Incentives and Stability of Internationa

NOTA DI
LAVORO
97.2011
Incentives and Stability
of International Climate
Coalitions:
An Integrated Assessment
By Valentina Bosetti, Fondazione Eni

Enrico Mattei and CMCC, Italy
Carlo Carraro, Fondazione Eni Enrico
Mattei, University of Venice, CEPR,
CESifo and CMCC, Italy
Enrica De Cian, Fondazione Eni Enrico
Mattei and CMCC, Italy
Emanuele Massetti, Fondazione Eni
Massimo Tavoni, Fondazione Eni
Climate Change and Sustainable Development Series
Editor: Carlo Carraro
Incentives and Stability of International Climate Coalitions:
An Integrated Assessment
By Valentina Bosetti, Fondazione Eni Enrico Mattei and CMCC, Italy
Carlo Carraro, Fondazione Eni Enrico Mattei, University of Venice, CEPR,
CESifo and CMCC, Italy
Enrica De Cian, Fondazione Eni Enrico Mattei and CMCC, Italy
Emanuele Massetti, Fondazione Eni Enrico Mattei and CMCC, Italy
Massimo Tavoni, Fondazione Eni Enrico Mattei and CMCC, Italy
Summary
This paper analyses the incentives to participate in and the stability of international climate
coalitions. Using the integrated assessment model WITCH, the analysis of coalitions’
profitability and stability is performed under alternative assumptions concerning the pure
rate of time preference, the social welfare aggregator and the extent of climate damages. We
focus on the profitability, stability, and “potential stability” of a number of coalitions which
are “potentially effective” in reducing emissions. We find that only the grand coalition under
a specific sets of assumptions finds it optimal to stabilise GHG concentration below
550 ppm CO2-eq. However, the grand coalition is found not to be stable, not even
“potentially stable” even through an adequate set of transfers. However, there exist
potentially stable coalitions, but of smaller size, which are also potentially environmentally
effective. Depending on the assumptions made, they could achieve up to 600 ppm CO2-eq.
More ambitious targets lead to the collapse of the coalition.
Keywords: Climate Policy, Climate Coalition, Game Theory, Free Riding
JEL Classification: C68, C72, D58, Q54
This paper is part of the research work being carried out by the Sustainable Development Programme
of the Fondazione Eni Enrico Mattei. The authors are grateful to OECD for financial support. In
particular, we are grateful to Romain Duval for extremely useful comments. The usual disclaimers
apply.
Address for correspondence:
Enrica De Cian
Fondazione Eni Enrico Mattei
Isola di San Giorgio Maggiore
30124 Venice
Italy
E-mail: [email protected]
The opinions expressed in this paper do not necessarily reflect the position of
Fondazione Eni Enrico Mattei
Corso Magenta, 63, 20123 Milano (I), web site: www.feem.it, e-mail: [email protected]
INCENTIVES AND STABILITY
OF INTERNATIONAL CLIMATE COALITIONS:
AN INTEGRATED ASSESSMENT
Valentina Bosettia, Carlo Carrarob, Enrica De Ciana*,
Emanuele Massettia and Massimo Tavonia
Abstract
This paper analyses the incentives to participate in and the stability of international climate coalitions.
Using the integrated assessment model WITCH, the analysis of coalitions’ profitability and stability is
performed under alternative assumptions concerning the pure rate of time preference, the social welfare
aggregator and the extent of climate damages. We focus on the profitability, stability, and “potential
stability” of a number of coalitions which are “potentially effective” in reducing emissions. We find that
only the grand coalition under a specific sets of assumptions finds it optimal to stabilise GHG
concentration below 550 ppm CO2-eq. However, the grand coalition is found not to be stable, not even
“potentially stable” even through an adequate set of transfers. However, there exist potentially stable
coalitions, but of smaller size, which are also potentially environmentally effective. Depending on the
assumptions made, they could achieve up to 600 ppm CO2-eq. More ambitious targets lead to the collapse
of the coalition.
a
Fondazione Eni Enrico Mattei and CMCC, Italy
b
Fondazione Eni Enrico Mattei, University of Venice, CEPR, CESifo and CMCC, Italy
JEL classification: C68; C72; D58; Q54.
Keywords: Climate policy; Climate coalition; Game theory; Free riding.
*Corresponding Author: [email protected]
This paper is part of the research work being carried out by the Sustainable Development Programme of
the Fondazione Eni Enrico Mattei. The authors are grateful to OECD for financial support. In particular,
we are grateful to Romain Duval for extremely useful comments. The usual disclaimers apply.
1
1. INTRODUCTION
The global public good nature of the climate change and its causes requires cooperation
among countries and broad-based participation of at least major economies is required for any
coalition to be environmentally effective. However, stable coalitions are generally small and
might well not address satisfactorily the environmental problem, especially when they deal
with a global externality as in the case of climate change (Barrett 1994, Carraro and
Siniscalco 1993, Asheim et al. 2006, Bréchet et al. 2011). When the costs and benefits of
international cooperation are large, as in the case of GHG emissions reduction, a stable
agreement is hard to achieve, reflecting the difficulty to provide sufficient participation
incentives to widely heterogeneous countries. Only when the benefits from cooperation are
small, stable coalition can succeed to sustain a large number of signatories (Barret 1994). But
in this case, they are not effective in reducing emissions considerably.
A successful international climate policy framework will have to meet two conditions, build a
coalition of countries that is potentially effective and give each member country sufficient
incentives to join and remain in this coalition. Such coalition should be capable of delivering
ambitious emission reduction even if some countries do not take mitigation action. In
addition, it should meet the target without exceedingly high mitigation costs and deliver a net
benefit to member countries as a whole. The novel contribution of this paper is mostly
methodological, but it also adds a better qualification of well-known results that are policy
relevant. The use of large scale macroeconomic models calibrated on historical data allow
generalising results that have been obtained under more specific, and at times simplifying,
assumptions and therefore it represents a complement to the broad literature on coalition
theory. To our knowledge only a few studies have used a similar approach to assess climate
coalitions. Bosello et la (2003) study the effects of different equity rules on the incentives to
cooperate using the dynamic integrated growth and climate model FEEM-RICE. Carraro et al
(2006) use a stylized integrated assessment simulation model to show how appropriate
transfers may induce almost all countries into signing a self-enforcing climate treaty.
2
Bréchet et al. (2011) compare different stability concepts using a modified version of
Nordhaus and Yang (1996) model . Nagashima et al. (2009) use the STACO model to
compare different transfer schemes and their impact on participation incentives, global
welfare and abatement efforts. However, the STACO model does not explicitly model the
nexus between economy, energy, and climate, but rather relies on reduced-form cost and
benefit functions. The ClimNeg World Simulation model used by Bréchet et al. (2011) and
Carraro et al. (2006), as well as the FEEM-RICE model, are a step ahead as they encompass
economics, climatic, and impact dimensions. However, they do not fully model the energy-
emission linkage and they neglect international interactions through the diffusion of clean
technologies developed for abatement purpose.
The modelling framework used in this paper represents a step-up over the models just
mentioned. The WITCH model, used in the present paper, has two major strengths in this
specific context. It belongs to the class of so-called integrated assessment models (IAMs), and
therefore it incorporates explicitly the gains from emission reductions in terms of avoided
climate change through regional damage functions that feed climate change back into the
economy. It has a game-theoretic structure. The 12 model regions and/or coalitions of regions
behave strategically with respect to all major economic decision variables, including emission
abatement levels, by playing a non-cooperative Nash game. Therefore, when deciding
whether or not to cooperate on GHG emission control, countries take into account how their
decisions affect all other countries, and whether these countries will cooperate or remain
outside the coalition. Mitigation options are fully modelled as investment choices in
alternative energy technologies, abatement in non-CO2 gases, and changes in deforestation
patterns. Moreover, technological change in energy efficiency and clean technologies is
endogenous and reacts to price and policy signals. Technological innovation and diffusion
processes are also subject to international spillovers; this means that the model can represent
multiple externalities, which can be partly internalized when coalitions are formed.
Bosetti et al. (2009) evaluated the potential environmental effectiveness of all 4069 coalitions
that would result from combination of the 12 regions. In particular, all coalition that could in
principle deliver the stabilisation targets commonly discussed in the policy arena were
identified. In the present paper we take stock of that analysis and, within the set of all possible
climate coalitions, we focus on those that have the potential to meet an ambitious enough
3
global mitigation target. These are coalitions whose global emission path would be consistent
with long-run stabilisation of global GHG concentration at 550 ppm CO2-eq, despite the BAU
emission pathway of non-participating regions. For this subset of coalitions, we evaluate
whether the welfare of each participating country is larger than the welfare it would obtain
from withdrawing from the coalition and free riding on other participants’ abatement efforts
(internal stability). We also checked whether there are international financial transfers that can
compensate for the free riding incentives (potential internal stability). Given the uncertainties
involved in predicting and valuing the future damages and risks from climate change, the
analysis is performed under four alternative combinations of damage and discount rate
assumptions. A low-damage case is based on the damage assessment in Nordhaus and Boyer
(2000), while a high-damage case incorporates the more recent, upward revisions made for
instance by Hanemann (2008) or Stern (2007). A low-discounting case assumes a (pure)
utility discount rate of 0.1%, in line with Stern (2007), while a high-discounting case takes the
3% value used in Nordhaus (2007). Finally we assess the effect of different weighting in the
aggregation of regions’ welfare and the effect this has on main findings.
Following a presentation of WITCH’s game-theoretic framework in Section 2, Section 3

assesses the basic individual incentives for countries to participate in coalitions. Section 4
brings together the incentive effects associated with damages and abatement costs to analyse
coalition formation and stability. Conclusions follow.
2. THE GAME THEORETIC STRUCTURE OF WITCH
The numerical analysis uses the WITCH model1, an energy-economy model that incorporates
a detailed representation of the energy sector into an inter-temporal growth model of the
global economy. The emphases on the energy sector and GHG mitigation options allows
technology-related issues to be studied within a general equilibrium framework characterised
by environmental (expected future climate change damages), economic (exhaustible natural
resources), and technology (knowledge and experience spillovers) externalities (Bosetti et al.
2006, 2008). Each region's economy is modelled in line with a Ramsey-Cass-Koopmans
model where the representative agent maximizes intertemporal welfare, by optimally
1
See www.witchmodel.org for model description and related papers.
4
choosing the investments path, given a production function for the final good, a budget
constrain and kinetic equations for capital accumulation. WITCH can simulate all degrees of
cooperation among the 12 macro-regions in which world countries are aggregated. The model
can run in a cooperative mode where global social welfare is maximised. In this case,
cooperation internalises the environmental externality. The model can also provides a
decentralised, or non-cooperative solution, by optimising the welfare of each individual
region, taking as given each other region’s choice. In between these two extremes it is
possible to model all possible combinations of smaller coalitions that coexist with non-
cooperating regions.
The scenarios obtained from the WITCH model are thus the outcome of a game in which
world regions interact in a setting of strategic interdependence.
Since players/regions are not symmetric in WITCH, both in their climate damages and
abatement costs, coalitions are characterized not only by their size but also by their
composition. The set of coalitions Γ is therefore composed of 4,095 possible combinations
among the 12 regions, including the grand coalition γGC, which comprises all players. When
formed, coalitions become players of the game. Regions that do not join the coalition are said
to behave as singletons or as free-riders.
The action of each player consists in choosing the path of investments in the economic
variables governing the economy, the energy sector, and technological progress. Investments
in the energy sector and in research and development (R&D) determine regional GHG
emissions. The accumulation of GHG emissions in the atmosphere and the effects on global
mean temperature are governed by a reduced form climate module. A climate change damage
function provides the climate feedback on the economic system, in the form of a GDP loss. In
this setting, cost-benefit analysis can be performed. Regions choose their investments trading
off the costs and the benefits, in terms of reduced damage, of their actions.
The outcome of the game for each region or group of regions is a consumption path over the
whole simulation horizon. Regions i=1,…,12 express their preferences over the outcomes of
the game using a monotonous, twice continuously differentiable utility function on discounted
per capita consumption. If t denotes time, ci,t per capita consumption, Wi the payoff of player
i defined as a function of utility of per-capita consumption, u, then
5
∂u(ci ,t (γ )) ∂ 2 u (c i ,t (γ ))
Wi (γ ) = ∑tu (ci ,t (γ ))
1
, with > 0 and < 0 , with γ being a coalition
(1 + δ t ) t ∂ci ,t ∂ c i ,t
2
in the set of all coalitions, γ ∈ Γ . In WITCH the utility function is the same for all players and
has logarithmic shape. The rate of pure time preference δt declines over the century.
Utility functions represent a complete preference ordering over a given set of goods or over an
aggregate consumption level. They can be used to assess if a given consumption level ci ,t is
preferred or not with respect to its alternative c~i ,t . Monotonicity implies that
ci ,t ≥ c~i ,t ⇔ u(ci ,t ) ≥ u (c~i ,t ) but does not allow to evaluate by how much ci ,t is preferred to c~i ,t :
utility is not a cardinal property. Therefore, utility cannot be compared directly among
players. This is an important aspect when studying the possibility of sustaining larger
coalitions by means of internal transfers.
When coalitions are formed, they act as players and choose actions to maximize joint welfare.
It is thus necessary to employ a social welfare aggregator that assigns a social preference to
every possible profile of individual preferences. We use the following social welfare
aggregator S:
S (γ ) = S (W1 (γ ), W2 (γ ), K , Wn (γ )) = ∑i =1 ∑tωi ,t u (ci ,t (γ ))

n 1
, (1)
(1 + δ t ) t
where ω i,t are weights that are used to aggregate regions. Weights, as the discount rate, can be
chosen on either normative or positive consideration. We experiment different weights and

explicitly look at their effect on results. In particular, we have considered weights equal to the
⎛ ∂u (ci ,t (γ )) ⎞
−1
inverse of marginal utility, ωi ,t = ⎜⎜ ⎟ and weights proportional to the population

⎝ ∂ci ,t (γ ) ⎠
⎟
share. The first set of weights represents a social welfare aggregation rule which is neutral
on the distribution of wealth among coalition members. The weights ω i, t linearize the
contribution of players utility functions to social welfare and avoids wealth transfers from
wealthy players to poor players and from the future to the present. Therefore, abatement effort
is distributed with the sole objective to minimize coalition’s emissions reduction costs, that is
to equalise marginal abatement costs across regions. This allocation of abatement effort, and
6
the uniqueness of the shadow value of carbon, is reminiscent of a decentralized solution in
which a global market for carbon is implemented. In fact, the social welfare aggregator that
we use produces the same actions as in a decentralized solution which internalizes the
environmental externality among coalition members and uses an international market of
carbon to distribute abatement effort.2 We also experiment with weights proportional to
population, which give more emphasis to developing countries, the countries that are going to
suffer more from climate change. This is then reflected on the overall environmental objective
of the coalition as well as in the distribution of the effort.
The WITCH model analysis assumes a non-cooperative, simultaneous, open membership,

Nash game with non-orthogonal free-riding, and allows for the possibility of international
transfers to enlarge climate coalitions. It is also possible to simulate issue linkage to increase
the profitability and stability of coalitions by letting regions to cooperate on externalities other
than the environmental one. In essence, the framework considers immediate, irreversible and
self-enforcing participation to climate change mitigation action, and abstracts from other
possible bargaining options such as delayed participation, renegotiation, sanctions or joint
negotiation in multiple areas (e.g. climate and international trade). In order to simulate
coalition formation, the model is solved as a one-shot meta-game. In the first stage players
decide on their participation and coalitions are formed. In the second stage players choose
their optimal emission levels internalizing only the environmental externality. The game is
then solved backward. In the second stage, coalition members maximize aggregate joint
welfare, whereas non participants behave as singletons and maximize individual welfare. The
equilibrium is found employing the γ-characteristic function approach (Chander and Tulkens,
1997): in the unique Nash equilibrium coalition members jointly play their best response to
non-coalition members, who adopt individually their best-reply strategies.
The game exhibits positive spillovers. When a new member joins the coalition all countries
outside the coalition are better off because they benefit from: (1) a better environment, (2)
technology spillovers (knowledge is not a club good) and (3) lower fossil fuel prices.
2
The weights ω i, t are often referred to as the Negishi weights, from Negishi (1960). Negishi weights have the
peculiar property of transforming a competitive economy problem into a centralized social planner problem and
are sometimes used by Integrated Assessment Models.
7
Let us now introduce some crucial definitions hat are later used in the paper to study
coalitions.
In order to exist, coalitions must be both profitable and stable (D’Aspremont et al., 1983;
D’Aspremont and Gabszewicz, 1986; Donsimoni et al., 1986; Carraro and Siniscalco, 1991).
A coalition γ ∈ Γ is said to be profitable if coalition members have a higher welfare than in a
scenario where the coalition is not formed (Nash equilibrium): Wi (γ ) ≥ Wi Nash ∀i ∈ γ . Where
Wi (γ ) is the welfare of player i that belongs to coalition γ , and Wi Nash is the welfare of
player i in the fully non-cooperative Nash equilibrium.
This is a necessary, but not sufficient condition for the coalition to be formed. A second
requirement concerns stability. A coalition γ is said to be stable if it is internally and
externally stable. A coalition is internally stable if signatory countries do not have the
incentive to defect and to behave non-cooperatively when other coalition members cooperate,
i.e. ∀i ∈γ Wi (γ ) ≥ Wi (γ \ i) , where Wi (γ \ i) denotes the welfare of player i when all
members but i are cooperating. A coalition is externally stable if there is no incentive to
enlarge the coalition by including non-signatory countries: ∀i ∉γ Wi (γ ) ≥ Wi (γ ∪{i}) .
Finally, a coalition γ is potentially internally stable if it can be turned into a stable coalition
through a set of self-financed transfers among coalitions members, μ = (μ1,t , μ2,t , K) with
1
∑∑ t i∈γ
μ i ,t
(1 + ri ,t ) t
= 0 ∀i ∈ γ , with ri,t denoting region i interest rate. This is the case when
coalition γ has sufficient resources to pay every member of γ its outside option.
3. ASSESSING THE DRIVERS OF PARTICIPATION IN INTERNATIONAL

CLIMATE COALITIONS
The incentives for main emitting countries to participate in climate coalitions ultimately
depend on a wide range of economic and political factors, not all of which can be captured by
model-based exercises. Bearing this caveat in mind, the analysis carried out in this paper
covers the major economic drivers of participation incentives, including the expected impacts
8
of climate change, the influence of distant impacts on current policy decisions (i.e. the
discount rate), and the costs of mitigation policies. This Section describes how each of these
three drivers are captured in the WITCH model analysis undertaken in this paper, and how
participation incentives vary across the main world regions.
3.1. Climate change impacts in WITCH
Adequate knowledge of climate change impacts is a prerequisite for well-informed climate

change mitigation policies. Alternative assumptions regarding such impacts can lead to
profoundly different policy insights, in terms of the outcome of cost-benefit analyses and the
incentives for individual regions to participate in climate coalitions.
Estimating the economic impacts of climate change raises a number of difficult issues. First,
the knowledge on the physical impacts of climate is limited, especially in relation to
nonmarket areas or impacts. Second, assigning monetary values to climate change damages is
particularly challenging. Third, the need to identify the global cost-benefit optimal emission
level requires defining an indicator of the global benefits of emission reduction in terms of
avoided damages. Therefore, impacts have to be aggregated across impacts, across regions,
which raises equity issues, and over time, which raises intergenerational issues.
Surveying the literature on impacts of climate change it is immediately clear damage

estimates vary widely and uncertainty in the size of economic impact affects mostly non-
market areas (Tol, 2002, Jamet and Jan Corfee-Morlot, 2009)3. Non-market impacts are either
not considered or underestimated. Most IAMs are indeed based on out-of-date evidence, and
most regional estimates are extrapolations from studies that have been carried out for one or
two regions, typically the United States (DICE/RICE, MERGE and PAGE) or the United
Kingdom (FUND). Most damage functions used in IAMs have not been updated according to
latest evidence on climate change, except for the PAGE model used in the Stern Review,
which takes into account new evidence on more rapid warming and large-scale changes to the
climate system (“system” surprise). As a consequence, previous modeling exercises exhibit
impacts that, on average, are quite small compared with the results described in the Stern
3
For a detailed survey of the literature on climate change impacts we refer to Appendix 3 in Bosetti et al
(2009b).
9
Review, but also compared with the latest estimates such as those reported in the UNFCCC
report (UNFCCC, 2007) or in the IPCC’s Fourth Assessment Report (IPCC, 2007).
Two alternative damage scenarios are considered here: i) a low damage scenario, embedded
in the basic version of the WITCH model, which in turn is based on the damage assessment
provided by Nordhaus and Boyer (2000); ii) a high damage scenario, which incorporates more
recent, higher damage estimates in the range of Stern (2007) and UNFCC (2007).
The WITCH model accounts for climate change damages, Ω , by means of regional functions
that describe reduced-form quadratic relationship between temperature, T, and gross world
product, GDP :
GDPi ,t
Ω i ,t =
1 + θ1iTt + θ 2 iTt 2
(2)
In the low damage scenario, for an increase in temperature below 3°C, climate change
impacts on GDP can be either positive or negative, depending on regional vulnerability and
geographic location. Above that level, damages are negative throughout the world and
increase in a quadratic relationship with temperature. The resulting pattern of regional
damages in a baseline as usual scenario shows higher estimated losses in developing
countries, in particular South Asia (including India) and Sub-Saharan Africa (Figure 1). These
two regions are expected to lose the most from climate change, especially because of higher
damages in agriculture and the increase of vector-born diseases (Sub-Saharan Africa) and
because of catastrophic climate impacts (South Asia including India). A recent review (Jamet
and Corfee-Morlot, 2009) also indicates Africa and South and Southeast Asia as the most
vulnerable regions, with GDP losses reaching more than 8% for a temperature increase above
pre-industrial levels between 2 and 2.5°C. Damage estimates for agriculture, coastal
settlements and catastrophic climate impacts are significant in Western Europe, resulting in
higher damages than in other developed regions. In China, Eastern EU countries, non-EU
Eastern European countries (including Russia), Japan-Korea, climate change up to 2.5°C
would bring small benefits, essentially because of a reduction in energy demand for heating
purposes (non-EU Eastern European countries including Russia) or positive effects on
agricultural productivity (China).
10
In the alternative, higher damage scenario the threshold above which impacts can only be
negative throughout the world is 1°C. Global climate damages are, by the end of the century
about twice as large as in the low damage scenario.
Figure 1. Regional damage functions in the baseline as usual of the WITCH model.
Low damage and high discount rate case.
1.00% Unite d State s
W e ste rn EU countrie s

0.00%
Easte rn EU countrie s

‐1.00%
A us‐Can‐Nzl
‐2.00%
P
D Japan‐Kore a
G
f
o ‐3.00%
ss
lo Non‐EU Easte rn Europe
ge
ta
n ‐4.00%
e
rc Middle East and North
e
P
A frica
‐5.00%
A frica
‐6.00%
South A sia
‐7.00% China
‐8.00% South East A sia

0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0
1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1
2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
Latin A me rica
Tim e
1. Korea is grouped with Japan, but is not an Annex I country. Source: WITCH model simulations.
Figure 2. Global damage functions in the WITCH model –
s 0 .0
s
lo
t
c
u
d - 1 .0
o
r
p
ld
r
o
w - 2 .0
s
s
ro
g
f
o - 3 .0
%
- 4 .0
- 5 .0
- 6 .0
- 7 .0
UNFCCC A R4
W IT C H L o w D a m a g e
- 8 .0
W I T C H H ig h D a m a g e
- 9 .0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5 3 .0 3 .5 4 .0
G lo b a l m e a n t e m p e ra t u re ( oC ) a b o v e p re - in d u s t ria l le v e l
Source: WITCH model simulations and UNFCCC (2007).
11
Figure 2 illustrates the time profile of the two climate damage scenarios comparing them to
climate change damages that can be extrapolated from the IPCC ranges reported in UNFCCC
(2007. The WITCH high damage function follows UNFCCC data quite closely until a 1.5°C
rise in global temperature, and increases more sharply beyond, moving closer to – but
remaining lower than – Stern’s (2007) estimates.
3.2. Discounting and pure rate of time preference
When analyzing the inter-temporal effects of climate change damages, the social discount rate
and, in particular, the pure rate of time preference plays a crucial role. There is a longstanding
controversy regarding the choice of the latter (Weitzman, 2001). Consistent with a long line
of economists (e.g. Ramsey, 1928; Harrod, 1948; Solow, 1974), Stern (2007) argues on
ethical grounds for a near-zero value, while others dismiss this assumption on the grounds
that it is inconsistent with actual individual behaviour (e.g. Nordhaus, 2007; Weitzman,
2007).
Aggregate discounted impacts are vastly increased if greater weight is assigned to the far
future, when damages are expected to be higher. Combining about hundred estimates from 27
studies to form a probability distribution for the marginal cost of carbon, Tol (2005) finds that
the median value of the social cost of carbon – an estimate of the marginal impact caused by
one additional ton of carbon – increases from $US7 to 39 per ton of carbon when the pure rate
of time preference declines from 3% to 0%, i.e. when it declines from the value used in
Nordhaus’ DICE/RICE model to that used in the Stern Review. This implies that cost-benefit
considerations would lead to very different abatement, depending on the value of pure rate of
time preference. Indeed, in our subsequent analysis we show that only if future damages are
given enough weight, i.e. a 0.1% pure rate of time preference is adopted, the grand coalition
endogenously achieve stabilisation targets that are in line with those discussed in the policy
debate, e.g. 550ppme (see Section 4).
In order to take into account the existing debate on the choice of the social discount rate, we
perform our analyses under two different assumptions regarding the pure rate of time
preference, namely 3% and 0.1%.4
4. Following Weitzman (2001), the pure rate of time preference is also assumed to be time-declining.
12
3.3. Abatement costs
The incentives to participate in climate coalitions are also shaped by mitigation policy costs.
One determinant of these costs is the shape of the aggregate marginal abatement cost curve
(MAC), which is a reduced-form relationship between the cost of abating an extra unit of
emissions and cumulative abatement. In WITCH, mitigation costs are the result of energy
switching, energy intensity improvement and innovation investments. It is possible to
compute and compare MACs across countries by running a range of global carbon tax
scenarios and report the resulting abatement. By repeating this exercise for a wide set of
carbon taxes, it is possible to draw a relationship between marginal abatement costs and
cumulative emissions abatement for the 12 regions of the model.5 Figure 3 reports the
marginal abatement costs as a function of relative abatement with respect to baseline.
WITCH’s MACs show the usual convex relation due to increasing marginal abatement costs.
A $US100 tax per ton of CO2eq achieves a cumulative CO2 abatement between 53% and
73%, depending on the region. China and the United States have relatively lower/flatter
marginal abatement curves, compared with other regions. MACs tend to become steep in all
regions beyond a tax of $US150 per ton of CO2eq.
One can also look at the cost, measured in terms of the discounted consumption loss,
alternative world carbon tax scenarios imply, Figure 4. Developing countries are found to
incur larger losses than their developed counterparts, due to their higher energy/carbon
intensity. Fossil fuel producers such as the Middle East and non-EU Eastern European
countries (including Russia) are the biggest losers, reflecting both terms-of-trade losses and
their very high energy/carbon intensity. Within the group of developed regions, Western
Europe and Japan-Korea would face smaller costs than the United States despite steeper
MACs, reflecting their lower energy/carbon intensity.
5. Each carbon tax path considered is not constant, but instead is assumed to grow over time at the rate at which
marginal abatement costs grow in the cooperative solution of the model. Obviously different dynamics of the tax
would imply different MACs. For the sake of the analysis though we find it useful to report representative
MACs. In order to focus only on abatement costs, all damages from climate change are excluded from this
exercise.
13
Figure 3. Marginal abatement cost curves (all GHGs included)
250
Panel A. Annex I regions1
Carbon tax, marginal abatement cost, 2005 $US
United States
200
Western EU countries
Eastern EU countries
Aus-Can-Nzl
150
Japan-Korea
Non-EU Eastern Europe
100
50
0
20.0 30.0 40.0 50.0 60.0 70.0 80.0
Cumulative emission reductions, %
250
Carbon tax, marginal abatement cost, 2005 $US
Panel B. Non-Annex I regions
Middle East and North Africa

200
Africa
South Asia
150 China
South East Asia
Latin America
100
50
0
20.0 30.0 40.0 50.0 60.0 70.0 80.0
Cumulative emission reductions, %
1. Korea is grouped with Japan, but is not an Annex I country.

Source: WITCH model simulations.
14
Figure 4. Discounted regional abatement costs under a range of world carbon tax scenarios
Discounted consumption loss2

1.0
Panel A. Annex I regions1
0.0
-1.0
-2.0
-3.0
-4.0
United States Western EU countries

-5.0
Eastern EU countries Aus-Can-Nzl
Japan-Korea Non-EU Eastern Europe

-6.0
20 40 60 80 100 120 140 160 180 200
Carbon Tax (2005 $US / t CO2 eq)
1.0
Panel B. Non-Annex I regions
0.0
-1.0
-2.0
-3.0
-4.0
Middle East and North Africa Africa

-5.0 South Asia China
South East Asia Latin America
-6.0
20 40 60 80 100 120 140 160 180 200
Carbon Tax (2005 $US / t CO2 eq)
1. Korea is grouped with Japan in the WITCH model, but is not an Annex I country.
2. Cumulative consumption gap relative to baseline in present value terms over 2015-2100, using a 3% annual discount rate.
The larger a region’s mitigation costs under a global carbon tax, the smaller its incentives to
participate in a climate coalition, ceteris paribus. One possible measure of free-riding
incentives that will be introduced and discussed below is the difference (in %) between a
region’s welfare (defined as the discounted sum of the logarithm of future domestic per-capita
consumption) if it free rides on a world coalition of acting countries, and its welfare if it
participates in that coalition. As Figure 5 shows, there is a strong positive relationship
15
between that synthetic indicator of free-riding incentives and the overall consumption loss
induced by a given world carbon tax – set here at $US100 per ton of CO2eq. This is because
climate coalitions are assumed to implement an efficient climate policy, i.e. to equalise
marginal abatement costs across all participating regions.6 As a result, countries that face
larger costs from a given world carbon price can expect to gain less from joining an
international coalition, and therefore have larger incentives to defect, ceteris paribus.
Figure 5. Mitigation policy costs and free-riding incentives
1.4 Middle East and North China

Free-riding incentive1
Africa
1.2
Non-EU Eastern
Europe
1.0
Aus-Can-Nzl Africa
South East Asia
0.8 Eastern EU countries
United States
Latin America
0.6
Japan-Korea South Asia
0.4
Western EU countries
0.2
0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
1. The free-riding incentive is computed as the difference in % between a region's intertemporal welfare if it withdraws from a
world coalition of acting countries (the so-called “grand-coalition”, see Section 3), and its intertemporal welfare if it participates in
the world coalition. A 0.1% annual pure rate of time preference is used to compute the present value of welfare.
2. Cumulative consumption gap relative to baseline in present value terms over 2015-2100, using a 3% annual discount rate.
4. ANALYSING COALITION FORMATION AND STABILITY
4.1. Cooperative versus non-cooperative outcomes
We start our analysis from the most environmental effective coalition, namely the grand
coalition. A world social planner maximizes the aggregate global welfare, which is defined as
6. This is a consequence of using Negishi weights to aggregate welfare.
16
the weighted sum of regional welfares, using the Negishi weights to ensure equal marginal
abatement costs worldwide. Later we see the effect of changing aggregating weights on
coalitions.7
We compare this optimal outcome with the non-cooperative outcome where each of the
12 regions is assumed to choose the optimal path of a set of choice variables (investments in
physical capital, in different energy technologies, in R&D, etc.) so as to maximize its own
social welfare function. In this framework, each region takes its decisions individually, given
the action of the other players. The outcome of this non-cooperative game is an open loop
Nash equilibrium.
Figure 6 shows the implications of the two solution concepts for global GHG emissions under
the different assumptions about the damage and discounting assumptions. In the non-
cooperative case, upper panel, world emissions grow throughout the century. Little abatement
is undertaken since individual regions do not internalise the negative externality they impose
on other regions, taking only into account the domestic ancillary benefits of their climate
policy. The choice of the pure rate of time preference or of the damage scenario, in this
instance, do not make much of a difference. In addition a higher discount rate implies not only
a higher weight on future damages – which should favour emission reductions – but also a
higher consumption path – which leads to higher emissions. When regions only see their own
damage, the relative strength of the second factor is larger, determining higher emissions.
If instead countries gather in a climate grand coalition, the environmental externality is fully
internalised, and emissions are reduced drastically (lower panel of Figure 6). In addition,
sensitivity to underlying assumptions is far greater than in the non-cooperative case. Under
these damage/discounting assumptions, the optimal emission path would be such to stabilise
long-run GHG concentration at about 546 ppm CO2-eq, when the pure rate of time preference
is 0.1%, and 676 ppm CO2eq when it is 3%. When looking at the relative effect of lower
7. In this analysis we assume cooperation is on the climate externality only. The WITCH model incorporates
other economic externalities related to the use of natural resources and to the production and diffusion of
knowledge and experience. However, this paper analyses the incentive to form climate coalitions, independently
from linkages with other issues. In that context, it is assumed that countries decide whether or not to cooperate
on the environmental externality only. Cooperation on technological externalities and on the use of natural
resources is not considered.
17
discounting and of higher damages, the former has a sizeable greater impact, in particular on
short term emissions.
Figure 6. Sensitivity of the cooperative and non-cooperative solutions to alternative damage

and discount rate assumptions
110.0
Panel A. Non-cooperative solution
Gt CO2 eq
100.0
90.0
80.0
70.0
60.0
50.0 Low damage - High discount rate

Low damage - Low discount rate
40.0 High damage - Low discount rate

High damage - High discount rate
30.0
2005
2010
2015
2020
2025
2030
2035
2040
2045
2050
2055
2060
2065
2070
2075
2080
2085
2090
2095
2100
80.0
Panel B. Cooperative solution
Gt CO2 eq
70.0
60.0
50.0
40.0
30.0
Low damage - High discount rate
Low damage - Low discount rate
20.0
High damage - Low discount rate
High damage - High discount rate
10.0
2005
2010
2015
2020
2025
2030
2035
2040
2045
2050
2055
2060
2065
2070
2075
2080
2085
2090
2095
2100
The difference between the cooperative and non-cooperative outcomes gives an indication of
the gains from cooperation, which are substantial. The theory of self-enforcing agreements
teaches us that the larger the potential benefits of cooperation, the stronger the incentive to
free ride (Barrett 1994). The damage and discounting assumptions also affect the size of
18
cooperation gains. A high discount rate reduces the benefits from cooperation, the more so
when damages are low.
4.2. Environmental effectiveness, profitability and stability of climate coalitions

With 12 regions the number of possible coalition is very large. However, we focus on a subset
of large coalitions that would have the potential to endogenously produce targets commonly
discussed in the policy arena (e.g. in line with the 600 and 550 ppm CO2 eq 2100 atmospheric
targets). We perform the cost-benefit analysis of each of these coalitions under different
assumptions on the pure rate of time preference, damage scenario, and weighting
assumptions.8 We first discuss the environmental effectiveness of this subset of coalition and
then look at their profitability and stability.
Among the larger coalitions, only the grand coalition is found to stabilise GHG concentration
below 550 ppm CO2-eq by 2100. This is illustrated in Table 1, which shows the
environmental performance of the grand coalition together with 6 large coalitions that have
the technical potential to meet similar targets at the 2100 horizon. The composition of these
coalitions suggests that large emitters such as China and India are to be included. We always
exclude Sub-Saharan Africa from all these sub-coalitions as it is realistic to assume that, if
any other region will stay out, then Sub-Saharan Africa will necessarily follow, invoking
equity arguments such as the right to grow. Even if only Sub-Saharan Africa behaves as a
singleton, the 550 ppm CO2-eq target in 2100 is no longer reached. Leaving an additional
region out of the coalition raises GHG concentration above the target 600 ppm.
8
In what follows we only report results for the high damage scenario which is most conducive to significant
emission reductions by the coalition considered. If a coalition is not environmentally effective in this case, it
cannot be either under lower damages.
19
Table 1. Analysis of the environmental achievements of potentially effective coalitions, cost-benefit mode,
high-damage/low-discounting case
Overall GHG concentration (ppm CO2 eq)

2050 2100
Non-participating regions:
None (Grand coalition) 507 546
Africa 518 603
Africa, Latin America 532 612
Africa, Non-EU Eastern Europe 531 603
Africa, Middle East and North Africa 529 609
Africa, South East Asia 526 598
Africa, South East Asia, Non-EU Eastern Europe 529 603
Two main factors explain the failure of smaller coalitions to achieve the 550 ppme
concentration target:9
1. Smaller coalitions internalise only a fraction of the environmental externality. The

fraction of damage that is not internalised drives a gap between the coalition and the
global marginal benefit of emission reduction. The larger this gap, the smaller the
emission cut that the coalition achieves. However, there is not a clear relationship
between coalition size, coalition marginal damage, and emission reductions, because
damages are not evenly distributed among world regions. This implies that the
inclusion in the coalition of regions facing larger damages from climate change
increases coalition’s marginal damage, and the optimal level of emission cuts, ceteris
paribus.
2. As coalitions get less inclusive, the number of free-riders obviously increases; these
countries might simply keep emission unchanged, or even undertake some small
abatement due to the lower price of carbon-free technologies, but most likely they
rather increase emissions and undo some of the coalition reduction.
9
One assumption is crucial here and may alter these results: the absence of negative emissions technologies, or
any other technology that might alter the climate (i.e. geo-engineering). If one assumed that by means of
bioenergy and carbon capture and sequestration, or direct CO2 capture, or other technologies that alter the
incoming solar radiation, cooperating countries could dramatically change the climate, than the requirements on
the dimension and composition of a coalition to endogenously produce the 550 ppm CO2eq by 2100 target
would be substantially different.
20
These two forces undermining the achievement of sub-coalitions can be illustrated by looking
at a coalition structure under which only Sub Saharan Africa does not participate and behaves
as a singleton (Figure 7). First, leaving out Sub-Saharan Africa means not internalising a large
share of the global damage (see Figure 1, together with South Asia, Sub Saharan countries
would suffer the largest damage). As a consequence, such coalition would achieve
significantly lower abatement effort compared to the grand coalition. Second, the emissions of
Sub Saharan Africa itself increase dramatically when it behaves as a singleton. It is worth
noting that, although the emissions of the Sub-Saharan African region are larger than in the
grand coalition, they are lower than in the non-cooperative scenario. Although Africa could
emit as much as in the non-cooperative scenario or even more, it does not find it optimal to do
so. The reason is the presence of international technology transfers from the cooperative bloc
that reduces the costs of carbon-free technologies outside the coalition itself.10
Figure 7. An illustration of free-riding incentives: the case of Africa

Difference in GHG emissions between the grand coalition without Africa and the grand coalition, %
4.0 70.0
% %
Other regions (All except Africa) (left scale)
3.5 Africa (right scale) 60.0
3.0
50.0
2.5
40.0
2.0
30.0
1.5
20.0
1.0
0.5
10.0
0.0 0.0
2015 2020 2025 2030 2035 2040 2045 2050
An important factor in shaping the effectiveness of coalition is the weight given to future
damages by setting the pure rate of time preference. A higher rate implies, for the grand
coalition, an addition of GHG concentration in the atmosphere equal to 125 ppm CO2-eq
10
Depending on the stringency of the target and the characteristic of the free riding countries either the energy
market or the innovation effect may prevail, see Bosetti and DeCian (2011) for a thorughout discussion of these
competing effects.
21
compared to the low discounting case approximately equivalent to an additional warming of
0.5°C. Smaller coalitions, already reaching more than 600 ppm CO2-eq with a low discount
rate, would lead to an increase of 80 ppm CO2-eq or more.
Finally, the coalition effectiveness is obviously affected by the weights used to aggregate
different regions when maximizing aggregate welfare. The results discussed so far are based
on Negishi weights, which implies a cost-efficient allocation of abatement within the
coalition. Table 2 looks again at the grand coalition and compares the solution based on
Negishi weights with that based on a weighting scheme more favourable to developing
countries, proportional to the population share. This aggregation scheme increases the
environmental effectiveness of the coalition in the short and medium term, and also in the
long term with a sufficiently high discount rate. However, with low discount rate the
population social welfare aggregator yields higher GHG emissions concentrations in 2100.
Table 2. Analysis of the environmental achievements of potentially effective coalitions, cost-benefit mode,
high-damage/low and high discounting using different weights
Overall GHG concentration (ppm CO2 eq)

2030 2050 2100
Negishi weight- 0.1% 480 507 546
Population weight - 0.1% 473 502 556
Negishi weight - 3% 506 574 672
Population weight- 3% 489 538 622
International climate coalitions need not only to be environmentally effective, but should also
be self-enforcing. In technical terms, this means that a coalition should be profitable and
stable, or at least potentially stable. As noted in Section 2, a coalition is profitable if each
cooperating player has a welfare larger than that she would get in the non-cooperative
scenario. A coalition is internally stable if the welfare of each participating region is larger or
equal to the welfare she would obtain from staying out of the coalition and free riding on
participants’ abatement efforts. As an example, when we check the stability of the grand
coalition we need to run simulations in which either of the 12 regions is assumed to deviate,
22
while the others continue to cooperate. By comparing discounted consumption in the
cooperation and non-cooperation case and we can then check whether and how many
countries have an incentive to abandon the grand coalition. Although not stable, a coalition
might be Potentially Internally Stable (PIS) if there is a transfer scheme that gives each
member at least her free-riding pay-off and shares the remaining surplus.
For different coalitions Table 3 reports profitability, internal and potential internal stability.
As expected, all large coalitions considered in the analysis are not profitable, as there is
always at least one region, namely China and Sub-Saharan Africa when it applies, which is
worse off than in the non-cooperative case.
Table 3. Profitability and Stability of potentially effective coalitions
Non-participating regions Profitability Internal Stability Potential Stability

Low PRTP (0.1%) - Negishi Weighted
None (Grand coalition) NO (Africa, China) NOT STAB (All) NOT PIS
Africa NO (China) NOT STAB (All) NOT PIS
Africa , Latin America NO (China) NOT STAB (All) PIS
Africa , Non-EU Eastern Europe NO (China) NOT STAB (All) PIS
Africa , Middle East and North Africa NO (China) NOT STAB (All) PIS
Africa , South East Asia NO (China) NOT STAB (All) PIS
Africa , South East Asia, Non-EU Eastern Eu NO (China) NOT STAB (All) PIS
High PRTP (3%) - Negishi Weighted
NOT STAB (All but China and
None (Grand coalition) YES Latin America) PIS
Low PRTP (0.1%) - Population Weighted
NO (All but Africa, South Asia,
None (Grand coalition) South-East Asia) NOT STAB (All but Africa) NOT PIS
In parenthesis countries for which coalitions are not profitable or internally stable
Last column in table 3 report results on whether coalition can be stabilized through transfers,
i.e. they are potentially internally stable. We do this by checking if the aggregate residual
surplus of consumption in the coalition is greater than the sum of the discounted consumption
gains that countries have when they free-ride . We find that the grand coalition is not PIS. The
aggregate, discounted surplus from cooperation is equal to 477 USD trillions over the 2005-
2100 time horizon, while the sum of the gains from free-riding is equal to 680 USD trillions.
We then test the coalition that includes all regions but Africa, the most environmental
effective of all partial coalitions. We find that this coalition also is not potentially internally
23
stable, but the gap is now only 2% of the aggregate discounted consumption gain of the
coalition. The smaller PECs are PIS. Therefore we find that the lowest level of GHG
concentration that can be achieved by a stable coalition is slightly above 518 ppm CO2-eq in
2050 and around 600 ppm CO2-eq in 2100.
To provide greater insight on internal stability, and how this interplays with crucial
assumptions on the aggregation and discounting choices, Figure 8 reports free-riding
incentives across regions, based on the difference in welfare between free riding on and
participating to the grand coalition. Let us start by looking at the blue bars, that refer to the
central case of 0.1% discounting and Negishi weights.
The Middle East-North Africa region, China, the rest of Africa, and non-EU Eastern
European countries are estimated to have the largest incentive to free ride. By contrast,
developed countries have the lowest free-riding incentives, with the exception of the
Australia-Canada-New Zealand region.
Figure 8. Estimated ranking of free-riding incentives across regions

% Difference in welfare per capita between free-riding on and participating in the grand coalition
42%
25%
20%
15%
10%
0.1% Negishi Weighted
5% 0.1% Population Weighted
3% Negishi Weighted
0%
‐5%
24
What happens if the pure rate of time preference used in the analysis is 3% instead of 0.1%
(see second row in Table 3)? Not only each region's welfare is larger in the grand coalition
than in the non-cooperative case, i.e. the grand coalition is profitable, but the incentives to
free ride are now much smaller, although the coalition is not stable, nor internally stable. We
find that the discounted surplus, in consumption terms, generated by cooperation (133 USD
trillions) is four times greater than the discounted aggregate surplus from free-riding (35 USD
trillions).
This result confirms a well-established result in coalition theory. When gains from
cooperation are large, as in the 0.1% case, free riding incentives are also likely to be high. On
the contrary, when gains from cooperation are small, as in the 3% case, free riding incentives
are reduced significantly. Figure 8 reiterates this by showing incentive to free ride when the
discount rate is higher, green bars There is a clear message that emerges from our analysis.
Cooperation is indeed possible, profitable and potentially stable, but only if the environmental
target is moderate (at least compared to what currently discussed in the policy debate), i.e.
around 600 ppm CO2-eq by 2100.
Finally, we test how using a different set of weights affects the results (see the last row in
Table 3). When using population-based weights, the grand coalition is profitable only for
Africa, South Asia, and South-East Asia and it is not internally stable (only Africa sees
welfare gains from cooperation that compensate incentives to deviate). Figure 8 (red bars)
shows that when population weights are used, the regional pattern of free-riding incentives is
reversed with respect to the cost-effective abatement allocation implemented through Negishi
weights. First, with population weights the abatement allocation is no longer the efficient one.
In addition, perceiving a larger damage, the grand coalition abates more. The effort is shared
across all members, which all abate more compared to the Negishi aggregation, with the
exception of Africa, South Asia, and South-East Asia. These regions have two characteristics.
Because of their high population they receive the largest weights in the social welfare
function, together with China. They have the largest benefits from cooperation, and in fact,
cooperation results profitable for them. This characteristics lead to a high benefit-cost ratio,
which explains their very low free riding incentives. This result suggests that a different
allocation of the effort can have important implication on countries incentives to participate.
25
5. Conclusions
This paper has studied the incentives to participate in and the stability of international climate
coalitions using the integrated assessment model WITCH. The WITCH model has a game-
theoretic structure in which different degrees of international cooperation can be simulated.
When countries decide whether or not to cooperate on GHG emission control, they take into
account how their decisions affect all other countries, and whether these countries will
cooperate or remain outside the coalition. The optimal level of abatement in each coalition has
been derived in a cost-benefit framework. The main incentives to participate in climate
coalitions mainly depend in WITCH on two major economic drivers, namely the abatement
costs incurred and the damages avoided both within and outside a coalition. We have
performed the analysis of coalitions profitability and stability under four alternative
combinations of damage and discount rate and assessed the effect of two different schemes of
aggregating welfare across countries. The high damage-low discount rate combination is the
most conducive to cooperation for emissions control because it increases in both directions
the size of expected present value of climate change damages. The low damage-high discount
rate combination is instead the less conducive to international cooperation. The weighting
scheme that is proportional to population size also increases the effort of the coalition by
giving greater weight to developing countries, where most of climate damages are projected
to occur. We have confined the analysis of profitability and stability only to the subset of
coalitions that could attain effective emission reductions, defined as those that have the
potential to stabilise global GHG concentration between 550 and 600 ppm CO2-eq. The focus
on a subset of policy-relevant coalitions is particularly convenient and avoids uninteresting
cumbersome numerical exercises. When account is made for free-riding behaviours of non-
participating regions, only a very broad international coalition excluding no region other than
Sub-Saharan Africa could achieve meaningful stabilisation targets by 2100.
Cost-benefit analysis suggests that only the grand coalition finds it optimal as a whole to
stabilise overall GHG concentration below 550 ppm CO2-eq in the high-damage/low-
discounting case. Smaller coalitions, including the grand coalition excluding Africa, achieve
less ambitious targets, above 600 ppm CO2-eq. This is because they do not fully internalise
the global environmental externality and allow a larger number of (non-participating)
countries to free ride. Although the grand coalition as a whole has an incentive to achieve the
26
550 ppm CO2eq target, it is not internally stable. Most regions gain more from non-
participation than from participation to the grand coalition. This is true also for smaller
coalitions. The grand coalition is not potentially internally stable (PIS) either, i.e. no set of
international financial transfers can be found that would offset the free-riding incentives of all
participating countries simultaneously. This is because the overall welfare gain from the grand
coalition relative to the non-cooperative outcome is not large enough to give each country her
free-riding pay off. After compensating all losers in the coalition, the remaining coalition
surplus is too small to offset free-riding incentives. The coalition that includes all regions but
Africa is also not PIS, but the gap between the consumption level after the redistribution of
cooperation gains is only 2% lower than the consumption level that regions achieve when
they do not cooperate. All other analysed coalition, that can attain around 600 ppm CO2-eq
are PIS.
There is a clear message that emerges from our analysis. Cooperation is indeed possible and
profitable, but only if the environmental target is moderate (at least compared to what
currently discussed in the policy debate). In fact, stability, even potential, is ensured only
when the coalition stabilises GHG concentrations in 2100 above 600 ppm CO2eq. Sensitivity
of environmental effectiveness, stability concepts, and free riding incentive to different
manners of weighting regional welfare in the coalition indicate that more ambitious targets
become optimal when regions with high damages have larger influence. We also show that
different allocation of the effort could have important implication on countries incentives to
participate.
Our findings are subject to a number of limitations. Even though some sensitivity analysis has
been carried out to assess the robustness of the main results, it should be acknowledged that
the model-based analysis relies on strong assumptions. In particular, there are wide
uncertainties in practice surrounding future emission trends,11 the market and non-market
impacts from climate change, the likelihood and effect of catastrophic risks, and the cross-
country distribution of these damages and risks.
Furthermore, the analysis focuses on immediate, irreversible and self-enforcing participation

to mitigation action, thereby abstracting from other possible bargaining options including
11. For instance, projected world BAU emission growth is somewhat higher in WITCH than in the OECD
model ENV-Linkages as featured in Burniaux et al. (2008) (100% versus 85% over the period 2005-50).
27
e.g. delayed participation, renegotiation, sanctions or joint negotiation in multiple areas
(e.g. linking climate and international trade negotiations. If feasible, these alternative
bargaining options which might yield different results (see Carraro and Massetti, 2010). For
instance, a major emitting country may have greater participation incentives than found here
if it expects its withdrawal to prevent the formation of any coalition.
The co-benefits from mitigation action, e.g. in terms of human health, energy security or
biodiversity, are not taken into account. Other studies suggest that such co-benefits are large,
although the participation incentives they provide are dampened by the fact that some of these
co-benefits could be reaped through direct policy action – in particular, local air pollution
might be reduced at a lower cost through direct policy action than through reductions in GHG
emissions (Bollen et al. 2009; Burniaux et al. 2008).
Removal of fossil fuel subsidies, one of the few policies to yield potentially both climate and
economic benefits, is also omitted from the analysis. Insofar as phasing out subsidies would
bring an economic gain and lower the carbon intensity of a number of (mainly developing)
countries, incentives to participate in international mitigation action could improve.
Another potential limitation of the analysis is to assume that even if a country benefits from
an international coalition relative to a BAU scenario, it will always prefer to free-ride if that
option is even more profitable. While this assumption merely derives from individual welfare
maximisation, current international redistributive policies such as official development aid
point instead to some degree of altruism. Against this background, there might be a possibility
for some countries to sign an agreement even if they could in principle gain more from free
riding on other countries’ abatement efforts. We test this possibility by computing the cost for
developed countries of using additional resources (additional to the coalition surplus) to
stabilise the grand coalition, i.e. to give each other region its free-riding pay off. These
calculations show that with a 3% loss in the discounted value of their consumption levels,
industrialised countries could stabilise the grand coalition in the high-damage/low-
discounting case, i.e. all other participating regions could be fully compensated for their free-
riding incentives through financial transfers, thereby bringing them into an agreement.
Finally, two crucial assumptions affect the results presented in the paper. The absence of
negative emissions technologies or any other technology that might alter the climate (i.e. geo-
28
engineering), and the absence of adaptation policies. If one assumed that by means of
bioenergy and carbon capture and sequestration, or direct CO2 capture, or other technologies
that alter the incoming solar radiation, cooperating countries could unilaterally change the
climate, than the requirements on the dimension and composition of a coalition to
endogenously produce the 550 ppm CO2eq by 2100 target would be substantially different.
Adaptation policies, by providing benefits that are local, at least within the boundary of
macro-regions considered in this model, could also change free riding incentives and thus the
willingness to cooperate on climate change mitigation as well.
29
References
Asheim, G., C.B. Froyn, J. Hovi and F.C.Menz (2006). “Regional versus Global Cooperation on
Climate Control” Journal of Environmental Economics and Management Vol 51: 93-109.
Barrett , S. (1994), “Self-Enforcing International Environmental Agreements”, Oxford Economic

Papers, Vol. 46, pp. 878-894.
Bollen, J., B. Guay, S. Jamet and J. Corfee-Morlot (2009), “Co-benefits of Climate change Mitigation
Policies: Literature Review and New Results”, OECD Economics Department Working Papers,
No. 692, OECD, Paris
Bosello, F., B. Buchner, Carraro, C. (2003). Equity, Development, and Climate Change Control,
Journal of the European Economic Association, Vol.1.
Bosetti V., C. Carraro, M. Galeotti, E. Massetti and M. Tavoni, (2006), "WITCH: A World Induced
Technical Change Hybrid Model", The Energy Journal, Special Issue. Hybrid Modeling of
Energy-Environment Policies: Reconciling Bottom-up and Top-down, 13-38
Bosetti, V., C. Carraro, E. Massetti and M. Tavoni (2008), “International Technology Spillovers and
the Economics of Greenhouse Gas Atmospheric Stabilization”, Energy Economics, 30(6): 2912-
2929.
Bosetti V., C. Carraro, R. Duval, E. De Cian, E. Massetti and M. Tavoni (2009). “The Incentives to
Participate in and the Stability of International Climate Coalitions: a Game Theoretic Approach
Using the WITCH Model,” OECD Economics Department Working Papers No. 702, June 2009.
Bosetti V., E. De Cian (2011), “A good opening: the key to make the most of unilateral climate
action,” Fondazione Eni Enrico Mattei Working Papers No. 82, June 2011.
Burniaux, J-M., J. Chateau, R. Duval and S. Jamet (2008), “The Economics of Climate Change
Mitigation: Policies and Options for the Future”, OECD Economics Department Working
Papers, No. 658, OECD, Paris.
Bréchet, T., F., Gerard, and H., Tulkens (2011), “Efficiency vs. Stability in Climate Coalitions: A
Conceptual and Computational Appraisal”, The Energy Journal, Vol. 32, No.1.
Carraro, C. and D. Siniscalco (1993), “Strategies for the international protection of the environment”,
Journal of Public Economics, Vol. 52: 309-328.
Carraro, C. and D. Siniscalco (1991), “Transfers and Commitments in International Negotiations”,

Paper prepared for the ESF Task Force 3 on Environmental Economics, in K.G. Maler, ed.,
International Environmental Problems: An Economic Perspective, Kluwer Academic
Publishers, Dordrecht.
Carraro, C. and E. Massetti (2010), “International Climate Change Negotiations: Lessons from
Theory,” in E. Cerdá and X. Labandeira, Eds, Climate Change Policies – Global Challenges
and Future Prospects, Edward Elgar.
Chander, P. and H. Tulkens (1997), “The Core of an Economy with Multilateral Environmental
Externalities”, International Journal of Game Theory, Vol. 26.
30
Carraro, C. and J. Eyckmans and M. Finus (2006), "Optimal transfers and participation decisions in
international environmental agreements," The Review of International Organizations, Vol. 1.
D’Aspremont, C.A. and J.J. Gabszewicz (1986), “On the Stability of Collusion”, in G.F. Matthewson
and J.E. Stiglitz, eds., New Developments in the Analysis of Market Structure, Macmillan, New
York.
D’Aspremont, C.A., A. Jaquemin, J.J. Gabszewicz and J. Weymark (1983), “On the Stability of
Collusive Price Leadership”, Canadian Journal of Economics, Vol. 16.
Donsimoni, M.P., N.S. Economides and H.M. Polemarchakis (1986), “Stable Cartels”, International
Economic Review, Vol. 27, pp. 317-327.
Hanemann, W.M. (2008), "What is the Cost of Climate Change?", CUDARE Working Paper
No. 1027, University of California, Berkeley.
Harrod, R. (1948), Towards a Dynamic Economics: Some Recent Developments of Economic Theory
and Their Application to Policy, MacMillan, London.
IPCC, (2007), “Climate Change 2007: Impacts, Adaptation and Vulnerability”, in M. Parry,
O. Canziani, J. Palutikof, P. van der Linden, C. Hanson (eds), Contribution of Working Group II
to the Fourth Assessment Report on Climate Change, Cambridge University Press, Cambridge.
Nagashima, M., R. Dellink, E. van Ierland, H.P. Weikard (2009). Stability of international climate
coalitsion – A comparison of transfer schemes, Ecological Economics, 68, 1476-1478.
Negishi, T. (1960), “Welfare Economics and the Existence of an Equilibrium for a Competitive
Economy”, Metroeconomica, Vol. 12, No. 2-3.
Nordhaus and Yang (1996) , “A regional dynamic general-equilibrium model of alternative climate-
change strategies”, American Economy Review, Vol. 86: 741-765.
Nordhaus, W. (2007), “A Review of the Stern Review on the Economics of Climate Change”, Journal
of Economic Literature, Vol. 45, No. 3.
Nordhaus, W.D. and J.G. Boyer (2000), “Warming the World: the Economics of the Greenhouse
Effect”, MIT Press, Cambridge, Massachusetts.
Ramsey, F. (1928), “A Mathematical Theory of Saving”, Economic Journal, Vol. 38, No. 152.
Solow, R.M. (1974), “The Economics of Resources or the Resources of Economics”, American
Economic Review, Vol. 64, No. 2.
Stern, N. (2007), The Economics of Climate Change: The Stern Review, Cambridge University Press,
Cambridge.
Tol, R.S.J. (2005), “The Marginal Damage Costs of Carbon Dioxide Emissions: An Assessment of
Uncertainties”, Energy Policy, No. 33.
UNFCCC (2007), “Investments and Financial Flows to Address Climate Change”, Background Paper
on Analysis of Existing and Planned Investments and Financial Flows Relevant to the
Development of Effective and Appropriate International Response to Climate Change.
31
Weitzman, M. (2007), “A Review of the Stern Review on the Economics of Climate Change”, Journal
of Economic Literature, Vol. 45, No. 3.
Weitzman, M. (2001), “Gamma Discounting”, American Economic Review, Vol. 91, No. 1.
32
NOTE DI LAVORO DELLA FONDAZIONE ENI ENRICO MATTEI
Fondazione Eni Enrico Mattei Working Paper Series
Our Note di Lavoro are available on the Internet at the following addresses:
http://www.feem.it/getpage.aspx?id=73&sez=Publications&padre=20&tab=1
http://papers.ssrn.com/sol3/JELJOUR_Results.cfm?form_name=journalbrowse&journal_id=266659
http://ideas.repec.org/s/fem/femwpa.html
http://www.econis.eu/LNG=EN/FAM?PPN=505954494
http://ageconsearch.umn.edu/handle/35978
http://www.bepress.com/feem/
NOTE DI LAVORO PUBLISHED IN 2011

SD 1.2011 Anna Alberini, Will Gans and Daniel Velez-Lopez: Residential Consumption of Gas and Electricity in the U.S.:
The Role of Prices and Income
SD 2.2011 Alexander Golub, Daiju Narita and Matthias G.W. Schmidt: Uncertainty in Integrated Assessment Models of
Climate Change: Alternative Analytical Approaches
SD 3.2010 Reyer Gerlagh and Nicole A. Mathys: Energy Abundance, Trade and Industry Location
SD 4.2010 Melania Michetti and Renato Nunes Rosa: Afforestation and Timber Management Compliance Strategies in
Climate Policy. A Computable General Equilibrium Analysis
SD 5.2011 Hassan Benchekroun and Amrita Ray Chaudhuri: “The Voracity Effect” and Climate Change: The Impact of
Clean Technologies
IM 6.2011 Sergio Mariotti, Marco Mutinelli, Marcella Nicolini and Lucia Piscitello: Productivity Spillovers from Foreign
MNEs on Domestic Manufacturing Firms: Is Co-location Always a Plus?
GC 7.2011 Marco Percoco: The Fight Against Geography: Malaria and Economic Development in Italian Regions
GC 8.2011 Bin Dong and Benno Torgler: Democracy, Property Rights, Income Equality, and Corruption
GC 9.2011 Bin Dong and Benno Torgler: Corruption and Social Interaction: Evidence from China
SD 10.2011 Elisa Lanzi, Elena Verdolini and Ivan Haščič: Efficiency Improving Fossil Fuel Technologies for Electricity
Generation: Data Selection and Trends
SD 11.2011 Stergios Athanassoglou: Efficiency under a Combination of Ordinal and Cardinal Information on Preferences
SD 12.2011 Robin Cross, Andrew J. Plantinga and Robert N. Stavins: The Value of Terroir: Hedonic Estimation of
Vineyard Sale Prices
SD 13.2011 Charles F. Mason and Andrew J. Plantinga: Contracting for Impure Public Goods: Carbon Offsets and
Additionality
SD 14.2011 Alain Ayong Le Kama, Aude Pommeret and Fabien Prieur: Optimal Emission Policy under the Risk of
Irreversible Pollution
SD 15.2011 Philippe Quirion, Julie Rozenberg, Olivier Sassi and Adrien Vogt-Schilb: How CO2 Capture and Storage Can
Mitigate Carbon Leakage
SD 16.2011 Carlo Carraro and Emanuele Massetti: Energy and Climate Change in China
SD 17.2011 ZhongXiang Zhang: Effective Environmental Protection in the Context of Government Decentralization
SD 18.2011 Stergios Athanassoglou and Anastasios Xepapadeas: Pollution Control with Uncertain Stock Dynamics:
When, and How, to be Precautious
SD 19.2011 Jūratė Jaraitė and Corrado Di Maria: Efficiency, Productivity and Environmental Policy: A Case Study of
Power Generation in the EU
SD 20.2011 Giulio Cainelli, Massimiliano Mozzanti and Sandro Montresor: Environmental Innovations, Local Networks
and Internationalization
SD 21.2011 Gérard Mondello: Hazardous Activities and Civil Strict Liability: The Regulator’s Dilemma
SD 22.2011 Haiyan Xu and ZhongXiang Zhang: A Trend Deduction Model of Fluctuating Oil Prices
SD 23.2011 Athanasios Lapatinas, Anastasia Litina and Eftichios S. Sartzetakis: Corruption and Environmental Policy:
An Alternative Perspective
SD 24.2011 Emanuele Massetti: A Tale of Two Countries:Emissions Scenarios for China and India
SD 25.2011 Xavier Pautrel: Abatement Technology and the Environment-Growth Nexus with Education
SD 26.2011 Dionysis Latinopoulos and Eftichios Sartzetakis: Optimal Exploitation of Groundwater and the Potential for
a Tradable Permit System in Irrigated Agriculture
SD 27.2011 Benno Torgler and Marco Piatti. A Century of American Economic Review
SD 28.2011 Stergios Athanassoglou, Glenn Sheriff, Tobias Siegfried and Woonghee Tim Huh: Optimal Mechanisms for
Heterogeneous Multi-cell Aquifers
SD 29.2011 Libo Wu, Jing Li and ZhongXiang Zhang: Inflationary Effect of Oil-Price Shocks in an Imperfect Market: A
Partial Transmission Input-output Analysis
SD 30.2011 Junko Mochizuki and ZhongXiang Zhang: Environmental Security and its Implications for China’s Foreign
Relations
SD 31.2011 Teng Fei, He Jiankun, Pan Xunzhang and Zhang Chi: How to Measure Carbon Equity: Carbon Gini Index
Based on Historical Cumulative Emission Per Capita
SD 32.2011 Dirk Rübbelke and Pia Weiss: Environmental Regulations, Market Structure and Technological Progress in
Renewable Energy Technology — A Panel Data Study on Wind Turbines
SD 33.2011 Nicola Doni and Giorgio Ricchiuti: Market Equilibrium in the Presence of Green Consumers and Responsible
Firms: a Comparative Statics Analysis
SD 34.2011 Gérard Mondello: Civil Liability, Safety and Nuclear Parks: Is Concentrated Management Better?
SD 35.2011 Walid Marrouch and Amrita Ray Chaudhuri: International Environmental Agreements in the Presence of
Adaptation
ERM 36.2011 Will Gans, Anna Alberini and Alberto Longo: Smart Meter Devices and The Effect of Feedback on Residential
Electricity Consumption: Evidence from a Natural Experiment in Northern Ireland
ERM 37.2011 William K. Jaeger and Thorsten M. Egelkraut: Biofuel Economics in a Setting of Multiple Objectives &
Unintended Consequences
CCSD 38.2011 Kyriaki Remoundou, Fikret Adaman, Phoebe Koundouri and Paulo A.L.D. Nunes: Are Preferences for
Environmental Quality Sensitive to Financial Funding Schemes? Evidence from a Marine Restoration
Programme in the Black Sea
CCSD 39.2011 Andrea Ghermanti and Paulo A.L.D. Nunes: A Global Map of Costal Recreation Values: Results From a
Spatially Explicit Based Meta-Analysis
CCSD 40.2011 Andries Richter, Anne Maria Eikeset, Daan van Soest, and Nils Chr. Stenseth: Towards the Optimal
Management of the Northeast Arctic Cod Fishery
CCSD 41.2011 Florian M. Biermann: A Measure to Compare Matchings in Marriage Markets
CCSD 42.2011 Timo Hiller: Alliance Formation and Coercion in Networks
CCSD 43.2011 Sunghoon Hong: Strategic Network Interdiction
CCSD 44.2011 Arnold Polanski and Emiliya A. Lazarova: Dynamic Multilateral Markets
CCSD 45.2011 Marco Mantovani, Georg Kirchsteiger, Ana Mauleon and Vincent Vannetelbosch: Myopic or Farsighted? An
Experiment on Network Formation
CCSD 46.2011 Rémy Oddou: The Effect of Spillovers and Congestion on the Segregative Properties of Endogenous
Jurisdiction Structure Formation
CCSD 47.2011 Emanuele Massetti and Elena Claire Ricci: Super-Grids and Concentrated Solar Power: A Scenario Analysis
with the WITCH Model
ERM 48.2011 Matthias Kalkuhl, Ottmar Edenhofer and Kai Lessmann: Renewable Energy Subsidies: Second-Best Policy or
Fatal Aberration for Mitigation?
CCSD 49.2011 ZhongXiang Zhang: Breaking the Impasse in International Climate Negotiations: A New Direction for
Currently Flawed Negotiations and a Roadmap for China to 2050
CCSD 50.2011 Emanuele Massetti and Robert Mendelsohn: Estimating Ricardian Models With Panel Data
CCSD 51.2011 Y. Hossein Farzin and Kelly A. Grogan: Socioeconomic Factors and Water Quality in California
CCSD 52.2011 Dinko Dimitrov and Shao Chin Sung: Size Monotonicity and Stability of the Core in Hedonic Games
ES 53.2011 Giovanni Mastrobuoni and Paolo Pinotti: Migration Restrictions and Criminal Behavior: Evidence from a
Natural Experiment
ERM 54.2011 Alessandro Cologni and Matteo Manera: On the Economic Determinants of Oil Production. Theoretical
Analysis and Empirical Evidence for Small Exporting Countries
ERM 55.2011 Alessandro Cologni and Matteo Manera: Exogenous Oil Shocks, Fiscal Policy and Sector Reallocations in Oil
Producing Countries
ERM 56.2011 Morgan Bazilian, Patrick Nussbaumer, Giorgio Gualberti, Erik Haites, Michael Levi, Judy Siegel, Daniel M.
Kammen and Joergen Fenhann: Informing the Financing of Universal Energy Access: An Assessment of
Current Flows
CCSD 57.2011 Carlo Orecchia and Maria Elisabetta Tessitore: Economic Growth and the Environment with Clean and Dirty
Consumption
ERM 58.2011 Wan-Jung Chou, Andrea Bigano, Alistair Hunt, Stephane La Branche, Anil Markandya and Roberta
Pierfederici: Households’ WTP for the Reliability of Gas Supply
ES 59.2011 Maria Comune, Alireza Naghavi and Giovanni Prarolo: Intellectual Property Rights and South-North
Formation of Global Innovation Networks
ES 60.2011 Alireza Naghavi and Chiara Strozzi: Intellectual Property Rights, Migration, and Diaspora
CCSD 61.2011 Massimo Tavoni, Shoibal Chakravarty and Robert Socolow: Safe vs. Fair: A Formidable Trade-off in Tackling
Climate Change
CCSD 62.2011 Donatella Baiardi, Matteo Manera and Mario Menegatti: Consumption and Precautionary Saving: An
Empirical Analysis under Both Financial and Environmental Risks
ERM 63.2011 Caterina Gennaioli and Massimo Tavoni: Clean or “Dirty” Energy: Evidence on a Renewable Energy Resource
Curse
ES 64.2011 Angelo Antoci and Luca Zarri: Punish and Perish?
ES 65.2011 Anders Akerman, Anna Larsson and Alireza Naghavi: Autocracies and Development in a Global Economy: A
Tale of Two Elites
CCSD 66.2011 Valentina Bosetti and Jeffrey Frankel: Sustainable Cooperation in Global Climate Policy: Specific Formulas
and Emission Targets to Build on Copenhagen and Cancun
CCSD 67.2011 Mattia Cai, Roberto Ferrise, Marco Moriondo, Paulo A.L.D. Nunes and Marco Bindi: Climate Change and
Tourism in Tuscany, Italy. What if heat becomes unbearable?
ERM 68.2011 Morgan Bazilian, Patrick Nussbaumer, Hans-Holger Rogner, Abeeku Brew-Hammond, Vivien Foster, Shonali
Pachauri, Eric Williams, Mark Howells, Philippe Niyongabo, Lawrence Musaba, Brian Ó Gallachóir, Mark
Radka and Daniel M. Kammen: Energy Access Scenarios to 2030 for the Power Sector in Sub-Saharan Africa
CCSD 69.2011 Francesco Bosello, Carlo Carraro and Enrica De Cian: Adaptation Can Help Mitigation: An Integrated
Approach to Post-2012 Climate Policy
ES 70.2011 Etienne Farvaque, Alexander Mihailov and Alireza Naghavi: The Grand Experiment of Communism:
Discovering the Trade-off between Equality and Efficiency
CCSD 71.2011 ZhongXiang Zhang: Who Should Bear the Cost of China’s Carbon Emissions Embodied in Goods for
Exports?
CCSD 72.2011 Francesca Pongiglione: Climate Change and Individual Decision Making: An Examination of Knowledge, Risk
Perception, Self-interest and Their Interplay
CCSD 73.2011 Joseph E. Aldy and Robert N. Stavins: Using the Market to Address Climate Change: Insights from Theory
and Experience
CCSD 74.2011 Alexander Brauneis and Michael Loretz: Inducing Low-Carbon Investment in the Electric Power Industry
through a Price Floor for Emissions Trading
CCSD 75.2011 Jean-Marie Grether, Nicole A. Mathys and Jaime de Melo: Unravelling the Worldwide Pollution Haven Effect
ES 76.2011 Benjamin Elsner: Emigration and Wages: The EU Enlargement Experiment
CCSD 77.2011 ZhongXiang Zhang: Trade in Environmental Goods, with Focus on Climate-Friendly Goods and Technologies
ES 78.2011 Alireza Naghavi, Julia Spies and Farid Toubal: International Sourcing, Product Complexity and Intellectual
Property Rights
CCSD 79.2011 Mare Sarr and Tim Swanson: Intellectual Property and Biodiversity: When and Where are Property Rights
Important?
CCSD 80.2011 Valentina Bosetti, Sergey Paltsev, John Reilly and Carlo Carraro: Emissions Pricing to Stabilize Global
Climate
CCSD 81.2011 Valentina Bosetti and Enrica De Cian: A Good Opening: The Key to Make the Most of Unilateral Climate
Action
CCSD 82.2011 Joseph E. Aldy and Robert N. Stavins: The Promise and Problems of Pricing Carbon: Theory and Experience
CCSD 83.2011 Lei Zhu, ZhongXiang Zhang and Ying Fan: An Evaluation of Overseas Oil Investment Projects under
Uncertainty Using a Real Options Based Simulation Model
CCSD 84.2011 Luca Di Corato, Michele Moretto and Sergio Vergalli: Land Conversion Pace under Uncertainty and
Irreversibility: too fast or too slow?
ES 85.2011 Jan Grobovšek: Development Accounting with Intermediate Goods
ES 86.2011 Ronald P. Wolthoff: Applications and Interviews. A Structural Analysis of Two-Sided Simultaneous Search
CCSD 87.2011 Céline Guivarch and Stéphane Hallegatte: 2C or Not 2C?
ES 88.2011 Marco Dall'Aglio and Camilla Di Luca: Finding Maxmin Allocations in Cooperative and Competitive Fair
Division
CCSD 89.2011 Luca Di Corato: Optimal Conservation Policy Under Imperfect Intergenerational Altruism
CCSD 90.2011 Ardjan Gazheli and Luca Di Corato: Land-use Change and Solar Energy Production: A Real Option
Approach
ERM 91.2011 Andrea Bastianin, Matteo Manera, Anil Markandya and Elisa Scarpa: Oil Price Forecast Evaluation with
Flexible Loss Functions
CCSD 92.2011 Elena Verdolini, Nick Johnstone and Ivan Haščič: Technological Change, Fuel Efficiency and Carbon
Intensity in Electricity Generation: A Cross-Country Empirical Study
CCSD 93.2011 Qin Bao, Ling Tang, ZhongXiang Zhang, Han Qiao and Shouyang Wang: Impacts of Border Carbon
Adjustments on China’s Sectoral Emissions: Simulations with a Dynamic Computable General Equilibrium
Model
ES 94.2011 Benno Torgler: Work Values in Western and Eastern Europe
CCSD 95.2011 Jean-Charles Hourcade, Antonin Pottier and Etienne Espagne: The Environment and Directed Technical
Change: Comment
CCSD 96.2011 Adamos Adamou, Sofronis Clerides and Theodoros Zachariadis: Oil Designing Carbon Taxation Schemes for
Automobiles: A Simulation Exercise for Germany
CCSD 97.2011 Valentina Bosetti, Carlo Carraro, Enrica De Cian, Emanuele Massetti and Massimo Tavoni: Incentives and
Stability of International Climate Coalitions: An Integrated Assessment

Incentives and Stability of Internationa

Uploaded by

Copyright:

Available Formats

Incentives and Stability of Internationa

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Incentives and Stability of Internationa

Uploaded by

Copyright:

Available Formats

NOTA DI

By Valentina Bosetti, Fondazione Eni

Keywords: Climate Policy, Climate Coalition, Game Theory, Free Riding

JEL Classification: C68, C72, D58, Q54

Address for correspondence:

Valentina Bosettia, Carlo Carrarob, Enrica De Ciana*,

Emanuele Massettia and Massimo Tavonia

Following a presentation of WITCH’s game-theoretic framework in Section 2, Section 3

2. THE GAME THEORETIC STRUCTURE OF WITCH

i defined as a function of utility of per-capita consumption, u, then

S (γ ) = S (W1 (γ ), W2 (γ ), K , Wn (γ )) = ∑i =1 ∑tωi ,t u (ci ,t (γ ))

chosen on either normative or positive consideration. We experiment different weights and

inverse of marginal utility, ωi ,t = ⎜⎜ ⎟ and weights proportional to the population

The WITCH model analysis assumes a non-cooperative, simultaneous, open membership,

3. ASSESSING THE DRIVERS OF PARTICIPATION IN INTERNATIONAL

3.1. Climate change impacts in WITCH

Adequate knowledge of climate change impacts is a prerequisite for well-informed climate

Surveying the literature on impacts of climate change it is immediately clear damage

1.00% Unite d State s

W e ste rn EU countrie s

Easte rn EU countrie s

‐8.00% South East A sia

Figure 2. Global damage functions in the WITCH model –

Source: WITCH model simulations and UNFCCC (2007).

3.2. Discounting and pure rate of time preference

Non-EU Eastern Europe

Cumulative emission reductions, %

Panel B. Non-Annex I regions

Middle East and North Africa

South East Asia

Cumulative emission reductions, %

1. Korea is grouped with Japan, but is not an Annex I country.

Discounted consumption loss2

United States Western EU countries

Japan-Korea Non-EU Eastern Europe

Middle East and North Africa Africa

Carbon Tax (2005 $US / t CO2 eq)

Figure 5. Mitigation policy costs and free-riding incentives

1.4 Middle East and North China

Discounted consumption loss2

4. ANALYSING COALITION FORMATION AND STABILITY

4.1. Cooperative versus non-cooperative outcomes

6. This is a consequence of using Negishi weights to aggregate welfare.

Figure 6. Sensitivity of the cooperative and non-cooperative solutions to alternative damage

50.0 Low damage - High discount rate

40.0 High damage - Low discount rate

Source: WITCH model simulations.

4.2. Environmental effectiveness, profitability and stability of climate coalitions

Overall GHG concentration (ppm CO2 eq)

Source: WITCH model simulations.

1. Smaller coalitions internalise only a fraction of the environmental externality. The

Figure 7. An illustration of free-riding incentives: the case of Africa

Source: WITCH model simulations.

Overall GHG concentration (ppm CO2 eq)

Source: WITCH model simulations.

Table 3. Profitability and Stability of potentially effective coalitions

Non-participating regions Profitability Internal Stability Potential Stability

Source: WITCH model simulations.

Figure 8. Estimated ranking of free-riding incentives across regions