Can EU carbon tax and emissions trading scheme coexist? 

Juan Ojeda Plasencia

Sections  

1. Introduction  
2. Efficiency Perspective and Incentive Effects 
3. Economic Justifiability  
4. WTO Admissibility Concerns considering the Differing Implications  

1. Introduction  

The goal of this paper is to examine whether the instruments that the EU has taken for tackling CO2  emissions in the internal policy (Emissions Trading Scheme) and in the external policy (Carbon Border Adjustment Mechanism) are coherent and induce similarly efficient behaviours in EU vs non-EU producers. Building on that judgment it will be possible to establish whether the later instrument is admissible in international trade law.  

At the intra EU level, the ETS regime follows the cap-and-trade system. This system has been in place since 2005 and has evolved gradually from a system in which free allowances dominated and their allocations were decided based on historical levels of emissions to a system in which most allowances are auctioned, and the total cap (where the EU approved further cuts) is defined by the performance of the 90th percentile most emission efficient plants in a sector (Lamy & Pons, 2022). There is one notable exception to this, the industrial sectors subject to high risk of carbon leakage, which up to 2026 will still enjoy a 100% free allowance allocation.  

Meanwhile, the current proposal of an external policy is the CBAM, a measure designed “to make more equitable the competition terms between the EU producers and their competitors outside the Union” (European Parliament, 2023). The CBAM starts being rolled out in the heaviest CO2 emitting industries (cement, iron, steel, aluminium, fertilisers, electricity, and hydrogen), with reporting duties as of October 2023 and first submissions of certificates in May 2026 (Regulation 2023/955, OJEU). The CBAM allows offseting the price of carbon charges previously paid (whether through taxes or through market instruments). The Commission has not established caps to the number  of certificates that can be surrendered (Bednarek, 2023). Meanwhile, the quantity of certificates needed is indirectly controlled by the importers (ultimately by selecting the producers with whom they contract), subject to a mandatory audit. Additionally, the price is determined by the Commission with reference to the weekly averages for the ETS scheme (which have surged strongly in last years, to a level of 70-90 eur/tonne).  

2. Efficiency Related Aspects  

On the one hand a well-designed ETS has certainly efficiency inducing advantages (Böhringer, Hoffmann, Manrique, 2006). First, it  allows for a trading of allowances between producers of different countries, which reduces the concern of carbon leakage intra-EU. This the case because it allows for the most cost-effective abatement to occur and marginal abatement costs equalization in the economy (Milliam & Prince, 1989). It can be argued that the system of trading favours optimal allocation of allowances (the companies who can reduce their emissions at cheaper costs will do so and achieve surpluses, which will then be bought by the companies most willing to pay for them, those with highest abatement costs). The administrative costs of running this system are lower than a proper EU tax or enforcing a simple pollution non-tradable permit system (which would require exact allocation of allowances). The incentive to cut emissions is given by the price signal of the allowance auctions (Rosendahl & Storrosten, 2011).   

Nevertheless, two realities trump optimal ETS implementation in the EU. Firstly, there is the issue of over allocation of allowances relative to the actual number of allowances needed (Fell, Hintermann & Vollebergh, 2015).  This issue stems from the grandfathering and time fixed benchmarking of free allowances. The use of historical data, which no longer corresponds with what can be considered to reflect a CO2 efficient production, has created a surplus stock of 1.1 billion allowances since the ETS was implemented. The problem is compounded by the bankability of allowances, which can be stored indefinitely by companies (Lamy & Pons, 2022). That opens the door to strategic behaviours, such as holding out (which drives up prices in expect of higher future returns) and delivering them to the market precisely once regulators approve further reductions in the cap, which are then neutralised, undermining credibility of CO2-reducing policies (Osorio et al, 2021). This considerably distorts price signals of CO2 allowances (e.g they provide prices that are too volatile to impact investor behaviour significantly). Over the period 2008-2020 the price of allowances remained steady, between 5-20 eur/tonne CO2  and according to Dechezleprêtre et al (2022), mostly below 10 eur/tonne which is basic for CO2 emission reduction. An auction price of 30 eur/tonne is generally considered to support investment into emissions innovation (Joltreau & Sommerfeld, 2019). 

On the other hand, the CBAM meets two of the criteria which would make a Pigouvian tax feasible and desirable (Shavell, 2011). Firstly, the accuracy of this tax is high, as the only measurable variable is the CO2 intrinsic to production (Scope 1). This charge is based on audited reports or subsidiary default standards which gives it objectivity (Regulation 2023/955, OJEU). Furthermore, the certificate price equals the weekly average price of the allowances market, which grants transparency. Certificate sales are expected to generate revenues between 5-14 bn euros (Oxford Analytica, 2021). Secondly, the concern of omitted variables is minor as what we are dealing with here is not per se dangerous activities, such that the level of care is not relevant (importing of products is necessarily polluting but it is not risky).   

When it comes to the administrative costs however, the declaration system is foreseeably burdensome, with monitoring costs estimated around 10-14 mn euros, as it envisages to make all importers report on the yearly number of emissions incurred for the products they are bringing to the internal market, on a plant-by-plant basis. The fact that the charge depends on self-reporting increases filing costs for companies (which stand at 0,37 euros per ton, more than triple the cost of using default values) and heightens the perceived opportunity to under report, which is strongly linked with tax avoidance (Kirchler, 2007). Additionally, phasing in the CBAM will require the establishment of EC databases, as well as three reporting exercises of certificate demand for suppletory calculations. Even before the system starts generating income, the customs and national administrations will incur those costs (Cosbey, Mehling, and Marcu, 2021). It should be noted that a great part of the agencies’ cost and of the administrative burden is due to the legislative choice among the Options in the July 2021 Commission proposal. The alternatives at hand were less costly than the chosen Option 4, the self-declaration of emissions based on a gradual CBAM implementation. For example, an ex officio appreciation of the emissions (Option 1) offers lower monitoring costs (EC estimates costs at 4-5 mn euro) and allows the administration to keep their own record of CBAM certificates due, which minimizes importers’ perceived margin for tax avoidance (Kirchler, 2007). In short, while theoretical instrument choice is appropriate, its practical implementation is generating distortions and lack of fit between measures. This detracts from their justifiability which will be now examined. 

3. Economic Justifiability of the Measures  

To begin, there is the issue of Carbon Leakage: the CBAM is based on the premise that the pollution costs of EU and non-EU producers are equalised to avoid offshoring production. Therefore, importers are charged to submit certificates based on the carbon emissions incurred for their purchases. Thus, those sourcing from producers who inherently incur more CO2  emissions are charged more than those who produce less (the Pigouvian component of the CBAM). Secondly, there is the bi tax effect, as rising private costs are offset by rising social costs (Buchanan, 1999). The diseconomies of the exporters translate into a heightened price of EU imports. This incentivises intra EU production in sectors covered, but already high EU allowance and consumer prices increase further. For example, in electricity, a CBAM sector, the pass throughs were estimated at 65% up to 110% of the cost (Hintermann, 2016 Fabra and Reguant, 2014). Hence, the CBAM is implicitly a tariff, with a trade-off between consumer’s welfare loss and covered EU producers’ surplus. The net effects of this trade-off are negative, however. Microeconomically, firms are asymmetrically impacted by ETS costs, which leads to different exposures to carbon leakage. At the meso level, the sectors covered by the CBAM are precisely those with a higher presence of free allowance allocation and most inefficient emissions technology (Dechezleprêtre, Nachtigall & Venmans, 2018). They are also high consumers of coal, a C02 heavy commodity by default. Indeed, the Manufacturing Sector merely reduced CO2 emissions by 2,1% since 2013 (Lamy & Pons, 2022). Macroeconomically, only a small part of differential emission costs entails an outward shift in production in the EU. Dechezleprêtre et al (2022) estimated that this effect would be lower than 0,7% of the costs.  

Alternatively, the gradual and limited rollout of the CBAM might lead to trade diversion. As only some products will be subject to certificates, there is the risk that the importers will replace CBAM subjected basic products with more valuable finished products downwards in the chain that will be exempt (Bellora & Fontagné, 2023). Higher prices will cause a real income loss, especially steep in Poland and Germany, EU’s largest coal importers and consumers (Moessner, 2022). Hence, imports of intermediate products are expected to drop by -8.3%  (-3% for finished goods) by 2040. Decarbonisation is discouraged as MS coal production increases to replace costlier imported electricity. Increasing prices of basic products will also make the EU exports of finished goods containing them more expensive. Exports of finished goods go down by -6% and for intermediate ones by -8.6% by 2040. Under the assumption that EU production standards for finished goods are more GHG efficient than the rest of the world, both effects lead to a net reduction in the value of EU production and a net increase in production in the rest of the world. In summary, possibly there will be higher global CO2 emissions and, certainly, a deterioration of the EU trade balance (Monjon & Quirion, 2011).  

4. WTO Admissibility Concerns considering the Differing Implications  

The coherence between the functioning of the ETS and the CBAM is a not only an economic question, but also conditions its international trade law admissibility (De Cendra, 2006). A first solution is compliance with article III(2) GATT. Said article requires that the products of any contracting party imported into the territory of any other are not imposed charges in excess of those applied to like domestic products (Krenek, Sommer & Schratzenstaller, 2020). A second solution is compliance with article XX GATT, so that the CBAM is granted an exemption. This requires that: (1) the external policy main aim is preserving exhaustible natural resources, (2) the measure itself contributes to such goal, and (3) there is an internal policy that jointly reduces production (emissions). Under requirement (2), the doctrine of the Appellate Body implies that the measure taken must be necessary part of a system with the internal policy (De Cendra, 2006). Thus, if the internal policy fails its environmental goal protection by itself, the CBAM is an inadmissible trade barrier. While both articles require policy coherence, the equivalence assumption is stronger for Article III(2) GATT. Article XX, however, requires extensive “good faith” renegotiations before the CBAM can be set in motion (likely to be contested by most non-EU countries), which is a disadvantage over Article III(2), conceived for unilateral measures. Therefore, for its lesser litigation and retaliation risks Article III(2) compliance is preferable.  

Hence, to ensure mild compatibility, ETS price signalling should be revised. First, the free allowance system and bankability clauses would need to be fully (and not gradually) phased out before the CBAM starts being levied, so that asymmetries in exposure to the price mechanism are fixed, and high carbon emitters start correcting their behaviour in earnest. As CBAM prices depend on ETS’, this ensures requisites 1 and 2 for compliance under Article XX GATT are met. Second, the current auctioning system leads to unreliable prices, which puts requisite 3 compliance in question. Preannouncing emissions caps makes bids susceptible to strategic behaviour. Firms adopt demand reduction strategies to underbid (Holt & Shobe, 2016). This inefficiency can be redressed by reversing auction procedures (Khezr & Mackenzie, 2021). Bidding should proceed first with indicative allowance ceiling. Then, based on revealed prices after the bid, the total number of emissions permits can be fixed. Only this way are marginal abatement costs revealed, so that the ETS can be considered to produce a reliable price signal, which effectively contributes to reduce emissions (Neuhoff, 2007).  

Additionally to the previous proposal, Article III(2) compliance requires CBAM reform, so that the charges are based on the best available technique standard of the WTO Appellate Body. In other words, the policy proposal would be to require certificate submission relative to the average level of intensity of emissions in the EU (default calculated emissions). This requirement maximizes compliance with the doctrine of the Appellate Body because it fulfils Art III(2) requirement of national treatment. It also avoids the problem of stablishing a CBAM in breach of previous trade treaties and the EU’s Generalised System of Preferences (Regulation 978/2012). Assuming EU producers to be the ones applying the best available abatement technologies in the market, this would give less developed countries products a  more equal chance to compete in the EU market, since importers would face lower tax bases respective to the ones considering declared emissions. Administratively, this policy is also cheaper, as previously argued.  Finally, trade diversion and real income issues are softened because import prices  will increase less (Bellora & Fontagné, 2023). Imports will decrease only by -3% on average and exports by -6,3% (for intermediate goods) and -2,6% (for final goods).