A 2018 screening study of EU-28 trade in cement, clinker and aluminium products (Healy, Schumacher and Eichhammer) found no evidence of carbon leakage for cement and clinker but inconclusive signals for unwrought non-alloyed aluminium, and explicitly left open which forces drove the observed trade-pattern changes: EU Emissions Trading System (EU ETS) carbon costs (direct for cement, indirect through electricity prices for aluminium), allocation-rule incentives that may reward strategic production, or confounders such as demand cycles, exchange rates and energy-price differentials. Later literature has not closed this gap: the 2026 systematic review of carbon-leakage evidence still finds the ex-post international-trade literature limited to a few economy-wide studies predating Phase III, and cites the 2018 paper as representative of the confounding critique. The open problem is to determine, with credible econometric identification at sector, member-state and plant level, whether and through which channels Phase III (2013-2020) EU ETS costs caused the observed trade and production shifts in the two sectors - separating direct cement cost effects (including strategic production to secure free allowances) and indirect aluminium electricity-cost effects from confounding drivers - and to cast the screening into a repeatable, statistically tested monitoring protocol.

Background

Carbon leakage denotes the situation in which climate policy in one jurisdiction causes emission-intensive production, and hence emissions, to shift to less regulated jurisdictions, offsetting domestic abatement. The EU Emissions Trading System (EU ETS) caps greenhouse-gas emissions of European installations; its third trading period, Phase III (2013-2020), replaced national allocation plans with a single EU-wide cap, made auctioning the default for allocating allowances, and gave allowances for free only to sectors deemed at significant risk of carbon leakage (an official list based on trade intensity and emissions intensity), using harmonized product benchmarks. For cement, the benchmark is defined per tonne of clinker, the intermediate product whose calcination accounts for most of cement's process emissions. Because free allocation is proportional to an installation's activity level, the Phase III rules include activity-level thresholds below which allocation is curtailed; Branger, Ponssard, Sartor and Sato (2015) showed these thresholds can reward strategic over-production, a mechanism invoked to explain part of the growth of EU clinker exports after 2013. Primary aluminium smelting is extremely electricity-intensive, so its direct process emissions are modest compared with the emissions embedded in purchased power; carbon pricing that raises electricity prices therefore acts on aluminium mainly as an indirect cost, and some member states grant compensation for such indirect costs as state aid. Against this background, Healy, Schumacher and Eichhammer (2018) built a descriptive screening of EU-28 trade 2000-2016 in cement products, clinker (Combined Nomenclature code 25231000; the Combined Nomenclature is the EU's product-classification system for trade statistics) and aluminium products, including unwrought non-alloyed aluminium (CN 76011000, i.e. primary smelted metal not yet worked into semi-finished products), using three indicators computed from public statistics (Eurostat PRODCOM production data, EU COMEXT and UN Comtrade trade data; PRODCOM is Eurostat's survey of the production of manufactured goods, COMEXT and Comtrade are the EU and United Nations international-trade databases): the ratio of extra-EU imports to domestic production, the share of intra-EU trade in total trade, and the EU share in selected third-country import markets. They found no leakage evidence for cement and clinker (extra-EU clinker imports fell from 15.7 Mt in 2007 to 0.7 Mt in 2016 while exports rose to 9.5 Mt), but sharply rising import penetration of unwrought non-alloyed aluminium after 2012, which they could not causally interpret. The prior econometric literature stops at Phase II: Branger, Quirion and Chevallier (2016) regressed cement and steel net imports on carbon prices and demand for 2005-2012, finding demand, not the carbon price, drives trade; Sartor (2013) found electricity prices critical for primary aluminium net imports during the first six and a half years of the EU ETS. Economy-wide ex-post studies of later vintages (Naegele and Zaklan 2019; Dechezleprêtre, Nachtigall and Venmans 2023) find little or no leakage or output loss, but use data that end before or early into Phase III and do not identify sector-level trade responses. What remains unresolved is precisely the causal question the screening raised: which drivers produced the observed Phase III trade patterns in these two sectors, and can such screening become a statistically grounded, repeatable monitoring instrument.

Problem Statement

Determine, with credible causal identification using public data (Eurostat PRODCOM production statistics, EU COMEXT and UN Comtrade trade data, the European Union Transaction Log - the public registry of installation-level allowance allocation and verified emissions - and electricity price series), whether, through which channels, and by how much the Phase III EU ETS (2013-2020) caused the observed changes in EU production and trade patterns in the cement and primary aluminium sectors, as distinct from confounding drivers (macroeconomic and demand cycles, exchange rates, transport costs, energy-price differentials, technological change, and domestic investment). This decomposes into linked components: (1) for cement and clinker, estimate the causal effect of direct carbon costs on EU production, imports and exports, and quantify the contribution of allocation-rule incentives (activity-level thresholds and strategic production to secure free allowances) relative to market forces in the observed growth of EU clinker exports; (2) for primary aluminium, in particular unwrought non-alloyed aluminium, estimate the causal effect of indirect EU ETS costs transmitted through electricity prices - distinguishing them from persistent international electricity-cost differentials and from national indirect-cost compensation - on EU producers' competitiveness and on the observed import growth, alongside any direct cost effect; (3) decompose the results across Member States and plants to control for border effects and intra-EU relocation; and (4) cast the screening into a repeatable monitoring protocol with statistical significance testing of indicator changes, so the assessment can be repeated regularly and extended to other sectors and finer product disaggregation. The research target is causal estimates with uncertainty for the direct cement channel, the allocation-gaming channel, and the aluminium indirect-electricity channel, obtained under identification that addresses the named confounders and member-state heterogeneity, together with a repeatable monitoring protocol; a purely descriptive indicator update without identification does not resolve the question.

References

  1. Healy, S., Schumacher, K., Eichhammer, W. (2018). Analysis of Carbon Leakage under Phase III of the EU Emissions Trading System: Trading Patterns in the Cement and Aluminium Sectors. Energies 11(5), 1231. DOI: 10.3390/en11051231. https://doi.org/10.3390/en11051231
  2. Branger, F., Ponssard, J.-P., Sartor, O., Sato, M. (2015). EU ETS, Free Allocations, and Activity Level Thresholds: The Devil Lies in the Details. Journal of the Association of Environmental and Resource Economists 2(3), 401-437. DOI: 10.1086/682343. https://doi.org/10.1086/682343
  3. Branger, F., Quirion, P., Chevallier, J. (2016). Carbon Leakage and Competitiveness of Cement and Steel Industries Under the EU ETS: Much Ado About Nothing. The Energy Journal 37(3), 109-136. DOI: 10.5547/01956574.37.3.fbra. https://doi.org/10.5547/01956574.37.3.fbra
  4. Sartor, O. (2013). Carbon Leakage in the Primary Aluminium Sector: What Evidence after 6.5 Years of the EU ETS? SSRN preprint. DOI: 10.2139/ssrn.2205516. https://doi.org/10.2139/ssrn.2205516
  5. Naegele, H., Zaklan, A. (2019). Does the EU ETS cause carbon leakage in European manufacturing? Journal of Environmental Economics and Management 93, 125-147. DOI: 10.1016/j.jeem.2018.11.004. https://doi.org/10.1016/j.jeem.2018.11.004
  6. Dechezleprêtre, A., Nachtigall, D., Venmans, F. (2023). The joint impact of the European Union emissions trading system on carbon emissions and economic performance. Journal of Environmental Economics and Management 118, 102758. DOI: 10.1016/j.jeem.2022.102758. https://doi.org/10.1016/j.jeem.2022.102758
  7. Venmans, F., Ellis, J., Nachtigall, D. (2020). Carbon pricing and competitiveness: are they at odds? Climate Policy 20(9), 1070-1091. DOI: 10.1080/14693062.2020.1805291. https://doi.org/10.1080/14693062.2020.1805291
  8. Chen, H., Di Gregorio, M., Paavola, J. (2026). Carbon leakage in emissions trading systems: a systematic review and meta-analysis of ex-ante and ex-post evidence. Climate Policy, 1-17. DOI: 10.1080/14693062.2026.2627750. https://doi.org/10.1080/14693062.2026.2627750
  9. Eichhammer, W., Friedrichsen, N., Healy, S., Schumacher, K. (2018). Impacts of the Allocation Mechanism Under the Third Phase of the European Emission Trading Scheme. Energies 11(6), 1443. DOI: 10.3390/en11061443. https://doi.org/10.3390/en11061443
  10. Saevarsdottir, G., Kvande, H., Welch, B. J. (2020). Aluminum Production in the Times of Climate Change: The Global Challenge to Reduce the Carbon Footprint and Prevent Carbon Leakage. JOM 72(1), 296-308. DOI: 10.1007/s11837-019-03918-6. https://doi.org/10.1007/s11837-019-03918-6
  11. Ying, Q., Li, M., Geng, Y. (2025). Unintended emissions: the carbon leakage effects of the third-phase EU ETS tightening on trade-exposed Chinese firms. Climate Policy, 1-15. DOI: 10.1080/14693062.2025.2545325. https://doi.org/10.1080/14693062.2025.2545325
  12. Ferrara, A. R., Giua, L. (2022). Indirect cost compensation under the EU ETS: A firm-level analysis. Energy Policy 165, 112989. DOI: 10.1016/j.enpol.2022.112989. https://doi.org/10.1016/j.enpol.2022.112989
  13. Böhringer, C., Fischer, C., Rosendahl, K. E., Rutherford, T. F. (2022). Potential impacts and challenges of border carbon adjustments. Nature Climate Change 12, 22-29. DOI: 10.1038/s41558-021-01250-z. https://doi.org/10.1038/s41558-021-01250-z
  14. Zeynalli, F. (2026). Does Free Allocation Slow Decarbonization? Installation-Level Evidence from the EU ETS, Phases III and IV. EarthArXiv preprint. DOI: 10.31223/x5249t. https://doi.org/10.31223/x5249t