Skip to main content


Does the international agreement on the OECD pillars mark the end of trade wars on digital taxes?

Speed read

The rise in both national and international digital tax measures has seen an increase in the use of trade law challenges, as the US seeks to put a stop to measures that it believes disproportionately target large US-headed groups. The international agreement in July on the OECD’s two-pillar approach is intended to take unilateral digital tax measures off the table and put an end to these trade wars. While we may see further challenges to any unilateral digital tax measures that are not withdrawn, we are beginning to see greater alignment between international trade and tax law in the digital tax sphere – with trade law more as a mechanism to help keep the peace, rather than being used to wage war.

Recent years have seen numerous attempts to levy more tax from the world’s biggest digital players. There have been a variety of drivers behind this, not least a desire to ensure that multinationals pay their ‘fair share’; and public pressure for action against ‘big tech’. But beyond this, there has been a growing sense that the existing tax system is not fit for purpose in an increasingly digitalised and globalised world and, more pragmatically, there is now also a need for more funds to aid recovery following the covid-19 pandemic. This combination of factors has resulted in an explosion of unilateral measures that seek to tax digital businesses in a number of different ways. This is in addition to multilateral efforts to address such issues at both an OECD and an EU level. At times, these measures have bumped into trade law, for instance the US s 301 investigations against the national digital services taxes imposed by France, Austria, India, Spain, Turkey, and the UK and the looming threat of retaliatory tariffs. In July, more than 130 members of the Inclusive Framework reached agreement on the OECD’s two pillar approach to tackling these issues. Does this mean the end of trade wars on digital taxes?

Unilateral measures

One of the key factors motivating jurisdictions to find global consensus along the lines of the OECD proposals has been the proliferation of unilateral digital taxes and the potential for multiple jurisdictions attempting to claim taxing rights in a variety of subtly different but overlapping ways over the same amounts. The July agreement on the OECD pillars includes a statement that the package will provide for ‘the removal of all digital service taxes and other relevant similar measures on all companies’. This agreement to withdraw unilateral measures is thought to have been fundamental to the US agreeing to come on board with the OECD’s plans.

What is less clear from the July agreement is exactly which measures will have to be withdrawn and by when. To the extent that some are retained permanently or perhaps retained for longer than might be considered desirable, could there still be scope for trade challenges? There has been a spectrum of responses to the July agreement from those jurisdictions with digital taxes. The French finance minister, Bruno Le Maire, for instance, was quick to reassure the US that the French national digital tax would disappear ‘as soon as the OECD tax is implemented’. However, it is unclear whether this enthusiasm stemmed from the understanding that the EU would be imposing its own digital levy, a position which is now in doubt. The UK has historically held its cards close to its chest, saying it would disapply the UK digital services tax if ‘an appropriate global solution’ were successfully agreed and implemented, without commenting on whether the OECD proposals provided such a solution. However, in the wake of the G7 agreement in June, the UK chancellor seems to have firmed up his stance, indicating a commitment to removing the tax ‘once a Pillar One solution is in place’. More concerning for the US might be the Austrian response, with finance minister Gernot Blümel pushing back on suggestions that the Austrian digital tax would be repealed in response to the July agreement. The Czech proposals for a new digital services tax, currently going through the legislative process, provide that if an agreement is reached at an OECD level then 2024 would be the last period for which the Czech tax would apply. With the OECD proposals due to be implemented in 2023, this raises the issue of when unilateral measures will have to be withdrawn by and whether this can be achieved quickly enough to avoid trade repercussions.

To the extent that unilateral measures are not withdrawn, or are not withdrawn quickly enough, it is clear that trade law could well have a role to play here. The conclusion of the s 301 investigations by the Office of the US Trade Representative (USTR) was that the digital taxes of France, Austria, India, Italy, Spain, Turkey and the UK were discriminatory against US businesses. The US suspended the additional 25 per cent. tariffs on imports from France until further notice and on imports from the other six countries until 29 November pending the OECD discussions, but if any of these countries do not respond in a way that satisfies the US, these trade sanctions may well be applied. In itself, this response to digital services taxes raises problems under WTO law. It is not possible for the US to impose retaliatory sanctions on other WTO members without a proper justification. Indeed, the WTO’s 1995 dispute settlement mechanism was designed specifically to constrain the use of s 301 during the 1980s and early 1990s.

A second question, however, is whether the US (or others) might seek recourse against digital services taxes under WTO law. This would, of course, depend on which measures the US was seeking to challenge but, depending on the design of the digital tax, it is not inconceivable that there could be a challenge under the General Agreement on Trade in Services (GATS) on the basis that such taxes are discriminatory. For this, a tax must apply differently to ‘like’ services and service suppliers from (or in) different countries. That depends on whether consumers consider them to be substitutable, and whether there are distinguishing features of the services and service suppliers meriting a differential tax burden. It is far from clear that the revenue earned by a given company is a sufficient basis for such differentiation.

Even if a digital tax applies differently to companies according to an arbitrary criterion such as company revenue, it will only be discriminatory if a greater proportion of affected companies are from a particular foreign country (for example, the US). Furthermore, while discrimination between foreign services and service suppliers from (or in) different countries will almost certainly violate the most favoured nation obligation (one of the two GATS non-discrimination obligations), discrimination between foreign services and service suppliers and domestic services and service suppliers will only violate the national treatment obligation (the second non-discrimination obligation) if the country imposing the tax has made a commitment not to discriminate against services and service suppliers in a given sector.

Finally, GATS sets out some exceptions to these non­discrimination obligations. One is for measures falling under double taxation agreements. However, digital taxes are generally designed not to do so. Another exception, and possibly the most important one, is for tax measures aimed at ensuring the equitable or effective imposition or collection of direct taxes in respect of services or service suppliers of other members. This exception would cover income taxes, including those based on revenue. But it would not cover indirect taxes levied on transactions. It is also limited in scope, providing cover only for violations of the national treatment obligation, not the most favoured nation obligation. It is therefore by no means impossible for a digital tax to survive GATS scrutiny, but it does need to be very carefully drafted for this to be the end result.

EU digital levy

And what of the EU digital levy? The EU has had various attempts to raise funds via digital taxes over the years. It has proposed a digital profits tax, an interim digital activities tax and, in November 2018, a digital services tax. This last proposal was taken off the table pending the OECD discussions. However, the covid-19 pandemic led to a pressing need for funds which mean that an EU digital levy was put back on the agenda and, in 2020, the EU heads of state asked the European Commission to put together a proposal which was due to be published in July 2021. Despite reassurances from the EU that this could comfortably sit side by side with the OECD proposals, the Commission came under considerable pressure, in particular from the US, to put its digital levy plan on hold so as to not jeopardise the international negotiations on the two-pillar approach.

The EU has not guaranteed that these plans will not re­surface. It has however learned some valuable lessons in terms of the design and structure of digital taxes and the latest iteration, the ‘digital levy’, is less likely to be discriminatory than previous proposals. For instance, previous proposals contained high revenue thresholds, excluded revenues primarily earned by European firms, and allowed VAT to be subtracted from the tax base. The US, with no equivalent of EU VAT, therefore had a number of legitimate grounds for claiming these proposals amounted to de facto discrimination against US companies. By way of contrast, indications are that the 2021 digital levy would be a low value, non-discriminatory tax on goods and services sold online, with a low (€50m) revenue threshold. So instead of taxing large multinationals, the aim would be to target a large number of digital sales. The Commission consultation on this levy earlier in 2021 noted that not only would the levy be consistent with the ongoing work at the OECD but also ‘will be compliant with WTO and other international obligations’. The EU does seem to have heeded the lessons of the past and, if crafted carefully, the EU digital levy could be safe from challenges under trade law.

OECD pillars

An area that does not appear to have been examined in great detail is whether the OECD proposals themselves could be subject to trade challenges. In terms of pillar one, there are strong indications that it will predominantly be US headed-groups that are brought into scope by the pillar one thresholds. However, it is hard to see a trade challenge coming on that front given that it was the US that spearheaded the latest compromise. Furthermore, it has been reported that US Secretary Yellen has indicated that pillar one would be revenue neutral for the US. This potentially reflects the fact that the US is itself a large market jurisdiction, meaning that a significant proportion of the profits subject to reallocation under pillar one may well end up being allocated to the US.

Is the position any different for pillar two? Until we have all the details, it is hard to be definitive on the potential for a trade challenge. However, the imposition of a minimum global tax rate will undoubtedly impact those jurisdictions that have chosen to implement a rate below the agreed minimum (or not to impose tax at all). That said, for there to be discrimination under WTO law, there would need to be discrimination against a group of ‘like’ companies from on country compared to the treatment of like companies from other countries. And, from what we know so far, it seems more likely that it would be multinationals with presence in pillar two participating jurisdictions that would be impacted as opposed to multinationals that only operate in jurisdictions that are not participating in pillar two. Those countries that have agreed to participate in pillar two are hardly likely to raise a challenge against it under trade law. And, as companies do not themselves have individual rights under trade law, as things stand, the potential for a trade challenge to pillar two seems limited.

Where does this leave us?

This is clearly not the end of the line for trade disputes in the digital tax sphere and we may well see further challenges to any unilateral digital tax measures that are not withdrawn (or not withdrawn quickly enough) in response to international implementation of the OECD’s two-pillar approach. However, those drafting tax rules are beginning to appreciate the importance of having an eye to international trade rules when designing digital taxes. International tax and trade law have historically operated.

in an uncoordinated manner, with one world having little regard for the other, but it is clear that these two worlds are no longer operating in isolation. In the area of digital taxes, we are beginning to see greater alignment between international trade and tax law. So, trade law is likely to continue to have a role to play in this area but perhaps more as a mechanism to help keep the peace, by shaping tax law in a way that does not give rise to trade challenges, rather than being used to wage war.

This briefing was originally published in Tax Journal on 17 September 2021.