The efficiency of unilateral climate policies may be hampered by carbon leakage and competitiveness losses. A widely discussed policy option to reduce leakage and protect competitiveness of heavy industries is to impose Border Carbon Adjustments (BCA) to non regulated countries, which remains contentious for juridical and political reasons. The estimation of carbon leakage as well as the assessment of different policy options led to a substantial body of litterature in energy-economic modeling. In order to give a quantitative overview on the most recent research on the topic, we conduct a meta-analysis on 25 studies, altogether providing 310 estimates of carbon leakage ratios according to different assumptions and models. The typical range of carbon leakage ratio estimates are from 5% to 25% (mean 14%) without policy and from -5% to 15% (mean 6%) with BCA. The output change of Energy Intensive Trade Exposed (EITE) sectors varies from -0.1% to -16% without BCA and from +2.2% to -15.5% with BCA. A meta-regression analysis is performed to further investigate the impact of different assumptions on the leakage ratio estimates. The decrease of the leakage ratio with the size of the coalition and its increase with the binding target is confirmed and quantified. Providing flexibility reduces leakage ratio, especially the extension of coverage to all GHG sources. High values of Armington elasticities lead to higher leakage ratio and among the BCA options, the extension of BCA to all sectors is in the meta-regression model the most efficient feature to reduce the leakage ratio. Our most robust statistical finding is that, all other parameters being constant, BCA reduces leakage ratio by 6 percentage points.
