A carbon tax may be the most efficient means to limit climate change - eventually mankind's largest externality. When discussing the effects of such a tax, today most climate economics research focuses on the demand, reducing the energy supply side to a static process. However, supply side effects can be crucial for the assessment of carbon emission reduction strategies: Along the claim H.-W. SINN entitled «Green Paradox», they imply that a realistic carbon tax introduced at a low initial level but rapidly increasing might be counterproductive for the climate, accelerating exploitation of the limited resources rather than delaying or reducing it. Owners of fossil fuel stocks optimise sales over time. Anticipating in early periods larger future taxes, they sell more of their fuels today rather than in future. Controversial, SINN's analysis impressively demonstrates the importance of supply side effects for greenhouse gas policy assessments. The analyses by SINN and subsequent contributors assume the debated policy to be the only relevant climate measure for all future. But avoiding a carbon tax today will not imply that no climate relevant development materializes ever. Rather, without substantial measures today, growing climate threats may increase the probability of future measures. This is relevant for the desirability of a carbon tax as resource owners anticipate also other potential future measures. We model the impact of a current carbon tax on global emissions, taking future climate measures into account: we assume that other climate measures, such as backstop technologies, global Kyoto-like demand cartels, carbon capture and storage systems, or alternative carbon taxes may be introduced in some future. The analysis is based on a dynamic multi-period model of the behaviour of forward looking resource owners which optimize sales inter-temporally. For generality, we neither assume a specific extraction cost curve nor use specific assumptions about the tax path or the demand function. We allow monopolistic and competitive sellers. In presence of an anticipated specific future regime change such as a backstop introduction, any positive worldwide tax bridging the time until the future measure, unambiguously reduces cumulative emissions not only in the long, but already in the medium-term. At least for limited tax levels, this holds also for regional taxes. We further consider a stochastic introduction of a backstop. Then, even the weak version of the Green Paradox, i.e. that taxes growing at a rate faster than the real interest rate increase current emissions, fails; such taxes even reduce current and near-term emissions. This analysis has important implications for general climate policy assessment. A decision on a particular climate policy will in general not be decisive for every other potential climate measure. Taking the possibility of alternative climate measures into account can therefore be necessary to prevent strongly biased results.
