We continue with part III of our series Carbon border tax explained. See Part I and Part II, if you have missed them.

In view of these benefits, why aren’t European politicians and EU officials all up for it? The answer is that the introduction of a Border Carbon Adjustment (BCA) has a number of caveats, most notably it’s technical complexity and the difficulty of designing a scheme that is compatible with internal trade law.

Designing the BCA so that it is WTO-proof

The World Trade Organisation’s mission is to regulate international trade between participating nations and ensure fair play between all parties.

One of the key principles of the international trading system is the non-discrimination of imported goods which must be treated no less favourably than domestically produced “like”, that is “comparable”, goods. “Like” is generally understood as “fulfilling the same purpose” and “competing with the domestic product”. Countries are allowed to impose taxes or charges on imports provided that the amount of the BCA imposed on the imported goods does not exceed the amount of the tax on the domestically-produced like products. In addition, they must show that the tax is imposed on a product (a so-called indirect tax) and not on a producer or manufacturer or their income (a direct tax).

Exceptions to the non-discrimination principle are allowed for the “protection of plant, animal and human life and health, and conservation of scarce natural resources” (Article XX General Exceptions of the GATT). At the same time, WTO law includes provisions that aim to ensure that Article XX is not used as a fig leaf for protecting an economy against competition. Most importantly, it includes the requirement that any policy may not represent arbitrary or unjustifiable discrimination between countries where the same conditions prevail or a disguised restriction on international trade. That means that if the EU revokes Article XX to justify the BCA it has to demonstrate that the measure genuinely aims at reducing carbon emissions for environmental and human health reasons and is not a protectionist measure in disguise.

The second important restriction imposed under WTO law on a carbon border tax is the subsidy law under the Agreement on Subsidies and Countervailing Measures (SCM) of the WTO. It regulates state aid (“subsidies”) by de facto prohibiting subsidies that are linked to promoting exports. Since EU products destined for exportation face a competitive disadvantage due to the higher production costs the EU could have an incentive to refund to EU producers the taxes paid. While experts opinions diverge whether such a rebate would be WTO-compatible, there are a number of practical problems that arise from giving a rebate to EU producers for products going into export. Firstly, under the SCM rules, other WTO members could also challenge the rebate if the amount higher than the domestic carbon tax paid or if the rebate is paid for exports to countries with a national carbon tax similar to the EUs, because it would be considered an illegal export subsidy. On the other hand, if exports to certain countries are excluded from the rebate, a number of other practical issues would emerge, which are essentially the same as those that apply when exempting certain importing countries from the tax. This is explained in the next point.

Achieving international and domestic acceptance of a carbon border tax

In the current climate of smouldering trade disputes, there is a real risk that the unilateral decision of the European Union to introduce a carbon border tax will provoke retaliatory action against the EU. Even though many experts are convinced that BCA can be designed in compliance with WTO rules, European diplomats warn against stoking disputes with major trading partners or within the bloc itself.

To design a policy that is perceived as fair and administrable, the EU could consider exempting en bloc from the BCA scheme those countries that have similarly strict regulations, e.g. parties to a multi-lateral climate change agreement or countries having defined a national cap on emissions or any other adequate and comparable measure. However, such a move would be in contradiction with the most-favoured-nation principle (Article I of the GATT), which requires the European Union to treat imports from all WTO members the same in terms of duties or fees. Another problem is that exempting certain markets would open the door for carbon leakage as non-exempted countries could ship products to exempted countries for re-labelling and re-export to avoid taxation. This would be difficult to control for the EU. On the other hand, not exempting parties to a multilateral climate change agreement could be considered a violation of the United Nations Framework Convention on Climate Change (UNFCCC) principle of common but differentiated responsibility as it would involve demanding more than is demanded by the treaty’s multilaterally agreed allocation of burdens. It especially raises the issue of discriminating low income and least developed countries, even though some experts argue that these countries would hardly be affected since almost none of them export the type of goods that are most likely targeted by a BCA. Furthermore, it will prove difficult to define what an “adequate” and “comparable” measure is that justifies an exemption.

The simplest version of the BCA, according to experts, would be a common domestic tax across the EU on certain carbon-intensive products, which could also apply to imports, instead of an adjustment mechanism that mirrors the carbon price in the EU’s ETS. But tax is a national competence in the EU so every EU member state would need to agree to and implement it independently.

Dealing with the technical complexity of quantifying the carbon intensity of imported goods

In order for the BCA to be compatible with WTO law it has to be reciprocal to an EU carbon tax and has to reflect the amount of carbon dioxide emitted during the production of the product. The EU could require goods for EU import to have a certificate that states the amount of carbon embodied in the product, its carbon footprint, and which has to be issued following an agreed standard methodology. Determining (or estimating) the carbon footprint of a given imported product is complex. It depends on the specific manufacturing plant, the energy sources used, and the production process. In the case of assembled goods, the carbon footprint would have to be determined for each part of the product. While this is not impossible and such methodologies exist already (e.g. for determining benchmarks for the EU’s Emissions Trading System), it is only useful if there is a way for the EU to verify the certificates veracity.

Experts therefore argue that any such system would need an appropriate alternative means to set the carbon content of an imported good if exporters are unable or unwilling to provide the necessary thirdparty verified data. There are several options how to do this. Reference values or benchmarks can be assumed which are either based on the 1.) average emission intensity of a product in the exporting country, 2.) average emission intensity in the EU, 3.) emission intensity of the best-available technology or 4.) emission intensity of the worst practice (or hybrids of all four).

Each method has its disadvantages:

• When assuming the average emission intensity of the exporting country, exporters who emit more have little incentive to improve their performance.
• When using the average emission intensity in the EU (with its higher standards) as benchmark, the policy would be less effective in preventing leakage. This is even more so true if the benchmark assumes that the best available technology has been used.
• Assuming worst practice would be the most effective benchmark for avoiding carbon leakage, but could come with a political price as it could be perceived as unfair by trading partners.

The best chance of withstanding WTO scrutiny, according to experts, have benchmarks assuming that the carbon content of the imported product is equal to the carbon content of the like product produced by the predominant method of production in the EU. Nonetheless, the benchmark method should only be used as a fallback if no actual thirdparty verified data can be provided.

Continued in Part IV: How should the carbon border tax be designed?.