Fighting Climate Change Is The Cheapest Option We Have Left, Modelling Shows

Climate change is a challenge for all life on Earth

By David Nield - 03. February 2020

You would think that ensuring the survival of the human race is something you can't put a price on, but one of the reasons that governments aren't always keen to take action on climate change is the economic costs of doing so.

But new research investigating the future costs of dealing with a warming planet shows just how counterintuitive that way of thinking actually is, because the longer we wait to take action, the more we're going to have to pay in the long run.

According to the study's calculations, the cheapest option at this point is to pay what it takes to limit the global temperature rise over the next century to 2 degrees Celsius – the same number that governments committed to with the Paris Agreement.

"To secure economic welfare for all people in these times of global warming, we need to balance the costs of climate change damages and those of climate change mitigation," says climate scientist Anders Levermann from the Potsdam Institute for Climate Impact Research (PIK) in Germany.

"Now our team have found what we should aim for."

To reach their figures, Levermann and his colleagues used the Dynamic Integrated Climate-Economy (DICE) computer simulation developed by the Nobel Laureate of Economics William Nordhaus, specifically built to look at the impacts of climate change.

The DICE model weighed up the costs of cutting greenhouse gas emissions (through a reduction in the use of coal) against the costs of further climate change – increasing weather extremes, reduced human labour capacity, and so on.

An earlier study from 2015 was used as a guide to how temperature and global gross domestic product (GDP) are connected, with climate change damage starting at zero from 2020 and projected all the way to 2100.

There are a lot of variables to consider – how consumption patterns might alter, the impact climate change might have on conflicts around the world, the effects of various tipping points that we haven't calculated for yet – but economically speaking, stopping warming at 2 degrees Celsius is our cheapest option.

"It is remarkable how robustly reasonable the temperature limit of more or less 2 degrees C is, standing out in almost all the cost-curves we've produced," says climate scientist Sven Willner from PIK.

The researchers say "fast and fundamental global action" is required to put a lid on rising temperatures around the globe, based on the best data we have so far. Putting action off to a future date is only going to become more and more costly.

Let's hope that the start of 2020 marks a real watershed in the world's attitude towards our climate crisis – with signs of gathering momentum now appearing – and that we can keep global warming down to 2 degrees Celsius or as close to it as possible. Economically or otherwise, it's the only option that makes sense.

"The world is running out of excuses to justify sitting back and doing nothing," says Levermann. "All those who have been saying that climate stabilisation would be nice but is too costly can see now that it is really unmitigated global warming that is too expensive."

"Business as usual is clearly not a viable economic option any more. We either decarbonise our economies or we let global warming fire up costs for businesses and societies worldwide."

The research has been published in Nature Communications.

Desert Sky (Jared Verdi/Unsplash)

Paris Climate Agreement passes the cost-benefit test

By Nicole GlanemannSven N. WillnerAnders Levermann 


The Paris Climate Agreement aims to keep temperature rise well below 2 °C. This implies mitigation costs as well as avoided climate damages. Here we show that independent of the normative assumptions of inequality aversion and time preferences, the agreement constitutes the economically optimal policy pathway for the century. To this end we consistently incorporate a damage-cost curve reproducing the observed relation between temperature and economic growth into the integrated assessment model DICE. We thus provide an inter-temporally optimizing cost-benefit analysis of this century’s climate problem. We account for uncertainties regarding the damage curve, climate sensitivity, socioeconomic future, and mitigation costs. The resulting optimal temperature is robust as can be understood from the generic temperature-dependence of the mitigation costs and the level of damages inferred from the observed temperature-growth relationship. Our results show that the politically motivated Paris Climate Agreement also represents the economically favourable pathway, if carried out properly.


The temperature targets as agreed upon in the Paris Climate Agreement1 result from a long and complex political process2. However, it is not clear whether the associated emission reduction efforts are economically favourable2,3. Although econometric analyses4,5,6,7,8 suggest large damages at higher temperatures, these have not yet been employed to derive the relative economic benefits of achieving these temperature targets2,3. In particular, estimates6,8 of observed temperature-induced losses in gross domestic product have not been accounted for in computations of the economically optimal policy pathways. Here we provide a macroeconomic assessment of these targets by accounting for recent estimates of warming-induced economic growth impacts, which are given by Burke et al.6,8 (BHM, hereafter). BHM have advanced prior knowledge4 on the relationship between temperature und economic growth by finding a universal non-linear relationship. Warming is shown to lead to a shift along the growth curve and to reduce growth beyond a certain temperature threshold.

So far, the BHM estimates have been shown to correspond to rather high social cost of carbon9, indicating that emission reduction should be stringent. However, the implications for optimal policy have only been investigated along predetermined scenarios of warming and economic growth6,8,9,10. Although such estimates are not without criticism9,11, it is a natural and necessary next scientific step to compare them to the costs of mitigating climate change (mitigation costs, hereafter) using an integrated assessment model (IAM). IAMs account for the diverse dynamic interactions between the economy and the climate12,13.

This comparison provides the end-of-century warming that is associated with the lowest total costs of damages and mitigation as employed in the IAM used (Fig. 1). Cost-benefit optimal warming is thus determined by the shape of the two cost curves. The mitigation-cost curves are characterized by two universal properties. First, they diverge at the present-day warming, in particular if negative-emission technologies are not available. Second, the mitigation costs decrease to zero for a warming scenario without any mitigation efforts. The damage-cost curve, on the other hand, is known to be zero without warming and to increase with rising temperatures. The level to which the damages rise without mitigation is subject to investigation. However, due to the divergence of the mitigation costs the economically optimal temperature becomes less sensitive to the exact level of damages once these have reached a certain level (Fig. 1). Here, we examine whether the damages that follow from extrapolating the observed relation of economic growth and temperature6,8 are beyond this level.


Fig. 1: Illustration of universality of the cost-benefit climate analysis.

Cumulative mitigation costs (green curve) and climate damages (black curve) as a function of Earthʼs warming level give the total climate costs (red curve). Mitigation costs diverge for present-day warming and converge to zero for unmitigated warming. The damages are zero for zero warming and increase with temperature. The characteristic steepness of the mitigation curve implies that beyond a certain damage level the economically optimal temperature (which minimizes the total costs) becomes insensitive to a further increase in damages. For example, increasing (black dashed) or decreasing (black dotted) the damage level by half of the initial damage level does not change the economically optimal warming level significantly (grey area).

To this end we incorporate the BHM estimates into one of the most prominent IAMs14,15,16, DICE-201316. With its simplicity, DICE allows assessing cost-benefit optimality in a scientifically highly transparent and controlled way. According to its original version, which has also been employed to advise US climate policy17,18,19 achieving the 2 °C target would cause mitigation costs significantly larger than the consequent avoided damages16,20,21. This result is largely due to a damage function that does not incorporate recent estimations of economic impacts13,22,23. Here, we update this function according to the BHM estimates8. As DICE searches for the economic growth path that maximises global welfare, the growth estimates cannot be implemented directly. As a solution to this problem we develop a novel procedure that preserves the growth model feature. In that, we iteratively adjust the damage function to reproduce the estimated temperature-induced growth relation in DICE-2013. For consistency with the BHM estimates, we design a scenario that emulates a future world in which key conditions are similar as in the past, i.e. the absence of climate policy.

We use this updated damage function to derive the cost-benefit optimal climate policy that begins with the year 2020. In this economically optimal scenario, mitigation is actively pursued to maximize global welfare. We continue holding the assumption of DICE-2013 that significant negative-emission technologies are not available in this century. We contrast the optimal policy with the business-as-usual (BAU) scenario, in which climate policy is absent. We find that under these conditions the 2 °C target as set by the Paris Climate Agreement gives the cost-benefit optimal pathway till the end of this century. We observe that this finding is largely robust to diverse uncertainties. Our results thus advocate for rapid and decisive implementation of the Paris Climate Agreement.


Cost-benefit optimal temperature

In our analyses, we account for uncertainty in the future temperature development by considering three alternative equilibrium climate sensitivity (ECS) values. In addition, we subject our results to extensive robustness tests. We examine the effects of uncertainty in the BHM estimates concerning the parameter values and the model specification. For this we adopt the bootstrapping approach from the original empirical study8 and use the resulting 1000 samples to derive a corresponding ensemble of damage functions. We also conduct a sensitivity analyses regarding social preferences for consumption changes24, alternative socioeconomic futures25, and mitigation costs.

We find that the 2 °C target represents the cost-benefit optimal temperature for the base calibration (Fig. 2a). This calibration involves the best estimate8 of the temperature–economic growth relation in the past and the original ECS value in DICE-2013 of 2.9 °C, which is at the centre of estimates for several decades26,27. Higher ECS values shift the level of target warming for which the mitigation-cost curve diverges to infinity to higher values (Fig. 1), i.e. they incur substantially higher mitigation costs. For ECS of 4 °C, for instance, the 2 °C target becomes too costly. Yet, with an optimal target warming of 2.4 °C the deviation from this target is not large. For smaller ECS values, e.g. of 2 °C, limiting warming further to well below 2 °C is economically optimal. Regardless of the exact ECS, the optimal mitigation efforts promise a significant damage reduction compared to the BAU scenario (~14% for ECS of 4 °C, ~10% for ECS of 2.9 °C, and ~8% for ECS of 2 °C). These efforts are, as also claimed by the Paris Agreement, ambitious (Article 3)1 and involve very stringent measures from the outset (Fig. 2c).


Fig. 2: Temperature increase, damage costs, and carbon emissions under cost-benefit optimal policy for three different climate sensitivities.

The black curves are associated with the original calibration of the climate sensitivity of 2.9°C; the blue curves with a 2°C climate sensitivity and the red curve with a 4°C climate sensitivity. The inset figures allow comparing the economically optimal temperature development and damage costs with their corresponding values in the BAU scenario.

Uncertainty in damage function

To examine the effects of uncertainty in the impact estimates, we use the cumulative GDP losses until 2100 (in 2005 $US) in the BAU scenario as a measure for the impact severity and pair them with the economically optimal end-of-century temperature (Fig. 3). The uncertainty in the damage costs, according to the empirical study6,8, is substantial with respect to the magnitude and sign of the warming impact and also implies large differences in our results. Nonetheless, the ensemble median of the optimal temperatures is only marginally higher than 2 °C for ECS of 2.9 °C, and well below 2 °C for ECS of 2 °C. This result is robust to alternative specifications of the bootstrapping approach8 (Supplementary Figs. 1 and 2) and to most alternative model specifications of BHM and the alternative econometric estimates by Dell et al.4 (Fig. 4 and Supplementary Figs. 36). Hence, the goal to limit warming to 2 °C or less is cost-benefit optimal for a wide set of damage functions. By contrast, the results of the original DICE versions16,21 deviate significantly from the computed likely range (Fig. 3).


Fig. 3: Relation between the cumulative GDP losses until 2100 (in 2005 $US) in the absence of climate policy and the economically optimal warming until the end of the century, given uncertainty in the estimates of the historical impact and uncertainty in the climate sensitivity value.

Scattered points give the uncertainty ensemble in the historical relation between temperature increase and economic growth for three different climate sensitivities; red points for 4 °C climate sensitivity, black points for the original climate sensitivity calibration in the DICE-2013R model, and blue points for 2 °C climate sensitivity. Each point depicts the DICE-2013 model output for a damage function calibrated according to one of the 1000 bootstraps of the historical regression. Curves in the main plot represent the best fit for the relation between cumulative damage costs and optimal warming. The histograms below and on the left give the frequency of the model results as well as the medians and likely ranges for each of the three climate sensitivities. The likely rage of optimal end-of-century warming is approximately located between 2.3 °C and 3.4 °C with a median of 2.5 °C for the climate sensitivity of 4 °C, between 1.8 °C and 3 °C with a median of 2.1 °C for a climate sensitivity of 2.9 °C and between 1.3 °C and 2.5 °C with a median of 1.7 °C for a climate sensitivity of 2 °C. The results of the original DICE versions are located outside the likely ranges as shown by the black brackets.



Fig. 4: Ensembles including the uncertainty in the estimates of the historical impacts according to BHM (blue bars) and some samples according to Dell et al.4 (DJO, red lines).

Specification of the estimates without (short-run (a, b)) and with (long-run (c, d)) the assumption that the influence of warming on economic growth is lagged and/or without (pooled (a, c)) and with (differentiated (b, d)) differentiating between impacts on poor and on rich countries. Each specification for BHM samples from 1000 bootstraps of the historical regression; samples for DJO include specifications with no lag (b) as well as 1-lag, 5-lag, and 10-lag specifications (d).

Uncertainty regarding preferences

We also test the sensitivity to two important preference parameters (Fig. 5). First, the ‘initial rate of social time preference’ (IRSTP) which reflects the preference for consumption at different points in time, with a higher value giving more emphasis to present rather than to future consumption; and second, the ‘elasticity of marginal utility of consumption’ (EMUC) which describes the preferences for more consumption, irrespective of its timing, and is interpreted as generational inequality aversion21. As these parameters crucially affect decisions of optimal mitigation and investment28, the implied growth effects are critical for our results. Taking the prescriptive viewpoint of calibrating IRSTP and EMUC24, we account for wide value ranges, including the base calibration in DICE-2013 and the suggestions by the IPCC-AR529. The latter proposes near-zero IRSTP values, which we interpret as values smaller than 1%. With the exception of a few unusual parameter values, this wide range of options leads to optimal warming of around 2 °C or lower (Fig. 5).


Fig. 5: Sensitivity of the economically optimal temperature in 2100 to alternative initial rates of social time preference and generational inequality aversion.

These simulations are based on the benchmark impact estimate as in Fig. 2 with an equilibrium climate sensitivity (ECS) of 2.9 °C. The unhatched box indicates the range of values recommended by the IPCC-AR5 report29. The black star depicts the DICE-201316 calibration. The red line marks the 2 °C isoquant.

Cost-benefit optimal temperature under SSP scenarios

Further tests also show robustness to alternative socioeconomic assumptions as described by the Shared Socioeconomic Pathways (SSPs)25 (Fig. 6). As the mitigation-cost function in DICE is strongly simplified, we investigate how our results change with functions that describe different technological possibilities in the future (Fig. 7). Similar to the differences between results for a range of damage functions, the uncertainty in mitigation costs reflects on the derived optimal warming level. Nevertheless, the mitigation costs deriving from the different SSPs tend to imply rather lower median optimal warming levels (1.8 °C, 1.9 °C, 2.0 °C).


Fig. 6: Economically optimal warming for SSP1, SSP2, SSP5 and DICE.

a The economically optimal temperature pathway for different socioeconomic conditions under the assumption that negative-emission technologies are not used within this century. b, c Recalibrated parameters in DICE to match the results of the REMIND model for the three SSPs. b shows the results for the total factor productivity (TFP) and c for the costs of mitigation. d Economically optimal warming in 2100 if negative-emission technologies are available in this century.


Fig. 7: Economically optimal temperature increase for alternative mitigation-cost functions.

The mitigation functions, which are sampled from the SSP fit, reflect different technological possibilities in the future as reflected by the SSPs. The dotted line shows the value for the benchmark estimate (DICE-2013).


Our findings build on the most recent empirical advances of impact estimates, which we consistently integrate in a dynamic IAM. These estimates are, however, not without critique, especially regarding the assumed functional relationship, the significance of using weather variables for insights into climate impacts and on other methodological challenges. In particular, using them in projections assumes that the historically observed temperature-impact link can be extrapolated into the future. Yet, this relation can change if further warming is associated with an unprecedented variation in climatic extremes for example with potential cascading effects30,31,32,33 or with the occurrence of devastating climatic tipping points34,35, or with significant changes in the societal response to warming. We also follow other studies using the estimates for projections6,8 to derive the benefits for smooth temperature paths without variability. The economic costs associated with temperature variability may, however, require even more stringent mitigation efforts.

Furthermore, assessing impacts in terms of GDP is an incomplete measure for the overall benefits of climate change mitigation as non-monetary losses such as loss of life and biodiversity are omitted. Unless adaptation to climate change becomes effective, most of these points suggest a strong underestimation of the mitigation efforts needed.

Similarly, a global analysis like ours, of course, neglects distributional issues as to who bears the burdens of damages as well as mitigation costs. Some specifications of the damage functions we employ here differentiate at least between two classes of income levels. Here, we have to make simplifying assumptions regarding shares of these classes to incorporate them into the one-region model, which constitute another source of uncertainty (Fig. 4). In general, a cost-benefit calculation has to be interpreted vary cautiously keeping ethical considerations in mind. Like other studies36 we use DICE as a parsimonious surrogate for more complex and spatially disaggregate IAMs. Future research should transfer our analysis to these IAMs to clarify questions of regional impact heterogeneity and to fully account for region-specific empirical estimates.

In our analysis, the leeway to reach the 2 °C target is considerably constraint by ruling out negative emissions in this century. Nonetheless, we show that, if future damages follow the same temperature dependence as historically observed, the overall damage costs will reach a level that renders 2 °C cost-benefit optimal. This result evolves as a direct consequence from the recently given empirical evidence attesting considerable marginal damage increases for higher temperatures and the universal functional behaviour of the mitigation costs in the vicinity of present-day temperatures (cf. Fig. 1).



Nature Communications volume 11, Article number: 110 (2020) Cite this article

Reporting summary

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

Data availability

The datasets generated and analysed during this study including data shown in the figures are available from the authors upon request.

Code availability

The code used in this study is available from the authors upon request.


  1. 1.

    United Nations Framework Convention on Climate Change. Paris Agreement. Article 2(a) (2015).

  2. 2.

    Hallegatte, S. et al. Mapping the climate change challenge. Nat. Clim. Chang. 6, 663–668 (2016).

  3. 3.

    The maximum climate ambition needs a firm research backing. Nature 537, 585–586 (2016).

  4. 4.

    Dell, M., Jones, B. F. & Olken, B. A. Temperature shocks and economic growth: evidence from the last half century. Am. Econ. J. Macroecon. 4, 66–95 (2012).

  5. 5.

    Auffhammer, M., Hsiang, S. M., Schlenker, W. & Sobel, A. Using weather data and climate model output in economic analyses of climate change. Rev. Environ. Econ. Policy 7, 181–198 (2013).

  6. 6.

    Burke, M., Hsiang, S. M. & Miguel, E. Global non-linear effect of temperature on economic production. Nature 527, 235–239 (2015).

  7. 7.

    Lemoine, D. & Kapnick, S. A top-down approach to projecting market impacts of climate change. Nat. Clim. Chang. 6, 51–55 (2016).

  8. 8.

    Burke, M., Davis, W. M. & Diffenbaugh, N. S. Large potential reduction in economic damages under UN mitigation targets. Nature 557, 549–553 (2018).

  9. 9.

    Ricke, K., Drouet, L., Caldeira, K. & Tavoni, M. Country-level social cost of carbon. Nat. Clim. Chang. 8, 895–900 (2018).

  10. 10.

    Ueckerdt, F. et al. The economically optimal warming limit of the planet. Earth Syst. Dyn. 10, 741–763 (2019).

  11. 11.

    Mendelsohn, R. Measuring weather impacts using panel data. In ASSA Annual Meeting 2017 (American Economic Association 2016).

  12. 12.

    Burke, M. et al. Opportunities for advances in climate change economics. Science 352, 292–293 (2016).

  13. 13.

    Diaz, D. & Moore, F. Quantifying the economic risks of climate change. Nat. Clim. Chang. 7, 774–782 (2017).

  14. 14.

    Waldhoff, S., Anthoff, D., Rose, S. & Tol, R. S. J. The marginal damage costs of different greenhouse gases: an application of FUND. Econ. Open-Access, Open-Assess. E-J. 8, 1 (2014).

  15. 15.

    Hope, C. The marginal impact of CO2 from PAGE2002: an integrated assessment model incorporating the IPCC’s five reasons for concern. Integr. Assess. 6, 19–56 (2006).

  16. 16.

    Nordhaus, W. The Climate Casino: Risk, Uncertainty, and Economics for a Warming World (Yale University Press, 2013).

  17. 17.

    Goverment, I. W. G. on S. C. of C. U. S. Technical Support Document: Social Cost of Carbon for Regulatory Impact Analysis Under Executive Order 12866. Social Cost of Carbon Estimates for Regulatory Impact Analysis: Development and Technical Assessment (2015).

  18. 18.

    Interagency Working Group on Social Cost of Carbon; United States Goverment. Technical Support Document: Technical Update of the Social Cost of Carbon for Regulatory Impact Analysis Under Executive Order 12866 (2016).

  19. 19.

    Greenstone, M., Kopits, E. & Wolverton, A. Developing a social cost of carbon for US Regulatory Analysis: a methodology and interpretation. Rev. Environ. Econ. Policy 7, 23–46 (2013).

  20. 20.

    Nordhaus, W. D. A Question of Balance (Yale University Press, 2008).

  21. 21.

    Nordhaus, W. D. Revisiting the social cost of carbon. Proc. Natl Acad. Sci. 114, 1518–1523 (2017).

  22. 22.

    Revesz, R. L. et al. Global warming: Improve economic models of climate change. Nature 508, 173–175 (2014).

  23. 23.

    Rose, S. K., Diaz, D. B. & Blanford, G. J. Understanding the social cost of carbon: a model diagnostic and inter-comparison study. Clim. Chang. Econ. 08, 1750009 (2017).

  24. 24.

    Stern, N. The Economics of Climate Change (Cambridge University Press, 2007).

  25. 25.

    Riahi, K. et al. The shared socioeconomic pathways and their energy, land use, and greenhouse gas emissions implications: an overview. Glob. Environ. Chang. 42, 153–168 (2017).

  26. 26.

    Roe, G. H. & Baker, M. B. Why Is Climate Sensitivity So Unpredictable? Science 318, 629–632 (2007).

  27. 27.

    IPCC. Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (Cambridge University Press, 2013).

  28. 28.

    Dasgupta, P. Discounting climate change. J. Risk Uncertain. 37, 141–169 (2008).

  29. 29.

    IPCC. Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (2014).

  30. 30.

    Hallegatte, S. An adaptive regional input-output model and its application to the assessment of the economic cost of Katrina. Risk. Anal. 28, 779–799 (2008).

  31. 31.

    Levermann, A. Make supply chains climate-smart. Nature 506, 27–29 (2014).

  32. 32.

    Wenz, L. & Levermann, A. Enhanced economic connectivity to foster heat stress-related losses. Sci. Adv. 2, e1501026 (2016).

  33. 33.

    Willner, S. N., Otto, C. & Levermann, A. Global economic response to river floods. Nat. Clim. Chang. 8, 594–598 (2018).

  34. 34.

    Lenton, T. M. et al. Tipping elements in the Earth’s climate system. Proc. Natl Acad. Sci. 105, 1786–1793 (2008).

  35. 35.

    Lenton, T. M. & Ciscar, J.-C. Integrating tipping points into climate impact assessments. Clim. Change 117, 585–597 (2013).

  36. 36.

    Lamontagne, J. R., Reed, P. M., Marangoni, G., Keller, K. & Garner, G. G. Robust abatement pathways to tolerable climate futures require immediate global action. Nat. Clim. Chang. 9, 290–294 (2019).

  37. 37.

    van Vuuren, D. P. et al. How well do integrated assessment models simulate climate change? Clim. Change 104, 255–285 (2011).

  38. 38.

    Hof, A. F. et al. The benefits of climate change mitigation in integrated assessment models: the role of the carbon cycle and climate component. Clim. Change 113, 897–917 (2012).

  39. 39.

    Glotter, M. J., Pierrehumbert, R. T., Elliott, J. W., Matteson, N. J. & Moyer, E. J. A simple carbon cycle representation for economic and policy analyses. Clim. Change 126, 319–335 (2014).

  40. 40.

    International Monetary Fund. World Economic Outlook Database. (2018).

  41. 41.

    Meinshausen, M. et al. The RCP greenhouse gas concentrations and their extensions from 1765 to 2300. Clim. Change 109, 213–241 (2011).

  42. 42.

    Olivier, J. G. J., Schure, K. M. & Peters, J. A. H. W. Trends in Global CO 2 and Total Greenhouse Gas Emissions: 2017 Report (2017).

  43. 43.

    Pindyck, R. S. The social cost of carbon revisited. J. Environ. Econ. Manag. 94, 140–160 (2019).

  44. 44.

    Nordhaus, W. D. & Moffatt, A. A survey of global impacts of climate change: replication, survey methods, and a statistical analysis. SSRN Electron. J. (2017).

  45. 45.

    Howard, P. H. & Sterner, T. Few and not so far between: a meta-analysis of climate damage estimates. Environ. Resour. Econ. 68, 197–225 (2017).

  46. 46.

    Fankhauser, S. & Tol, R. S. J. On climate change and economic growth. Resour. Energy Econ. 27, 1–17 (2005).

  47. 47.

    Moore, F. C. & Diaz, D. B. Temperature impacts on economic growth warrant stringent mitigation policy. Nat. Clim. Chang. 5, 127–131 (2015).

  48. 48.

    Dietz, S. & Stern, N. Endogenous growth, convexity of damage and climate risk: how Nordhaus’ framework supports deep cuts in carbon emissions. Econ. J. 125, 574–620 (2015).

  49. 49.

    Guivarch, C. & Pottier, A. Climate damage on production or on growth: what impact on the social cost of carbon? Environ. Model. Assess. 23, 117–130 (2018).

  50. 50.

    Newell, R. G., Prest, B. C. & Sexton, S. E. The GDP-Temperature Relationship: Implications for Climate Change Damages. Working Paper 18-17 REV (2018).

  51. 51.

    Hsiang, S. Climate econometrics. Annu. Rev. Resour. Econ. 8, 43–75 (2016).

  52. 52.

    Letta, M. & Tol, R. S. J. Weather, climate and total factor productivity. Environ. Resour. Econ. 73, 283–305 (2019).

  53. 53.

    GISTEMP Team. GISS Surface Temperature Analysis (GISTEMP), version 4. NASA Goddard Institute for Space Studies. (2011).

  54. 54.

    Lenssen, N. J. L. et al. Improvements in the GISTEMP Uncertainty Model. J. Geophys. Res. Atmos. 124, 6307–6326 (2019).

  55. 55.

    O’Neill, B. C. et al. The roads ahead: Narratives for shared socioeconomic pathways describing world futures in the 21st century. Glob. Environ. Chang. 42, 169–180 (2017).

  56. 56.

    Kriegler, E. et al. Fossil-fueled development (SSP5): an energy and resource intensive scenario for the 21st century. Glob. Environ. Chang. 42, 297–315 (2017).

  57. 57.

    Leimbach, M., Kriegler, E., Roming, N. & Schwanitz, J. Future growth patterns of world regions—a GDP scenario approach. Glob. Environ. Chang. 42, 215–225 (2017).

  58. 58.

    Luderer, G. et al. Description of the REMIND model (Version 1.6). SSRN Electron. J. (2015).

  59. 59.

    Ha-Duong, M., Grubb, M. J. & Hourcade, J.-C. Influence of socioeconomic inertia and uncertainty on optimal CO2-emission abatement. Nature 390, 270–273 (1997).

  60. 60.

    Richels, R. G. & Blanford, G. J. The value of technological advance in decarbonizing the U.S. economy. Energy Econ. 30, 2930–2946 (2008).

  61. 61.

    Hogan, W. W. & Jorgenson, D. W. Productivity trends and the cost of reducing CO2 emissions. Energy J. 12, 67–85 (1991).

  62. 62.

    Broome, J. The ethics of climate change. Sci. Am. 298, 96–102 (2008).

  63. 63.

    Evans, D. J. The elasticity of marginal utility of consumption: estimates for 20 OECD countries. Fisc. Stud. 26, 197–224 (2005).

  64. 64.

    Goulder, L. H. & Williams, R. C. The choice of discount rate for climate change policy evaluation. Clim. Chang. Econ. 3, 1250024 (2012).

Download references


This research has received funding from the Horizon 2020 Framework Programme of the European Union (grant agreement no. 641811).

Author information

Author notes
  1. These authors contributed equally: Nicole Glanemann, Sven N. Willner.


  1. Potsdam Institute for Climate Impact Research, 14473, Potsdam, Germany
    • Nicole Glanemann
    • , Sven N. Willner
    •  & Anders Levermann
  2. WHU—Otto Beisheim School of Management, 56179, Vallendar, Germany
    • Nicole Glanemann
  3. Institute of Physics, Potsdam University, 14476, Potsdam, Germany
    • Anders Levermann
  4. Columbia University, New York, NY, 10027, USA
    • Anders Levermann


A.L. and N.G. designed the study. N.G. and S.W. developed the methods and carried out the computation. N.G., S.W., and A.L. analyzed the results and wrote the manuscript.

Corresponding author

Correspondence to Sven N. Willner.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Reporting Summary

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Glanemann, N., Willner, S.N. & Levermann, A. Paris Climate Agreement passes the cost-benefit test. Nat Commun 11, 110 (2020).

Download citation

Share this article

Anyone you share the following link with will be able to read this content:



By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Download PDF