Coalition Formation and International Trade in Greenhouse Gas Emissions

The success of any international climate change agreement depends on abatement targets and the incentives for countries to participate. International emissions trading is frequently cited in the policy literature as a means to making headway on both issues. With trading, more stringent targets become less costly, and agreements can be profitable for the developing world by allowing them to capture the rents from permit sales. While numerous market equilibrium studies have measured the potential efficiency gains from international permit trade, these analyses provide limited insight into the strategic feasibility of such agreements. For example, while a comprehensive global agreement delivers efficient abatement, it is highly unlikely that any global agreement would be self-enforcing. The larger an abatement coalition, the stronger is the economic incentive to defect. The present paper formulates a computable general equilibrium model of self-enforcing carbon abatement agreements. In our setting permit endowments (and therefore abatement targets) are determined endogenously as equilibrium behavior within a candidate coalition. The model provides a representation of the global economy in 7 sectors and 6 regions. Preferences and technology are calibrated to a 1998 base year, so the model offers a consistent, empirically-based representation of the economic effects of carbon abatement policies. We run the model over a 20 year time horizon, and use annual equilibrium calculations to identify those coalitions in which member states have no incentive to defect. We find that the most effective coalitions, those which result in the lowest levels of global emissions, are sub-global and typically involve the US, Europe or Japan (high income–high abatement cost countries) in coalition with China. Abatement targets may improve if certain, willing countries are omitted from the agreement. Terms of trade effects in energy markets actually give permit sellers, like China, an incentive to pursue more stringent abatement targets. However, even the best self-enforcing agreements provide only modest levels of abatement relative to the global first-best. In our reference specification we find that abatement provided by non-cooperative global agreements amounts to only 30% of the first-best emissions cutback. Lastly, we find no evidence in our model calibrated to a various baseline growth paths that the prospects for an effective agreement will improve over the next twenty years.