Quantifying climate loss and damage consistent with a social cost of carbon
Article Date: 25 March 2026
Article URL: https://www.nature.com/articles/s41586-026-10272-6
Article Image: (no image provided)
Summary
This Nature paper presents a formal, quantitative framework to calculate loss and damage (L&D) from CO2 emissions that is aligned with the social cost of carbon (SC-CO2). The authors decompose total L&D into: (1) historical damages already realised from past emissions (HD-CO2); (2) future damages that will result from past emissions (FD-CO2); and (3) future damages from present/future emissions (SC-CO2). They combine emissions inventories, the FaIR reduced-complexity climate model, CMIP6 pattern scaling and an econometric temperature–GDP growth relationship (with lagged effects) to translate emissions into local economic impacts and present discounted damages. The estimates show past emissions have already caused very large economic losses and that much larger damages are still expected in future unless emissions are removed or warming is avoided.
Key Points
- The authors define L&D as the net present value of economic and non-economic impacts attributable to greenhouse-gas emissions, net of adaptation.
- They compute three components: HD-CO2 (discounted historical damages), FD-CO2 (future damages from past emissions) and SC-CO2 (future damages from a marginal present emission).
- Method combines emissions inventories, FaIR to get global mean temperature change, CMIP6 pattern scaling to get local changes, and an empirically estimated temperature–GDP growth function with lags.
- Under a conservative 2% discount rate and impacts ending 2100, the SC-CO2 is estimated at about $1,013 per tonne — substantially higher than many recent bottom-up estimates.
- FD-CO2 typically far exceeds HD-CO2: e.g. a 1 Gt CO2 pulse in 1990 causes ~$184 per tonne in discounted damages by 2020 but ~$1,840 per tonne for 2021–2100 (2% discount rate).
- Attribution examples: one extra long‑haul flight per year for 2010–2020 → ~$165 in past damages and ~$25,000 in discounted future damages; several celebrities’ private flights in 2022 each produce >$1M in future discounted damages.
- Firm and country results are large: Saudi Aramco’s 1988–2015 emissions → ~$3 trillion cumulative damages by 2020 and ~$64 trillion through 2100; the US emissions 1990–2020 → ~$10.2 trillion in cumulative damages by 2020.
- Results are sensitive to discount rates, time horizon, damage function choices and adaptation assumptions; many non‑GDP channels (health, ecosystems, sea-level rise, extreme events) are not fully captured and would raise total damages.
Content summary
The paper formalises L&D accounting so it can be computed consistently with SC-CO2 methods. It treats an emitted unit of CO2 as creating a stream of (often negative) value for populations worldwide. The authors present an implementable pipeline: emissions → FaIR → pattern scaling (CMIP6) → country-level temperature changes → temperature–GDP growth regression (including lags) → discounted damage sums (HD-CO2, FD-CO2, SC-CO2).
Key empirical findings: earlier emissions generate larger cumulative damages because of longer exposure and compounding growth effects. Future damages from past emissions are often an order of magnitude (or more) greater than damages already realised. The spatial pattern shows mid-latitude and tropical countries suffer the largest relative economic harm; large economies register the largest absolute losses.
The paper quantifies damages at multiple scales: individual behaviours (flights, diet, driving), firm-level (carbon majors’ scope 1+3 emissions), and bilateral country-to-country attribution. It discusses mitigation options including direct monetary compensation, carbon dioxide removal (CDR), solar radiation management and adaptation, emphasising that timing and feasibility matter (CDR effectiveness declines with delay).
Context and relevance
This work bridges climate attribution, econometric damage estimation and policy-relevant accounting. By aligning L&D with SC-CO2 frameworks the paper provides tools usable in policy, litigation and international finance discussions. The high estimated SC-CO2 and massive firm/country attributions underscore the economic scale of historical responsibility debates and strengthen the scientific basis for considering financial transfers, liability, or large-scale mitigation/removal investments.
Why should I read this?
Quick and blunt: if you care about climate policy, justice, litigation risk, corporate exposure or the numbers behind “who owes what”, this paper matters. It gives a repeatable way to translate emissions into dollarised harms (and shows those harms are huge). Read it to get the methods, the headline estimates and the caveats — it’ll save you time and give you figures to cite in policy or risk analyses.
Limitations & caveats
- The primary channel is aggregate GDP growth; many harms (mortality, cultural loss, ecosystems, sea-level impacts, some extremes) are under‑represented, so total damages are likely understated.
- Estimates depend strongly on discount rate, damage horizon and modelling choices; legal and ethical questions about who owes what are beyond the quantitative framework.
- CDR, SRM and adaptation are discussed but their feasibility, scale and distributional effects remain unresolved.