The CBAM evolves: the EU’s Omnibus Regulation and anti-shuffling measures

As the world’s first Carbon Border Adjustment Mechanism (CBAM), the EU's CBAM will enter its formal implementation phase in October. However, some aspects of the mechanism remain under discussion. On 26 February 2025, the European Commission proposed an Omnibus Regulation aimed at simplifying compliance procedures under the CBAM. More recently, Climate Commissioner Wopke Hoekstra mentioned their intention to introduce measures to address ‘resource shuffling.’ This blog examines the context of these developments and their implications.

1. CBAM Omnibus

A key element of this proposal is the establishment of a De Minimis threshold such that traders importing less than 50 tonnes of net mass per year of CBAM-covered goods are exempt from reporting. This threshold is designed to balance administrative efficiency with environmental objectives: chosen to leave more than 99% of embedded emissions subject to CBAM-payments, it exempts over 90% of importers from reporting. This calculation is based on weight and emissions data from 1 October 2023 to 30 September 2024 and is laid out in detail in an excellent staff paper. We have two key concerns with this approach.

First, the Commission's decision to adopt a mass-based rather than an emissions-based threshold is notable. This approach retains a conceptual link to emissions—albeit in an approximate manner—it offers a more direct proxy for embedded emissions than product value, which was the original (and is still the UK) basis for the threshold. However, this method introduces an additional layer of approximation, as emissions per tonne of product vary substantially across different goods,1 a limitation acknowledged by the Commission itself. The primary justification for this choice is administrative simplicity. However, this simplification comes at the expense of accuracy.

A better alternative would involve publishing default values for emissions intensity2 by CN code3 and allowing the system to automatically estimate total emissions based on reported quantities. This would make an emissions-based threshold as straightforward as a mass-based one, provided the data infrastructure was in place. By streamlining emissions calculation, this method would better support the objectives of the CBAM policy.

Second, the methodology used to establish the 50-tonne threshold also presents challenges. The threshold was determined by computing total emissions based on customs data for net mass and applying estimated emissions intensity without mark-ups. (Mark-ups are intended to capture the extent to which products from each country exceed the default values.) The intent was to ensure that the threshold captures at least 99% of total emissions. However, this approach assumes that mark-ups are applied proportionally across all products and countries, which may not hold in practice. While the staff paper acknowledges this limitation, it remains a critical issue that warrants further scrutiny. If mark-ups vary across product categories and exceed 1, actual emissions will be higher than those calculated without mark-ups. Additionally, the country composition of imports could influence overall emission estimates. To be fair, however, ignoring the mark-ups in the calculations also impacts the emission-based threshold that we proposed.

The Omnibus Regulation introduces many other simplifications, of which two significant ones are:

1. Importers can choose freely between reporting actual emissions or using default values, with no justification required.

2. Default values for calculating payments will be based on the average emission intensity of the ten highest-emission countries for which data are available. (Note, these are not the defaults, the Commission’s figures, used in the calculations above, but will be published in due course.)

The effectiveness of this system depends heavily on data collection, particularly from high-emission (‘dirty’) countries. If sufficient data from these countries is unavailable, the default values may be lower than actual emissions, making them an attractive option for importers seeking to minimise their declared emissions. This could reduce incentives for firms to provide actual emissions data, potentially undermining the environmental ambition of the CBAM.

2. Anti-shuffling measures

Simultaneously with the Omnibus reforms to reduce the burden of the CBAM on traders, the EU is also contemplating a move in the opposite direction. Andy Bounds of the FT reports that Climate Commissioner Wopke Hoekstra intends to introduce measures to prevent ‘resource shuffling’ whereby foreign firms send their cleaner output to the EU (and so incur lower CBAM charges) and continue to use ‘dirty’ methods for goods sold elsewhere. He proposes that, in the calculation of payments, the carbon intensity of exports to the EU will be based on the average for the firm’s total output (or on a national average) rather than that of just the exports to the EU. He says this strengthens the incentive for foreign firms to clean up their production. We see several disadvantages of this approach.

First, Hoekstra has to extract intensity data on total output from the foreign firm; it might not be willing and would certainly incur greater costs than reporting just on exports. If the proposal is restricted to iron and steel, it may be feasible – quite a lot is known about steel plants around the world from the World Steel Association – but it is not clear whether that is what Hoekstra means. But if this approach is implemented only for iron and steel, it will lead to inconsistencies across CBAM products. Over 42% of EU importers dealing with CBAM-covered goods trade other products and they will face different rules from steel traders.

Second, the move takes the EU away from ‘merely’ equalising the cost of emissions embodied in imports and in EU production. The latter is a perfectly reasonable public finance objective that stretches back at least to Adam Smith. But charging firms for emissions that neither occur nor are consumed in the EU, introduces extra-territoriality. It is almost ‘Trumpian’ – ‘we don’t like what you do, so we’re imposing a tax’. The CBAM’s consistency with WTO rules is uncertain, but the EU has made a huge effort to align it to the maximum extent. Hoekstra’s proposal completely undermines that effort and surrenders any ‘moral high ground’ that the EU might claim in justifying the CBAM.

Resource-shuffling undoubtedly weakens the EU’s objective of the CBAM inducing decarbonisation abroad. Overseas decarbonisation would suit most of us, but that does not make it a legitimate target for EU policies; in most international regimes, such extraterritoriality is frowned upon. Given that it is likely to stir up a sufficiently large swarm of objections that the CBAM’s viability is threatened, we would avoid it. Third, it probably reduces the incentive for foreign firms to decarbonise. They may be willing to pay €X to halve the carbon intensity of exports to the EU and so halve their CBAM payment. But if, say, sales to the EU are only 10% of their output, averaging means the investment will now reduce intensity and payment by just 5% (10% of half) and so not be worthwhile. Many firms make little effort to decarbonise, so diluting the incentive on sales to the EU could mean the difference between something and nothing - i.e. undermine their incentive to experiment and learn about low carbon methods at all.