Following the EU’s failed attempt to adopt the European Commission’s (EC’s) proposed EU-widedigital services tax (DST) earlier this year, France recently introduced a 3% tax on revenues generated from certain digital activities exceeding €750 million worldwide and €25 million in France. Several other EU Member States, including Austria, Belgium, the Czech Republic, Italy, Spain and the UK, are considering introducing their own DSTs. Aside from their financial impact, these measures are likely to be problematic for targeted companies, since they would likely require companies to track and attribute revenues in new ways and could result in double taxation.
In addition to these concerns, national DSTs may run afoul of EU State aid rules on the grounds that they arguably target large, mainly U.S.-based digital companies and thereby confer a selective advantage on smaller, domestic companies. France and other Member States would likely argue that their DSTs are analogous to progressive tax systems that have been upheld by the European Courts, but it is unclear whether these parallels will hold. Member States would also likely argue that their DSTs are closely modelled on the EC’s own proposed directive, but national measures have the potential to distort competition among Member States in a way that an EU-wide measure would not.
Meanwhile, EC President-elect Ursula von der Leyen has put an EU DST back on the EU agenda, instructing Commissioners-designate Gentiloni and Vestager to propose an EU DST if an international approach is not agreed by the end of 2020. Unilateral Member State measures may be seen as an obstacle to agreeing an EU-wide approach.
This article briefly discusses how DSTs work, using the EU’s proposed directive as an example, summarizes the legal test for determining whether such measures may constitute State aid, anddiscusses the potential implications.
On 21 March 2018, the EC published a proposal for an EU-wide DST of 3% on gross revenues (net of value-added and similar taxes) derived in the EU from the provision of certain digital services characterized by “user value creation” (i.e., where user participation plays a central role in the creation of value for the company). The EC proposal was based on the premise that traditional corporate tax rules fail to recognise the new ways in which profits are created in the digital world, in particular the role that users play in generating value for digital companies.The EU DST was intended as a temporary measure, pending an OECD agreement on a broader international framework for taxing digital companies. Similarly, many Member States stress the “temporary” nature of their taxes.
The EU DST covered the following digital services: (i) advertising (placing on a digital interface of advertising targeted at users of that interface); (ii) intermediation services (offering digital platforms that expedite the interaction between users and the transfer of goods and services between users); and (iii) selling of user data. On the other hand, revenues from the supply of digital content/solutions and online retail activities were excluded from the scope of the proposed DST. The EU DST would only have applied to groups with total global revenues over €750 million and EU-wide taxable revenues over €50 million. Consequently, only a limited number of companies were likely to be affected by the DST (an early EC draft specifically identified examples of businesses that would generate "advertising" revenue (Facebook, Google AdWords, Twitter, Instagram and "free" Spotify) and "intermediation services" revenue (Airbnb and Uber). Many of these are U.S.-based, although policymakers reject suggestions that these taxes are deliberately targeted at U.S. businesses.
Since the EU could not reach an agreement on the EC’s DST proposal, France and certain other Member States decided to introduce their own DSTs (though closely modelled on the EC’s proposal).
As noted above, while an EU directive could not be challenged as State aid, unilateral measures by Member States could be. Since these measures are not covered by existing exemptions and have not been notified to the EC for approval under Article 108(3) of the Treaty on the Functioning of the European Union (TFEU), they would be illegal under EU law if they were determined to constitute State aid.
For a national measure to be classified as State aid, the measure must (i) confer a selective advantage on the recipient; (ii) be granted by the State or through State resources; (iii) be liable to affect trade between Member States; and (iv) distort or threaten to distort competition. When tax measures are challenged as potential State aid, the key issue is usually whether the measure confers a selective advantage.
The Commission typically assesses the selectivity of a tax measure by means of a three-step analysis: (i) identifying the “system of reference”, that is the “normal” tax system applicable in the Member State concerned; (ii) assessing whether the relevant tax measure “derogates” from that system insofar as it differentiates between undertakings in a comparable factual and legal situation; and (iii) if there is derogation, verifying whether that derogation may nevertheless be justified by the nature or the general scheme of the reference system.
In cases involving Spain in 2016 and 2018, however, the Court of Justice treated the measure in question as its own reference system, analysing whether the measure had the effect of benefitting certain undertakings and not others, although all those undertakings are in a comparable factual and legal situation in the light of the objective pursued by the tax measure. An examination of this kind was also recommended by Advocate General Kokott in her recent opinion in a Hungarian case concerning turnover-based tax for telecommunications undertakings. By contrast, in the Fiat and Starbucks judgments on the Commission’s finding that individual tax rulings constituted State aid, the General Court did not contest the application of the standard three-step analysis.
Are National DSTs State Aid?
Under the Commission’s traditional analysis, a tax measure’s selectivity often turns on the choice of the reference system. In the case of DSTs, one possibility is that the reference system is the national corporate income tax system (although DSTs are taxes on gross revenues, not profits, as would be normal for a corporate income tax regime). In that case, the DSTs’ application only to certain types of revenues would need to be justified by objective differences between the companies to which the DST applies and others, or based on the nature or overall structure of the national tax system. Under this approach, DSTs would arguably confer a selective advantage on companies that derive revenues from sources other than those caught by the DST or that do not meet the digital revenue thresholds. Under this approach, it would be hard to argue that the selectivity was justified if the digital turnover thresholds de facto exempt domestic companies.
If a DST is treated as its own reference system, a State aid challenge may be more difficult. The key question would be whether the DST creates an unjustified selective advantage for some companies, compared to others in a comparable factual and legal situation, having regard to the DST’s objective of taxing value generated for digital companies by users in the relevant Member State. In this regard, DSTs could be argued to create a selective advantage for companies that benefit from user generated value but are not covered by the DST, and/or for companies that fall below the thresholds.
Member States could try to rely on recent judgments involving progressive turnover taxes in Poland and Hungary. In those cases, the General Court held that progressive taxation above a certain threshold – even a high threshold – does not in itself imply the existence of a selective advantage. The Polish and Hungarian cases’ value as precedents with regard to DSTs may be questioned, however, because both decisions are under appeal and because the objective pursued was arguably different. The objective of the Polish and Hungarian measures was redistributive, and the General Court found that such an objective is compatible with a progressive structure, because it can be reasonably assumed that undertakings with high turnover might have proportionately lower costs and, therefore, proportionately greater disposable revenue. It is not clear that the same argument would apply to DSTs, whose main objective is to tax user-generated value. Indeed, the same global turnover threshold – €750 million – has been proposed in other EU initiatives, such as the proposed directive on a Common Consolidated Corporate Tax Base, which have nothing to do with user value creation.
Member States could seek to rely on other objectives to try to justify the turnover thresholds of their unilateral DSTs. For example, the EC cited protecting the integrity of the single market, preventing national tax base erosion, ensuring a level playing field for businesses, and fighting against aggressive tax planning. Even if those were generally acceptable justifications, however, it may be difficult for Member States to demonstrate that the criteria used in unilateral DSTs are consistent with those aims.
Unilateral DSTs like France’s appear to apply to a relatively small group of international, mainly U.S.-based, companies. Not surprisingly, such taxes have proved highly controversial. Apart from their direct cost to digital companies, they would likely be difficult and expensive to implement, since the concept of user-generated value is unclear and does not correspond to normal revenue and cost categories. DSTs may also create double-taxation concerns, if revenues are subject to DST in two different Member States and/or DST is not fully creditable against corporate income tax.
In addition to these concerns, national DSTs may be open to challenge under EU State aid rules. The EC may find itself in an awkward position if it seeks to challenge them, since national DSTs are closely modelled on the EC’s own DST proposal. On the other hand, since President-elect von der Leyen has made an international or EU-level DST a priority for the next EC mandate, the EC may object to Member States taking unilateral action on DSTs. In any event, the multinationals most likely to be affected by national DSTs may file complaints or take other action that could leave the EC little choice.
|Jay Modrall, Partner,
Norton Rose Fulbright Brussels LLP
|James R. Modrall is an antitrust and competition lawyer based in Brussels. He joined Norton Rose Fulbright LLP in September 2013 as partner, having been a resident partner in a major US law firm since 1986. A US-qualified lawyer by background, he is a member of the bar in New York, Washington, D.C. and Belgium.With 27 years of experience, he is a leading advisor for EU and international competition work, in particular the review and clearance of international mergers and acquisitions.Mr Modrall also has extensive experience with EU financial regulatory reform, advising the world’s leading private equity groups in connection with the new EU directive on alternative investment fund managers and leading banks and investment firms on EU initiatives including EU regulation of derivatives, EU reforms in financial market regulation and the creation of a new EU framework for crisis management, among others.
Mr. Modrall’s native language is English, and he is fluent in Italian and proficient in Dutch and French.
|Violetta Bourt, Senior Associate,
Norton Rose Fulbright Brussels LLP
|Violetta Bourt is an antitrust and competition lawyer based in Brussels.Violetta’s practice covers all aspects of EU competition law, with a focus on EU and international merger control. Violetta has particular experience in notifying mergers and joint ventures to the European Commission, as well as coordinating the notification of international transactions with multiple national competition authorities. Violetta also advises clients on abuse of dominance cases, anti-competitive agreements, global competition compliance issues, and other competition and general EU matters.Violetta joined the practice as an associate in 2014, after six years in a major US law firm in Brussels. She received an LLM degree from the London School of Economics in 2007, and a law degree, magna cum laude, from the University of Louvain (UCL, Belgium) in 2006.Her native language is Russian/Ukrainian, and she is fluent in English and French.