The European Union’s Carbon Border Adjustment Mechanism (EU CBAM) puts a carbon price on emissions embedded in certain products imported into the EU. Its effects will depend not only on how carbon-intensive exports are, but also on the particular products countries export and the price of those products. For South Africa, this makes industrial upgrading, cleaner electricity, and credible emissions measurement increasingly important for export competitiveness.

On 1 January 2026, the European Union’s Carbon Border Adjustment Mechanism (EU CBAM) entered its definitive phase. The policy is intended to ensure that carbon-intensive products imported into the EU face a carbon cost comparable to that paid by producers within the EU. It currently covers selected products in carbon-intensive sectors2.

But it will not hit all exporters equally – and this matters for South Africa.

For exporters of the relevant products, carbon emissions increasingly become a cost in accessing the EU market. The more emissions associated with the production process, the greater the liability associated with exporting those products to the EU. This has understandably raised concerns among emerging economies. Their production processes are often more carbon-intensive, while firms may have less access to the finance and technologies needed to decarbonise. Existing research consequently suggests that the costs of the EU CBAM could fall disproportionately on emerging economies.

But looking only at countries or broad industries misses an important part of the story. Our research shows that the impact of the EU CBAM can differ substantially between products within the same industry and between countries exporting the same product.

A carbon price at the border

The basic principle behind the EU CBAM is relatively simple. EU producers already pay for their emissions through the EU Emissions Trading System. The EU CBAM extends a comparable carbon cost to imports, reducing carbon leakage incentives – e.g., EU firms relocating carbon-intensive production abroad or replacing EU production with emissions-intensive imports.

However, what sounds simple as a concept is more complicated when applied. EU CBAM liability depends on the quantity imported, the emissions embedded in producing the goods, the ability of firms to monitor, report and verify these emissions (and the default emission factors applied if they cannot), the EU carbon price, carbon prices already paid in the exporting country, and the pace of the removal of free emissions allowances received by EU producers.

Competitiveness depends on more than just embedded emissions

An important element of the policy is that embedded emissions are measured per tonne of product imported into the EU, not per unit of value. This distinction sounds technical, but it has important economic consequences. As an example, suppose two exporting firms sell the same type of steel product, and producing one tonne generates exactly the same emissions. Both firms face the same carbon cost per tonne of steel. But suppose one exporter sells its steel for EUR 500 a tonne while the other sells for EUR 1 000. A carbon cost of EUR 100 raises the price of the first product by 20%, but the second by only 10%. The carbon emissions and associated costs are identical, but the effect on competitiveness is not.

This means the impact of the EU CBAM depends not only on carbon intensity, but also on product price. Lower-priced exports may experience a much larger percentage price increase for the same emissions intensity.

Why sector averages are inadequate

Much of the existing research estimates EU CBAM impacts using aggregated sectors. Our research instead examines trade at the detailed product-level used in EU customs data. Ideally, plant- and product-level data would be used. As these are unavailable, product-level estimates of emission factors for major exporters to the EU are used, alongside EU trade data3.

Our study analyses 200 iron & steel and 27 aluminium products, all covered by the EU CBAM. The countries for which the data are available (16 countries, with 8 that export both products, the EU, and a ‘Rest of World’ average) account for more than 85% of extra-EU iron & steel imports and 86% of aluminium imports.

The findings reveal that product-level differences hidden by sector averages are substantial. For example, estimated total emissions (direct emissions from the production process plus indirect emissions from the electricity consumed in production) for South African iron & steel products range from 0.74 to 7.47 tonnes of carbon-dioxide equivalent per tonne of product (average of 5.4). For aluminium, the total emission factor has a narrower (16 to 18) range, but a higher average (17.6). Aggregate sector-level emission factors cannot account for this variation and so knowing the average emissions intensity of “South African iron & steel or aluminium” tells us relatively little about the exposure of a particular exporter or product.

What the simulated results mean for EU imports

We use a product-level trade model4 to estimate how EU imports could respond to the price increases created by the EU CBAM. A carbon price of EUR 80 per tonne carbon-dioxide equivalent is applied to embedded emissions, and the resulting carbon cost is expressed as a percentage of each product’s import price (estimated tariff-equivalent). These price increases are used in a multi-country partial equilibrium trade model, which captures two important trade effects:

Direct effect: higher prices reduce EU demand for imports

Diversion effect: differences in carbon costs between suppliers can change their relative competitiveness, and so EU buyers may shift towards products from countries facing smaller percentage price increases

The results should not be treated as forecasts, but as simulations showing how the EU CBAM could reshape trade under a common set of assumptions. When only direct emissions are included, EU imports of covered iron & steel products decline by around 15%, while aluminium imports fall by around 5%. If we include indirect emissions, these decline to 17.2% and 13.1%, respectively. These aggregate effects, however, mask substantial differences across countries and products.

For the median iron & steel product in the lower-income group, EU imports fall by 23.4%, compared with 14.7% among middle-income exporters, and only 3.1% among higher-income exporters. The same pattern, although less pronounced, appears in aluminium. Imports of the median product in the lower-income group fell by 6%, compared with 5.4% for the middle-income group and 2.8% for the higher-income group.

This does not mean that GDP per capita itself determines EU CBAM exposure. Rather, income is associated with factors that shape the overall impact - including production technologies, emissions intensity, and product prices. Higher-income economies6 in our sample tend to export higher-priced product variations with lower embedded emissions, while several lower-income emerging economies combine more emissions-intensive production with lower-priced exports. Moreover, the inclusion of indirect emissions in the simulations widens the distribution of import changes and amplifies the apparent bias against emerging economies in the lower- and middle-income groupings.

South Africa’s electricity problem could become a trade problem

At present, the EU CBAM applies only to direct emissions for iron & steel and aluminium products. But the incorporation of indirect emissions is being considered. This decision is particularly important for South Africa, as our estimates show that the inclusion of indirect emissions substantially increases potential exposure of iron & steel and aluminium exports. This is because of the carbon intensity of the country’s electricity system. Across the products analysed, South Africa records the highest average price increases once both direct and indirect emissions are included.

The aluminium results are particularly striking. Under direct emissions alone, simulated EU imports from South Africa decline by about 7%. Once indirect emissions are added, the estimated decline rises to approximately 75%. This is not a prediction that South African aluminium exports will fall by 75%. Firms can change technologies and switch to renewable energy sources, trade patterns can adjust, and the eventual EU rules may differ from our scenario. However, the results do illustrate the scale of the exposure created when the country’s electricity system becomes part of the carbon content of its exports.

Electricity-sector decarbonisation is therefore no longer only an environmental issue. It is increasingly an industrial policy and trade competitiveness issue.

What does this mean for policymakers in South Africa?

The first implication is that emission reduction efforts cannot simply be left to individual exporters. Firms have strong incentives to reduce their own emissions through cleaner technologies and more efficient production, but these investments are expensive. Industrial policy can help overcome these financing and coordination constraints through instruments such as concessional finance, investment incentives, research and development support, and recycling carbon-tax revenues into industrial decarbonisation.

Second, with the inclusion of indirect emissions in time, South Africa’s electricity transition will matter for exporters to the EU. Expanding lower-carbon generation and grid capacity can reduce the indirect emissions embedded in locally produced goods.

Third, domestic carbon pricing presents an opportunity rather than simply an additional cost. Where carbon prices paid in the exporting country are recognised by the EU, they can be deducted from the EU CBAM liability. A credible domestic carbon-pricing system therefore allows South Africa to retain carbon revenues domestically rather than effectively transferring them to Europe. Recycling those revenues into industrial decarbonisation could reinforce this benefit.

Finally, exporters need credible systems for measuring, reporting and verifying their emissions. Firms that cannot demonstrate their actual emissions can be required to use conservative default values, potentially increasing their liability significantly. Building the institutions and technical capacity required to measure carbon accurately is therefore becoming part of export infrastructure.

To conclude, the EU CBAM will not affect every South African exporter in the same way. Exposure depends on what a firm produces, how it produces it, and the price at which its products compete. That makes the policy challenge more complicated than simply asking whether South Africa is “exposed” to the EU CBAM. The more useful question is which products and firms are exposed, why, and what can be changed? The answer increasingly links trade policy, industrial policy, carbon pricing, and the electricity transition. As carbon increasingly becomes part of the price of participating in international markets, those policy areas can no longer be treated separately.

Written by Lawrence Edwards, Nicola Wills, and Brett Cohen, Econ3x3