Impact Analysis Report on Optimising the CBAM De Minimis Threshold

Impact Analysis Report on Optimising the CBAM De Minimis Threshold

ERCST’s Impact Analysis Report on optimising the CBAM de minimis threshold tests whether the current 50-tonne cumulative threshold strikes the best balance between administrative relief and emissions coverage, and how alternative designs would compare.

Key ERCST observations:

  • The study tests six threshold designs against a dataset of 63,405 import shipments from downstream manufacturers, principally in the automotive and machinery and equipment sectors. Administrative relief is multi-dimensional. A threshold can relieve the burden on importers or the burden along the value chain (supplier relationships), and these are not the same.
  • The current 50-tonne rule relieves importers well: it takes 54% of the importers in the dataset out of scope while retaining 99.5% of embedded emissions. However, it does little for the value chain, leaving more than 90% of supplier relationships in scope.
  • Only a combined importer-and-pair design relieves both. It adds a 1-tonne threshold per importer-supplier relationship on top of the 50-tonne rule, exempting 56% of importers and 74% of relationships while retaining 98.5% of emissions.
  • Shipment-based designs (value or weight per consignment) exempt more shipments but lose more emissions coverage. They are also the most exposed to circumvention through shipment splitting.
  • The value-chain burden will grow over time. For every supplier relationship in scope, the third-country installation must determine and communicate its embedded emissions, and that cost rises several-fold as CBAM moves from default to actual values.
  • The 50-tonne rule should be complemented, not replaced, with a relationship-level threshold. In fragmented sectors, where most importers already fall below 50 tonnes, a pair-based threshold applied on its own may be the more relevant instrument. The precise calibration is set out in the accompanying Policy Proposal.