What Economic Models Say About CBAM’s Impact on Europe

Yes—if your work requires you to explain CBAM's economic effects, decide which commercial exposures deserve deeper analysis, challenge a headline, or prepare a sourced management/client briefing. No—if you need an individual company liability, supplier quote, legal conclusion or deterministic forecast. Those require company data and separate tools.

The practical message is more useful than either “CBAM barely matters” or “CBAM will transform everything.” Aggregate model effects are small, but cost, trade and value-added channels are uneven. Directly covered sectors receive some leakage protection while still facing free-allocation withdrawal, carbon costs, imported-input costs and export-market pressure. Downstream sectors can carry a large share of the aggregate effect because they are a much larger part of the economy.

The definitive regime started in 2026, but the strongest official quantitative record still combines transitional reporting data, adopted legal schedules and ex-ante models. The Commission can report quantities, Registry activity and estimates; those observations do not isolate what CBAM caused in European output, prices or jobs.C-001

For the observed quantitative baseline, use CBAM by the Numbers. This page owns the interpretation of economic models—not the Registry dataset itself.

Model percentages become misleading when their contracts disappear. The OECD reports value added, not GDP. The Commission's 2021 assessment reports a 2030 Fit for 55 option comparison. The World Bank Economic Exposure Index is estimated excess carbon payments divided by GDP—not a GDP-loss forecast. The Commission's 2025 downstream assessment measures marginal proposal effects against a baseline that already contains CBAM.C-002C-018

In the OECD model, the Full Policy Mix leaves aggregate EU all-industry value added at −0.29% relative to the 2019 baseline—the same displayed aggregate as removing relevant free allowances without CBAM. That does not make CBAM economically irrelevant: the mechanism improves some covered-sector outcomes while indirect losses spread across much larger non-covered sectors.C-004

The Commission's earlier JRC-GEM-E3 assessment also produced small macro differences across options: EU-27 GDP was approximately −0.22% to −0.23% versus baseline in 2030. But the Commission explicitly linked that contraction mainly to the wider MIX constraints required to meet the 2030 climate target—not to a standalone CBAM shock.C-011

CBAM is imposed at the import boundary, but economic incidence can travel through covered inputs, domestic production costs, substitution, demand and exports. OECD estimates that non-covered sectors account for 83% of the European value-added loss in its free-allocation-removal-plus-CBAM decomposition.C-006

Average sector price increases caused by introducing CBAM are mostly modelled between 0% and 0.6%. Basic metals and fabricated metals show larger effects, followed by electrical equipment, machinery and motor vehicles. The model assumes 100% cost pass-through; the values are not observed supplier pass-through coefficients. Use them to prioritise metal-intensive bills of material for evidence collection.C-007

The historical product-level screen is deliberately separate from the OECD model. It asks which 2019 manufacturing product groups cross a leakage-risk threshold at two carbon prices. It does not show that 13% or 26% of manufacturing relocated, nor that those goods are current CBAM scope.

“Covered by CBAM” is not the same as “protected from every economic channel.” The removal of free allocation raises costs for EU producers. CBAM can reduce import-side leakage, but covered producers may consume covered imports and remain exposed on export markets. OECD's modelled EU covered-sector value-added changes run from roughly −0.01% to −1.44%; it cuts the basic-metals loss from almost −3% without CBAM to around −1%, but non-metallic minerals receive less protection because only part of the broad model sector is covered.C-008

For current legal scope, use the goods checker and the current-versus-proposed downstream guide. Model sectors are broader than CN-coded legal scope.

The Commission's 2021 preferred transitional Option 4 modelled aggregate CBAM-sector imports 11.9% below baseline in 2030. Fertilisers showed the largest selected sector change at −26.41%; aluminium was −4.41%. The same assessment warned that resource shuffling and cleaner foreign production could materially reduce those import effects.C-012

The import result does not mean equivalent EU production growth. Demand can fall, inputs can be substituted and trade can reroute. The assessment also found weaker CBAM-sector export performance versus the MIX scenario, driven mainly by free-allocation loss and higher domestic prices. CBAM prices imports; it does not automatically compensate EU exports in third-country markets.

The Commission's 2025 downstream-extension assessment uses an updated JRC-GEM-E3 model and a baseline that already contains Fit for 55, REPowerEU, existing CBAM and the adopted Omnibus I simplifications. It therefore measures only the additional effect of three proposed downstream extensions. Every value in the table below is a 2035 result. It is not a newer estimate of total CBAM impact and the options are not adopted law.

The 2035 results explain an apparent contradiction: downstream imports can decline while basic-material imports rise slightly, as higher protected EU downstream output increases demand for inputs.C-023

JRC-GEM-E3 described overall employment effects as limited, but its Option 4 sector results do not share one sign. Covered sectors aggregate to +0.32%, while downstream sectors aggregate to −0.43%; iron and steel is +0.22%, cement −0.48%, fertilisers +2.59% and aluminium +0.89% in that option.C-013

Selected consumer-price effects were also small in the historical option comparison. Relative to full auctioning without CBAM, Option 4 ranges from no displayed change for heating/cooking appliances to +0.27% for fuels and power. This is not total inflation: it isolates one policy design comparison inside the 2030 assessment.C-014

The linked EUROMOD analysis described the overall distributional effect as regressive but very small, with differences across countries, income deciles and policy options. That finding is useful for deciding whether distributional sensitivity belongs in a policy brief. It is not precise enough to estimate a household bill today.

OECD's country decomposition covers value added in CBAM industries—not whole-country GDP. Results are mostly between +0.1% and −6%, with more pronounced adverse effects in Bulgaria, Cyprus, Estonia, Greece and Poland than in EU14 countries. Revenue recycling is excluded from this decomposition.C-009

Do not publish “Country X loses 6% of GDP.” The OECD figure is a change in covered-industry value added. For external-country exposure metrics and observed origin volumes, keep the World Bank and Commission populations separately labelled.

In OECD's all-industry comparison, higher ETS prices produce a 7.4% leakage rate. Removing relevant free allowances without CBAM raises that to 19.2%. Adding CBAM changes it to −12.4% because non-EU emissions also fall in the model. Adding CBAM increases global emissions reductions by 36% against the no-CBAM policy step—not against a no-policy world.C-010

The mechanism is trade reallocation towards less emissions-intensive suppliers. The result depends on historical intensities, modelled substitution and the absence of some future adaptations. It should support a counterfactual discussion, not a claim that a measured 2026 leakage rate is negative.

Most compelling model results represent full policy implementation. Adopted law phases CBAM in as the corresponding EU ETS free allocation falls: 97.5% is retained in 2026, 51.5% in 2030 and 0% in 2034. The CBAM-covered shares are the complements—2.5%, 48.5% and 100%.C-017

This matters twice. First, the cost channel grows over time even if emissions and certificate prices were unchanged. Second, firms, suppliers and governments can adapt during the ramp, while OECD deliberately applies full implementation immediately to the 2019 economy. Use the phase-in guide for the legal schedule and the cost calculator for a deterministic input-based planning estimate.

The quantitative spine is the OECD 2025 supply-chain paper and the Commission's 2021 impact assessment. Source PDFs were downloaded from official institutional hosts, checksum-frozen and extracted locally. Tables with difficult text-layer wrapping were checked against rendered source pages. CBAM Pulse created new visual treatments rather than copying institutional chart artwork.

Every promoted number carries a metric, unit, baseline, policy case, population or horizon, evidence class, locator and limitation through the adjacent text, figure note or claim appendix. Commas/decimal points were normalised for display; no hidden precision was created. OECD prose rounding of −1.06% is preserved rather than silently replacing it with Table 6's −1.07%.

The models are not averaged. Value added is not relabelled GDP. The World Bank exposure index is not relabelled GDP loss. Broad model sectors are not relabelled current legal CN scope. Historical Option 4 is not relabelled the final 2026 mechanism. Modelled emissions/leakage are not relabelled observations.

GSC showed no 90-day query/page rows containing “economic” or “impact” at the research cut-off. The article therefore exists as a useful, shareable professional reference and canonical evidence interpretation—not because current first-party search demand proves the topic. Update triggers include a Commission ex-post economic evaluation, a materially updated institutional model, realised definitive-regime sector evidence, or a legal phase-in change.

For the current operating picture, read The State of CBAM 2026. For official quantities and reporting patterns, use CBAM by the Numbers.