Any policy that aims at reducing GHG emissions by way of modulating the structure of an economy will entail resource reallocation and therefore an implicit economic cost. In this paper, we present a novel answer to this question using positive and normative analyses in such a way that they complement one another. From a positive perspective, we first propose a new look at the analysis of sectors' distributed GHG forward emissions on the basis of absolute rather than marginal effects. Using this information, we then move to a normative viewpoint using an environmental extended input-output linear programming system and compute lower bounds for the potential gross and net output losses for each production unit when facing emissions reduction targets, such as those proposed by the European Union in their 20-20-20 Directive. The originality of our approach relies on two aspects, namely, the introduction of an Armington assumption to link domestic and imported output and that, differently to previous works, total final demand drives the optimal adjustments to reach emissions cuts while minimizing output losses. Our empirical exercise compares the results of these normative and positive analyses for the six largest economies in the European Union.