Jump to ContentJump to Main Navigation
The Design of Climate Policy$

Roger Guesnerie and Henry Tulkens

Print publication date: 2008

Print ISBN-13: 9780262073028

Published to MIT Press Scholarship Online: August 2013

DOI: 10.7551/mitpress/9780262073028.001.0001

Show Summary Details
Page of

PRINTED FROM MIT PRESS SCHOLARSHIP ONLINE (www.mitpress.universitypressscholarship.com). (c) Copyright The MIT Press, 2018. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in MITSO for personal use. Subscriber: null; date: 24 October 2019

Heterogeneity of Countries in Negotiations of International Environmental Agreements: A Joint Discussion of the Buchner-Carraro, Eyckmans-Finus, and Chander-Tulkens Chapters

Heterogeneity of Countries in Negotiations of International Environmental Agreements: A Joint Discussion of the Buchner-Carraro, Eyckmans-Finus, and Chander-Tulkens Chapters

(p.187) 9 Heterogeneity of Countries in Negotiations of International Environmental Agreements: A Joint Discussion of the Buchner-Carraro, Eyckmans-Finus, and Chander-Tulkens Chapters
The Design of Climate Policy

Sylvie Thoron

The MIT Press

Abstract and Keywords

Different countries have different preferences for the type of measures that should be applied to prevent climate change. This chapter analyzes the heterogeneity of countries in negotiations of international environmental agreements. It discusses the problems posed by the definition of stability in coalition formation and the problem of participation implicit in the negotiations on climate change.

Keywords:   climate change, international environmental agreements, stability, coalition formation, participation

An important characteristic of the Kyoto protocol is that the different partners are treated very differently. For the first time an international environmental agreement (IEA) sets not only a global target and a timetable but also very precise country-specific targets. In previous IEAs the objective was to reach a consensual common target as, for example, in the Protocol on the Reduction of Sulphur Emissions by at Least 30 Percent. The Montreal Protocol on Substances That Deplete the Ozone Layer also fixes a uniform target for industrialized countries but introduces a differentiation in the timetable with a ten-year grace period for developing countries. In the case of the Kyoto Protocol the differentiation is much more marked. The developing countries are not committed to any binding target for the first period (until 2012) and nothing is specified for the following period. The global target of a decrease of greenhouse gases by 5.5 percent is supposed to be achieved by assigning specific targets to the different countries that vary from a reduction of 8 percent to an increase of 10 percent. Furthermore in the same Protocol the flexibility mechanisms, the EU bubble, the principles of joint implementation, emissions trading, and the clean development mechanism allow the exploitation of differences and complementarities among countries. Indeed the Kyoto protocol is clearly the result of a negotiation between partners whose differences, far from being ignored in order to reach a consensus on a common target, have occupied a central position in the discussions.

It is not surprising, in this context, that the theorists who wanted to analyze the IEA, in general, and the climate change negotiations, in particular, were not satisfied with a theory of coalition formation that (p.188) limits its analysis to a symmetric framework. Although I do not want to advocate the latter choice at this point, I will try to give a historical explanation. The integration of externalities and the exploitation of efficiency gains are two important economic justifications for cooperation. Chander and Tulkens explain (in chapter 8 of this volume) how these two arguments have been developed in different literatures. The problem of externalities has been analyzed by economists in the framework of Samuelson’s public good theory (1954) or as developments of the Coase theorem (1961). In parallel and exactly during the same period, the cooperative game theory literature was analyzing superadditive games (Shapley 1953, which introduced the Shapley value; Gillies 1959, which developed the concept of the core). This second literature has always been criticized because of its focus on the problem of how to exploit efficiency gains inside a coalition, while leaving aside the incentive problem generated by the persistent externalities between coalitions. Thrall and Lucas (1963) proposed a framework to deal with both problems simultaneously but the cooperative game theory literature did not take up this aspect. Interest in the Thrall and Lucas proposal was revived in the 1990s, with the development of the theory of coalition formation.

It is notable that both literatures then focused on disparities between actors. The Coase theorem explained how polluters and pollutees could reach an efficient outcome by bargaining. The aim of the solution concepts proposed by cooperative game theory was to solve the problem of sharing a coalition’s worth between heterogeneous players. Why then, did the theory of coalition formation not deal with the problem of heterogeneity? Because the original aim of this theory was to explain how the existence of these externalities or spillovers could lead to inefficient outcomes. The Coase theorem claimed the emergence of an efficient outcome, and in cooperative game theory efficiency is assumed. By contrast, this new literature focused on the sources of inefficiencies. Noncooperative game theory focused on the problem of compliance to agreements and the theory of coalition formation on the problem of participation in agreements. I will not address the problem of compliance here. After all, to quote Abram Chayes and Antonia Chayes (1991: 311), “International lawyers and others familiar with the operations of international treaties take for granted that most states comply with most of their treaty obligations most of the time.”

Conversely, participation is often a serious problem for IEA, and this has been illustrated by the negotiations on climate change. However, (p.189) now that we know that the incentives are not conducive to the formation of the grand coalition, even in a symmetric framework, several questions remain to be answered. Is the situation better or worse off in a nonsymmetric framework? Did the theory of coalition formation propose solutions to the participation problem? If the answer is in the affirmative, are these solutions relevant for the heterogeneous case? Finally, are there solutions specific to heterogeneous cases? These are the kind of questions addressed in the three papers on international environmental negotiations of this volume.

9.1 Sources of Heterogeneity

In this volume, Eyckmans and Finus (chapter 6), on one hand, and Buchner and Carraro (chapter 7), on the other, use modified versions of the same RICE model (i.e., regional integrated model of climate and the economy). This model, proposed by Nordhaus and Yang in 1996, is a regional, dynamic, general equilibrium model of the economy which integrates economic activity with the sources, emissions, and consequences of greenhouse-gas emissions, and consequences of greenhouse-gas emissions and climate change. It divides the global economy into six different regions: USA (United States), JPN (Japan), EU (European Union), CHN (China), FSU (former Soviet Union), and ROW (rest of the world). The novelty of the RICE model in comparison with previous models of global warming is to allow nations to adopt different strategies.

In the CLIMNEG world simulation model, Eyckmans and Finus introduce differences in discount rates. Discount rates for developing regions are higher than those for developed countries. The countries also differ in energy efficiency. USA, JPN, and EU have steep marginal abatement costs, while CHN and ROW have flat marginal abatement costs. This means that energy-efficient regions face higher costs when cutting back emissions. The countries are also more or less vulnerable or more or less sensitive to climate change. Damage functions are particularly steep in EU and ROW, to a lesser extent in USA and JPN, and relatively flat in FSU and CHN.

The FEEM-RICE model used by Buchner and Carraro incorporates an endogenous technical change that yields endogenous growth and improves the emission or output ratio. The Kyoto Protocol constitutes the starting point: the European Union, Japan, and Russia are committed to complying with their Kyoto targets, and there is a market for (p.190) pollution permits. The authors present different scenarios and explain in each how the countries can move away from the situation. In each scenario two coalitions are formed. Each coalition can create a market for pollution permits. On each market, the countries draw benefits from their complementarities. Big polluters with stringent targets can buy pollution permits from small or “not yet so big” polluters with mild targets. Despite their mild targets, less energy-efficient regions have an incentive to reduce their emissions. By selling their own emission permits to more energy-efficient regions, they generate funds that can be used to invest in new technologies.

9.2 Transfers and Stability

Eyckmans and Finus (EF) pose the question as to the links between the stability of agreements and transfers among partners. As already discussed above, in most of the literature in economics and cooperative game theory dealing with the formation of agreements, efficiency is a basic criterion. With this as a principle, the outcome should be calculated to satisfy efficiency in a first step. The different partners’ contributions are chosen to maximize the sum of their welfares. As a consequence in the framework described above, for example, it may be that energy-inefficient regions facing smaller abatement costs will have to make bigger reductions in pollution. Then, in a second step, different normative criteria can be used to justify transfers between partners. Consent on criteria is necessary because the different partners will not sign an agreement they consider to be unfair. The difficulty is that the definition of fairness is not unique. In particular, much depends on which countries’ characteristics are to be considered. There is a whole literature on this delicate question to which Eyckmans and Finus refer.

Eyckmans and Finus define a series of transfer schemes. Each of these guarantees to a partner at least the welfare level it would have had if no agreement were signed. They differ in the way the costs and benefits from partnership are shared: a normative criterion is used to define a country-specific coefficient that is applied to the difference between the sum of welfares before and after the agreement implementation. In one of these schemes, a solution originally proposed by Chander and Tulkens (1997)—the CT solution—each country’s contribution to pollution abatement is proportional to the cost that it incurs as a result of the damage from climate change. As a consequence those (p.191) who care more about climate change, or are more vulnerable to it, must contribute more. From a normative point of view, this is debatable. It depends on what kind of good the climate is. Given an objective that is to determine a contribution for each country, let us consider a parallel with private goods. Suppose that climate can be considered a luxury good: in this case the preference for the environment, in general, and climate stability, in particular, increases with wealth. It may then be justified to ask richer countries that have a higher preference for protecting the climate to contribute more. However, climate is more often considered an inferior good:1 those who have the highest demand for protecting the climate are those who are not wealthy enough to be able to protect themselves against climate change. In this case it is more difficult to justify, from a normative point of view, the idea that developing countries that are more vulnerable to climate change should contribute more.

However, the CT solution is also attractive because it can be interpreted in a positive framework. In a purely positive view, the stability of an agreement is not guaranteed by its fairness but rather by the fact that each partner has an incentive to sign, given the alternative and considering its own interest. This depends on each country’s anticipation about what would be its ex-partners’ reactions if that country decided to rescind its agreement. The literature on coalition formation explains how the assumption about the reactions of the ex-partners after a deviation determines the outcome. Suppose that each partner anticipates that a deviation would provoke a complete dismantling of the agreement. Call this anticipation assumption 1. Now suppose that after a deviation, a country anticipates that its ex-partners will continue with the agreement. These countries will just adjust their contributions so that the exiting country’s welfare is no longer taken into account. Call this assumption 2. When externalities are positive, as in the case of an abatement game, the dismantling of the original coalition is a bigger threat than a simple adjustment of the coalition’s abatement (see Hart and Kurz 1983; Yi 1996; Thoron 2000). As a consequence under assumption 2 it is more difficult to attain stability than under assumption 1.

Eyckmans and Finus propose to test the claim that transfers can increase coalition stability. They consider stability under assumption 2. Each transfer scheme guarantees the partners at least a welfare level that is the same as that in the situation without any agreement. Hence, when transfers are added, the coalitions are stable under assumption 1. (p.192) For example, Chander and Tulkens (1997) prove that the CT solution belongs to the core that is also defined under assumption 1. However, there is no direct reason why these coalitions should still be stable under assumption 2, unless the normative criterion used to share the surplus plays a role in stability, which is the case with the CT solution. Indeed the countries that care more about climate change are also those that are least likely to withdraw. If they contribute more, those that care less contribute less and this favors the attainment of stability.

However, this relationship is the exception rather than the rule. Other than by simple coincidence, there is no link between an allocation that satisfies a normative criterion and an allocation that satisfies a stability criterion. I will go even further: there is in fact a certain contradiction between the normative and the positive approaches. On the one hand, the normative approach starts from an efficient outcome and organizes transfers on the basis of a normative criterion. On the other hand, the positive approach describes the outcome of a negotiation, whether or not it is efficient, and considers that the allocation of payoffs is endogenous to this negotiation process. In the first case, the efficiency is a basic criterion, in the second case, it is neither a starting point, nor necessarily an outcome.

I explained above the problem posed by the definition of stability in normal form games of coalition formation. In the literature on coalition formation, other types of models are proposed. Bloch (1996) and Ray and Vohra (2001) used extensive form games to represent the negotiation process. They proved the inefficiency of the outcome when there are positive externalities. In order to focus on the problem of heterogeneity, Maskin (2003) uses these extensive form games to define a Shapley value for games with externalities. Here again, Maskin’s conclusion is that this new value does not satisfy the efficiency axiom. This result is due, though, to the introduction of another axiom, namely one that allows for communication breakdown between players, as pointed out by Chander and Tulkens. The other common feature of these positive models is that the payoffs are generated during the negotiation with only the objective of stability. For example, in an extensive form game, when a player makes a proposal to another player, it is just sufficient to convince that player to become a partner.

The conclusion is not that normative criteria can or cannot be used as arguments during a negotiation. Two problems prevent my giving a clear answer to this question. First, it is theoretically difficult to combine (p.193) the normative and positive approaches. Maybe the reason is that the basis of positive models, the maximization of individual payoffs, is too restrictive. Second, the empirical evidence is lacking, other than simple declarations that normative criteria have some relevance for the outcome of a negotiation.2

9.3 Membership Rule and Climate-Blocs

When the internal or external stability criterion is applied in a symmetric framework, it has been proved that equilibrium always exists (d’Aspremont et al. 1983). The agreement generated at equilibrium can be restricted, as in the extreme case it corresponds to the trivial situation where the agreement is signed by a unique player (!), but equilibrium exists. It has also been proved that such an equilibrium is robust against deviations by coalitions (Thoron 1998). This case is no longer true when players are not symmetric. To illustrate this point, consider the following example: it may be that FSU wants to join coalition (EU, JPN) and that USA wants to join (EU, JPN, FSU). Both coalitions are externally unstable. However, it may be that FSU does not want to leave (EU, JPN, FSU) but would want to leave (EU, JPN, FSU, USA). In comparison with the symmetric case, the fact that one country wants to join the coalition no longer means that the extended coalition is internally stable. As a consequence the equilibrium may not exist, and the membership may cycle. Eyckmans and Finus propose to reduce this instability by using an exclusive membership rule.

The difficulty, in the asymmetric case, is that the different countries have different preferences for the type of measures that should be applied to prevent climate change. Indeed economic theory has taught us that two kinds of problems threaten the efficient provision of a public good: the incentive to free ride and the difficulty with choosing the public good for partners that have different preferences for it. While the recent theory of coalition formation has focused on the first problem, which can be studied in a symmetric framework, the second problem, which only occurs when there are asymmetries, is the main object of an older literature on coalition formation: the theory of jurisdiction formation and the theory of clubs. When no agreement is possible with an open membership rule, this means that preferences are too different for all countries to reach a common agreement. Then, as a consequence of what would be the equivalent of a Tiebout mechanism, (p.194) different agreements may emerge. The countries that manage to reach an agreement will have similar preferences. However, the literature on jurisdiction formation, which tried to model the Tiebout mechanism, showed that the existence of equilibrium cannot be proved in this framework. Furthermore the conclusion that can be drawn from the theory of clubs is that this problem of nonexistence can indeed be solved by using an exclusive membership rule.

Buchner and Carraro in this volume use another argument to justify the formation of climate-blocs: the exploitation of complementarities on markets for pollution permits. We know the advantages of a market for pollution permits: the gains in efficiency generated by the reallocation of the original targets through a market mechanism. These efficiency gains increase with the differences between countries (more precisely the differences in energy efficiencies). As a consequence emissions trading has been proposed as the tool necessary to make the joint implementation and clean development mechanisms work. However, it does not seem to be the case that the same socially beneficial incentives explain why the countries would prefer to form subglobal blocs. In Buchner and Carraro’s chapter, if some coalition structures turn out to be better, from a total welfare point of view (see table 8.2), this is due to the assumption that if China and the United States were to enter a coalition, they would accept stringent targets (−10 percent for China). In fact the argument is that countries would have a stronger incentive to accept such targets if they had the possibility to form subglobal blocs. However, in each of the coalition structures the authors consider, there is at least one country in the coalition that is worse off than in the current Kyoto scenario. To be convincing, this kind of argument has to be based on a stability analysis.

Furthermore I am not sure that the possibility for the countries to form parallel climate-blocs is the solution to the participation problem, and this is for two reasons. First, we know that the formation of a coalition structure implies a social loss in comparison with the formation of the grand coalition. Of course, the formation of climate-blocs is only proposed because the grand coalition cannot form. But, and this is the second point, there are arguments in favor of the formation of a unique subglobal bloc. Indeed a country that did not sign or ratify the Kyoto Protocol cannot benefit from the existing market for pollution permits. In other words, the existence of a unique market generates negative externalities and likely incentives to join this subglobal bloc. Even if this perspective is too optimistic, in my opinion it is better than a situation (p.195) where a country becomes stuck in a collectively suboptimal stable coalition structure, if indeed such a structure exists.

Even in the symmetric case several coalitions can exist simultaneously. When the assumption that only one coalition can be formed is dropped, the driving forces are the same as in the restricted framework. Because the game is superadditive, the countries have an incentive to cooperate; however, they also have an incentive to free-ride, which generates the nonparticipation problem. Now the different coalitions free-ride on each other. Because the externalities are positive, within a given structure the smaller coalitions obtain considerable benefits from the existence of the bigger ones. In Ray and Vohra’s model (2001) the coalitions form sequentially. When the number of negotiating partners is intermediate, two coalitions are formed. The biggest coalition forms first, the second coalition forms afterward in order to free-ride on the previous one. When the externalities are negative, the outcome is very different. In this case Yi (1996) showed that the grand coalition is always an equilibrium.

When describing the formation of climate-blocs, Buchner and Carraro draw a parallel with the formation of the World Trade Organization. Indeed coordination and information problems can make the formation of the grand coalition, in one step, difficult. This is why, in the case of the organization of free trade, countries started by signing binary agreements. However, it seems to me that there is a fundamental difference between this last case and negotiations on climate change. Thanks to negative externalities, the incentives made for a move in the right direction to reach the final outcome of the formation of the World Trade Organization. This will not work in the same way when considering negotiations on global warming. In this case, because the externalities are positive, the emergence of several agreements does not necessarily mean that we are moving in the right direction. It may mean that we have reached a suboptimal equilibrium in which one coalition free-rides on the other.

An optimistic point of view would be that there are other driving forces that can create negative externalities between blocks and that markets for permits constitute a source of these externalities. Another force is the emergence of social norms. An argument developed by Finnemore and Sikkink (1998) is that at a certain “tipping point” in a norm’s evolution, a “norm cascade” takes place, and then states join the coalition in large numbers because of pressure from other states and nonstate actors. In this case the disparities among countries can (p.196) help the process, since the countries that are more committed can, in the end, convince the others to follow their lead.

9.4 Conclusion

The theory of coalition formation has been developed essentially in a symmetric framework. It explains that the problem of participation inherent to the negotiations on climate change, is generated by the existence of positive externalities and an incentive to free-ride. A conclusion would then be that the countries involved are unable to reach the social optimum because the incentive to free-ride is too strong. As a result they may form subgroups, the climate-blocs, and reach a suboptimal situation in which one coalition free-rides on the other. However, the forces that can change the externalities can help offset these incentives.

The asymmetries among potential partners introduce new aspects. On the one hand, the differences between the countries’ evaluations of climate change costs and abatement benefits produce new difficulties. The impossibility of reaching a consensus among asymmetric parties is another explanation for the emergence of climate blocs of similar countries. On the other hand, Buchner and Carraro argue that technology differences can also be considered as an advantage when they take the form of “complementarities” among countries. In this case they provide an explanation for the formation of climate-blocs of countries that are different. I understand the arguments in favor of a second best in which countries with similar perceptions associate. I am less convinced by a second best generated by collusion on the markets for pollution permits. Furthermore, even in the first case, the difficulty is to disentangle two explanations: free riding or efficient exploitation of similarities.

In conclusion, even if the outcome turns out to involve differential treatment of the countries and, in the extreme case, the formation of several blocs, I would still recommend a global framework for negotiations. In the case of negotiations on climate change, all the countries felt themselves obliged, at least initially, to negotiate together under the guidance of the UN. These lengthy negotiations provided conditions for a more creative approach. Countries had to find other ways to differentiate themselves from each other within a single agreement. In the framework of the Kyoto Protocol they came up with the country-specific targets and the flexibility mechanisms. The advantage (p.197) of this process is that it is a better guarantee that countries do not free ride but work to increase efficiency when they want to differentiate themselves from each other.

This work was done while the author was a visitor at the Institute for Advanced Study at Princeton, and she would like to thank the Institute for its hospitality.


Bibliography references:

Bloch, F. 1995. Endogenous structures of association in oligopolies. Rand Journal of Economics 26: 537–56.

Bloch, F. 1996. Sequential formation of coalitions with externalities and fixed payoff division. Games and Economic Behavior 14: 90–123.

Chander, P., and H. Tulkens.1995. A core-theoretic solution for the design of cooperative agreements on transfrontier pollution. International Tax and Public Finance 2(2): 279–94.

Chander, P., and H. Tulkens.1997. The core of an economy with multilateral environmental externalities. International Journal of Game Theory 26: 379–401.

Chayes, A., and A. H. Chayes.1991. Compliance without enforcement: State regulatory behavior under regulatory treaties. Negotiation Journal 7: 311–31.

Maskin, E. 2003. Bargaining, coalitions and externalities. Presidential address to the Econometric Society European Meeting, Stockholm. Mimeo.

Hart and Kurz.1983. Endogenous formation of coalitions. Econometrica 51(4): 1047–64.

Kauppi, H., and M. Widgrén.2004. What determines EU decision making? Needs, power or both? Economic Policy (July): 221–66.

Lucas, W. F., and J. C. Maceli.1978. Discrete partition function games. In P. C. Ordeshook, ed., Game Theory and Political Science. New York: New York University Press, pp. 191–213.

Ray, D., and R. Vohra.1999. A theory of endogenous coalition structure. Games and Economic Behavior 26: 286–336.

Ray, D., and R. Vohra.2001. Coalitional power of public goods. Journal of Political Economy 109(6).

Roth, A. E. 1998. The Shapley Value: Essays in Honor of Lloyd S. Shapley. Cambridge: Cambridge University Press.

Shapley, L. S. 1953. A value for n-person games. In Kuhn and Tucker, eds., Contributions to the Theory of Games, vol. 2. Princeton: Princeton University Press, pp. 307–17.

(p.198) Schelling, T. C. 2006. Strategies of Commitment and Other Essays. Cambridge: Harvard University Press.

Thoron, S. 1998. Formation of a coalition proof stable cartel. Canadian Journal of Economics 1 (February).

Thoron, S. 2000. Conjectures about the reaction of a frustrated coalition. Working paper GREQAM 31A00.

Thrall, R. M., and W. F. Lucas. 1963. n-Person games in partition function form. Naval Research Logistics Quarterly 10: 281–98.

Yi, S.-S. 1997. Stable coalition structure with externalities. Games and Economic Behavior 20: 201–37.


(1) . See, for example, the interpretation given by Schelling (2006) in the chapter What makes greenhouse sense? of his book Strategies of Commitment.

(2) . See the article by Kauppi and Widgren (2004) on the difference between declared and real arguments underlying the European budget allocation.