A near-term to net zero alternative to the social cost of carbon for setting carbon prices

Nature Climate Change (2020), by Noah KaufmanAlexander R. BarronWojciech KrawczykPeter Marsters & Haewon McJeon Cite this article

Abstract: The social cost of carbon (SCC) is commonly described and used as the optimal CO2 price. However, the wide range of SCC estimates provides limited practical assistance to policymakers setting specific CO2 prices. Here we describe an alternate near-term to net zero (NT2NZ) approach, estimating CO2 prices needed in the near term for consistency with a net-zero CO2 emissions target. This approach dovetails with the emissions-target-focused approach that frames climate policy discussions around the world, avoids uncertainties in estimates of climate damages and long-term decarbonization costs, offers transparency about sensitivities and enables the consideration of CO2 prices alongside a portfolio of policies. We estimate illustrative NT2NZ CO2 prices for the United States; for a 2050 net-zero CO2 emission target, prices are US$34 to US$64 per metric ton in 2025 and US$77 to US$124 in 2030. These results are most influenced by assumptions about complementary policies and oil prices.

**

Economists overwhelmingly support pricing CO2 emissions1. How much to charge for each ton of emissions is perhaps the most important element of a carbon pricing policy, yet little consensus exists among economists about the appropriate level for CO2 prices2.

To find optimal CO2 prices, economists have long focused on a metric called the social cost of carbon (SCC), an estimate of the marginal damages of an additional ton of CO2 emissions. However, the SCC cannot be credibly estimated with sufficient precision to provide practical assistance to policymakers setting CO2 prices. The SCC approach is also disconnected from real-world policy discussions that position CO2 prices as one element of a strategy to avoid the risks of exceeding thresholds of global warming.

In the face of these constraints, this paper introduces an alternative approach. It starts with policymakers selecting a net-zero CO2 emissions target informed by the best available science and economics. Then, near-term to net zero (NT2NZ) CO2 prices are combined with a broader policy strategy to achieve an emissions pathway consistent with the net-zero target in the near term, when the projections of energy–economic models are most useful.

Helping policymakers set CO2 prices

In textbooks, optimal CO2 prices are identified with perfect precision. Net benefits to society are largest if the government taxes an activity that creates a negative externality (such as CO2 emissions) at the rate that equalizes the marginal social benefits and marginal social costs of emissions reductions3.

Economists have long recognized that, in the real world, approaches for developing optimal policies can be constrained by various uncertainties and measurement difficulties4,5, including imprecision, ambiguity, intractability and indeterminacy6. Therefore, in addition to maximizing net benefits, analysts should strive to identify CO2 prices using approaches with the following attributes:

  • Credible precision. The approach produces a range of CO2 prices that provides policymakers with valuable information to incorporate into policy design decisions. For example, if adding complexity to a model injects disproportionate uncertainty, it can be more informative to use a simpler framework7.
  • Transparency. The approach enables policymakers to understand the causes of relatively high or low estimates within the range of CO2 prices identified.
  • Consistency with policy objectives. The approach produces CO2 prices that fit the objectives of the policy. The most common rationale for a CO2 price is to reduce emissions as part of a broader global response to the risks of climate change.

Policymakers setting CO2 prices are also concerned with additional factors that are outside our scope, such as public health benefits from reducing air pollution, energy security, competitiveness, the expected actions of other jurisdictions and (perhaps most importantly) political viability.

Challenges with setting CO2 prices using SCC estimates

The SCC, commonly used as the optimal CO2 price that maximizes net benefits to society8, is estimated with projections of the following parameters: global emissions over the next few centuries, the effects of emissions on temperatures and other climate impacts, and the impacts of climate change on the economy and human welfare, using economic methods that aggregate centuries of impacts into a single value representing the net benefits of emissions reductions9.

Unfortunately, the degree of uncertainty in SCC estimates spans virtually any conceivable stringency level for a CO2 pricing policy. Meta-analyses find recent SCC estimates that range from under US$0 per ton of CO2 to over US$2,000 per ton (excluding outliers still leaves a range of hundreds of dollars)8,10.

SCC estimates will continue to improve9,11,12, but methodological advancements are unlikely to narrow the range of SCC estimates much. After all, large uncertainties come from parameters that are inherently uncertain, such as the appropriate discount rates9, risk aversion levels13, issues around inequality14 and attempts to assign monetary values to non-economic climate damages15. In addition, methodological improvements often involve incorporating new uncertain elements that were omitted from previous estimates, which can widen the range of SCC estimates16,17.

The advantage of the SCC approach is that it attempts to perfectly maximize net benefits to society; however, the SCC approach provides limited practical value to policymakers setting specific CO2 prices due to its difficulty satisfying the desired attributes identified in the previous section. First, the range of SCC estimates is too wide to credibly support the use of any single CO2 price. (In contrast, the US government developed SCC estimates for use in the separate context of benefit–cost analysis, where a wide range can be incorporated9.) Second, the differences in SCC estimates hinge partly on assumptions that are not usually transparent to policymakers (such as the value placed on future generations by discount rates). Third, the SCC approach is disconnected from constraining global warming beyond specific levels—the goal of most policymakers setting CO2 prices. For example, William Nordhaus’s 2018 Nobel Prize Lecture shows an optimal pathway of 4 °C of warming by the mid-2100s as the implication of his SCC estimates18, an outcome far outside the bounds of the Paris Agreement’s goals (which other analyses have found would pass a cost–benefit test19). Helping policymakers set CO2 prices in practice therefore necessitates an approach that balances benefits and costs only imperfectly, such as the NT2NZ approach described in the next section.

The classic alternate approach in the face of this sort of uncertainty is a cap-and-trade programme, which involves mandating emissions targets. We assume that policymakers have other reasons for selecting a price instrument, such as the interactions with complementary policies, a desire for certainty in business planning, concerns about market manipulation or political economy considerations20.

The NT2NZ approach

NT2NZ CO2 prices are designed to accommodate uncertainties and measurement difficulties and to align with real-world policy objectives. They are estimated using the following four steps:

Step 1: select a net-zero CO2 emissions date

While the international climate change negotiations have focused primarily on temperature targets, policymakers are increasingly shifting to net-zero emissions targets for both substantive and political reasons21,22,23,24. A global net-zero CO2 emissions target has a science-based rationale: surface temperatures will continue to increase until the sources and sequestration of CO2 are equal, at which point temperatures will roughly stabilize25,26 (reductions in non-CO2 GHGs and land-use change emissions are required for full stabilization24, but we focus on CO2 emissions, which are the bulk of what would be covered by a CO2 price). Views differ on what threshold of global warming should not be exceeded, but failing to achieve net-zero CO2 emissions implies ever-increasing temperatures, which will eventually reach unacceptably high levels regardless of one’s threshold. Net-zero targets also naturally scale to any jurisdiction, because achieving global net zero requires, on average, all jurisdictions to achieve net zero.

Policymakers must balance a range of factors to set net-zero targets, including the risks of even-higher temperature changes and the additional costs of decarbonizing faster. International climate agreements recognize that countries have the responsibility to decarbonize at different paces27, which means that jurisdictions will set different net-zero target dates. Like estimating an SCC, setting a net-zero target involves judgements about concern for future generations, willingness to tolerate risks and aversion to inequality (among other factors). Under the NT2NZ approach, these trade-offs are made by the governing officials selecting the target.

Step 2: select an emissions pathway to the net-zero target

An infinite number of pathways are conceivable between current emissions levels and a future net-zero target. Some frameworks emphasize the benefits of reducing near-term disruption and enabling innovation to bring down technology costs, which argue for a slower initial rate of reductions28. Other frameworks emphasize the importance of near-term deployments in reducing costs and avoiding technology lock-in29,30,31, as well as benefits from reducing cumulative emissions and respecting intergenerational equity.

Policymakers, weighing these considerations as well as technical and political constraints, may choose a straight-line pathway to net zero for simplicity and transparency, or a different trajectory that fits the circumstances of the jurisdiction (such as a developing country with a peak-and-decline pathway)32.

Step 3: estimate CO2 prices consistent with the emissions pathway in the near term

Energy–economic models can be used to estimate the CO2 prices required to reduce emissions on a desired pathway under a given set of assumptions about future technologies, prices and behaviour33. Unlike the SCC approach, energy–economic models enable analysts to combine CO2 prices with other policy measures to overcome multiple market barriers to emissions reductions.

While net-zero emissions is the long-term goal, the NT2NZ approach focuses on the near term (the next decade, for example). Models that simulate economic and energy systems are built using historical data on production, consumption and market dynamics, which may be a reasonable assumption in the near term. After all, most energy technologies and consumer behaviours evolve relatively slowly. But such models become less useful as the time horizon of the exercise lengthens34. Changes in technologies, preferences and policies will inevitably impact energy systems in unexpected ways in a rapidly decarbonizing country—just as advancements in solar energy technologies and the shale revolution in the United States were almost entirely unforeseen decades ago35.

Focusing on the near term means that CO2 price estimates should not be unduly influenced by assumptions about the highly uncertain long-term evolution of technologies and behaviour. Analysts often use models with ‘foresight’, which means that decisions are contingent on assumed future changes to the energy system. For example, if the costs of breakthrough low-carbon technologies are assumed to remain prohibitively high, models with foresight may suggest that the cost-effective pathway involves higher near-term CO2 prices, so that additional emissions reductions can be achieved in sectors where stock turnover is slow. In contrast, if the costs of breakthrough technologies are assumed to fall precipitously, the same model may suggest a cost-effective pathway that involves relatively less near-term mitigation (see Supplementary Information, Appendix 2).

Step 4: periodically update Steps 1–3

Knowledge about climate science and the costs of mitigation technologies will continue to change rapidly, especially given the substantial social and economic shifts associated with decarbonization. This calls for an adaptive management strategy36 whereby the analysis described above is repeated periodically. Using emissions outcomes and other relevant metrics (such as deployment rates of low-carbon infrastructure and progress in hard-to-decarbonize sectors30), both the CO2 price pathway and the broader climate policy strategy can be revised and extended, capturing the most up-to-date information.

Adaptive management can enable jurisdictions to stay close to the desired emissions pathway without making policy details contingent on assumptions about highly uncertain long-term variables. Various mechanisms for periodically updating policies have been proposed in recent years37, including CO2 prices that are contingent on emissions outcomes (that is, if emissions exceed target levels, price increases accelerate in future years).

Illustrative NT2NZ CO2 prices for the United States

We demonstrate the approach described above to produce illustrative NT2NZ CO2 prices for the United States. For this analysis we use the 50-state version of the Global Change Assessment Model (GCAM-USA), an integrated assessment model of energy–economy–environment systems, and data available as of 2019 (Methods). Real-world policy design should be informed by multiple analytic tools using the most up-to-date available information32.

We begin with three straight-line annual emissions pathways from current (2020) levels to net-zero CO2 emissions targets in 2060, 2050 and 2040 (Fig. 1a) to reflect a range of emissions pathways discussed in recent years by US policymakers. In the absence of a consensus favouring larger or smaller near-term emissions reductions38,39, an illustrative straight-line emissions pathway for the United States may be appealing due to its simplicity and transparency. These pathways correspond to 2030 CO2 emissions of 35%, 42% and 57% below 2005 levels, respectively. Consistent with both economic theory and policy practice, we assume that the CO2 price is surrounded by complementary policies that address separate market failures40: efficiency policies, air pollution regulations and early-stage support for the deployment of low-carbon technologies (such as electric vehicles).

figure1
Fig. 1: US CO2 emissions pathways to net zero and associated NT2NZ CO2 prices.

For our benchmark scenario, we find NT2NZ CO2 prices in 2025 of US$32, US$52 and US$93 per metric ton (in 2018 dollars) for consistency with net-zero targets in 2060, 2050 and 2040, respectively. The NT2NZ CO2 prices in 2030 are roughly two times larger (Fig. 1b), reflecting a much higher annual growth rate than typical CO2 price estimates based on the SCC or rising at the rate of interest (see Supplementary Information, Appendix 2).

For each emissions pathway, we show a range of NT2NZ CO2 prices from sensitivity scenarios that intend to capture uncertainty in influential model inputs (Methods). Figure 2 shows that compared with our benchmark scenario (with a 2050 target), more stringent and successful complementary policies (that is, air quality regulations that lead to higher coal retirements, more aggressive energy efficiency measures and more aggressive early-stage deployment support for certain low-carbon technologies) lowered the CO2 prices by US$10–US$20 per ton. The prices rise by about the same amount with less aggressive complementary policies. Changing the future oil price trajectory from a pathway to either US$45 or US$176 per barrel in 2030 leads to a swing in CO2 prices of US$40 per ton in 2030. Figure 2 also displays the impacts on NT2NZ CO2 prices from changing inputs related to natural gas prices, innovation and economic growth. Appendix 1 in the Supplementary Information provides additional results.

figure2
Fig. 2: US NT2NZ CO2 prices for a net zero by 2050 pathway: comparison of sensitivity scenarios to the benchmark scenario.

For a 2050 target, the range of NT2NZ CO2 prices is largely consistent with the range of CO2 prices in legislation proposed to the US Congress in 2019 (Fig. 1). However, the prices are on the lower end of the range of CO2 prices that global energy–economic models have identified as consistent with constraining average global temperature increases to 1.5 and 2 degrees Celsius41 for at least three reasons. First, NT2NZ CO2 prices will differ by jurisdiction, and the United States has a large amount of coal-fired electricity generation that can be replaced at a relatively low cost. Second, while many studies assume that the CO2 price is implemented without other policies, we assume that multiple policies are implemented to address multiple market barriers to emissions reductions. Third, the actors within GCAM do not have foresight, so their energy consumption is based on their myopic vision of current market conditions and not based on long-term projections of technological progress (see Supplementary Information, Appendix 2 and Extended Data Fig. 1).

Discussion and policy implications

Economists have long referred to the SCC as the optimal CO2 price. The use of the SCC as a CO2 price has become commonplace in recent years, including in federal carbon tax proposals and state-level subsidies for clean electricity generation in the United States (see Supplementary Information, Appendix 3). Using the NT2NZ approach instead offers several advantages.

First, CO2 prices can be estimated with more precision. All of the uncertainties in the CO2 prices estimated using the NT2NZ approach—such as near-term clean energy innovation and fossil fuel prices—are also uncertainties using the SCC approach. But the NT2NZ approach avoids much larger uncertainties, including assigning monetary values to climate change damages. The NT2NZ approach focuses on the most important and (relatively) better understood aspects of the problem.

Second, while any approach can be transparent, the NT2NZ approach focuses on changes to the energy system in the near future, which should enable policymakers to better understand the rationale for selecting CO2 prices of varying levels. An NT2NZ analysis enables external stakeholders to assess whether a jurisdiction’s policies are consistent with its targets and how the outcomes are influenced by key assumptions about technology costs and energy prices (Fig. 2). While the full range of our estimates is ~US$40 per ton for any given target, policymakers are not likely to regard all sensitivities as equally likely and can choose within that range accordingly. By comparison, the complexity of modelling the SCC can make it more difficult to communicate why estimates differ (often by hundreds of dollars per ton under different plausible assumptions about discounting, adaptation and so on).

Third, the NT2NZ approach is more consistent with the objectives of most policymakers implementing CO2 prices. While economists have traditionally recommended a harmonized global CO2 price to maximize efficiency, national annual emissions targets have long been the lingua franca of international negotiations on climate change. The illustrative example in the previous section shows how US policymakers could use the NT2NZ approach to help design a federal carbon price as an element of its 2030 Nationally Determined Contribution to the Paris Agreement. Other countries, groups of countries or subnational jurisdictions could do the same. The NT2NZ approach also enables the CO2 price to be one piece of a broader policy strategy to address multiple market failures around climate, which better aligns with economic theory compared with a CO2-price-alone approach40. It also better aligns with the real-world practice of combining CO2 prices with a range of other (often sectoral)42 policies, recognizing that preparation for deep decarbonization31 requires adaptive management using an ecosystem of policies. For example, passenger vehicle decarbonization might include goals and policy measures to encourage electric vehicles, charging infrastructure, fuel economy and modal shifts away from single-occupancy vehicle travel43.

Pairing a long-term emissions target with a set of iterative near-term policies is not novel. The United Kingdom, for example, has adopted a national target of net-zero GHG emissions by 2050 and sets five-year carbon budgets to act as stepping-stones44. Indeed, the Paris Agreement encourages this framework by calling on nations to produce long-term development strategies for low GHG emissions and near-term Nationally Determined Contributions that are updated every five years26.

The NT2NZ approach does not attempt to set CO2 prices by perfectly balancing costs and benefits, so it does not satisfy the definition of an ‘optimal CO2 price’ found in economics textbooks. Instead, it enables policymakers to consider both qualitative and quantitative information about climate science and economics when selecting a net-zero target. Models can then estimate the CO2 prices for cost-effective reductions in the near term, when their projections are most useful, with the understanding that policy details can be updated as uncertainties are resolved.

We describe our results as illustrative because no single model should inform policy setting. The NT2NZ approach should be implemented across a suite of models to further characterize uncertainties, identify results that are robust across methods34 and explore how prices are best combined with other policies. Future work should also examine the roles of the domestic land-use sink and non-CO2 GHG emissions.

Methods

Overview of the model

To numerically estimate CO2 prices using the NT2NZ approach, we use GCAM-USA. GCAM is an economy-wide global integrated assessment model representing the energy and land sectors linked with a climate model; GCAM is used to explore the interactions of emissions-reducing investments and activities across the US and global economies. We begin with the same version of the model and the standard assumption set used for the US Mid-century Strategy (MCS)45. The technical assumptions used in the US MCS are documented in the MCS technical appendix46. We have also made updates and changes reflecting the developments in the energy markets and policies as of late 2019, when this analysis was conducted (before the COVID-19 crisis). These are described in detail below. More detail on GCAM can be found in the Supplementary Information.

Key updated assumptions

Most of the assumptions follow the benchmark scenario described in the MCS technical appendix. We updated key assumptions to align the model with the changing market dynamics in the past several years, using the sources in Supplementary Table 3. Given the uncertain and influential nature of these assumptions, our sensitivity scenarios are designed by selecting low, benchmark and high assumptions for each.

Innovation and low-carbon energy technologies

To reflect the rapidly falling costs of renewable energy, we have updated the renewable energy costs to those from the National Renewable Energy Laboratory’s (NREL) Annual Technology Baseline47. To reflect uncertainties in the evolution of future energy markets as well as concerted research, development and deployment policies to further reduce low-carbon energy costs, we conduct a sensitivity analysis using the high, medium and low estimates of solar, wind and nuclear costs. The detailed cost assumptions are summarized in Supplementary Table 4.

Oil and gas prices

To reflect future uncertainty in oil and gas prices, we have updated the oil and gas price trajectories consistent with the Energy Information Administration’s (EIA) oil and gas price trajectories from Annual Energy Outlook 2019 (ref. 48). Each five-year price point in GCAM is based on a three-year running average of prices to avoid short-term fluctuations. We conduct a sensitivity analysis using the high, medium and low estimates of the oil and gas price trajectories. The detailed assumptions are summarized in Supplementary Table 5. We note that oil prices have fallen considerably since our analysis was performed, and we probably will not know how projections for 2025 or 2030 will be influenced until after the immediate COVID-19 crisis subsides.

Population and GDP

We conducted a sensitivity analysis using high, medium and low estimates of population and GDP. Medium population projections are based on US Census Bureau National Population Projections Tables49. The high population growth scenario assumes 0.1% per year additional growth, and the low population growth scenario assumes 0.1% per year less growth. The GDP variations are based on Annual Energy Outlook 2019, which provides high, medium and low variants of GDP growth projections. The COVID-19 pandemic will considerably depress GDP growth in the near term and has the potential to do so through 2025—future analysis with updated projections from the EIA (or another source) should take this effect into account. The detailed population and GDP data used are tabulated in Supplementary Table 6.

Complementary policies

We developed three complementary policy scenarios (detailed below), both to reflect policies needed to overcome market barriers left unaddressed by CO2 prices and to reflect the uncertain future of sectoral policies that influence CO2 emissions.

Coal retirements

Coal-fired power plants release not only CO2 emissions but also various other air pollutants, such as particulate matter and ozone. Policymakers have long recognized the need for regulations to protect constituents from the harmful impacts of these pollutants. Coal-fired power plants have been rapidly retiring in the United States due to changing market dynamics and concerted efforts to reduce emissions of CO2 and other air pollutants. To capture the potential impact of future environmental regulations on coal generation, we developed high, medium and low coal-retirement scenarios. These coal-retirement pathways assume no CO2 emission mitigation policy; they thus serve as a starting point for our CO2 price scenarios. The low trajectory tracks the EIA Annual Energy Outlook 2018 coal-retirement reference case projections48. The medium (benchmark) trajectory roughly tracks the US Environmental Protection Agency Integrated Planning Model May 2019 reference case50 (which also assumes no new policy). The high trajectory tracks the Integrated Planning Model reference case to 2021 and then assumes a trajectory that is consistent with the Enhanced Engagement scenario in the America’s Pledge Report51 (this post-2021 rate is consistent with a capacity retirement rate between the medium and average retirement rates since 2012). The impact of these coal-retirement trajectories on generation is shown in Table M5. In addition to scheduled retirements, GCAM allows the power plants to prematurely retire when they are no longer profitable in the market. Sufficiently high CO2 prices can therefore force coal power plants out of the market (Supplementary Fig. 3).

Energy demand

A CO2 price alone is often insufficient to encourage consumers to take advantage of all cost-effective opportunities to reduce energy usage due to market barriers including informational failures and consumer short-sightedness. To reflect uncertainty in the future growth of energy demand, including policies to reduce the rate of energy demand growth, we developed high, medium and low scenarios for future demand growth in key energy sectors using assumptions from the Smart Growth scenario of the MCS report45. Specifically, our low and medium energy demand scenarios assume less vehicle travel and greater efficiency in the buildings sector than does the Benchmark scenario. The detailed energy demand differences are shown in Supplementary Table 7.

Early-stage deployment support for low-carbon technologies

In addition to the lack of a CO2 price, numerous market barriers stand in the way of a large-scale shift to products that do not directly burn fossil fuels (such as electric vehicles and electric heat pumps). To reflect the highly uncertain future of electrification, including policies designed to support electrification, we develop high, medium and low sensitivity scenarios for electrification in buildings and transportation, drawn from the NREL Electrification Futures Study52. The electrification rates are shown in Supplementary Table 7.

Land-sector sink assumptions

We develop three net-zero emissions targets (2040, 2050 and 2060) (Fig. 1). In our analysis, net CO2 emissions are constrained to linearly decrease to zero in the specified target years. In 2030, this formulation corresponds to an emissions range of 3,400 MtCO2 to 2,200 MtCO2 (35–57% below the 2005 emission level). For simplicity, we assume the CO2 emissions from land-use change are constant at current levels: 714 MtCO2 (ref. 53). The US MCS has considered multiple possible pathways for CO2 emissions from land-use change, but given the large fluctuations in year-by-year estimates for land-use change emissions, we did not find other estimates to be of additional value that warrants the additional complexity. A decrease in the land sink over time would require additional emissions reductions from fossil fuel combustion beyond what we model here.

Reporting Summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability

All data generated or analysed during this study are from publicly available sources and are either included in this article (and its Supplementary Information files) or available from the corresponding author on reasonable request. Source data are provided with this paper.

References

  1. 1.Akerloff, G., Auman, R., Baily, M. & Bernanke, B. Economists’ statement on carbon dividends. Wall Street Journal (16 January, 2019).
  2. 2.Espagne, E., Pottier, A., Fabert, B. P., Nadaud, F. & Dumas, P. SCCs and the use of IAMs: let’s separate the wheat from the chaff. Int. Econ. 155, 29–47 (2018).Article Google Scholar 
  3. 3.Pigou, A. C. The Economics of Welfare (Macmillan, 1932).
  4. 4.Baumol, W. J. On taxation and the control of externalities. Am. Econ. Rev. 62, 307–322 (1972).Google Scholar 
  5. 5.Koomey, J. Moving beyond benefit–cost analysis of climate change. Environ. Res. Lett. 8, 041005 (2013).Article Google Scholar 
  6. 6.Smith, L. A. & Stern, N. Uncertainty in science and its role in climate policy. Phil. Trans. R. Soc. A 369, 4818–4841 (2011).Article Google Scholar 
  7. 7.May, R. M. Uses and abuses of mathematics in biology. Science 303, 790–793 (2004).CAS Article Google Scholar 
  8. 8.Tol, R. S. J. The economic impacts of climate change. Rev. Environ. Econ. Policy 12, 4–25 (2018).Article Google Scholar 
  9. 9.Valuing Climate Damages: Updating Estimation of the Social Cost of Carbon Dioxide (National Academy of Sciences, Engineering, and Medicine, 2017); https://doi.org/10.17226/24651
  10. 10.Wang, P., Deng, X., Zhou, H. & Yu, S. Estimates of the social cost of carbon: a review based on meta-analysis. J. Clean. Prod. 209, 1494–1507 (2019).Article Google Scholar 
  11. 11.The Social Cost of Carbon Initiative: Resources for the Future (RFF, 2020); https://www.rff.org/social-cost-carbon-initiative/
  12. 12.Social Cost of Carbon. Climate Impact Lab http://www.impactlab.org/research-area/social-cost/ (2020).
  13. 13.Kaufman, N. The bias of integrated assessment models that ignore climate catastrophes. Clim. Change 110, 575–595 (2012).Article Google Scholar 
  14. 14.Dennig, F., Budolfson, M. B., Fleurbaey, M., Siebert, A. & Socolow, R. H. Inequality, climate impacts on the future poor, and carbon prices. Proc. Natl Acad. Sci. USA 112, 15827–15832 (2015).CAS Article Google Scholar 
  15. 15.Pindyck, R. S. Climate change policy: what do the models tell us? J. Econ. Lit. 51, 860–872 (2013).Article Google Scholar 
  16. 16.Ricke, K., Drouet, L., Caldeira, K. & Tavoni, M. Country-level social cost of carbon. Nat. Clim. Change 8, 895–900 (2018).CAS Article Google Scholar 
  17. 17.Daniel, K. D., Litterman, R. B. & Wagner, G. Declining CO2 price paths. Proc. Natl Acad. Sci. USA 116, 20886–20891 (2019).CAS Article Google Scholar 
  18. 18.Nordhaus, W. Climate change: the ultimate challenge for economics. Am. Econ. Rev. 109, 1991–2014 (2019).Article Google Scholar 
  19. 19.Glanemann, N., Willner, S. N. & Levermann, A. Paris Climate Agreement passes the cost–benefit test. Nat. Commun. 11, 110 (2020).CAS Article Google Scholar 
  20. 20.Metcalf, G. E. Paying for Pollution: Why a Carbon Tax Is Good for America (Oxford Univ. Press, 2019).
  21. 21.Geden, O. An actionable climate target. Nat. Geosci. 9, 340–342 (2016).CAS Article Google Scholar 
  22. 22.Fay, M. et al. (eds) in Decarbonizing Development: Three Steps to a Zero-Carbon Future 55–76 (World Bank, 2015); https://doi.org/10.1596/978-1-4648-0479-3_ch3
  23. 23.Getting to Net-Zero Emissions: Lessons from Latin America and the Caribbean (Inter American Development Bank & Deep Decarbonization Pathways for Latin America and the Caribbean, 2019); https://doi.org/10.18235/0002024
  24. 24.Rogelj, J. et al. Zero emission targets as long-term global goals for climate protection. Environ. Res. Lett. 10, 105007 (2015).Article Google Scholar 
  25. 25.Matthews, H. D. & Caldeira, K. Stabilizing climate requires near-zero emissions. Geophys. Res. Lett. 35, L04705 (2008).Article Google Scholar 
  26. 26.Zickfeld, K., MacDougall, A. H. & Matthews, H. D. On the proportionality between global temperature change and cumulative CO2 emissions during periods of net negative CO2 emissions. Environ. Res. Lett. 11, 055006 (2016).Article Google Scholar 
  27. 27.The Paris Agreement (UNFCCC, 2016); https://unfccc.int/process-and-meetings/the-paris-agreement/the-paris-agreement
  28. 28.Wigley, T. M. L., Richels, R. & Edmonds, J. A. Economic and environmental choices in the stabilization of atmospheric CO2 concentrations. Nature 379, 240–243 (1996).CAS Article Google Scholar 
  29. 29.Hammitt, J. K. Evaluation endpoints and climate policy: atmospheric stabilization, benefit–cost analysis, and near-term greenhouse-gas emissions. Clim. Change 41, 447–468 (1999).Article Google Scholar 
  30. 30.Vogt-Schilb, A., Meunier, G. & Hallegatte, S. When starting with the most expensive option makes sense: optimal timing, cost and sectoral allocation of abatement investment. J. Environ. Econ. Manage. 88, 210–233 (2018).Article Google Scholar 
  31. 31.Iyer, G. et al. Measuring progress from nationally determined contributions to mid-century strategies. Nat. Clim. Change 7, 871–874 (2017).Article Google Scholar 
  32. 32.Waisman, H. et al. A pathway design framework for national low greenhouse gas emission development strategies. Nat. Clim. Change 9, 261–268 (2019).Article Google Scholar 
  33. 33.McFarland, J., Fawcett, A. A., Morris, A. & Reilly, J. Overview of economy-wide U.S. carbon tax strategies: results from EMF 32. Clim. Change Econ. 9, 1840002 (2018).Article Google Scholar 
  34. 34.Barron, A. R., Fawcett, A. A., Hafstead, M. A. C., McFarland, J. & Morris, A. C. Policy insights from the EMF 32 study on U.S. carbon tax scenarios. Clim. Change Econ. 9, 1840003 (2018).Article Google Scholar 
  35. 35.Reed, A. et al. Interrogating uncertainty in energy forecasts: the case of the shale gas boom. Energy Transit. 3, 1–11 (2019).Article Google Scholar 
  36. 36.Conroy, M. J. & Peterson, J. T. in Decision Making in Natural Resource Management: A Structured, Adaptive Approach 192–231 (John Wiley & Sons, 2013); https://doi.org/10.1002/9781118506196.ch7
  37. 37.Brooks, S. A. & Keohane, N. O. The political economy of hybrid approaches to a U.S. carbon tax: a perspective from the policy world. Rev. Environ. Econ. Policy 14, 67–75 (2020).Article Google Scholar 
  38. 38.Tol, R. Timing of greenhouse gas emission reduction. Pac. Asian J. Energy 10, 63–68 (2000).Google Scholar 
  39. 39.Grubb, M. & Wieners, C. Modeling Myths: On the Need for Dynamic Realism in Dice and Other Equilibrium Models of Global Climate Mitigation Working Paper No. 112 (Institute for New Economic Thinking, 2020); https://doi.org/10.36687/inetwp112
  40. 40.Stiglitz, J. E. Addressing climate change through price and non-price interventions. Eur. Econ. Rev. 119, 594–612 (2019).Article Google Scholar 
  41. 41.IPCC: Summary for Policymakers. In Special Report on Global Warming of 1.5°C (eds Masson-Delmotte, V. et al.) (WMO, 2018).
  42. 42.Vogt-Schilb, A., Hallegatte, S. & de Gouvello, C. Marginal abatement cost curves and the quality of emission reductions: a case study on Brazil. Clim. Policy 15, 703–723 (2015).Article Google Scholar 
  43. 43.Global EV Outlook 2019 (IEA, 2019); https://www.iea.org/reports/global-ev-outlook-2019
  44. 44.The Climate Change Act 2008 (2050 Target Amendment) Order 2019 (UK Parliament, 2019).
  45. 45.United States Mid-century Strategy for Deep Decarbonization (White House, 2016); https://unfccc.int/files/focus/long-term_strategies/application/pdf/mid_century_strategy_report-final_red.pdf
  46. 46.United States Mid-century Strategy for Deep Decarbonization—Technical Appendix (White House, 2016); https://unfccc.int/files/focus/long-term_strategies/application/pdf/us_mcs_documentation_and_output.pdf
  47. 47.Annual Technology Baseline (ATB) from the National Renewable Energy Laboratory (NREL) (NREL, 2019); https://atb.nrel.gov/
  48. 48.Annual Energy Outlook 2018 (EIA, 2018); https://www.eia.gov/outlooks/aeo/
  49. 49.2017 National Population Projections Datasets: Projections for the United States: 2017 to 2060 (US Census Bureau, 2017); https://www.census.gov/data/datasets/2017/demo/popproj/2017-popproj.html
  50. 50.Incremental Documentation for IPM Platform v6 May 2019 Reference Case (US EPA, 2019); https://www.epa.gov/airmarkets/incremental-documentation-ipm-platform-v6-may-2019-reference-case
  51. 51.Fulfilling America’s Pledge: How States, Cities, and Business Are Leading the United States to a Low-Carbon Future (America’s Pledge Initiative on Climate, 2018); https://www.americaspledgeonclimate.com/fulfilling-americas-pledge/
  52. 52.Jadun, P. et al. Electrification Futures Study: End-Use Electric Technology Cost and Performance Projections through 2050 (NREL, 2017); https://www.nrel.gov/docs/fy18osti/70485.pdf
  53. 53.Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990–2017 (US EPA, 2019).

Download references

Acknowledgements

We thank K. Gillingham, G. Heal, A. Fawcett, J. Larsen, S. Sayre, J. Goffman and P. Kelleher for comments on an early version of this manuscript. This work was made possible by support from the Center on Global Energy Policy at the School of International and Public Affairs of Columbia University.

Author information

Affiliations

  1. SIPA Center on Global Energy Policy, Columbia University, New York, NY, USANoah Kaufman & Peter Marsters
  2. Environmental Science and Policy, Smith College, Northampton, MA, USAAlexander R. Barron
  3. Joint Global Change Research Institute, University of Maryland, College Park, MD, USAWojciech Krawczyk & Haewon McJeon

Contributions

N.K., A.R.B., H.M. and P.M. contributed to the design of the research. W.K. and H.M. conducted the modelling. All authors contributed to the analysis of the results. N.K., A.R.B., H.M. and P.M. wrote the manuscript.

Corresponding author

Correspondence to Noah Kaufman.