Conference Agenda

Estimating the cost from a ton of CO2 to society requires connecting a model of the climate system with representation of the economic and social effect of changes in climate and the valuation and aggregation of diverse, uncertain impacts across both time and space. The literature on this cost, termed the social cost of carbon (SCC) is large and growing, with substantial differences in underlying assumptions both across and within studies. Significant prior work has focused on better constraining parameter values such as climate sensitivity, the discount rate, and the damage function. A growing literature, however, has also examined the effect of varying more fundamental structural elements of the models supporting SCC calculations. These structural model choices - including the introduction of climate or economic tipping points, changing the structure of economic preferences, and the persistence of climate damages - have often been analyzed in piecemeal, uncoordinated fashion, leaving their relative importance unclear. Here we perform a comprehensive synthesis of the evidence on the SCC, combining 1823 estimates of the SCC from 147 studies published between 2000 and 2020 with a survey of the authors of these studies. We find that the distribution in published and expert SCCs are both wide and substantially right-tailed. Survey evidence suggests that experts believe there to be a substantial downward bias in published SCC values. Analysis of the drivers of variance in the distribution reveals that structural variation across SCC models is important, particularly the persistence of climate damages. We estimate a random forest model based on SCC variation in the literature and combine this with expert assessment to generate a 'synthetic SCC' distribution integrating over expert assessments of uncertainty in model structures and the discount rate as well as parametric and residual uncertainty represented in the literature. This distribution has a mean of $467 per tCO2 for a 2020 pulse year (5%–95% range: $14–$1379). There is thus a substantial and varied body of evidence pointing towards using a high SCC in policy-making.

Integrated assessment models (IAMs) play a key role in climate policy analysis, not least for estimating the social cost of CO2 emissions (SCC). The recent literature has documented that structural modelling changes---such as updates to Earth system processes, inclusion of tipping points, consideration of limited substitutability of ecosystem services or natural capital, or allowing for the persistence of damages to economic output---are an important driver of SCC estimates. Yet, most existing IAM applications only assess the effect of single modifications. Here, we build a modular IAM that is able to systematically combine several of these modifications. This allows us to study how different structural modelling choices interact in shaping SCC estimates, and other key aspects of climate policy paths, and to compute a compound ``structural-interactions SCC''. We find that substantial nonlinear interaction effects arise as soon as the model features growth-based damages.

What are the economic benefits of methane action? Using the META model, we estimate the Global Methane Pledge would avoid 0.13°C of warming in 2050 and reduce climate damages by 11% or more than $1trn per year in 2050. The NPV of methane action over the period 2020-2050 is between $3.4-$10.2trn, with a benefit-cost ratio of methane action of more than three. We find methane action to provide larger relative benefits in low-income than high-income countries, and we quantify how methane action reduces climate tail risk, including under tipping points. Lastly, we estimate a SC-CH4 of more than $7,000/tCH$.

Social cost of carbon (SCC) research has paid increasing attention to climate tipping points and feedback mechanisms. The weakening of the Atlantic Meridional Overturning Circulation (AMOC) is currently treated as a global benefit, as it would lower Northern Hemisphere surface temperatures and thereby offset parts of global warming and its associated economic damages. We add to the literature on economic impacts of AMOC weakening by, for the first time, adding a second impact channel which acts through carbon cycle changes. A weaker AMOC directly leads to a reduced export of carbon-rich surface waters to the deep ocean, such that, conversely, more carbon remains in the atmosphere and acts to increase global temperatures and associated economic damages. By drawing on carbon cycle feedback and freshwater hosing experiments, we provide climate modelling evidence on the magnitude of this AMOC-induced carbon feedback, and develop an emulator with which to include these estimates into a simple integrated assessment model. Based on these IAM calculations calibrated to climate modelling results, we find that carbon cycle feedbacks lead to an SCC increase of around 1%, which is in the same order of magnitude as the SCC decrease caused by AMOC-induced temperature changes. Taking into account this carbon effect could thus flip the overall economic effect of AMOC weakening from a net benefit into a net cost.

While financial investors are increasingly alert to the economic threats of climate change, most academic and regulatory assessments of financial risk have not accounted for climate tipping points. Here, we combine recent advances in the integrated assessment modeling of tipping points with dividend discount modeling for major stock indices to assess index-specific risk exposures to climate change damages. We find that for the MSCI World, a globally diversified stock index, tipping points increase the expected loss due to climate change damages under SSP2-4.5 by 48\% (USD 0.2 trillion)---a magnitude comparable to moving from meeting the Paris targets to the "hothouse world" scenario RCP8.5. The reason is that investment horizons are more affected by near-term risks of tipping points than by long-term differences in mitigation outcomes. Risk increases are driven by methane-related tipping points (permafrost thaw and ocean methane hydrates) and ice sheet disintegration, with the highest increases for investments in emerging markets with extensive coastal areas, such as India or Indonesia. The absolute magnitude of financial risks varies substantially across damage functions and assumptions regarding damage persistence. However, the relative importance of tipping points is robust across different damage specifications and investor discount rates. Therefore, our results call for integrating tipping points into climate scenario analyses in the financial sector and climate risk stress tests by regulators.