17
government. The table also gives information on the value of each of the components of
equation (15). Note that the share of the separable component (which does not induce
any feedback) in total damage is important, reaching 70% and more except for O3. The
parameters of the utility function, the health production function and the production
function are calibrated such that the values of Table 2 are obtained in the reference
equilibrium. Note also that the calibration ensures that ex-ante the total MWTP remains
the same in both versions of GEM-E3.
3. SIMULATION RESULTS
In this section we assess the importance of introducing these three feedback
effects in the GEM-E3 model by comparing, for a scenario aiming at reaching the EU
Kyoto target, the standard GEM-E3 model and the new version of the model in which
the feedback effects of air pollution are incorporated.
3.1. Scenario description
We compare the two models for a domestic CO2 tax that aims to reach the Kyoto
target of the EU given the burden sharing agreement within the EU. These targets are
presented in the first column of Table 3. The Kyoto target and the burden sharing
agreement refer to all greenhouse gases (GHG). Since GEM-E3 considers, at this stage,
only CO2 emissions, these targets need to be translated in terms of CO2 reductions. The
EU and national CO2 targets for 2010 have been taken from ECOFYS et al. (2001) with
minor adaptations (second column of Table 3). This implicitly assumes that the relative
CO2 and other GHG abatement costs do not change with respect to the baseline. The
reference scenario derived with GEM-E3 is a business-as-usual scenario in which no