




































Ecology, Economy and Society–the INSEE Journal 3 (1): 11–30, January 2020 

 
THEMATIC ESSAY  
 

Has Economics Caught Up with Climate Science?  
 

Shreekant Gupta   
 

Oh, East is East, and West is West, and never the twain shall meet, 
… 
But there is neither East nor West, ... 
When two strong men stand face to face, tho‘ they come from the ends of the 
earth! 

   The Ballad of East and West (Rudyard Kipling, 1889) 

 
Abstract: Whereas scientific evidence points towards substantial and urgent 
reduction in greenhouses gas (GHG) emissions, economic analysis of climate 
change seems to be out of sync by indicating a more gradual approach. In 
particular, economic models that use benefit cost analysis, namely, integrated 
assessment models (IAMs) have been criticised for being conservative in their 
recommendations on the speed of reducing GHG emissions and the associated 
levels of carbon taxes. This essay focuses on a prototypical IAM, namely, 
Nordhaus‘ DICE model to argue the schism between science and economics is 
more apparent than real. Analysis of the DICE model suggests extreme climate 
scenarios can be captured through alternative specifications of the damage function 
(the impact of temperature on the economy). In particular, damage functions that 
extend the standard quadratic representation are highly convex (Weitzman 2012). 
Thus, they are able to capture climate tipping points as well as ―fat tail‖ risks 
originating from uncertainty with regard to equilibrium climate sensitivity. 

 

1. INTRODUCTION 

A fundamental question in climate policy is by how much should 
greenhouse gases (GHGs) be reduced and how fast?1 Economists attempt 

                                                        
 Department of Economics, Delhi School of Economics, University of Delhi, Delhi 
110007; sgupta@econdse.org. 

Copyright © Gupta 2020. Released under Creative Commons Attribution-Non-commercial 
4.0 International licence (CC BY-NC 4.0) by the author. 

Published by Indian Society for Ecological Economics (INSEE), c/o Institute of Economic 
Growth, University Enclave, North Campus, Delhi 110007.  

ISSN: 2581-6152 (print); 2581-6101 (web). 

DOI: https://doi.org/10.37773/ees.v3i1.86  

https://doi.org/10.37773/ees.v3i1.86


Ecology, Economy and Society–the INSEE Journal [12] 

to answer this question using the framework of benefit-cost analysis (BCA) 
since the cost of reducing GHG emissions has to be incurred now, whereas 
the benefit of doing so will accrue in future, though perhaps not so distant 
future. Put differently, the economic gain from business-as-usual (BAU) is 
now whereas the pain, that is damages due to climate change, will be down 
the road. However, the nature, magnitude and timing of these damages is 
uncertain and sometimes even unknown. Further, some of these damages 
are irreversible. Thus, economic analysis of climate change is an exercise in 
intertemporal BCA but with the added dimensions of risk, uncertainty and 
irreversibility. The centrality of BCA in economic analysis of climate change 
was epitomised by the award of the Nobel Prize in economics last year to 
one of its primary exponents, William Nordhaus. 2  While BCA is the 
reigning orthodoxy in climate change economics, critics argue it ignores or 
does not adequately address what climate science is telling us about risk, 
uncertainty, and irreversibility. According to them this severely limits policy 
recommendations emanating from BCA, if not making them downright 
incorrect and misleading (see for instance Weitzman 2009 and Stern 2014).3 

Critics contest the mainstream view among economists (based on BCA 
models) that mitigation of GHGs must start gradually and then be ramped 
up. The so-called ‗policy ramp‘, argues climate policy, should lead to 
―modest rates of emissions reductions in the near term, followed by sharp 
reductions in the medium and long term‖ (Nordhaus 2007).4 Critics argue 
this does not square with what climate science is telling us about tipping 
points and other abrupt changes in climate. According to them economic 
analysis based on BCA does not find the need for immediate and deep cuts 
in GHG emissions and this is problematic. 

A more fundamental methodological critique of BCA is by Weitzman 
(2009), who argues climate science poses an existential question for BCA, 
namely, in the presence of potentially catastrophic and irreversible damages 

                                                                                                                            
1 In this essay, I cast the analysis at a global aggregate level and thus abstract from issues of 
burden sharing across nations. Issues of inter-generational burden sharing are of course 
unavoidable. In fact, intertemporal trade-off is at the heart of the problem being studied. 
2 In this essay BCA is used in a broad sense as some overall economic analysis focused on 
maximizing welfare. Thus, it overlaps with an integrated assessment model (IAM) such as 
the celebrated Dynamic Integrated Climate Economy (DICE) model of Nordhaus and we 
can treat the two terms as interchangeable. 
3  These two citations are representative of several papers and one book by these two 
economists, who have been the most prominent critics from within the profession of BCA 
as applied to climate change. 
4 The climate policy being alluded to is carbon prices which should start low and eventually 
get ratcheted up to high levels. By corollary, emissions reductions should be modest in the 
near term and increase gradually. 



[13] Shreekant Gupta 

BCA is not even meaningful. According to him uncertainty permeates 
climate science especially with regard to equilibrium climate sensitivity 
(ECS).5 Despite decades of research, science has not been able to pin down 
this key parameter and there is a significant downside risk of it being very 
high. Thus, ―the main purpose of keeping GHG concentrations down is 
effectively to buy insurance against catastrophic global warming‖ 
(Weitzman 2011, 279). 

The problem, however, of accepting this argument is that there is ―little role 
for economics or any analysis of trade-offs or assessing costs and benefits 
because these don‘t matter when the science is so clear and the future of 
mankind is at stake‖ (McKibbin 2014, 560). Unfortunately, in the messy real 
world one cannot sidestep economics – the scale and pace at which 
countries and the world as a whole are reducing GHG emissions reflects 
trade-offs, if not explicitly then implicitly. While the ―deep structural 
uncertainty‖ (a la Weitzman) surrounding ECS has an important bearing on 
climate policy, I argue BCA models of climate change can in principle 
incorporate such uncertainty. It does not make them irrelevant. In 
particular, I show how BCA can incorporate the possibility of catastrophic 
climate damages and thus remains a useful framework for climate policy. 

The purpose of this essay is not to defend BCA as an end in itself. Indeed, 
BCA has several ethical and methodological problems embodied in it (Stern 
2014). Nonetheless, I argue BCA can lead us to the very conclusion, 
namely, immediate and large reductions in GHG emissions (henceforth 
‗urgent action‘) that climate science may suggest. With this limited purpose 
in mind, the next section describes economic models of climate change that 
use the BCA framework, the so-called integrated assessment models 
(IAMs). For expository purposes, I consider a simple and highly aggregated 
IAM, namely, the Dynamic Integrated Climate Economy (DICE) model of 
William Nordhaus. I identify the key features of this model that are 
influenced by climate science. This is followed by a discussion of specific 
issues in climate science that have a direct bearing on economic analysis, 
namely, uncertainty regarding equilibrium climate sensitivity (ECS), non-
linear climate damages and tipping points. The fourth section shows how 
these aspects can be captured in IAMs (DICE in particular) and can lead to 
emissions trajectories consistent with ‗urgent action‘. The final section 
concludes the essay. My basic message is from Kipling‘s poem whose first 

                                                        
5 ―The transient climate response (TCR) is the temperature change at the time of CO2 
doubling‖ and the ―equilibrium climate sensitivity‖, T2x, is the temperature change after the 
system has reached a new equilibrium for doubled CO2, i.e., after the ―additional warming 
commitment‖ has been realised‖ (Comín, Francisco and MA Rodríguez-Arias 2003, 21). As I 
show later, there is a non-negligible probability of ECS being very high. 



Ecology, Economy and Society–the INSEE Journal [14] 

line is widely cited but whose third and fourth lines, though less well 
known, are more applicable in the context of this essay. 

 

2. ECONOMIC MODELING OF CLIMATE CHANGE 

The standard economic approach to modeling climate change is through 
integrated assessment models. As the phrase suggests, these are scientific 
models that combine knowledge from several domains into one framework 
in order to better understand a problem that has multiple dimensions.6 This 
is particularly true for climate change, which is closely connected with 
geophysical sciences and economic activities. Thus, IAMs of climate change 
integrate geophysical stocks and flows, especially of GHGs, with economic 
stocks and flows so that all key endogenous variables can be analysed 
simultaneously. ―IAMs generally do not pretend to have the most detailed 
and complete representation of each included system. Rather, they aspire to 
have, at a first level of approximation, a representation that includes all the 
modules simultaneously and with reasonable accuracy‖ (Gillingham et al. 
2018).The first climate-economy IAMs were essentially energy models that 
included a carbon emissions module and later a small climate model. There 
are five key links that map anthropogenic climate change in an IAM: 

a) from ‗people‘ (producers/consumers) to emissions of GHGs 
b) from emissions to stocks of GHGs 
c) from GHG stocks to changes in temperature 
d) from rising temperature to climate change more broadly 
e) from climate change to human (economic) impact 

In this section, I focus on Nordhaus‘ DICE model as a prototypical IAM. 
Without going into too much technical detail, I describe heuristically its 
main features especially how it models climate damages.7 While there are 
several widely used IAMs (see Gillingham et al. 2018 for a recent 
comparison), the DICE model is a simple yet elegant construct that goes to 
the heart of the policy question stated at the beginning of this essay. The 
genesis of DICE model can be traced back to Nordhaus‘ papers in 1977 
and was first articulated in its current from by Nordhaus in 1992.8 DICE 
belongs to a class of IAMs known as benefit-cost (BC) IAMs as contrasted 
to detailed process (DP) IAMs (see Weyant 2017) for details). While both 

                                                        
6 In effect the term ―integrated assessment‖, of course, is generic and can apply to a range of 
contexts. Here it is used explicitly in the context of climate-economy models. 
7  The complete DICE model can be found in Nordhaus (2008) and also at 
http://webdice.rdcep.org/static/docs/Equations_141227.pdf 
8 See, Nordhaus (1977 a, b) and Nordhaus (1992 a, b). For a history of the evolution of the 
DICE model see Newbold (2010). 

http://webdice.rdcep.org/static/docs/Equations_141227.pdf


[15] Shreekant Gupta 

classes of models analyse climate-economy interactions, the former are 
simpler and present a highly aggregate representation of costs of GHG 
mitigation and of climate damages.9 The primary motivation for BC IAMs 
is to ―compute the optimal trajectory of global GHG emissions, and the 
corresponding prices to charge for those emissions‖ (Weyant 2014, 381). 
Together these constitute what is often termed as ‗optimal‘ climate policy. 
An ‗optimal‘ emissions trajectory is one that equates the marginal 
(discounted) benefits of avoided climate damage with the marginal 
(discounted) costs of GHG mitigation. Put differently, this time path of 
GHG emissions maximises the discounted present value of global welfare.10 
A benefit-cost IAM such as DICE calculates this time path and also the 
shadow prices of emissions, namely, the social cost of carbon (SCC).11 

2.1. Overview of the DICE model 

The DICE model views climate change within the framework of 
neoclassical economic growth theory. In the Ramsey–Cass–Koopmans 
(RCK) optimal growth model (Ramsey 1928; Cass 1965; Koopmans 1965) 
society invests in capital goods by reducing consumption today in order to 
increase consumption in future. The main decision in each time period is 
how much to consume and save. The DICE model ―extends this approach 
by including the ‗natural capital‘ of the climate system as an additional kind 
of capital stock. In other words, it views concentrations of greenhouse 
gases as negative natural capital, and emissions reductions as investments 
that raise the quantity of natural capital (or reduce the negative capital). By 
devoting output to emissions reductions, economies reduce consumption 
today but prevent economically harmful climate change and thereby 
increase consumption possibilities in the future‖ (Nordhaus 2013, 1080). 
An increase in concentration of GHGs has a negative impact on future 
economic output because of its influence on the global mean surface 
temperature (GMST or ATT  or simply T ).12 The fraction of output ( Y ) 

                                                        
9 For example, in the DICE model the world is taken as one region. DICE also has a 
regionally disaggregated companion model, namely, the Regional Integrated Climate 
Economy (RICE) model (Nordhaus and Yang 1996). In its most recent version, the RICE 
model divides the world into 12 regions of which India is one (Nordhaus 2010). 
10 Some readers may find this framing problematic. Indeed, there are a number of critiques 
of IAMs some methodological and others more broad ranging (e.g., Ackerman et al. 2009; 
Pindyck 2013, 2017; Stern 2013). Again, my objective is not to defend IAMs as an 
epistemology. It is to demonstrate that IAMs like DICE with suitable modification can result 
in a big bang reduction in GHG emissions. 
11 The only GHG that is controlled in DICE is emissions of CO2 from industries. CO2 
from land-use change (e.g., deforestation) and other GHGs are treated as exogenous trends. 
12 Captured through equilibrium climate sensitivity (ECS) parameter. Also note ATT  actually 

refers to a change in temperature from pre-industrial baseline. 



Ecology, Economy and Society–the INSEE Journal [16] 

Figure 1: Schematic illustration of the DICE model 

 
Note: . The dark blue arrows represent the economic component of the model. 
Red arrows show how the economy impacts climate and vice versa. Green 
arrows illustrate the effect of climate policy. 
Source: Adapted from Wieners (2018) 

lost in each time period is captured through a Hicks-neutral damage 
function (  ). By devoting some portion of economic output to GHG 
abatement ( )(investment in natural capital), future temperature increases 
and associated climate damages ( ) can be avoided. The net output in each 
period ( Q ) then is divided between consumption and investment in 

physical capital (figure 1). 

Following Nordhaus (2008), net output at time t, )t(Q is gross output )t(Y  

scaled by climate damages )t(  and minus abatement expenditures )t(  

(percentage of output spent on reducing GHG emissions). )t(Q  is further 

allocated between consumption )t(C  (broadly defined) and investment 

)t(I .13 

 

                                                        
13 It should be emphasized that consumption is broadly defined to include marketed goods 
and services and also non-market goods and services especially environmental amenities. 
Similarly, damages reflect ―damages in various economic sectors, notably agriculture, 
farming, forestry, tourism, water, energy and real estate (human settlements), as well as 
impacts on human health and ecosystems. These are due to a number of mechanisms 
involving an increase in average temperature, sea level rise, and (extreme) weather patterns 
and events like rainfall, storms, heat waves and hurricanes‖ (Botzen and van den Bergh 2012, 
373). 



[17] Shreekant Gupta 

)t(Y)]t()[t()t(Q  1  (1) 

 1)t(L)t(K)t(A)t(Y  (2) 

])t(T)t(T/[)t( ATAT
2

2111   (3) 

)t(I)t(C)t(Q   (4) 

Here, the damage function )t(  represents one minus the fraction of 

aggregate output lost due to climate change, i.e., net output. For T = 0, 
)t(  = 1 (no climate damage) whereas for large temperature changes )t(  

approaches zero (maximum damage). 14  It is evident then the damage 
function plays a central role in the DICE model and in BCA more 
generally. It maps the impact of increase in temperature due to an increase 
in GHG concentration into lost output. As I show below, alternative 
functional forms of the damage function can lead to very different results. 

The climate module in DICE tracks stocks and flows of carbon in 3 
reservoirs: lower atmosphere, shallow ocean, and deep ocean. ―The climate 
equations are a simplified representation that includes an equation for 
radiative forcing and two equations for the climate system. The radiative 
forcing equation calculates the impact of the accumulation of GHGs on the 
radiation balance of the globe. The climate equations calculate global mean 
surface temperature ( ATT ) and the average temperature of the deep oceans 

for each time-step. These equations draw upon and are calibrated to large-
scale general circulation models of the atmosphere and ocean systems‖ 
(Nordhaus 2008, 36). 

DICE generates an optimised path of savings and reduction in GHGs over 
a planning horizon of several centuries. The objective function that is 
maximized is the discounted sum of future utility from consumption.15 The 
utility function is of the constant relative risk aversion (CRRA) form where 

                                                        
14The DICE damage function is calibrated to damages in the range of 2 to 4°C which as we 
shall see later is problematic. 

15 More formally, the objective function is 




maxT

t

)t(R)].t(L),t(c[UW

1

 where U is utility, c(t) is 

consumption in period t and L(t) is the population in period t. 
t)(

)t(R



1

1  is the discount 

factor determined by the pure rate of time preference   (also known as the utility discount 

rate). Optimal policy is the path of emissions reductions that maximizes the objective 
function and the carbon tax that achieves those reductions. 



Ecology, Economy and Society–the INSEE Journal [18] 

future utility is discounted at a constant pure rate of time preference  .16 

DICE can also be thought of as an optimal control model where GHG 
emissions are control variables,17 whereas change in global mean surface 
temperature (T) is a state variable. T is a key variable that describes the 
climate system and a change in T depicts (one dimension) climate change. 
Through policy instruments, such as carbon taxes, policymakers can 
influence GHG emissions. 

Three key features of the DICE model drive the ‗policy ramp‘ 
recommendation that was mentioned earlier: (i) the discount rate (not to be 
confused with the pure rate of time preference or the utility discount rate),18 
(ii) ECS and (iii) damage function )t( . While the choice of discount has 

profound implications for ‗urgent action‘ (or the lack of it), the latter two 
are the ones most relevant for this essay. We focus on them in the next 
section. 

 

3. WHAT DOES CLIMATE SCIENCE TELL US? 

Recall five key links that map anthropogenic climate change in an IAM: 

a) from ‗people‘ (producers/consumers) to emissions of GHGs 
b) from emissions to stocks of GHGs 
c) from GHG stocks to changes in temperature 
d) from rising temperature to climate change more broadly 
e) from climate change to human (economic) impact 

Climate science is deeply embedded in links (b), (c) and (d). While there is 
not much debate about quantification of (b), considerable uncertainty 
surrounds (c) or ECS. I discuss this in detail below and the interlinked issue 
of ―fat tails‖. The DICE model suffers from two major problems vis-à-vis 
climate science. First, it does not handle well the uncertainty vis-à-vis ECS 
and its implications for catastrophic outcomes. Second, DICE skips link (d) 
and goes directly from (c) to (e) via the damage function )t( described 

                                                        
16 Also known as the constant intertemporal elasticity of substitution (CIES) utility function 






1

1c
)c(U which can be represented by logarithmic utility clog)c(U  when  = 1. 

17 In DICE this is the emissions control rate, i.e., the fraction of emissions reduced by a climate 
policy, for example, carbon taxes. Under business as usual scenario emissions control rate is 
set to zero. 
18 The discount rate gr  (Ramsey Rule) where  is the pure rate of time preference 

(utility discount rate),  is the parameter of the CRRA/CIES utility function and g  is the 

growth rate of consumption. 



[19] Shreekant Gupta 

earlier in equation (4). This is a potential problem since it is (d) that links 
rising temperatures to climate impacts such as sea level rise and a greater 
frequency of extreme weather events. It is also here that tipping points and 
nonlinearities can be captured.19 In the absence of this link, increases in 
temperature directly enter (e) and the climate damage function in the DICE 
model gets seriously flawed. But all is not lost. In the next section I show 
how DICE can be modified to address these shortcomings. Before doing 
that, it is important to discuss the key uncertainty in climate science, 
namely, climate sensitivity and its corollary, ―fat tails‖. 

3.1. Equilibrium climate sensitivity (ECS) and the problem of “fat 
tails” 

ECS and TCR (see, footnote 5) are useful metrics summarising the 
temperature response of the climate system to an externally imposed 
radiative forcing (RF).20  While TCR is a short-run concept, ECS is the 
eventual temperature response to increases in GHG concentrations. More 
precisely, ECS is the equilibrium change in GMST following a doubling of 
the atmospheric carbon dioxide concentration (for details see Stocker et al. 
2013, TFE.6). It measures how sensitive global average temperature is to 
changes in CO2 concentration in the long run. ―It is perhaps the most 
studied and most frequently quoted summary statistic in all of climate 
science‖ (Heal and Millner 2014, 121). The problem, however, as 
mentioned at the beginning of this essay, is that despite decades of research, 
science has not been able to pin down this key parameter and there is a 
significant downside risk of it being very high. Figure 2 depicts various 
estimates of the probability distribution for climate sensitivity (Millner, 
Dietz and Heal, 2013). Rather than going into technical detail on why they 
differ, note instead that all of them indicate ―it is very unlikely that climate 
sensitivity is less than 1°C. In addition, a lot of the  weight  in  most  of  the  

                                                        
19 A tipping point is an irreversible change such as the collapse of the Western Antartic or 
Greenland ice sheets or the melting of the permafrost. See, Lemoine and Traeger (2016) for 
a discussion especially in the context of DICE. Nonlinearities imply climate damage 
functions do not follow neat quadratic, exponential, or other smooth functional forms. A 
very recent paper in Nature shows tipping points are much more imminent than previously 
believed (Lenton et al. 2019). 
20  Radiative forcing or climate forcing is the difference between insolation (sunlight) absorbed by the 

Earth and energy radiated back to space. Changes to Earth's radiative equilibrium, that cause 
temperatures to rise or fall over decadal periods, are called climate forcings. Positive radiative 
forcing means Earth receives more incoming energy from sunlight than it radiates to space. This 
net gain of energy will cause warming. Conversely, negative radiative forcing means that Earth loses 
more energy to space than it receives from the sun, which produces cooling. A system in thermal 
equilibrium has zero radiative forcing. (Wikipedia contributors 2020) 



Ecology, Economy and Society–the INSEE Journal [20] 

Figure 2: Estimated probability density functions for climate sensitivity from a 
variety of published studies collated by Meinshausen et al. (2009) 

 
Source: Millner, Dietz and Heal (2013). 

distributions is in the 2 to 4.5°C range, which is the IPCC‘s official ―likely‖ 
range for climate sensitivity. The disagreement between the estimates 
occurs in the upper tails of the distributions, that is, the data do not limit 
the estimates of the high end of climate change well at all (Allen et al. 2006; 
Roe and Baker 2007) which means that we understand little about the 
likelihood of worst case outcomes‖ (Heal and Millner 2014, 123). In 
graphical terms, most of the climate sensitivity distributions are right 
skewed – implying realisations of higher temperatures are more likely than 
low ones. In other words, the long right tails of these distributions have 
non-negligible probability. Weitzman picturesquely phrases this as ―extreme 
disaster lurking in the distant tails of distributions‖ (Weitzman 2007, 17). 

A revised estimate of ECS, based on different and newer datasets and 
expert judgement, is found in the IPCC Fifth Assessment Report (FAR) of 
2013. Using IPCC terminology for uncertainty, it is likely (66% or greater 
probability) that ECS will be in the 1.5°C to 4.5°C range. It is extremely 
unlikely (up to 5% probability) to be less than 1°C and very unlikely (up to 
10% probability) that it will be greater than 6°C. This is shown in figure 3 
where a log-normal distribution is fitted around the ―likely‖ range for 



[21] Shreekant Gupta 

Figure 3. Eventual global average warming due to a doubling of carbon dioxide 

 
Source: Adapted from Wagner and Weitzman (2015) 

climate sensitivity in IPCC FAR. To show why these probabilities are 
disconcerting we must turn our attention to so-called ―fat tails‖. 

3.2. “Fat tails” and their consequences for climate policy 

As mentioned earlier, climate sensitivity is a key indicator of the eventual 
temperature response to greenhouse gas (GHG) changes. It is likely to be in 
the range 1.5°C to 4.5°C with a best estimate of 3°C but values substantially 
higher than 4.5°C cannot be excluded. This leads us to the issue of fat tailed 
probability distributions. Whereas in a thin tailed distribution, such as the 
normal distribution, the probability of reaching the extremes of the tails 
converges to zero relatively quickly, for a fat tailed distribution the 
probability in the extreme ends converges to zero very slowly. As stated 
earlier, Weitzman was the most vocal proponent of ―fat tails‖ as applied to 
climate policy. Figure 4 shows the difference in probability for each 
standard deviation (―sigma‖) from the mean between a thin tailed normal 
and a fat tailed Pareto distribution (Nordhaus 2011). It is evident that 
assuming a thin tail distribution when it is fat tailed can lead to a serious 
underestimate of probabilities in the tails. 

As an illustration of fat tails in the context of climate change, consider the 
following point from Wagner and Weitzman (2015). Though global average 
warming of 5 or even 6°C is horrifying and unimaginable, when we 
combine ECS from IPCC with a likely 700 ppm scenario (IEA 2013) this 
doomsday scenario has a greater than 10 percent of occurring (figure 5).21 
The show the consequences of this scenario  with  just  one  fact — the last 
time   global   average   temperatures  were  about  2  to  3.5°C   above   the  

                                                        
21Like figure 3, this one fits a log normal distribution around the IPCC‘s (2013) ―likely‖ 
range for climate sensitivity. But in the specific context of GHG concentrations of 700 ppm 
CO2e. 



Ecology, Economy and Society–the INSEE Journal [22] 

Figure 4: Illustration of tails for a normal distribution and a Pareto distribution 
with scale parameter a = 1.5. 

 
Note 1: Sigma = standard deviation from mean 

Note 2: Each curve shows the probability that the variable will be greater than 
the sigma shown on the horizontal axis 

Source: Adapted from Nordhaus (2011) 

pre-industrial level (3 million years ago) sea levels were up to 20 meters 
higher than today (IPCC 2013)!22 

The implications of fat tails are extremely serious. Proponents of this view 
argue against framing climate policy in terms of BCA. Instead, they view 
climate change as a risk management problem (Risky Business Project 2014; 
Wagner and Weitzman 2015). According to them ―average projections are 
bad enough, but it‘s the small-probability, high impact events that ought to 
command particular attention. That possibility all but calls for a 
precautionary approach to climate policy‖ (Convery and Wagner 2015, 308). 
As mentioned earlier, Weitzman through his numerous papers and a book 
made the most persuasive case for going beyond BCA. He argued climate 
change is among a small list of potentially catastrophic low-probability, high 
impact events. Thus, it merits special attention that standard BCA (namely, 

                                                        
22 A 2 to 3.5°C warming above pre-industrial levels must be juxtaposed against the fact 1°C 
warming has already occurred. So another 1 to 1.5°C above current levels could potentially 
tip us over the precipice. 



[23] Shreekant Gupta 

IAMs such as Nordhaus‘ DICE model) cannot offer. In his view aggressive 
action on GHG mitigation is like buying insurance against climate 
catastrophe (which has a non-negligible probability of occurring). 

 

4. CAN ECONOMIC MODELING OF CLIMATE CHANGE 
CATCH UP WITH CLIMATE SCIENCE? 

How damaging are non-linearities and tipping points in climate change (as 
captured in fat tails) for BCA? Like most IAMs, DICE is a deterministic 
model. In order to address uncertainties (or surprises) about future costs 
and benefits, it uses ―best guesses‖ (expected values) over a hypothesized 
probability distribution.23 In other words, DICE attempts to incorporate 
abrupt climate change by calculating the expected value of low-probability, 
high-cost catastrophic damages. This is done by running Monte Carlo 
simulations after probability distributions have been assigned to various 
parameters. As Ackerman et al. (2010) put it, ―DICE addresses catastrophic 
risk in theory, only to turn it into a deterministic guess in practice‖ 
(Ackerman et al. 2010, 1658).24 Focusing attention on the damage function 

                                                        
23 In particular, it makes the standard assumption that the climate sensitivity parameter is 3 
(midpoint of the IPCC range). 
24 Cai and Lontzek (2019) develop a stochastic dynamic programming version of the DICE 
model, namely, DSICE which can, inter alia, model the impact of uncertainty about climate 
tipping points on economic policy of climate change. In an earlier paper using DSICE they 
show the uncertainty associated with the timing of stochastic tipping points indicates carbon 
taxes have to increase by at least 50% compared to the deterministic DICE 
model (Lontzek, Cai, Judd and Lenton 2015). For a rapid, high-impact tipping event, these 

Figure 5: Eventual global average warming based on passing 700 ppm CO2e 

 

Source: Adapted from Wagner and Weitzman (2015) 



Ecology, Economy and Society–the INSEE Journal [24] 

and following Ackerman et al. (2010) the DICE damage function (equation 
3) can be rewritten as 

N

N

aT

aT
d




1
 

(5) 

where d is climate damages as a fraction of world output and T  is the 
increase in temperature from the base year. The use of equation (5) 
prevents climate damages from exceeding the value of world output.25 This 
would be logical if damages only reduced current income as DICE assumes. 
However, more realistically if we assume climate damages also include loss 
of capital assets, damages can exceed 100% of annual output. The exponent 
N measures the speed with which damages increase as temperatures rise. 

As can be seen from figure 6 for the quadratic formulation used by 
Nordhaus (N = 2), damages rise gradually and less than half of global 
output is lost till T = 19°C — ―far beyond the temperature range that has 
been considered in even the most catastrophic climate scenarios‖ 
(Ackerman et al. 2010, 1660). In contrast, as N increases, half of world 
output is lost at temperatures of about 7°C for N= 3; 4.5°C for N= 4; or 
3.5°C for N= 5, implying a sense of urgency. As N tends to infinity, Eq. (5) 
approaches a vertical line. ―This would be the appropriate shape for the 
damage function under the hypothesis that there is a threshold for an 
abrupt world-ending (or at least economy-ending) discontinuity…Thus 
choosing a larger N (―closer to infinity‖) means moving closer to the view 
that complete catastrophe sets in at some finite temperature threshold. 
Choosing a smaller N means emphasizing a gradual rise of damages rather 
than the risk of discontinuous, catastrophic change‖ (Ackerman et al. 2010, 
1660). Below I discuss the implications of a highly convex damage function 
(footnote 27). 

It is evident from this illustration that IAMs such as DICE are a tool and 
do not provide answers independent of the assumptions built into them. 
DICE can essentially yield results in line with Weitzman by manipulating 
the parameter space, modifying the damage function or introducing 
endogeneity in the model. For example, Ackerman et al. (2010) examine the 
implications of incorporating a fat-tailed probability distribution for ECS 
and modifying the damage function in the DICE model (see above). They 
conclude,  

                                                                                                                            
taxes should increase by more than 200%. They conclude the discount rate to delay 
stochastic tipping points is much lower than that for deterministic climate damages. In a 
different line of research, Kelly and Kolstad (1999) show how Bayesian learning can 
influence policy in a model with uncertainty. 
25 Unlike the case where d = aTN. 



[25] Shreekant Gupta 

Figure 6: Damage function exponents: DICE and variants  

 
Source: Ackerman et al. (2010) 

[I]f either the damage function exponent remains at or near the default value of 
2, or climate sensitivity remains at or near the default value of 3, then DICE 
projects relatively little economic harm. With plausible changes in both 
parameters, however, DICE forecasts disastrous economic decline and calls for 
rapid mitigation. The bad news is that the optimal policy recommended by a standard 
IAM such as DICE is completely dependent on the choice of key, uncertain parameters. The 
good news is that there is no reason to believe that sound economics, or even the choice of 
established, orthodox models, creates any grounds for belittling the urgency of the climate crisis. 
(Ackerman et al. 2010, 1664; emphasis added) 

I conclude this section by citing a recent significant paper by Dietz and 
Stern (2015), whose title says it all.26 They show if DICE takes into account 
three essential features of the climate problem, namely, endogeneity of 
economic growth, highly convex damage function and climate risk (i.e., 
high values of ECS), ―optimal policy‖ of DICE calls for strong controls. 
Dietz and Stern (2015) extend earlier work such as Ackerman et al. (2010) to 
incorporate endogenous drivers of growth and allow climate change to 
adversely affect these drivers. This is in sharp contrast to existing IAMs that 
are based on the RCK optimal growth model, where the major driver of 
growth is exogenous improvements in productivity and where climate 
change only impacts current output. Next, they assume the damage 
function is highly convex for a large increase in temperature like 6°C, but 

                                                        
26  ―Endogenous Growth, Convexity of Damage and Climate Risk: How Nordhaus‘ 
Framework Supports Deep Cuts in Carbon Emissions‖  



Ecology, Economy and Society–the INSEE Journal [26] 

not for smaller increases. 27  Consideration of some of the science, for 
example, on tipping points, leads them in this direction. Finally, they allow 
for explicit and large climate risks by allowing the possibility of high values 
of climate-sensitivity. 

 

5. CONCLUDING THOUGHTS 

Without exaggeration climate change poses an existential threat to human 
civilization. What is worse, this threat is more imminent than previously 
believed—the phrase ―climate emergency‖ is now part of the discourse. 
Climate science almost on a daily basis provides new evidence of this. 
Global mean temperatures are already 1.1°C more than pre-industrial levels 
and business as usual climate scenarios are truly horrifying. The period 
2015–2019 is on track the warmest five-year period ever recorded. 
Undoubtedly, deep and immediate cuts in GHG emissions are required. 
The question then is whether standard economic analysis that weighs 
benefits and costs up to the task? The answer is it is possible. Economic 
models are useful analytical tools but what they produce depends on what 
goes into them. By incorporating more realistic assumptions and giving up 
some very limiting ones, and by incorporating science more carefully as they 
build their models, economists have the potential to be in sync with calls 
for action. Economics ‗done right‘ has the potential to catch up with 
climate science. 

 

ACKNOWLEDGMENTS 

I would like to thank two anonymous reviewers for helpful comments and 
Sujayata Choudhry for her assistance. Kanchan Chopra provided the 
encouragement and impetus to write this essay and Kuntala Lahiri-Dutt 
nudged me along. Journal office was of immense help with the figures. I am 
grateful to all of them. This is to also acknowledge permission granted by 
Springer and Elsevier towards construction of figures 2 and 6, and William 
Nordhaus for figure 4.  

 

 

                                                        
27  The damage function taken from Weitzman (2012) captures tipping points better. 

Weitzman introduced convexity in the DICE damage function by adding a third term to the 

quadratic specification by Nordhaus: ])t(T)t(T)t(T/[)t( .
ATATAT

7546
3

2
2111  . The 

exponent of the third term is chosen such that at T = 6, 50 percent of output is lost. 



[27] Shreekant Gupta 

REFERENCES 

Ackerman, Frank, Stephen J. DeCanio, Richard B. Howarth, and Kristen Sheeran. 2009. 
―Limitations of Integrated Assessment Models of Climate Change.‖ Climatic Change 95: 
297-315. https://doi.org/10.1007/s10584-009-9570-x 

Ackerman, F., Elizabeth A. Stanton, and Ramón Bueno. 2010. ―Fat Tails, Exponents, 
Extreme Uncertainty: Simulating Catastrophe in DICE.‖ Ecological Economics 69 (8): 
1657–1665. https://doi.org/10.1016/j.ecolecon.2010.03.013 

Allen, Myles et al. 2006. ―Observational Constraints on Climate Sensitivity.‖ In Avoiding 
Dangerous Climate Change, edited by Hans Joachim Schellnhuber, Wolfgang Cramer, 
Nebojsa Nakicenovic, Tom Wigley, and Gary Yohe. Cambridge, UK: Cambridge 
University Press 

Botzen, W.J. Wouter, and Jeroen CJM. van den Bergh. 2012. ―How Sensitive is 
Nordhaus to Weitzman? Climate Policy in DICE with an Alternative Damage 
Function.‖ Economics Letters 117: 372-374. 
https://doi.org/10.1016/j.econlet.2012.05.032 

Cai, Yongyang, and Thomas S. Lontzek. 2019. "The Social Cost of Carbon with 
Economic and Climate Risks," Journal of Political Economy 127 (6): 2684-2734. 
https://doi.org/10.1086/701890 

Cass, David. 1965. ―Optimum Growth in an Aggregative Model of Capital 
Accumulation.‖ The Review of Economic Studies 32 (3): 233-240. 
https://doi.org/10.2307/2295827 

Comín, Francisco A. and Miguel Angel Rodríguez-Arias. 2003. ―What we know about 
the Climate System? A brief review of current research,‖ in Global Climate: Current 
Research and Uncertainties in the Climate System, edited by Xavier Rodö and Francisco A. 
Comín. New York: Springer. https://doi.org/10.1007/978-3-662-05285-3_1 

Convery, Frank J., and Gernot Wagner. 2015. ―Reflections–Managing Uncertain 
Climates: Some Guidance for Policy Makers and Researchers.‖ Review of Environmental 
Economics and Policy 9 (2): 304-320. https://doi.org/10.1093/reep/rev003 

Dietz, Simon, and Nicholas Stern. 2015. ―Endogenous Growth, Convexity of Damages 
and Climate Risk: How Nordhaus‘ Framework Supports Deep cuts in Carbon 
Emissions.‖ The Economic Journal 159 (583): 574–620. 
https://doi.org/10.1111/ecoj.12188 

Gillingham, Kenneth, William Nordhaus, David Anthoff, Geoffrey Blanford, Valentina 
Bosetti, Peter Christensen, Haewon McJeon, and John Reilly. 2018. ―Modeling 
uncertainty in integrated assessment of climate change: A multimodel 
comparison.‖ Journal of the Association of Environmental and Resource Economists 5 (4): 791-
826. https://doi.org/10.1086/698910 

Heal, Geoffrey, and Antony Millner. 2014. ―Uncertainty and decision making in Climate 
Change Economics.‖ Review of Environmental Economics and Policy 8 (1): 120-137. 
https://doi.org/10.1093/reep/ret023 

https://doi.org/10.1007/s10584-009-9570-x
https://doi.org/10.1016/j.ecolecon.2010.03.013
https://doi.org/10.1016/j.econlet.2012.05.032
https://doi.org/10.1086/701890
https://doi.org/10.2307/2295827
https://doi.org/10.1007/978-3-662-05285-3_1
https://doi.org/10.1093/reep/rev003
https://doi.org/10.1111/ecoj.12188
https://doi.org/10.1086/698910
https://doi.org/10.1093/reep/ret023


Ecology, Economy and Society–the INSEE Journal [28] 

Intergovernmental Panel on Climate Change (IPCC). 2001. Climate Change 2001: The 
Scientific Basis. Contribution of Working Group 1 to the Third Assessment Report of the 
Intergovernmental Panel on Climate Change. Cambridge, UK: Cambridge University Press. 

Intergovernmental Panel on Climate Change (IPCC). 2013. ―Summary for 
Policymakers,‖ in 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, UK: Cambridge University Press. 

Intergovernmental Panel on Climate Change (IPCC). 2014. Climate Change 2014: 
Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the 
Intergovernmental Panel on Climate Change. IPCC, Geneva. 
https://doi.org/10.1017/CBO9781107415416 

International Energy Agency (IEA). 2013. World Energy Outlook. Available at 
http://www.wec-france.org/DocumentsPDF/Evenements/06-12-13_AIE.pdf 

Kelly, David L., and Charles Kolstad. 1999. ―Bayesian learning, growth, and pollution.‖ 
Journal of Economic Dynamics and Control 23(4): 491–518.Koopmans, Tjalling C. 1965. ―On 
the Concept of Optimisation Economic Growth‖ in The Economic Approach to 
Development Planning. 225-287. Chicago: Rand McNally. https://doi.org/10.1016/S0165-
1889(98)00034-7 

Lemoine, Derek, and Christian P. Traeger. 2016. ―Economics of Tipping the Climate 
Dominoes.‖ Nature Climate Change 6 (5): 1-23. https://doi.org/10.1038/nclimate2902 

Lenton, Timothy M., Johan Rockström, Owen Gaffney, Stefan Rahmstorf, Katherine 
Richardson, Will Steffen, and Hans Joachim Schellnhuber. 2019. ―Climate Tipping 
Points - Too Risky to Bet Against.‖ Nature 575: 592-595. 
https://doi.org/10.1038/d41586-019-03595-0 

Lontzek, Thomas S., Yongyang Cai, Kenneth L. Judd, and Timothy M. Lenton. 2015. 
―Stochastic Integrated Assessment of Climate Tipping Points Indicates the Need for 
Strict Climate Policy.‖ Nature Climate Change 5: 441–444. 
https://doi.org/10.1038/nclimate2570 

McKibbin, Warwick. 2014. [Review of the book ―Climate Economics: The State of the 
Art.‖ by Frank Ackerman and Elizabeth A. Stanton] Journal of Economic Literature 52 (2): 
559-562. https://doi.org/10.1257/jel.52.2.535.r14 

Meinshausen, Malte, Nicolai Meinshausen, William Hare, Sarah C. B. Raper, Katja 
Frieler, Reto Knutti, David J. Frame, and Myles R. Allen. 2009. ―Greenhouse-gas 
Emission Targets for Limiting Global Warming to 2 °C.‖ Nature 458 (7242): 1158-1163. 
https://doi.org/10.1038/nature08017 

Millner, Antony, Simon Dietz, and Geoffrey Heal. 2013. ―Scientific Ambiguity and 
Climate Policy.‖ Environmental and Resource Economics 55 (1): 21-46. 
https://doi.org/10.1007/s10640-012-9612-0 

Newbold, S. C. 2010. ―Summary of the DICE Model.‖ Improving the Assessment and 
Valuation of Climate Change Impacts for Policy and Regulatory Analysis. 
https://yosemite.epa.gov/ee/epa/eerm.nsf/vwan/ee-0564-114.pdf/$file/ee-0564-
114.pdf 

https://doi.org/10.1017/CBO9781107415416
http://www.wec-france.org/DocumentsPDF/Evenements/06-12-13_AIE.pdf
https://doi.org/10.1016/S0165-1889(98)00034-7
https://doi.org/10.1016/S0165-1889(98)00034-7
https://doi.org/10.1038/nclimate2902
https://doi.org/10.1038/d41586-019-03595-0
https://doi.org/10.1038/nclimate2570
https://doi.org/10.1257/jel.52.2.535.r14
https://doi.org/10.1038/nature08017
https://doi.org/10.1007/s10640-012-9612-0
https://yosemite.epa.gov/ee/epa/eerm.nsf/vwan/ee-0564-114.pdf/$file/ee-0564-114.pdf
https://yosemite.epa.gov/ee/epa/eerm.nsf/vwan/ee-0564-114.pdf/$file/ee-0564-114.pdf


[29] Shreekant Gupta 

Nordhaus, William D. 1977a. ―Strategies for the Control of Carbon Dioxide.‖ Cowles 
Foundation Discussion Paper No. 443, Yale University.  

Nordhaus, William D. 1977b. ―Economic Growth and Climate: The Carbon Dioxide 
Problem.‖ The American Economic Review 67 (1): 341-346. 

Nordhaus, William D. 1992a. ―The ‗DICE‘ Model: Background and Structure of a 
Dynamic Integrated Climate-economy Model of the Economics of Global Warming.‖ 
Cowles Foundation Discussion Paper No. 1009, Yale University. 

Nordhaus, William D. 1992b. ―An Optimal Transition Path for Controlling 
Greenhouse Gases.‖ Science 258 (5086): 1315-1319. 
https://doi.org/10.1126/science.258.5086.1315 

Nordhaus, William D. 2007. ―A Review of the Stern Review on the Economics of 
Climate Change.‖ Journal of Economic Literature 45: 686-702. 
https://doi.org/10.1257/jel.45.3.686 

Nordhaus, William D. 2008. A Question of Balance: Weighing the Options on Global Warming 
Policies, New Haven, CT: Yale University Press. 

Nordhaus, William D. 2010. ―Economic Aspects of Global Warming in a post-
Copenhagen Environment.‖ Proceedings of the National Academic of Sciences 107 (26): 11721-
11726. https://doi.org/10.1073/pnas.1005985107  

Nordhaus, William D. 2011. ―The Economics of Tail Events with an Application to 
Climate Change.‖ Review of Environmental Economics and Policy 5 (2): 240-257. 
https://doi.org/10.1093/reep/rer004 

Nordhaus, William D. 2013. ―Integrated Economic and Climate Modeling‖ in Handbook 
of Computable General Equilibrium Modeling. First Edition, 1069-1131. Amsterdam: 
Elsevier. https://doi.org/10.1016/B978-0-444-59568-3.00016-X 

Nordhaus, William D., and Zili Yang. 1996. ―A Regional Dynamic General Equilibrium 
Model of Alternative Climate Change Strategies.‖ American Economic Review 86 (4): 741-
765. 

Pindyck, Robert S. 2013. ―Climate Change Policy: What Do the Models Tell 
Us?‖ Journal of Economic Literature 51 (3): 860-72. https://doi.org/10.1257/jel.51.3.860 

Pindyck, Robert S. 2017. ―The Use and Misuse of Models for Climate Policy.‖ Review of 
Environmental Economics and Policy 11 (1): 100-114. https://doi.org/10.1093/reep/rew012 

Ramsey, F. P. 1928. ―A Mathematical Theory of Saving.‖ Economic Journal 38 (152): 543-
559. https://doi.org/10.2307/2224098 

Risky Business Project. 2014. ―Risky Business: The Economic Risks of Climate Change 
in the United States.‖  
https://riskybusiness.org/site/assets/uploads/2015/09/RiskyBusiness_Report_WEB_
09_08_14.pdf 

Roe, Gerard H., and Marcia B. Baker. 2007. ―Why is Climate Sensitivity So 
Unpredictable?‖ Science 318 (5850): 629-632. https://doi.org/10.1126/science.1144735 

https://doi.org/10.1126/science.258.5086.1315
https://doi.org/10.1257/jel.45.3.686
https://doi.org/10.1073/pnas.1005985107
https://doi.org/10.1093/reep/rer004
https://doi.org/10.1016/B978-0-444-59568-3.00016-X
https://doi.org/10.1257/jel.51.3.860
https://doi.org/10.1093/reep/rew012
https://doi.org/10.2307/2224098
https://riskybusiness.org/site/assets/uploads/2015/09/RiskyBusiness_Report_WEB_09_08_14.pdf
https://riskybusiness.org/site/assets/uploads/2015/09/RiskyBusiness_Report_WEB_09_08_14.pdf
https://doi.org/10.1126/science.1144735


Ecology, Economy and Society–the INSEE Journal [30] 

Stanton, Elizabeth A., Frank Ackerman and Sivan Kartha. 2009. ―Inside the Integrated 
Assessment Models: Four issues in climate economics.‖ Climate and Development 1 
(2):166-184. https://doi.org/10.3763/cdev.2009.0015 

Stern, David I., Frank Jotzo, and Leo Dobes. 2013. ―The Economics of Global Climate 
Change: A Historical Literature Review.‖ CCEP Working Paper No. 1307. Available at 
https://ageconsearch.umn.edu/record/249412/files/ccep1307.pdf.  

Stern, Nicholas. 2014. ―Ethics, Equity and the Economics of Climate Change Paper 1: 
Science and Philosophy.‖ Economics and Philosophy 30 (3): 397-444. 
https://doi.org/10.1017/S0266267114000297 

Stocker, Thomas F., Qin Dahe, Gian-Kasper Plattner. 2013. ―Technical Summary.‖ In 
Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth 
Assessment Report of the Intergovernmental Panel on Climate Change, edited by Stocker, T.F., D. 
Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and 
P.M. Midgley. USA: Cambridge University Press. 

Wagner, G., and Martin L. Weitzman. 2015. Climate Shock. USA: Princeton University 
Press. https://doi.org/10.1515/9781400865475 

Weitzman, Martin L. 2007. ―Role of Uncertainty in the Economics of Catastrophic 
Climate Change.‖ AEI-Brookings Joint Center Working Paper 07-11. Available at 
https://pdfs.semanticscholar.org/9ebd/87c605d31a59545cca2699a584068515522c.pdf. 

Weitzman, Martin L. 2009. ―On Modeling and Interpreting the Economics of 
Catastrophic Climate Change.‖ The Review of Economics and Statistics 91 (1): 1–19. 
https://doi.org/10.1162/rest.91.1.1 

Weitzman, Martin L. 2011. ―Fat Tailed Uncertainty in the Economics of Catastrophic 
Climate Change.‖ Review of Environmental Economics and Policy 5 (2): 275-292. 
https://doi.org/10.1093/reep/rer006 

Weitzman, Martin L. 2012. ―GHG Targets as Insurance Against Catastrophic Climate 
Damages.‖ Journal of Public Economic Theory 14 (2): 221–44. 
https://doi.org/10.1111/j.1467-9779.2011.01539.x 

Weyant, John. 2014. ―Integrated Assessment of Climate Change: State of the 
Literature.‖ Journal of Benefit-Cost Analysis 5 (3): 377-409. https://doi.org/10.1515/jbca-
2014-9002 

Weyant, John. 2017. ―Some Contributions of Integrated Assessment Models of Global 
Climate Change.‖ Review of Environmental Economics and Policy 11 (1): 115-137. 
https://doi.org/10.1093/reep/rew018 

Wieners, Claudia Elisabeth. 2018. ―God does not play DICE – but Bill Nordhaus does! 
What can models tell us about the economics of climate change? ‖ Climate: Past, Present 
& Future (blog) December 3, 2018. 
https://blogs.egu.eu/divisions/cl/2018/12/03/god-does-not-play-dice-but-bill-
nordhaus-does-what-can-models-tell-us-about-the-economics-of-climate-change/ 

Wikipedia contributors. 2020. ―Radiative forcing.‖ Wikipedia, The Free Encyclopedia. 
Accessed January 26, 2020. 
https://en.wikipedia.org/w/index.php?title=Radiative_forcing&oldid=936693346  

https://doi.org/10.3763/cdev.2009.0015
https://ageconsearch.umn.edu/record/249412/files/ccep1307.pdf
https://doi.org/10.1017/S0266267114000297
https://doi.org/10.1515/9781400865475
https://pdfs.semanticscholar.org/9ebd/87c605d31a59545cca2699a584068515522c.pdf
https://doi.org/10.1162/rest.91.1.1
https://doi.org/10.1093/reep/rer006
https://doi.org/10.1111/j.1467-9779.2011.01539.x
https://doi.org/10.1515/jbca-2014-9002
https://doi.org/10.1515/jbca-2014-9002
https://doi.org/10.1093/reep/rew018
https://blogs.egu.eu/divisions/cl/2018/12/03/god-does-not-play-dice-but-bill-nordhaus-does-what-can-models-tell-us-about-the-economics-of-climate-change/
https://blogs.egu.eu/divisions/cl/2018/12/03/god-does-not-play-dice-but-bill-nordhaus-does-what-can-models-tell-us-about-the-economics-of-climate-change/
https://en.wikipedia.org/w/index.php?title=Radiative_forcing&oldid=936693346

