On the political economy of international climate agreements

Global cooperation is crucial to tackle climate change. However, cooperation on such a scale has proven to be difficult given the public good nature of the problem. Abatement of greenhouse gases (GHGs) results in benefits at a global scale. This translates into strong free-rider incentives that threaten the success of an international agreement. However, even if current policy approaches are successful and they achieve a meaningful reduction of GHGs, the effects of past emissions on the atmosphere will still be noticeable for hundreds of years from now. Although these effects are both negative and positive, on a global scale, it is likely that the positive effects (e.g. increase in crop yields as a result of higher temperatures and more precipitation, fewer cold related deaths, less energy demand for heating and an increase in amenity and recreational values) have a lower impact than the negative effects (e.g. rise in sea levels, retreat of glaciers, increase in heat stress, vector-borne diseases and loss of lives due to an increase in extreme weather events such as floods and storms).It is recognized that climate change is a global challenge. Countries are sovereign and most of the time act following self-interest. An agreement to tackle climate change, then, should shape the incentives of countries in such a way that cooperation is fostered. The success (i.e. the level of reduction of GHG emissions) of an international climate agreement (ICA) depends to large extent on its design. ICAs' design faces two problems: enforceability and political feasibility. First, an agreement has to be self-enforcing because there is no supranational authority that can enforce and sanction anICA. Second, an agreement has to be politically feasible and recognize that the position of the government at the international negotiations tables is influenced by the national political actors - e.g. ministries, lobbies and the general electorate.The objective of this thesis is to analyze the impact of design characteristics and national political actors on the potential stability and success of ICAs. In this thesis, the cost-benefit, game theoretical and political economy strands of the literature are combined to attain this objective. For this purpose, I use the STAbility of COalitions (STACO) model, which combines a game theoretical module with an empirical module. STACO analyzes ICAs in a 'cartel formation' setting (i.e. only one agreement can be signed at a time). The game theoretical module considers coalition formation. The empirical module considers that governments base their membership decision on a net benefit function. This function comprises benefits of abatement in the form of reduced damages (focusing on emissions of CO ₂ ).Impact of design characteristicsChapter 4 focuses on the effects of considering ICAs as exclusive membership agreements. I assume that coalition members decide on the accession of new members following two rules: unanimity and simple majority voting. I find that in an exclusive membership agreement, global cooperation is difficult to attain. The grand coalition is not stable regardless of the membership rule and the voting rule used for accession. I find that partial cooperation is stable but only under exclusive membership. However, stable coalitions are rather small and do not improve significantly upon the situation without an agreement.From the stable coalitions, the most successful in terms of gains of cooperation (i.e. the difference in global payoff between full and no cooperation) is a coalition between the energy exporting countries (EEX) andChina. This coalition achieves 11.7% of the possible gains of cooperation (i.e. the difference in payoff between the situation with no cooperation and the grand coalition) and it is stable under unanimity voting. The results suggest that it is worthwhile to consider the possibility of having anICAbased on exclusive membership rules. Furthermore, these results give a rationale for the frequent application of unanimity voting in international policy-making and suggest that when countries have a veto power this would not be a major obstacle to achieve larger participation in a stableICA.Emission permits are considered crucial instruments for addressing many environmental problems. However, there is no clear consensus about the most successful way of allocating emissions permits. Often, it is argued that a distribution of permits following equity considerations may be desirable. In Chapter 5, I test seven different permit trading schemes. Each one reflects a different initial allocation of emission permits. I classify the schemes in two categories: pragmatic and equitable. The pragmatic schemes are close to the current status quo and allocate permits according to a uniform emission reduction from a reference emission level. The equitable schemes are motivated by different notions of fairness cited in literature.I find that, without permit trading, the gains of partial and full cooperation are unevenly distributed among coalition members. This is caused by the heterogeneity among regions. Depending on the permit scheme this inequality might be reduced - to some extent. I find that the pragmatic schemes help to reduce the uneven distribution of gains of cooperation. In contrast, equitable schemes only aggravate it. I find stable coalitions only under the pragmatic schemes. The coalition between the European Union (EU-15) andChinais the most successful in terms of abatement and it is stable under the allocation of permits proportionally to BAU-emissions. This coalition achieves almost 25 % of the gains of full cooperation. These results, thus, do not support the conjecture that equity may enhance the success and participation of anICA. A pragmatic scheme, though apparently less fair, may be more successful in attracting more countries to cooperate.Most of the studies on ICAs assume that abatement decisions of members and non-members are efficiently chosen. However, this contrasts with the evidence on international environmental agreements. Many of these agreements specify a uniform quota for all participants. In Chapter 6, I study the effect that a uniform abatement quota has on the participation and effectiveness of ICAs. I analyze four agreement designs, representing different forms of choosing the level of abatement of coalition members. Firstly, the 'Reference Design' that implies a cost-effective allocation of abatement within the coalition and a coalitional optimal aggregate abatement level. Secondly, the 'Joint Quota' design assumes that countries maximize their coalitional payoff but with a restriction. The restriction is that all coalition members have to reduce emissions by a quota from their level of emissions when no agreement is signed. The last two agreement designs assume a bargaining process where each coalition member maximizes its own payoff considering that each member has to make a quota proposal for the coalition. Thus, thirdly, the 'median quota proposal' design assumes that members agree on the median proposal. Fourthly, the 'lowest quota proposal' design assumes that members agree on the lowest proposal. Additionally, I consider the possibility of trading quotas among coalition members.The results of Chapter 6 show that the quota agreement designs help to improve participation in stable agreements. I find that each of the three quota designs result in one stable coalition at least. Quotas foster participation because they represent a less uneven distribution of abatement burdens. From the three quota designs the most successful (in terms of global net benefits) is the 'Lowest quota proposal' design. In this setting, I find that the coalition of EU-15,China,Indiaand rest of the World (ROW) is stable. This coalition reaps 19% of the gains from cooperation. Furthermore, I find that when quotas may be traded among coalition members the design that results in the most successful coalition is the 'Median quota proposal'. In this case, a coalition amongUSA, EU-15,ChinaandIndiais stable. This coalition reaps 40% of the gains from cooperation. These results suggest that the inefficiency of quotas might be compensated through an increase in participation and the success of anICA.Impact on national political actors

In Chapter 7, I study the effect of lobby groups on the size and stability of ICAs. Thus, I depart from the standard assumption on coalition formation analysis about the behavior of governments. I consider that governments are not simply welfare maximizers but that they consider the political pressure of their national lobbies. Thus, a government bases its decision about participation and level of abatement considering both the net benefits from abatement and the lobby contributions. I assume that there are only two lobbies: industry and environmentalist. I consider two types of environmentalist lobbies, supergreen and green. A supergreen lobby is interested in the global effects of the abatement policies. A green lobby is only interested in the regional effects of abatement policies. As for the industry lobby, I assume that the industry is always harmed by their government abatement decisions given the associated abatement costs.

I find that, in the absence of an agreement, supergreen lobby contributions may help to foster an increase in abatement efforts, but only when supergreen and industry lobby contributions are considered. Furthermore, I find that full cooperation is not stable and, although partial cooperation is stable, it does not help much to tackle climate change. I find that the success of stable coalitions, in terms of global abatement and payoff, depends on whether there is a supergreen or a green lobby. There is a stable coalition betweenJapanand EU-15 with supergreen and industry lobby contributions. This stable coalition achieves little in terms of gains of cooperation - it shows an improvement of only 2 %. However, when there are green and industry contributions the stable coalition falls short of achieving the abatement of the case without contributions. Finally, I find that contrary to intuition, industry contributions are compatible with participation in anICA. For instance, EU-15 receives industry contributions in the stable coalition.

It has been argued that there may be another reason for unsuccessful IEAs, namely that voters support candidates whose environmental preferences differ from their own. Voters choose a 'less green' government - i.e. strategically delegate their decision power - because it gives them a better bargaining position at the negotiations- i.e. lower abatement targets and associated costs and the possibility of receiving a larger compensation if transfers would be available. In Chapter 8, I study the effect of this strategic delegation behavior on the effectiveness and stability of anICA. I assume that there are different types of politicians, corresponding to the different attitudes towards the environmental problem of the voters. Furthermore, I assume that voters can elect the type of politician that will represent them at the international negotiations. I study a model of two asymmetric countries, for a pure public good and analyze stability. I find that when countries act as singletons leakage effects (i.e. the fact that the increase in abatement of one country is offset by a decrease in abatement of the other country) are the main cause of strategic voting. Moreover, I find that when countries sign an IEA, voters always have an incentive to elect strategically their government but that the resulting agreement improves in environmental terms (i.e. abatement levels) upon the case without cooperation. Finally, I find that strategic voting undermines the success of an IEA but not because voters elect a 'less green' government but because strategic voting makes IEAs unstable.

From the analysis that I present in this thesis it is clear that anICAneeds to be designed to attain a meaningful participation (i.e. that members actually reduce GHGs emissions). Thus, an agreement would better tackle climate change if it is deep (i.e. that attains a significant reduction of GHGs) rather than wide (i.e. that is signed by a large number of countries). The design of an agreement (considering, for instance, membership issues and instruments such as tradable permits and quotas) is a crucial point on this respect. Furthermore, it is important to have a better understanding of the underlying political process of the negotiation of ICAs. The political economy aspects of these agreements are essential to understand the underlying incentives of countries to participate. I highlight the relevance of national political actors in the outcomes of the negotiations. Lobby groups and voters certainly have an influence, direct and indirect, on the terms and success of anICA. Moreover, the analysis of the decision-making process in ICAs and IEAs needs to consider that their success is related to the extent to which the government can be influenced. At the international level, lobby groups have a clear influence on the outcome of the negotiation, whereas voters have only an influence at national or local level. Policy-makers may perceive this and act accordingly not only to maximize the overall welfare of the nation but also to maximize their private benefit (reflected not only in monetary terms but in the many forms of political support).