Unilateral climate policy can be detrimental to global climate protection. Our objective is to provide insight into such a policy, to quantify the risk of carbon leakage, and to investigate the effects related to potential anti-leakage measures. We analyze existing definitions of carbon leakage and propose an alternative, rigorous one, which is different in three respects. The definition is then tested using computable general equilibrium analysis of the global economy and decomposition analysis. We identify a list of parameters that affect not only the magnitude but also the sign of the carbon leakage rate. Manipulating elasticities of substitution suggests that carbon leakage can be either positive or negative. Computable general equilibrium models, which are widely applied, including by the European Commission in this area, should be transparent, and their assumptions call for careful validation. We find that emission limits are properly distributed between sectors covered by the European Union Emissions Trading System and other sectors for the first commitment period (ended in 2012) but not for the second one (ending in 2020), where the target for the non-trading sectors should be reduced relative to the target for the trading sectors in order to equlize marginal abatement costs.
Keywords: Carbon leakage; Computable general equilibrium modeling; Decomposition analysis; European union policy.