Bibliografisk note

Funding Information:
The authors are grafeful for valuable comments from two anonymous reviewers on earlier versions of the article. They also acknowledge funding support from five Danish funding sources ( Svineafgiftsfonden , Mælkeafgiftsfonden , Kvægafgiftsfonden , Fjerkræafgiftsfonden , and Norma og Frode Jakobsens Fond ). These funding bodies have no influence on the research activities and the contents of this article. All remaining errors belong to the authors.

Funding Information:
As the CBAMs partially shift the abatement burdens to non-EU region (i.e. lower increases in non-EU emissions as shown in Table 5), global emission reductions become larger under EUGD_BCA1 and EUGD_BCA2 compared to EUGD. Similar to the EU, non-EU countries may be motivated to retaliate by imposing their own CBAMs that target EU EITE exports, aimed at supporting their own EITE sectors. The effects of retaliations are captured in the EUGD_BCA2_R1 and EUGD_BCA2_R2 scenarios. Compared to EUGD_BCA2 (on which the retaliatory CBAMs are imposed), non-EU EITE outputs rise while sectoral emissions increase, thereby resulting in rising leakage rates and lower global EITE emissions reductions (Table 5).Our modeling results – based on a recursive dynamic CGE model and a dynamic baseline towards 2030 – indicate that the EU's unilateral climate policies contained in its Green Deal would result in carbon leakage to the rest of the world, if not supported by preventative measures. The economy-wide carbon leakage rate in the EU Green Deal scenario is 46%, close to the higher end reported in systematic literature reviews and in meta-analyses (Branger and Quirion, 2014a; Carbone and Rivers, 2017; Yu et al., 2021). However, the leakage rates reported in this study are comparable to what have been reported in the more recent literature that quantifies impacts of climate ambitions by the EU (with leakage rates between 15% and 76%; see Appendix A7). Such a wide range is due to a variety of reasons. For example, compared to earlier studies, our research focuses on a 2030 time horizon (implying a smaller EU share of global GHG emissions) with higher EU climate policy ambition than what is assumed in earlier studies. This points to higher emission leakage rates (Boeters and Bollen, 2012; Böhringer et al., 2014; Botteon and Carraro, 1998; Branger and Quirion, 2014a; Carbone and Rivers, 2017). Additionally, the two-pronged EU climate policy modeled in this study (mimicking the EU ETS and the EU ESD) might lead to higher leakage rates, when compared to a uniform carbon price within the EU (Paltsev, 2001). The inclusion of non-CO2 emissions could also worsen the leakage effects (Bednar-Friedl et al., 2012). Finally, no technological spillovers from the EU to the rest of the world are modeled, as for example done by Gerlagh and Kuik (2014), which have been also shown to reduce leakage rates.The seemingly beneficial effects to the EU of imposing an aggressive CBAM, however, are based on the assumption that such a design can be implemented without major costs. In fact, several studies suggest that the administrative burdens can be substantial. For example, Böhringer et al. (2022) note that requiring individual data from producers and third-party verification might be achievable for direct emissions, but seems highly expensive (if feasible at all) for indirect emissions. Further research is therefore needed to quantify practical implementation costs of the EU CBAM. Moreover, these results also need to be evaluated against possible reactions from the rest of the world. In particular, the implementation of an ‘aggressive’ CBAM might lead to retaliatory actions or might be accommodated with increased mitigation efforts by non-EU countries. Two alternative retaliatory measures are therefore evaluated in this study: a CBAM on EU EITE exports based on the average emission intensities in the EU's EITE sectors, and a CBAM on EU EITE exports based on sectoral emission intensities in individual EU members. Both measures increase carbon leakages from the EU to the rest of the world, negatively affect EU EITE exports and outputs, and result in higher global emission, compared to the ‘aggressive’ EU CBAM scenario without international retaliations, thereby pointing to the need for global cooperation. However, in such scenarios, carbon leakages and EITE output losses are lower than in the scenario where the EU Green Deal is implemented without supporting trade measures and in the scenario with an EU CBAM in line with WTO's non-discrimination principle. In contrast, if the rest of the world increases its climate ambitions to prevent aggregate carbon leakage from the EU (e.g. in the electricity and transport sectors), the economic outcomes for the EU EITE sectors would be similar to the EU Green Deal scenario. In this scenario, global emission reduction is the highest among all scenarios considered.Realizing the potentially differential exposures to carbon leakages and competitiveness losses to EITE industries across EU member states, we also report and analyze within-EU results. Member states and groups of member states with higher emission intensities in the energy and EITE sectors are affected the most by stringent EU ETS targets, with larger reductions in EITE outputs and GHG emissions as compared to average changes at the EU level. In contrast, member states with lower emission intensities suffer limited losses or can even increase their EITE outputs. According to our results, alternative CBAM measures and reactions from non-EU trade partners, while affecting EU-wide outcomes, do not significantly impact the distribution of emission reductions and output changes across EU member states. Therefore, trade policy measures supporting unilateral decarbonization can reduce carbon leakage and improve aggregate economic outcomes, but cannot alleviate distributional consequences across EU members, especially if the rest of the world counters the EU's CBAM with retaliatory measures. These unequal reductions in EITE emissions and outputs should therefore be dealt with by using complementary domestic measures within the EU.The authors are grafeful for valuable comments from two anonymous reviewers on earlier versions of the article. They also acknowledge funding support from five Danish funding sources (Svineafgiftsfonden, Mælkeafgiftsfonden, Kvægafgiftsfonden, Fjerkræafgiftsfonden, and Norma og Frode Jakobsens Fond). These funding bodies have no influence on the research activities and the contents of this article. All remaining errors belong to the authors.

Publisher Copyright:
© 2023 The Authors