The European Union's Carbon Border Adjustment Mechanism (CBAM) aims to avoid carbon leakage by pricing the production emissions of imported goods. Currently, the CBAM applies to iron and steel, cement, aluminium, fertilizers, electricity and hydrogen. As the European Union considers extending the CBAM to chemicals by 2030, its effectiveness in this complex industry remains uncertain. Here we assess how well the CBAM would capture emissions in the chemical industry by using the life-cycle assessment data of 4,470 global chemical production facilities, focusing on the key products ethylene and polyethylene. We find that the current CBAM framework would cover 50-60% of production emissions for these products, leaving substantial upstream emissions unpriced. This limited emission coverage risks compromising the policy's climate ambition and distorts market incentives. To improve effectiveness, the CBAM would need to expand to include fossil feedstocks and refinery products, and implement high fallback default values for embodied emissions to incentivize data reporting. Our findings highlight the need for detailed supply chain tracking and early engagement with chemical industry stakeholders to ensure that the CBAM supports a fair and effective net-zero transition.

Increased climate action is necessary to limit the consequences of global warming. To mitigate climate change, countries set nationally determined contributions under the Paris Agreement. While the Paris Agreement facilitates policy coordination through periodic reporting and progress assessment, climate policy implementation remains largely decentralized. This decentralization can result in unilateral climate policies that financially burden producers, potentially shifting production to regions with less stringent regulation -- a phenomenon called carbon leakage. Carbon leakage can increase global emissions and compromise the competitiveness of industries in climate-ambitious countries. To mitigate carbon leakage, multiple countries consider Carbon Border Adjustment Mechanisms (CBAMs) to establish a level playing field for domestic industries by taxing imported goods based on their carbon footprint.

In 2023, the European Union (EU) introduced a CBAM for major industry sectors at risk of carbon leakage to complement the EU Emissions Trading System (ETS). While the ETS regulates emissions from emission-intensive industries within the EU, the CBAM extends carbon pricing to certain imported goods. However, its scope is more limited, currently covering iron and steel, cement, fertilizers, aluminium, electricity and a few other products, including hydrogen. Moving from its current transitional phase to full effect by 2026, the EU CBAM functions similarly to an import tax proportional to the greenhouse gas emissions caused during the production ('embodied emissions') of imported goods.

Note that while the EU legislation uses the term 'embedded' emissions, we refer to embodied emissions throughout this work to avoid confusion with end-of-life carbon release, particularly in the context of life-cycle assessment (LCA).

To prevent carbon leakage, the CBAM must accurately distinguish between high-emission and low-emission imports and charge producers accordingly. Inaccurate determination of embodied emissions could undermine the goal of the CBAM, potentially incentivizing EU producers to relocate to regions with looser emission standards and increasing global production emissions.

Accurate determination of embodied emissions is thus crucial for the effectiveness of CBAMs. From an environmental standpoint, the ideal policy would account for actual cradle-to-border life-cycle emissions for each import. However, achieving this theoretical ideal requires prohibitively extensive primary data. As an alternative, each imported product can be assigned a default embodied emission value, at the cost of reduced accuracy and diminished CBAM effectiveness.

The adopted EU CBAM stipulates that embodied emissions for imports are calculated using facility-specific emission data ('baseline calculation', Fig. 1). Specifically, baseline embodied emissions are the sum of direct emissions from the facility ('scope 1') and embodied emissions of any CBAM-covered precursor inputs into the facility, thereby including some non-energy-related supply chain emissions ('scope 3') in the calculation. In addition, emissions from energy supply ('scope 2') are accounted for in the fertilizer and cement sectors. Ideally, embodied emissions from CBAM-covered precursors should be determined using the same facility-specific baseline calculation. However, if facility-specific data for precursors are unavailable, up to 20% of the reported embodied emissions may be based on country-specific default values published by the European Commission. These fallback default values currently represent average embodied emissions for product groups. After the transitional period, all fallback values will be increased by a mark-up, whose size is still to be determined ('fallback calculation', Fig. 1). The EU Commission's proposed Omnibus I package (February 2025) would remove the 20% fallback cap, allowing more flexible use of fallback values.

The EU CBAM currently does not include the organic chemical and polymer industries despite the fact that these industries are emitting 20% of the global industrial greenhouse gases and are regarded as at risk of carbon leakage by the EU. While the EU debated expanding the CBAM scope to a number of organic base chemicals and polymers, these products were ultimately excluded owing to technical limitations that prevent accurate calculation of their embodied emissions. The rationale of the EU for the initial exclusion thus focuses on the feasibility of embodied emissions calculation at the beginning of the chemical supply chain. By contrast, the European Chemical Industry Council raises concerns about potential circumvention issues further downstream in the supply chain.

Despite these concerns, the European Commission will reassess an extension of the CBAM to the chemical and polymer sectors as part of the full CBAM review at the end of the transitional period, that is, by the end of 2025, with the goal of including these sectors by 2030. While qualitative studies on the CBAM classify the polymer and chemical industries as at risk of carbon leakage, and particularly complex, current quantitative studies on CBAM design bypass the issue of embodied emissions determination by focusing on broad economic assessments across all sectors and assuming one uniform product per sector. Although these higher-level models offer valuable insights into overall trade impacts, industry competitiveness, welfare impacts in individual countries and carbon leakage mitigation rates, they lack the granularity needed to capture product-specific differences in embodied emissions -- an aspect critical for accurately evaluating the effectiveness of the CBAM in the chemical sector.

Here we quantitatively assess whether the current EU CBAM is likely to accurately account for the emission intensity of non-EU chemicals and polymers. We focus on the beginning of the chemical supply chain, investigating products that were considered for CBAM inclusion during policy negotiations, in which the EU anticipates challenges in embodied emissions calculation. We quantify the performance of the EU CBAM in determining embodied emissions in imports of polyethylene and ethylene, the most-produced polymer and basic chemical, respectively. Using life-cycle assessment data for 4,470 chemical production facilities worldwide as a benchmark for embodied emissions, we study the capacity of the EU CBAM to calculate accurate embodied emissions for beginning-of-supply-chain chemicals.

Our analysis shows that the EU CBAM extended to the chemical sector covers approximately 50-60% of production chain emissions for ethylene and polyethylene production and qualitatively identifies emission intensity for polyethylene. However, we find perverse incentives in ethylene reporting. To improve the accuracy of embodied emissions calculations, the CBAM would need to include fossil feedstocks and refinery products and implement high fallback default values. We showcase that addressing these areas could substantially improve the CBAM emission coverage for beginning-of-supply-chain chemicals and polymers.