Measuring the Welfare Cost of Climate Change Policies: A Comparative Assessment Based on the Computable General Equilibrium Model GEMINI-E3

Measuring the welfare cost of climate change policies is a real challenge, as evidenced by the very large range of results obtained in the literature for comparable scenarios, such as for instance the implementation of the Kyoto Protocol. Even the sign is a topic for debate as the “double-dividend” theory aims at pointing that the cost may be negative, yielding a welfare gain in addition to the environmental benefit. Effectively, the Greenhouse Gas Effect differs significantly from more familiar externalities such as classical pollutions, even those that have been handled at the global level, as for instance the CFC. Aside from the global impact of the externality, differences include a very long lag between emissions and effects, a very great number of disseminated sources, and the non-existence –at least up to nowof direct “anti-pollution” means or end-of-pipe controls. In the foreseeable future, only indirect policies can be contemplated, aimed at favoring through taxation or change in the price-structure all possible substitutions, and if possible the least costly, in order to reduce emissions. The cost of these policies, which is basically linked to distortions created in the economy –coming, in some circumstances, on top of already existing distortions – is well known in the economic literature as the “Deadweight Loss of Taxation”. But in actual situations its measurement may be very difficult to perform with a good precision. The analysis is moreover complicated by the spill-over effects channeled through international markets and yielding, according to the cases and to the countries, a gain or a loss form “terms of trade”. Finally, is at stake not only the total welfare cost –an indicator of the lower limit of the environmental gain which justifies to implement the contemplated policybut also the marginal cost, which is relevant for a country in determining whether to purchase or to sell, and in which amount, in an eventual market of tradable permits. This marginal cost is not necessarily –and is not generallyequal to the carbon price (though often in the literature, and in particular in the presentation of results obtained with numerical models, the two are undistinguished). The present paper, which is based on classical economic concepts and theoretical developments presented in the literature and in other papers by the authors, aims at showing how they can be implemented in an actual and comprehensive representation of the world economy. Computable General Equilibrium models are obvious candidates for this task. The model GEMINI-E3, which has been specifically designed in this perspective, and benefits now from a continuous eight years experience, has been selected for this presentation and for comparison with results of other models or studies, but it is to be viewed mainly in its pedagogical aspect. Numerical results depend on assumptions which it would be too long to present in detail and argue. Qualitative results –in particular the structure of costs and the relative importance of their componentsis the important lesson which may hopefully be expected.