For the majority of its existence, the social cost of carbon was the kind of figure found in academic papers and regulatory impact assessments; it was meticulously calculated, constantly contested, and mostly unseen by those in charge of actual capital. It provided an answer to a question that was genuinely difficult to put into practice: how much economic damage results from releasing one more tonne of CO2 into the atmosphere? Climate models, damage functions, discount rates, and projections over centuries had to be integrated in order to provide the answer, and the final figure varied so significantly depending on the assumptions made that its detractors could always find an excuse to reject it. For a long time, the argument over the appropriate number was a handy way to avoid using it.
That is starting to shift. Not consistently, and not without opposition, but in ways that aren’t as prevalent as they were five years ago in corporate disclosure regulations, portfolio construction, and climate litigation tactics. In a paper published in Nature in March 2026, a group of Stanford researchers advanced the discussion beyond its previous level by developing a framework that maps the social cost of carbon onto individual nations and emitters, essentially creating a ledger of climate debt with real names in the columns.
The study’s numbers are startling. By 2020, the discounted global economic damage from one tonne of CO2 emissions in 1990 was about $180. Between now and 2100, that same tonne will result in an additional $1,840 in damage—a tenfold multiple—due to the way warming lowers growth rates as well as economic levels. This means that the harm compounds rather than hits once and fades. Using a two percent discount rate and conservative assumptions, the researchers’ headline figure for the social cost of carbon was $1,013 per tonne. That is at least five times greater than the estimate used by the federal government of the United States prior to the metric’s widespread abandonment. Under the current administration, the United States has also withdrawn from the Paris Agreement, withdrawn from the Loss and Damage Fund established at COP27, and left both the UNFCCC and the IPCC. These actions, when viewed in the context of the Stanford study’s figures, take on a certain historical significance.
IMPORTANT INFORMATION TABLE — SOCIAL COST OF CARBON (SCC) & CLIMATE FINANCE
| Category | Details |
|---|---|
| Concept | Social Cost of Carbon (SCC) = the monetized economic damage caused by emitting one additional tonne of CO₂ |
| EPA Updated Estimate (2023) | ~$255 per tonne in 2025, rising to ~$370 per tonne by 2050 |
| Stanford/Nature Study (March 2026) | SCC of $1,013 per tonne under conservative assumptions (2% discount rate, impacts through 2100) — at least 5× higher than recent US federal estimates |
| Lead Researchers | Marshall Burke, Mustafa Zahid, Noah S. Diffenbaugh, Solomon Hsiang — Stanford University Doerr School of Sustainability |
| Key Finding | One tonne of CO₂ emitted in 1990 caused ~$180 in discounted global damages by 2020, but will cause an additional $1,840 through 2100 — a tenfold gap |
| Flight Example | A single long-haul flight annually for a decade causes ~$25,000 ($6,000–$77,000) in future damages by 2100 |
| US Total Damages Caused | ~$10.2 trillion in cumulative global economic harm (1990–2020) — single largest national emitter by damage |
| China Total Damages | ~$8.7 trillion cumulative global damage (1990–2020) |
| US Damage to India | ~$500 billion ($180–$1,300 billion uncertainty range) from US emissions since 1990 |
| US Damage to Brazil | ~$330 billion ($110–$820 billion uncertainty range) from US emissions since 1990 |
| Climate Finance (2023) | Total climate finance flows reached $1.9 trillion |
| Wall Street Application | Asset managers using SCC to price “shadow cost” of carbon in portfolios, stress-test climate risks, and evaluate stranded asset exposure |
| Discount Rate Controversy | A 1% vs. 5% discount rate changes the entire calculation — lower rates treat future suffering as equivalent to present suffering |
| Current US Federal Policy | Trump administration dramatically reduced the US SCC estimate and withdrew from the Loss and Damage Fund, Paris Agreement, UNFCCC, and IPCC |
| Loss & Damage Fund | Agreed at COP27; remains largely unfunded despite formal establishment |
If one is paying close attention, the paper’s bilateral damage accounting is where things really get awkward. According to the study, U.S. emissions since 1990 have damaged Brazil’s economy by about $330 billion and India’s economy by about $500 billion; even the lower bound is a significant amount. Over the course of the period, U.S. emissions were responsible for approximately $10.2 trillion in global damage, making them the largest national source of quantifiable climate harm. China came in second with $8.7 trillion. Third place went to the European Union. Large emitters were not the nations most negatively impacted relative to their own GDP. These countries were low-income and tropical, and they were also the least accountable for the damaging emissions.
Marshall Burke and his associates are cautious about the legal ramifications of this. The study makes it clear that damage estimates do not establish what is owed between parties; that is a moral and legal issue outside the purview of the analysis, in accordance with Article 8 of the Paris Agreement. However, because climate litigation has been growing quickly and legal claims need some quantifiable basis to proceed, it is important to make the accounting precise. For precisely this type of harm, major fossil fuel companies have been facing lawsuits from cities and states all over the United States as well as plaintiffs in several foreign jurisdictions. Although the Stanford framework does not establish liability, it does establish a methodology for determining what a liability might look like. Once this methodology is published in a peer-reviewed journal of the caliber of Nature, it becomes part of the evidentiary landscape in ways that are difficult to reverse through lobbying.
In financial circles, there is a perception that the social cost of carbon is gradually but significantly rising in prominence, going from being a policy input utilized by regulators to something that asset managers are actively integrating into their own risk models. The reasoning is simple. If a company has significant fossil fuel reserves that are implicitly valued at the current market assumption, it will also have a liability if those assets become stranded as global carbon pricing tightens. Acknowledging that some of the earnings projections currently included in energy sector valuations are based on the implicit assumption that emissions will remain effectively free—which may or may not be true over a pension fund’s twenty-year investment horizon—is accomplished by pricing the shadow cost of carbon into a portfolio. With differing degrees of conviction, investors appear to think that some form of carbon pricing is on the horizon and that timing rather than direction is the key.
The discount rate controversy deserves more public attention than it typically receives because it is doing enormous work inside these calculations. Whether you use one percent or five percent to translate future damages back into present value reshapes the entire cost-benefit calculus of climate action. A high discount rate makes future suffering look cheap from today’s vantage point, reducing the apparent urgency of action; a low rate treats the wellbeing of people who haven’t been born yet as roughly comparable to the wellbeing of people alive today. The Stanford team presents their results across a range of discount rates precisely because this choice cannot be settled by science — it is, ultimately, a question about how much we weight the future, which is an ethical and political determination wrapped in the language of economics.
It’s hard not to notice the gap between where this analysis now stands and where official policy sits in some of the world’s largest economies. The U.S. federal government has not merely declined to adopt the higher SCC estimates — it has actively retreated from the frameworks that would use any SCC at all. The adoption of these figures in domestic policy may be slowed by this retreat, but the numbers themselves remain unchanged. The debt that Burke and his associates are figuring out is not a product of policy. According to their framework, it is an accounting of harm that has already been done, harm that keeps getting worse even if no one in Washington is currently entering it into a spreadsheet.
