VoxEU Column Environment

Nations can work together for a better planet, but cooperation on stringent climate targets will be difficult

International agreements on ways to tackle climate change are in a depressing deadlock. This column argues that part of the problem is aiming too high. It suggests that slow and gradual progress towards controlling climate change is the only way forward. But this is at least better than waiting for an ideal solution that may never come.

There is a major lesson emerging from the recent history of environmental negotiations: international cooperation comes into force when the agreement is non-onerous for signatories. The Kyoto protocol is a good example. It is a nearly global agreement, but it contains very mild commitments for signatory parties. In fact, climate negotiations dealing with a post-Kyoto agreement and focused on more ambitious targets have not been very successful.

Long-term aspirations are shared (precisely because they are aspirational), but short-term quantitative targets tend to either bring discussion to a halt or to create a situation of mutual conditional commitments. This is of no surprise for economists. Coalition theory yields a clear prediction: when dealing with a global externality such as climate change, stable coalitions are generally small and/or are ineffective, (see, for example, Barrett 1994). High and uncertain mitigation costs on the one hand, and large, uncertain and non-homogenously distributed damages on the other hand, only exacerbate the problem. Yet, 20 years of very slow progress after the Rio Conference (with the notable exception of some key areas such as Europe) have not dashed all hopes for a more promising future: the Durban Action Platform signed last December has set forth a new and tight timeline for designing a new international agreement over the next 3 years, which would enter into force from 2020. This suggests that continued research in this field is warranted.

In order to keep the problem tractable, theory has often made various simplifying assumptions. Some of these assumptions – such as the symmetry of players and the presence of a single externality – can really be a crucial limit to the analysis. In a recent study, we adopt a numerical approach based on WITCH, an integrated model of the economic and the climate systems (Bosetti et al. 2006), to deal with asymmetric players and a realistic set of assumptions on the crucial economic variables at stake. In this study, we assess the incentives of asymmetric countries to form coalitions for fighting global warming and we measure the environmental effectiveness of successful coalitions.

Climate international cooperation: The best is the enemy of the good

It is a standard working hypothesis to assume that a coalition of countries chooses the optimal path of future emissions that equalises coalition-wide discounted abatement costs and benefits from reduced climate damages over a long time horizon. The fact that these are standard assumptions does not mean that they are universally accepted. There is a lively and still unsettled debate on what discount rate should be used and how countries with very different levels of development should be weighted, two necessary ingredients of cost-benefit analysis.1 There is also an intense debate on marginal abatement costs and future climate change impacts. All these factors definitely play an important role in determining the coalition’s optimal emission pathway. However, this debate is truly meaningful only insofar countries are willing to cooperate in a global climate coalition.

Indeed, our results suggest that even extreme assumptions about impacts, social welfare aggregators, and pure rate of time preferences have very limited influence on the optimal level of mitigation action when countries are not internalising other countries’ damages. Even under assumptions that magnify future impacts of climate change – high damages, low discount rates2 – countries have a small incentive to reduce emissions if they act in their own self interest (Figure 1).

Figure 1. Global GHG emissions in a decentralised solution under different assumptions about climate damages and pure rate of time preference

Therefore, a coalition of cooperating countries is necessary to control greenhouse gas (GHG) emissions and the related climate change impacts. However, under what conditions will countries accept to cooperate on climate change control? The usual argument goes as follows. Countries in the coalition benefit from the reduction of emissions, but they would be even better off if they were to stop their own mitigating effort, exit the coalition, and benefit from lower warming due to others' efforts. In other words, the climate coalition is not stable. Does this conclusion hold for all possible coalitions? And for all mitigation targets? And for all discount rates and expected damages from climate change? The goal of our study is to analyse coalition stability under different scenarios and check the stability of coalitions of different size and membership.

Our conclusions are not encouraging. If we focus on coalitions that are meaningful from a political point of view – i.e. those which include at least all high-income regions – then whatever the level of damages and of the discount rate, and even for non ambitious climate targets such as 550 ppm CO2-eq, economic incentives do not lead to the formation of a stable climate coalition. There may be other non-economic, social, political, or equity concerns that motivate the formation of a climate coalition. But from an economic viewpoint there is no reason to expect an effective climate coalition to form in the coming years – not only the grand coalition, even smaller coalitions.

Some positive messages can nonetheless be inferred from our analysis: some coalitions become stable if, through transfers, benefits from cooperation can be shared (Table 1). This is obviously not a new result in theoretical papers. It is however comforting to find that this holds in a large model that simulates a number of complex energy-economy-climate interactions among asymmetric players. Obviously, the size of the coalition that can be stabilised through transfers depends on assumptions made on the expected damages and on the discount rate. Under assumptions that magnify damages (high damages and low discounting), we find that a coalition including all regions but Latin America and Sub-Saharan Africa – the two regions we identify as having the largest incentives to defect from cooperation – would be stable if transfers were allowed. The grand coalition can also be stabilised through an appropriate transfer mechanism, but only for low climate damages and/or high discounting.

Table 1. Coalition stability with and without transfers

Low discount rate if not differently specified

Whether considering partial coalitions and more aggressive assumptions about future damages, or considering the grand coalition and less aggressive assumptions, the stable agreement results in concentrations of greenhouse gases around 600 ppm-eq. Stability with transfers would thus come at an extra ‘cost’ of about 50 ppm CO2-eq with respect to a Pareto optimal case, taking us further away from stringent climate objectives.3

These results indicate that cooperation on a climate agreement is in principle possible, though it requires transfers that are politically hard to sell. The resulting emission reductions would be less stringent than those compatible with ambitious climate objectives that are often debated in the climate discussion. A slow and gradual progress towards climate change control seems to be the only feasible pathway, given today ‘s information about damages and mitigation costs.

These findings have important implications for the next round of climate negotiations, now synthesised in the Durban Action Platform. Cooperation can in principle be achieved by the potential treaty that might emerge after 2020, but it is unlikely that it will lead to a drastic fall in greenhouse gas pollution. An intermediate target is more likely to be enforced. Still, such intermediate target could bring significant emission reductions compared to the business-as-usual, with potentially large side benefits in terms of local pollution, more sustainable use of fossil fuels, and innovation in green technologies.

We are deeply concerned by uncontrolled global warming. We believe that we should not leave emissions to grow unchecked in the hope that a better agreement materialises. With our work, we find evidence that Voltaire’s witticism “the best is the enemy of the good” well fits the climate change problem and could provide a pragmatic solution to the current climate deadlock.


Barrett , S (1994), “Self-Enforcing International Environmental Agreements”, Oxford Economic Papers, 46:878-894.
Bosetti, V, C Carraro, E De Cian, E Massetti, M Tavoni (2011), “Incentives and Stability of International Climate Coalitions: An Integrated Assessment”, FEEM Working Paper No.97.
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
Carraro, C and D Siniscalco (1993), “Strategies for the international protection of the environment”, Journal of Public Economics, 52:309-328.
Nordhaus, WD and JG Boyer (2000), Warming the World: the Economics of the Greenhouse Effect, MIT Press, Cambridge, Massachusetts.
Nordhaus, W (2007), “A Review of the Stern Review on the Economics of Climate Change”, Journal of Economic Literature, 45(3).
Stern, N (2007), The Economics of Climate Change: The Stern Review, Cambridge University Press.
Weitzman, M (2001), “Gamma Discounting”, American Economic Review, 91(1).

1 See for example Weitzman 2001, Stern 2007, and Nordhaus 2007.
2 Our 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, among others, in Nordhaus (2007).
3 It must also be noted that we already assume a great deal of cooperation in our analysis because world countries are aggregated in twelve regional blocks that behave as a single negotiating block. This might be a valid assumption for the European Union and regions that match with a single country (USA, China), but it might be optimistic for large regions such as Latin America and South-East Asia.


1,785 Reads