The debate over the costs of GHG emission reduction has become more complex recently as disagreements over the existence of economic and environmental double dividends have been added to discussions over the existence of a negative cost potential, We argue that basic assumptions about economic efficiency the (sub-)optimality of the baseline and the rate of technical change are more important than model structure, and we underline the importance of the timing of decisions for determining the costs, Moreover the use of a single baseline ＇no policy＇ scenario and several policy intervention scenarios may be fundamentally misleading in the longer term simply because the very idea of a business as usual scenario is deeply problematic, Ultimately the debate turns on political judgments about the desirability of alternative development paths.