.
- Socioeconomic and emissions projections
- Physical climate system
- Monetized climate damages
- Discounting
Figure 1. Modular Framework for Estimating the Social Cost of Carbon (SCC)
Resources for the Future (RFF) created the Social Cost of Carbon Initiative in 2017 to advance research that addresses the NASEM recommendations. This effort, which focuses on improving the scientific quality and transparency surrounding SCC estimates, involves a network of partners—RFF; the University of California, Berkeley; Harvard; Princeton; the University of Washington; and others. RFF’s research efforts have fully implemented most of the NASEM recommendations by doing the following: implementing a transparent, open-source computational framework; developing country-level GDP per capita, global population, and global emissions accounting for future policies and dependencies between the variables; incorporating an updated climate model used for SCC calculations; and providing an improved discounting framework. Additional work will assemble new climate damage functions from the best available literature is rapidly nearing completion.
Getting to Know the Known Unknowns
We will take a focused look at the detailed methods and results of one facet of RFF’s efforts: building a new set of long-term projections for regional GDP, based on statistical evidence and an elicitation of expert predictions for the major developed economies. We’ll also summarize the results of our work to generate projections for global population growth, CO₂ emissions, and temperature. The comprehensive methodological underpinnings and results of this work will be published in our contribution to the Brookings Papers on Economic Activity series. And further details about determining an appropriate discount rate for the SCC can be found in this article.
Country-Level GDP Per Capita Projections
We quantified uncertainty about economic growth by generating a distribution of long-run growth projections at the country level, following a three-step approach: We started with statistically generated projections of country-level economic growth derived in recent research. Statistically generated projections based on historical data will be more informative about economic growth in the near term than in the very long term; so, with an eye toward the long term, we augmented the statistical evidence with an Economic Growth Survey that we implemented through a formal elicitation of experts. Through this second step of conducting the survey, we quantified expert uncertainty about future economic growth, focusing in particular on the very long term. In the third and final step, we combined the expert uncertainty with the statistically generated trajectories to provide projections over a nearly 300-year time horizon that reflect information from both the statistical method and the experts.
The Economic Growth Survey involved ten experts, selected based on their expertise in the fields of macroeconomics and economic growth and by the recommendations of their peers. Consistent with the statistics-based forecasts, the experts demonstrated a substantial range of uncertainty in future economic growth, beyond that typically represented in scenarios that have been employed for energy and climate analysis. However, as shown by the individual expert predictions, and as expressed in verbal comments during the elicitation, most expert participants did not expect long-run future growth to equal that projected by a purely statistical model of the past 100 years (Figure 2), which predicted nearly 2 percent growth. Combining the expert judgment with statistical projections reduced the expected future growth rate in the resulting data set.
Figure 2. Distributions of Future Average GDP per Capita Growth for the Major Developed Economies, Based on a Systematic Survey of Experts and Econometric Sources
Circles indicate medians; lines indicate the 1st, 5th, 95th, and 99th percentiles of each distribution
The expert responses exhibited considerable diversity in their characterization of economic growth, with some of the widest ranges driven by their explicit inclusion of events that are not present nor fully realized in the historical record of economic growth on which statistical projections are based. When considering the major developed economies, most experts expected a deceleration of growth in global GDP per capita, and possibly very low future growth rates (Figure 3), relative to what purely historical statistical evidence suggests.
Figure 3. Multi-century Projections for Average Growth in GDP per Capita, Based on Econometric Sources and Systematic Surveys of Experts
Data combine statistical and expert-based projections. Shaded areas represent 90 percent and 98 percent prediction intervals
When asked to identify their primary drivers of the potential low-growth outcomes, the experts most commonly cited climate change, followed by world conflict, natural catastrophes, and global health crises. In identifying the primary drivers of high growth, experts most often cited the rapid advancement of technology, followed by regional cooperation and advances in medical science.
Global Population Projections
To generate very long-term projections of population, we extended the existing statistical approach used by the United Nations for its official population forecasts through 2100, modifying the methodology with input from a panel of nine leading demographers. The resulting distribution of trajectories predicts a peak in median world population at about 11 billion in the middle of the next century (Figure 4), with a decline from the peak to about 7.5 billion by 2300, accompanied by wide uncertainty.
Figure 4. Multi-century Projections for World Population, Based on Econometric Sources and Systematic Surveys of Experts
Data combine statistical and expert-based projections. Shaded areas represent 90 percent and 98 percent prediction intervals
Global Emissions Projections
We additionally quantified the uncertainty surrounding four categories of future greenhouse gas emissions—including uncertainty about future climate policy and economic growth—through a formal expert survey. The median of the resulting trajectories suggests an approximately 60 percent reduction in CO₂ emissions by 2100 (Figure 5), with wide uncertainty that includes the possibility of net-zero global emissions as well as substantial increases from current levels. In general, the expert projections suggest that the shared socioeconomic pathways are relative outliers in their emissions projections through 2100, and the full range is largely inconsistent with the enforced requirement of shared socioeconomic pathways to converge to zero emissions in 2250.
Figure 5. Multi-century Projections for Carbon Dioxide Emissions, Based on Econometric Sources and Systematic Surveys of Experts
The data combine statistical and expert-based projections. Shaded areas represent 90 percent and 98 percent prediction intervals
We next ran an updated climate model with samples from the emissions projection data to evaluate projected changes in temperature. Our results suggest a median global temperature increase of 2.6° Celsius from the preindustrial level by 2100, with a continued increase in temperature through 2300. The results indicate a roughly 20 percent likelihood of staying below a temperature increase of 2° Celsius by 2100. Our experts suggested that negative CO₂ emissions, from afforestation or direct air capture, could play a significant role in the future and allow for temperature pathways that peak and then decline.
Our approach to improving long-run projections of economic growth, global population growth, and global emissions, and combining these variables into a consistent set of interrelated projections, fully implements the near-term NASEM recommendations for the update of the SCC currently in progress. By providing these projections at the country level, our research takes significant additional steps toward meeting longer-term recommendations of the NASEM, as well.
Looking Forward to a New Social Cost of Carbon
The SCC is a vitally important metric that can guide climate policy. As such, its calculation must be supported by the best available science—including the explicit incorporation of uncertainty. Socioeconomic uncertainty, along with discounting that varies in tandem with socioeconomic uncertainty, are important drivers of the SCC. Our work provides the means and opportunity to incorporate those relationships and uncertainties into ongoing updates to this important metric. The research of RFF and our collaborators fully implements the NASEM near-term recommendations for three of the four modules described above, and we’re rapidly nearing the completion of updates for the final module.
More generally, SCC-related research is an ongoing endeavor, and the SCC should be updated at regular intervals as the scientific frontier advances on multiple fronts, as recommended by NASEM. Our work speaks directly to those NASEM recommendations and provides tools that the US government can use to improve upon the simple, deterministic approaches to socioeconomic projections and discounting methodologies employed to date. Our work demonstrates how to better reflect the interrelated uncertainties that surround the future trajectories of population, income, emissions, climate, and discount rates. Preliminary results suggest that accounting for these uncertainties is likely to increase the value of the SCC considerably.
A summary of the research efforts that have informed this article will be published in the Brookings Papers on Economic Activity series.
The work of Raftery and Ševčíková was supported by NIH grant R01 HD070936 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development. RFF’s work under the Social Cost of Carbon Initiative has been supported by the Alfred P. Sloan Foundation, the Hewlett Foundation, and individual donors.
A version of this article appeared in print in the Fall 2021 issue of Resources magazine.
