The European Commission's Digital Services Tax Proposal: a policy and legal evaluation

(1)

The European Commission’s Digital Services Tax Proposal:

a policy and legal evaluation

Laura Simmonds

Supervisor: Professor D. Weber Submitted: 27 July 2018

(2)

Table of Contents

Introduction ... 3

Chapter I The evaluation framework ... 6

Section 1 Policy and legislative context ... 6

Section 2 The specific objectives of the DST ... 8

1 - Cross jurisdictional scale without mass ... 8

2 - Reliance upon intangible assets ... 12

3 - Data and user participation ... 15

Section 3 General principles of taxation ... 17

Interim conclusion ... 18

Chapter II Assessment of the DST according to the evaluation framework ... 19

Section 1 Objective scope ... 19

Comments ... 21

Section 2 Subjective scope ... 24

Comments ... 24

Section 3 Territorial scope ... 26

Comments ... 27

Section 4 Chargeability and rate ... 28

Comments ... 28

Section 5 Reporting obligations and Administrative Cooperative ... 29

Comments ... 30

Section 6 Impact on digital businesses ... 30

Chapter III Implementing the DST in the international and EU tax law context ... 33

Section 1 Multi-sided digital platforms ... 33

Section 2 The DST and ‘Covered Taxes’ ... 36

Section 3 The DST and the principle of non-discrimination in EU law ... 39

Section 4 The DST and double tax relief in the EU context ... 42

Conclusion ... 47

(3)

Introduction

In 2015, the OECD (Organisation of Economic Cooperation and Development) concluded that it would be ‘difficult, if not impossible, to ring-fence the digital economy from the rest of the economy for tax purposes’1

At the same time, it concluded that the digital economy presented some features that pose particular challenges for taxation and which exacerbate base erosion and profit shifting (hereinafter ‘BEPS’) issues. In the wake of the OECD’s report entitled Addressing the Tax Challenges of the Digital Economy (hereinafter ‘the Final 2015 Report’) the discussion and controversy around the taxation of the digital economy have continued and initiatives have multiplied. 2

In September 2017, the European Commission published its communication on a Fair and Efficient Tax System in the European Union for the Digital Single Market 3 in which it argued that the current international tax rules were no longer fit to address the rapid digitalisation of the economy, resulting in digital companies not paying taxes where profits and value are generated. According to the Commission, this leads to a distortion of competition between traditional businesses and digital ones. 4

In light of these findings, the Commission called for fundamental changes to the international tax framework and urged the OECD to put forward meaningful policy options. In the absence of such proposal, the Commission indicated that it would put forward legislative proposals in spring 2018.

In March 2018, the OECD Interim Report on the Tax Challenges Arising from Digitalisation 5 (hereinafter ‘the Interim Report’) arrived with no concrete policy options because of a lack of consensus on the merits of, or the need for interim measures’. 6

The Interim Report, as this paper will explore, did however take some important steps towards better understanding the challenges of taxing the digital economy.

1

OECD (2015), Addressing the Tax Challenges of the Digital Economy, Action 1 - 2015 Final Report, OECD/G20 Base Erosion and Profit Shifting Project, OECD Publishing, Paris. http://dx.doi.org/10.1787/9789264241046-en

2

Inter alia, The Italian Google tax and the Indian equalization levy.

3

Communication from the Commission to the European Parliament and the Council, Fair and Efficient Tax System in the European Union for the Digital Single Market, COM(2017) 547 final, Brussels, 21.9.2017

4_{Commission, supra note 3, page 6.}

5

OECD (2018), Tax Challenges Arising from Digitalisation – Interim Report 2018: Inclusive Framework on BEPS, OECD/G20 Base Erosion and Profit Shifting Project, OECD

Publishing, Paris. http://dx.doi.org/10.1787/9789264293083-en

6

(4)

In reaction, on 21 March 2018, the Commission published two proposals for taxing the digital economy in European Union (hereinafter the ‘EU’). The focus of this paper is the interim solution for taxing certain digital services (the Digital Services Tax, hereinafter the ‘DST’). 7

While the author of this paper is of course aware of other similar initiatives such as the Italian Google tax and the Indian equalisation levy, these will not be discussed. This is because these taxes differ in some important aspects from the DST and comparisons are of limited usefulness in understanding the workings of the DST. The second proposal introducing a permanent establishment based on significant digital presence 8 is also outside the scope of this paper because that proposal is intended to be a long term solution which could be introduced into the wider corporate income tax system of Member States. As we will see, the DST does not fit into the corporate income tax structure.

This paper evaluates the Digital Services Tax from policy and legal perspectives and asks whether the DST meets its policy objectives in a manner which is compatible with international tax law and EU tax law principles. Ultimately, the objective of this paper is to determine whether the adoption of the DST is advisable from a policy perspective and from international and EU tax law perspectives.

Many Member States of the EU, including Ireland, Luxembourg, the Netherlands 9 and the Nordics states 10 have already expressed their strong opposition to the DST and their preference for a global solution led by the OECD. In the face of such strong and early opposition, the DST may never reach the statute book. However, an in-depth analysis and evaluation may still be useful in order to understand the reasons for such a failure and to inform future, perhaps more successful, proposals.

Any evaluation requires a set of criteria. To this effect, Chapter I first considers the policy and legislative context of the DST by considering the findings of the BEPS Project and the EU’s response to the former’s recommendations. Secondly, the specific objectives of the DST will be considered. This chapter limits itself to considering the difficulties from a direct tax perspective although extensive work has been done in the field of indirect taxation. This is because the DST is primarily

7

Proposal for a Council Directive on the common system of a digital services tax on revenues resulting from the provisions of certain digital services, COM(2018) 148 final, Brussels 21.3.2018.

8

Proposal for a Council Directive laying down rules relating to the taxation of a significant digital presence, COM (2017) 147 final.

9

Smyth, P.; Netherlands and Luxembourg join Ireland as wary of digital tax, The Irish Time [online]. Available at: https://www.irishtimes.com/news/world/europe/netherlands-and-luxembourg-join-ireland-as-wary-of-digital-tax-1.3437129, viewed on 7 June 2018.

10

Reuters, Nordic countries oppose EU plans for digital tax on firms’ turnover, Thompson Reuters [online]. Available at:

https://uk.reuters.com/article/uk-eu-digital-tax/nordic-countries-oppose-eu-plans-for-digital-tax-on-firms-turnover-idUKKCN1IW333, accessed on 7 June 2018.

(5)

designed as a proxy for corporate income taxation, as this paper will demonstrate. Finally, the chapter puts forward some general principles which should be adhered to when designing taxes.

Chapter II of this paper sets out the objective, subjective and territorial scopes of application of the DST in detail and examines whether these meet the objectives and principles set out previously in Chapter I. This will also be an opportunity to point out some of the design issues of the DST.

Finally, Chapter III will concentrate on an evaluation of the DST in light of international tax law and EU tax law principles. Only a selected set of issues will be explored, namely the taxation of multisided digital interfaces, the position of the DST under Double Tax Conventions (hereinafter ‘DTC’), the discriminatory nature of the DST and question of double tax relief.

(6)

Chapter I: The evaluation framework

This first Chapter aims to set the scene for an evaluation of the Commission’s proposal for a Digital Service Tax. To do so, Section 1 will outline the policy and legislative context in which the DST must be considered, and the work of the OECD and the EU to combat BEPS. In a second step, Section 2 will focus on the specific objectives of the Commission’s proposal, with illustrations drawn from the taxation of digital economy in the EU context. Finally, Section 3, identifies some general principles against which the DST may be evaluated.

Section 1 Policy and legislative context

Before exploring the objectives of the DST, it is necessary to consider the broader context in which the proposal is made. This context includes the work on BEPS undertaken by the OECD and the EU’s response to and endorsement of the BEPS Reports recommendations. This work is relevant to the DST’s context because the first difficulty identified by the OECD in its Final 2015 Report was the fact that the digital economy exacerbates many BEPS strategies. These strategies were described as follows:

Minimising the income allocable to entities located in market jurisdictions

To achieve this, local activities of an entity or subsidiary may be structured in such a way as to generate little taxable profits. This is achieved by contractually allocating functions, assets and risks in such a manner as to minimise taxation in the market jurisdiction for example by using a PE or subsidiary to perform marketing or technical support. 11

Maximising deductions in market jurisdictions

A common BEPS practice is to ‘maximise the use of deductions for payments made to other group companies in the form of interest, royalties, service fees, etc…’ 12 As we will see below, this is particularly true in the digital economy for royalty payments for the use of intangibles and intellectual property rights.

Avoiding withholding taxes

The Final 2015 Report identified the use of shell companies and conduit companies by digital economy companies to avoid withholding taxes.13 However, in the context of the EU, withholding taxes have largely been abolished and will not be central to this paper’s discussion.

11

OECD, supra note 1, paragraphs 186 to 188.

12

OECD, supra note 1 paragraph 189.

13

(7)

Use of preferential regimes

The Final 2015 Report identifies the possibility, for the digital economy, of assigning rights in intangibles and their related returns to a particular jurisdiction in order to benefit from a preferential IP regime (for example: patent box regimes). For the OECD, ‘heavy reliance in the digital economy on intangibles as source of value may exacerbate the ability to concentrate value-creating intangibles in that way’. 14

Use of hybrid mismatches

Finally, the Final 2015 Report identified, but did not expand upon, the possibility for digital economy businesses to use hybrid entities and payments in order to generate deductible payment either in a market jurisdiction or intermediate jurisdiction that are not taxable in the country of the recipient. 15

At the time of the Final 2015 Report, the OECD anticipated that wide adoption of the BEPS recommendations would eliminate many of these BEPS strategies associated with the digital economy. In fact, the Final 2015 Report suggested that all the recommendations of the BEPS project would address, to varying degrees, the difficulties of taxing the digital economy. 16

The EU implemented many of the BEPS recommendations in Anti-Tax Avoidance Directives (hereinafter ‘ATAD 1’17

and ‘ATAD 2’18) which introduced an interest limitation rule based on an EBIDTA ratio, 19 a general anti-abuse rule20, controlled foreign companies rule 21, and hybrid mismatch rules. 22 In addition, many Member States have adopted BEPS recommendations in their domestic laws or in their DTCs through the Multilateral Instrument (MLI). However, the EU has not addressed other relevant issues such as changes to the definition of PE in DTCs or changes to transfer pricing rules.

Despite the flurry of reform and legislation, dissatisfaction continues to surround the taxation of the digital economy and it was acknowledged that further worked was

14

OECD, supra note 1, paragraph 193.

15

16

OECD, supra note 1, paragraphs 369 and 370.

17

Council Directive (EU) 2016/1164 of 12 July 2016 laying down rules against tax avoidance practices that directly affect the functioning of the internal market, OJ L 193, 19.7.2016, p. 1– 14, (hereinafter ‘ATAD 1’)

18

Council Directive (EU) 2017/952 of 29 May 2017 amending Directive (EU) 2016/1164 as regards hybrid mismatches with third countries, OJ L 144, 7.6.2017, p. 1–11, (hereinafter ‘ATAD 2’)

19

ATAD 1, supra note 17, Article 4.

20

21

22

(8)

required. 23 As the OECD and the Commission explain, key features of the digital economy unrelated to BEPS strategies challenge the application of traditional corporate tax rules. The objective of the DST is therefore to address these key features.

Section 2 The specific objectives of the DST

The recitals and Explanatory Memorandum to the DST proposal are clear that the ultimate objective is to ‘ensure a fair and effective taxation’ of the digital economy. 24 According to these, the current corporate tax rules do not achieve this objective because they do not address some of the key characteristics of the digital economy. In the Commission’s view, the DST ‘addresses in an interim way the problem that the current corporate tax rules are inadequate for the digital economy.’ 25

In sum, the DST proposal is a response to a series of perceived problems with the current corporate tax rules.

In addition to the BEPS risks associated with the digital economy, the perceived problem, according to the Commission, is the corporate tax rules’ inability to handle the following characteristics of the digital economy: the ability to conduct activities remotely and with limited or no physical presence, the importance of intangible assets and the contribution of end-users to value creation. 26 The Commission’s assessment echoes the OECD’s conclusions in the Interim Report that identified the following key features of the digital economy: (1) cross jurisdictional scale without mass, the global reach of business functions and activities; (2) reliance upon intangibles assets, including intellectual property rights and (3) user participation. This far at least, there is agreement on the nature of the problems.

Each of these features poses specific problems for the current corporate tax rules. Relying on the work of the OECD and the Commission, this section will attempt to explain these difficulties while also providing concrete examples drawn from the Apple and Amazon state aid investigations and Google France’s judgment. These examples will also illustrate many of the BEPS strategies identified in Section 1. (1) Cross-jurisdictional scale without mass

The proposal refers to ‘the failure of current rules to capture the global reach of digital activities where physical presence is not a requirement anymore in order to be able to supply digital services’27

The OECD, more succinctly, describes this as

23

OECD, supra note 1, page 384.

24

Commission, supra note 7.

25

Explanatory Memorandum, Proposal for a Council Directive on the common system of a digital services tax on revenues resulting from the supply of certain digital services, COM(2018) 148 final, Brussels 21.3.2018.

26

ibidem.

27

(9)

jurisdictional scale without mass’ or the capacity of businesses to access a jurisdiction’s market without or with minimum physical presence in that jurisdiction. Scale without mass is made possible by the advent of remote information technology and in particular of the internet. This is twofold: first, interaction with the market has become dematerialised in the sense that the sale of goods or services may now occur online, contracts are concluded at a distance while delivery may occur digitally or goods may be delivered by a third party. Secondly, the organisation and structure of these businesses is strongly centralised. As a result of increased means of communication, these companies may centralise their strategic and management functions in a jurisdiction (perhaps a low tax jurisdiction) and maintain more limited functions in other jurisdictions from which they nevertheless produce revenue by interacting remotely.

From a taxation standpoint, cross-border scale without mass challenges the central theories of residence and source and their manifestations in concepts such as permanent establishment and profit attribution according to the arm’s length principle. 28

In particular, ‘scale without mass had led to an increasing share of the profits from cross border activities not being taxed in the market jurisdiction’ 29

because activities can be conducted remotely and centralised in a single entity located abroad. At the same time, entities in market jurisdictions may escape permanent establishment status or may, in effect, be allocated very little taxable profits.

Illustration 1.1: Amazon

Amazon 30 provides a good example in the EU context. Amazon’s EU operations were structured as follows: Lux SCS (a hybrid entity) was the sole shareholder of LuxOpCo. LuxOpCO functioned as the ‘headquarters of the Amazon group in Europe and the principal operator of Amazon’s European online retail and service businesses as carried out through the EU websites’. 31

LuxOpCo made all the strategic, pricing, marketing and promotional decisions for Amazon’s Europe operations. In addition, inventory was held and sold by LuxOpCo which assumed all the risks in that regard. Downstream from LuxOpCO operated local EU subsidiaries. These subsidiaries were given low value functions such as marketing or customer service. They assumed no risk for the goods or services sold on Amazon’s local websites and did not book the associated revenues.

Amazon’s structure illustrates well the digital economy’s capacity to access a market with limited physical presence: purchases made online on amazon.co.uk were handled remotely by LuxOpCo rather than by the local subsidiary. At the same time,

28

29

30

‘Amazon’ refers to Amazon.com, Inc. and all companies controlled by Amazon.com, Inc.

31

Commission decision of 4.10.2017 on state aid SA.38944 (ex 2014/NN) implemented by Luxembourg to Amazon, C(2017) 6740 final, Brussels 4.10.2017, paragraph 108.

(10)

LuxOpCO was able to make all the strategic and management decisions for the whole Amazon EU operations through technology, remote software editing, remote monitoring, remote management of Amazon’s websites, email, etc… Finally, LuxOpCo managed ‘all aspects of the order fulfilment business’. 32

Finally, Amazon’s structure illustrates a more common BEPS problem identified by the BEPS Action Plan 33 and in Section 1, namely the use of hybrid entities. 34 Indeed, LuxSCS was a Luxembourg limited partnership with two non-resident partners (US residents) treated as transparent for tax purposes in Luxembourg. While the state aid investigation makes no mention of it, LuxSCS was most likely treated as opaque by the US tax rules as a result of a ‘check the box’ election. Therefore, any income received by LuxSCS was not taxed in Luxembourg and achieved long-term deferral of tax in the US.

Illustration 1.2: Apple

The Commission’s investigation into Apple’s tax structure also provides interesting information about how digital economy businesses can achieve scale without mass in the EU. 35 Apple designs, manufactures and markets, inter alia, smartphones, personal computers, related software, third-party digital content and applications. These sales occur through Apple’s retail stores, online stores, direct sales and third-party resellers.36

Apple’s activities in Europe were structured around a number of Irish companies and branches. For instance, ASI, a non-resident Irish incorporated company, was the seller for the European region but it did not declare permanent establishments in European jurisdictions where the goods are sold. 37 ASI’s sales were concluded through authorised representatives in market jurisdictions. As a result, much like Amazon, Apple was able to avoid creating a taxable presence in market jurisdiction in Europe by using authorised representatives and a system of ‘authorised signatures’. This is facilitated by the use of the Internet, automated transmission of contracts and orders etc…

32

ibid, paragraph 108: ‘Fulfillment services refer to the process initiated by Amazon when an order for a product is received including warehousing, finding the item ordered, packaging it, and dispatching it directly or through third parties.

33

OECD (2013), Action Plan on Base Erosion and Profit Shifting, OECD Publishing, Paris, https://doi.org/10.1787/9789264202719-en.

34

For the purposes of this discussion, I will not discuss all the aspects of Amazon’s tax structure but limit myself to the use of a hybrid.

35

Opening decision State aid SA.38373 (2014/C) (ex 2014/NN) (ex 2014/CP) – Ireland alleged aid to Apple, C(2014) 3606 final, Brussels, 11.06.2014 and Commission Decision of 30.8.2016 on State aid SA.38373 (2014/C) (ex 2014/NN) (ex 2014/CP) implemented by Ireland to Apple, Brussels, 30.8.2016 C(2016) 5605 final.

36

Commission Decision of 30.8.2016 on State aid SA.38373 (2014/C) (ex 2014/NN) (ex 2014/CP) implemented by Ireland to Apple, Brussels, 30.8.2016 C(2016) 5605 final paragraphs 40 to 42.

37

(11)

While the Commission’s state aid decision centers around the arm’s length nature of the remuneration of ASI’s branch (and another Apple company’s branch, AOE) it is interesting to note that the revenue from European sales was booked by ASI which was not resident in Ireland for tax purposes nor, according to the decision, anywhere else.38Its immediate parent company, Apple Operations International is also tax resident no-where. 39 It is reasonable to conclude that those revenues escaped taxation in the EU altogether.

Illustration 1.3: Google

Google’s business and tax structure also illustrates scale without mass. Google generates revenue in the EU by providing online advertising in several ways and well as a variety of goods and services. In one case 40, advertisers will remunerate Google to display ads, according to key words and cookies, to web users on its ubiquitous search engine (Google Search). 41 Alternatively, website operators may sign up to its Adsense program whereby Google places advertisers’ ads on partner websites. The website operators are remunerated in exchange of making their website available and Google takes a share of that revenue. In this case, Google acts as an intermediary between advertisers and website operators.

The Paris administrative tribunal’s judgment 42

reveals some interesting information about how Google operates in the EU. The French tax authorities alleged that Google’s local entity in France (SARL Google France) was a dependent agent of Google Ireland Limited, leading to the presence of a permanent establishment in France that could be subject to source taxation.

Pursuant to a Marketing and Services Agreement (hereinafter ‘MSA’), Google Ireland Limited remunerated SARL Google France on a cost-plus basis in consideration for all services, consulting, recommendation and assistance required by Google Ireland Limited. These services included marketing services, market and strategy analysis, and providing demos of Google’s services to potential clients in France. The MSA also made clear that SARL Google France did not have the authority to conclude contracts in the name of Google Ireland Limited. Both Google Ireland Limited and SARL Google France are subsidiaries of Google Inc., an US resident entity.

The tax authorities argued that, in fact, Google France’s employees actively engaged in client recruitment, development of sales and extensive contract negotiations with

38

ibid, paragraph 54.

39

This outcome is the result of a mismatch between Irish and US legislation that will not be discussed here, as it does not pertain specifically to any aspect of the digital economy.

40

Adword.

41

Cellan-Jones, R.; How does Google make money, I-Wonder [online]. Available at:

http://www.bbc.co.uk/guides/z9x6bk7, viewed: 15th April 2018.

42

(12)

clients. In effect, the tax authorities argued, Google France did have the power to conclude contracts in the name of Google Ireland Limited and was not a true ‘commissionaire’. However, the tribunal took into consideration that the general contractual terms were set by Google Ireland limited, Google Ireland Limited always countersigned the contracts with clients albeit electronically 43 and that prices were not set by Google France because they were determined according to an automated online biding system. As such, SARL Google France did not constitute a permanent establishment of Google Ireland Limited and therefore escaped tax liability in France on those activities. 44

Google’s structure also illustrates how digital economy companies can avoid taxation in market jurisdictions by contractually assigning limited functions to their local entities, notably through commissionaire arrangements. In addition, remote means of communication and the possibility to conclude contracts at a distance facilitate the circumvention of dependent agent PE status of those local entities. In addition, the use of automated online systems (such as the Google’s price fixing auction) contributes to minimising local entity functions. As a result, Google is able to obtain revenue from advertising activities in market jurisdictions such as France without a significant presence there.

As we have seen in the cases of Amazon, Apple and Google new forms of technology permitting remote communications and automated processes create opportunities for businesses to achieve scale in market jurisdictions without creating a taxable presence.

(2) Reliance upon intangible assets

The second common characteristic of digital companies is a heavy reliance on intangible assets and in particular intellectual property rights. Both the Commission in the Explanatory Memorandum and the OECD in its Final Report identified this. 45 The OECD concluded that intangibles were a core value driver for the digital economy and identified software as of particular relevance.

From a corporate tax perspective, this results in the shifting of profits through the transfer of intangible assets. Indeed, the report identified the easy manipulation of the ownership of intangibles for tax purposes, thanks to their mobility (intellectual property rights are assignable and transferred by contract). The Interim Report noted that legal ownership could be vested in an entity that was incapable, in practice, of developing, enhancing, maintaining or protecting the intangible in question. This assignment of ownership results in holding companies (often in a low-tax jurisdiction

43

In French in the judgment: ‘contreseing électronique’.

44

All these elements are taken from the judgment of the administrative tribunal, recitals 8 to 13 in particular.

(13)

or benefiting from a preferential regime) receiving royalty payments while taxable profits are minimised by royalty payment deductions in market jurisdictions. In other words, royalty payments can be used to erode taxable profits in market jurisdictions while escaping taxation in the parent or intermediate jurisdiction through the use of hybrids or preferential regimes.

Illustration 2.1: Amazon

Amazon, once again, provides a good illustration since a reseller such as Amazon needs patents and technology to operate its business. These technologies, supported by trademarks and copyrights, will be protected by intellectual property rights. 46 Intangibles are an important value-driver for Amazon. This affects Amazon’s tax structure in the following manner:

In 2005 LuxSCS entered into a Buy-in Agreement and Cost-Sharing-Agreement with two US based companies pursuant to which LuxSCS obtained the right to exploit and licence IP rights related to the use of operating the European websites. It also acquired the trademarks and IP rights to the European websites from a local EU company. Subsequently, LuxSCS entered into a Licence Agreement with LuxOpCo pursuant to which LuxOpCo irrevocably obtained the exclusive right to develop, enhance, and exploit the intangibles for operating the EU website in return for a royalty payment.47 This royalty was calculated according to LuxOpCo’s operating profit and was deductible from LuxOpCo’s tax base.

LuxSCS, while the legal owner of the intangibles, was in practice incapable of performing any functions in their regard, as it possessed no staff or premises. Finally, as a hybrid LuxSCS’s royalty income escaped taxation in Luxembourg and achieved long term deferral in the US.

Illustration 2.2: Apple

Apple’s structure also proves this point. Apple Inc (the US parent of ASI and AOE) entered into a cost-sharing agreement (CSA) with ASI and AOE. Under that agreement, Apple Inc, ASI and AOE agreed to participate in the R&D efforts and to share the costs and rights relating to the development of intangibles and intellectual property rights for the Apple group. In addition, ASI and AOE obtained the beneficial ownership, for territory outside the Americas, of the intangible property developed as a result of the R&D conducted under the CSA. However, the Commission’s decision points out that ASI and AOE had no employees of their own, no role in the management of Apple’s IP and that employees of ASI and AOE’s branches only participated in intangible development activities to the extent that they related to routine localisation and product testing.48

46

47

Much of the state aid investigation and decision focuses on the valuation of that royalty payment and whether the royalty price was truly at arm’s length.

48

(14)

It appears that ASI and AOE, despite the terms of the CSA, did not in fact have the capacity in terms of people or assets to perform any functions in relation to the intangibles used by Apple and did not, in fact, perform any of these functions.

While the CSA has limited impact for taxation in Europe, it was ‘a key component to Apple’s ability to lower its US taxes’. 49

The CSA, by allocating beneficial ownership and costs according to level of product sales made in the Americas (45%) or offshore (55% to AOE and ASI), 50 allows Apple to divert a significant portion of the profits generated by the intangibles created by Apple employees in the US. As a result, the allocation of profits generated from the intangibles does not reflect the fact that Apple Inc conducts 95% of the R&D in relation to those intangibles while ASI and AOE do not have the capacity to contribute to the development of intangibles. 51

To conclude, Apple’s business model relies heavily on intangibles and its tax structure makes use of those intangibles and related cost-sharing agreement in order to shift its profits from the US to two non-resident Irish corporation that escape taxation altogether in the EU.

Illustration 2.3: Google

Google’s tax structure in the EU also exemplified how reliance on intangibles can be used for tax purposes in combination with traditional BEPS strategies. It employed a structure commonly known as the ‘Double Irish Dutch sandwich’ as follows:

Google Ireland Holding (GIH) entered into a cost sharing agreement with Google’s US based entity pursuant to which GIH obtained the right to exploit Google’s intellectual property (IP) rights for the EMEA area (Europe, Middle East and Africa). Then, as described by the French judgment, GIH licenced the IP rights to the Google Netherlands Holding that then sub-licenced them to Google Ireland Limited. 52 Pursuant to those licencing agreement, according to the French administrative court, Google Ireland Limited (an Irish resident entity) made royalty payments to Google Netherlands Holding (a Dutch resident entity) which in turn transferred most of these payments to Google Ireland Holding, a Bermuda resident entity. 53 Since Ireland and Bermuda do not have a tax treaty, the sole purpose, a priori, of the Dutch resident entity is to benefit from the 0% withholding tax on outbound royalty payments in the Netherlands. In addition, the royalty payments from GIL to the Dutch conduit

49

Levin, Carl; McCain, John (May 2013), Memorandum: Offshore profit shifting and the U.S. tax code - Part 2 (Apple Inc.) (memorandum of the Permanent Subcommittee on Investigations) page 28.

50

Commission, supra note 36, paragraph 121.

51

Levin and McCain, supra note 49.

52

Van den Hurk, H.T.P.M, ‘Starbucks versus the People’, Bulletin of International Taxation, vol. 68, no.1 [online]. Available at: https://online.ibfd.org/document/bit_2014_01_int_3

53

(15)

company benefit from the Interest and Royalty Directive 54 prohibition on withholding taxes. 55

This structure achieves many of the BEPS strategies described above in Section 1 and is made possible by Google’s strong reliance on IP to add value to its business. In particular, the cost-sharing agreement allows Google US entities to shift profits out of the US, despite the fact that object of the agreement (the IP rights) have been developed in the US and the fact that GIH possesses no assets or persons capable of contributing to the development of the IP.

Secondly, the IP licence payments minimise profits in market and intermediary jurisdictions by generating deductions from the tax base while the use of a Dutch conduit entity successfully makes use of the abolition of withholding taxes on royalties in the EU 56 and of the absence of withholding taxes on royalties in the Netherlands. Finally, there is no pick-up of GIH’s royalty income in the EU through the exploitation of Ireland’s tax residency rules. 57

As demonstrated, the digital economy’s heavy reliance on intangible provides it with many opportunities minimise its profits in market and intermediary jurisdictions through the use of licence agreements and royalty payments.

(3) Data and user participation

The Explanatory Memorandum and the recitals identify ‘the contribution of end-users in value creation’ as a key difficulty for the corporate tax rules. The Commission draws particular attention to the misalignment between value creation and profit location in the case of businesses models heavily reliant on user participation. 58 The DST’s recitals are particularly focussed on user generated content and data. Indeed, the DST targets services ‘where users contribute significantly to the process of value creation’ 59

or ‘largely reliant on user data’. 60 It is submitted that addressing

54

Council Directive 2003/49/EC of 3 June 2003 on a common system of taxation applicable to interest and royalty payments made between associated companies of different Member States, OJ L 157, 26.6.2003, p. 49–54.

55

Van den Hurk, supra note 52.

56

Supra note 54.

57

GIH can be described as a hybrid entity, a common BEPS strategy. GIH is incorporated in Ireland but managed in Bermuda. As a result of Irish domestic law, it is not subject to tax in Ireland. Furthermore, under US Treasury Regulation 301.7701-2, GIH was declared as transparent for US tax purposes and was attributed the activities of GIL and Dutch Holding. As result, GIH’s activities are not limited to received passive income and it escapes US CFC legislation. The result is no taxation in Ireland and long term deferral in the US.

58

Commission, supra note 25, recital 2.

59

60

(16)

the failure of the corporate tax rules to capture the value of users is the primary objective of the DST proposal.

In its 2015 Final Report, the OECD had already identified reliance on data and user participation as a key feature of the digital economy. 61 In 2018, it refined its analysis with the following findings:

Data is increasingly being used by businesses to improve products and services. However, the degree of importance of data and user participation may vary across different business models and companies. 62

The OECD identifies certain business lines that particularly benefit from data and user participation. First, ‘datasets allow digitalised businesses to better target online advertisements to specific user groups’ 63

and second, social networks for which user-participation is ‘central feature of the business’. 64

Indeed, some social media platforms such as Facebook, Youtube, Instagram or Twitter largely depend on users generating content for them in the form of video, blogs, discussion form posts, digital images, audio files, and other forms of media that are created by customers or end-users.’65

As regards the value, for taxation purposes, of data and user participation, the OECD makes clear there is no consensus on this issue and further research and thinking will be needed. 66 On the one hand, some countries believe that data and users are crucial value drivers for businesses in the digital economy, particularly for online advertising and social media platforms. On the other hand, some countries argue that users and digital businesses enter into barter transactions where users provide data and content in exchange for providing the platform and its services for free. These transactions could, in principle, be subject to income tax or could be viewed as inputs in the same way as electricity or broadband access. Finally, some countries do not attribute any value to data. 67

The Commission has clearly come down on one side of the debate. In its paper ‘Fair and Efficient Taxation System in the European Union for the Single Digital Market’ 68

the Commission noted that the generating of value by intangibles, data and knowledge presented a main policy challenge. The Commission, in its Memorandum

61

62

63 ibid, paragraph 140. 64 ibid, paragraph 144. 65 ibid, paragraph 151. 66 ibid, paragraph 161. 67 ibid, paragraphs 158 to 161. 68

(17)

to the DST takes the view that ‘user generated content and data collection have become core activities for the value creation of digital business’. 69

As regards the impact of data on taxation, the Commission argues that the current corporate tax rules result in a misalignment of value creation and taxes because ‘permanent establishment and transfer pricing rules […] do not take into account the user contribution in the allocation of taxation profits, which results in a mismatch with the value creation’. 70

However, the corporate tax implications of data and user participation are not immediately apparent in the tax structures discussed above and do not seem to play any direct role in the tax planning of digital enterprises. However, based on the OECD’s findings, the following can be noted: Google, as provider of online advertising collects and analyses user data in order to improve the advertising services it sells advertisers. 71 Amazon can use user browsing data, user content such as reviews and ratings to ‘stimulate sales, customise services and improve customer targeting’. 72

To conclude, the specific objectives of the DST are to address what prior reforms have failed to do effectively by tackling the key features of the digital economy which pose a challenge to the traditional corporate income tax rules. These are: achieving mass without scale, heavy reliance on intangible and data are all considerations of the DST. In addition, it appears that addressing the value of data and user-generated content is a particularly important objective for the DST.

Section 3 – General principles of taxation

The above sections identified the context of the DST and its specific objectives. This section asks what principles must governments consider when introducing a tax or reforming a tax system? Following centuries of discussion, doctrine has settled on a number of accepted criteria which may take different expressions. For the purposes of this discussion, the principles of equity, neutrality and administrative efficiency, as described by the 1987 Meade Committee, 73 will be discussed. It is submitted that adhering to these principles should be a general objective of the DST proposal.

69

Commission, supra note 25, page 1.

70

Commission Communication, Impact Assessment accompanying the Proposal for a Council Directive on the common system of a digital services tax on revenues resulting from the provision of certain digital services, SWD (2018) 18/final2, Brussels 21.3.2018, page 16.

71

Kofler, G.W; Mayr, G & Schlager, C.; ‘Taxation of the Digital Economy: A Pragmatic Approach to Short-Term Measures’, European Taxation, vol.58, n.4 [online]. Available at:

https://online.ibfd.org/document/et_2018_04_int_1

72

Commission, supra note 5, paragraph 154.

73

Institute of Fiscal Studies, ‘The structure and reform of direct taxation’, Report of a Committee chaired by J.E Meade, 1987.

(18)

First, the term equity can encompass the principles of fairness and equal treatment. Equal treatment demands that equal circumstances be treated equally 74 while fairness is often equated with the taxpayer’s ability to pay. According to horizontal equity, two taxpayers with the same ability should bear the same burden and vertical equity requires the taxpayer with greater ability to bear a higher burden. 75

A further principle is the economic principle of neutrality according to which a tax should not ‘interfere with decision making’. Tax law drafters should strive to ensure neutrality so that ‘rational business and commercial decisions can be made without the influence of the tax consequences flowing from them’. 76 In order words, tax should not favour one type of investment or activity over another.

However, in some cases, the tax law drafters specifically want to discourage certain activities which they deem harmful for society as a whole or which produce negative externalities which need to be priced. This is for example the case of excise duties on tobacco, alcohol and petroleum products. In those cases, a biased result is the objective.

Finally, the principle of administrative efficiency encompasses the requirement of simplicity, ease of understanding and absence of excessive administrative costs for the tax authority and taxpayer. In particular it must be clear what is taxable and what is not as well as the amount of tax which should be paid on each taxable object. 77 Interim Conclusion

To conclude, this section explained and illustrated the challenges the digital economy poses for the current corporate tax rules in the EU. These challenges are digital economy businesses’ capacity to achieve scale without mass, heavy reliance on intangibles and on user-generated content and data. Failure to address these key features of the digital economy leads to, according to the Commission, to digital business models facing a lower effective average tax burden than traditional businesses 78 creating unfairness and a distortion of competition. In this context, the DST’s objective is to correct these deficiencies by addressing these features. Doing so, it is argued, must also take account of the general principles of taxation identified in Section 3.

74

Thuronyi, V., Tax Law design and Drafting, Volume 1, International Monetary Fund, 1996, page 5.

75

Miller, A; Oats, L.; Principles of international taxation, Fifth Edition, Bloomsburry Professional Ltd, 2016, page 6.

76

ibid.

77

Meade Report, supra note 73, page 19.

(19)

Chapter II Assessment of the DST according to the evaluation framework

In this second chapter, the Digital Services Tax will be explained in detail and measured against the objectives and principles identified in Chapter I. The impact of the DST on the digital economy businesses discussed in Chapter I will be briefly discussed.

First, the DST has two remarkable design features: it is based on revenue rather than profits and, secondly, it allocates taxable revenue according to an allocation key based on location of users. As such it may be described as a destination-based revenue tax on certain digital services. These elements will be discussed in more detail below. Section 1: Objective scope

The DST targets the revenue of certain digital services. Article 3 provides that revenues resulting from the provision of certain services shall qualify as ‘taxable revenues’ for the purposes of the DST. These services can be broadly categorised as online advertising, multi-sided digital interfaces and the transmission of data.

According to Article 3, paragraph (1)(a), online advertising services are ‘the placing on a digital interface of advertising targeted at users of that interface’. Article 2 defines ‘digital interface’ as ‘any software, including a website or part thereof and applications, including mobile applications, accessible by users’ 79

and ‘user’ means any individual or business. 80 Article 3(3) provides that it is irrelevant whether the digital interface is the entity responsible for placing the advertisement on the interface and that in the event the owner and the entity responsible are not the same, the owner of the interface will not be liable for the DST in ‘order to avoid possible cascading effects and double taxation’. 81

The second category of targeted services is multi-sided digital interfaces. Article 3, paragraph 1(b) defines this service as ‘the making available to users of a multi-sided digital interface which allows users to find other users and to interact with them, and which may also facilitate the provision of underlying supplies of goods and services directly between users. 82 This definition is intended to catch two types of services: social media digital platforms and intermediation services as will be explained below. The key element in both these services, according to the Commission, is the multi-sided digital interface’s ‘capacity to enable users to find other users and interact with

79

Commission, supra note 7, Article 2, paragraph 5.

80

Commission, supra note 7, Article 2, paragraph 4.

81

Commission, supra note 7, Recital 11.

82

(20)

them.’ 83

On this basis, the Commission considers that suppliers of communication (instant messaging services, e-mail services) or financial services (e-payment services) should be excluded from the scope of the Article 3(1)(b) on the grounds that these usually require users to have previously communicated by other means. 84 The proposal defines the first type of service on a multi-sided digital interface as ‘a service consisting in the making available of a multi-sided digital interface through which users can upload and share digital content with other users.’ 85

In other words, the tax targets the providers of services more commonly known as ‘social media platforms’ whereby the content is created by users either in the form of pictures, videos or text. Examples of these service providers are Instragram, Snapchat, Facebook or Twitter.

However, the proposal distinguishes this first type of multi-sided digital interface from digital content interfaces which are excluded from the scope of the DST on the grounds that the value of these services lies in the digital content itself rather than in the user-participation in the interface. 86 Digital content is defined in Article 2(5) as ‘data supplied in digital form, such as computer programmes, applications, music, videos, texts, games and any other software, other than the data represented by a digital interface’. From this definition, it appears that providers of video and music streaming services such as Netflix or Spotify are not covered by the DST. This Article also excludes ‘communication services’ from the scope of the DST. 87

The second limb of the definition refers to ‘intermediation services’ as multi-sided digital interfaces that facilitate the provision of underlying supplies of goods or services directly between users.

First, the DST does not target revenue generated by users of these platforms. Recital 13 makes clear that the revenues generated by users supplying and purchasing goods or services via digital interfaces are not taxable amounts under the DST. It is the revenues generated by the digital platform itself that are targeted by the DST proposal. In addition, the sale of goods or services by the provider of the interface, for his own account, is outside the scope of Article 3(1)(b).

Certain intermediation services are excluded from the scope of the DST: the supply of a trading venue or systematic supplies by a trading venue or a systematic internaliser of any services as referred to in points (1) to (9) of Section A of Annex I to Directive

83

84_{Commission, supra note 7, recital 12.} 85

86

87

(21)

2014/65/EU 88 and the supply by a regulated crowdfunding service provider of any of the services referred to in points (1) to (9) of Section A of Annex I to Directive 2014/65/EU. 89

Finally, the DST targets services consisting of the transmission of data collected about users and generated from users’ activities on digital interfaces. 90

This service is construed narrowly as only covering data generated from users’ activities in digital interfaces which is then monetised. The tax is not intended to cover the simple transmission of data, the use of data for internal purposes of a business or shared for free. 91

1.1 Comments on the objective scope

First, it is evident that the definition of advertising services will affect a wide array of websites and applications, beyond the advertising appearing in search engines such as Google Search and other search engines. Indeed, this definition of services includes any website or application that displays advertising such as newspapers or blogs. If these are members of Google’s Adword network for instance, Google is considered to have provided the service and is liable for the tax.

The wording of the DST may run counter to the principle of neutrality as explained in Chapter 1. First, the DST may impact the price of such services, as it is passed on to final consumers. For instance, depending on the elasticity of the demand for Google’s advertising services (and we may assume it is inelastic because of Google’s dominant position on the market), 92 Google can pass on the incidence of the DST to its advertising customers leading to an increase in the cost of online advertising in the EU.

Second, the distinction between the owner of the interface and the advertising provider may create unexpected market distortions. For example, if the owner of an interface is currently managing his own online advertising, for instance an online newspaper or magazine, the DST may encourage him to outsource his advertising to a third party such as Google who, according to the recitals, will bear the incidence of

88

Directive 2014/65/EU of the European Parliament and of the Council of 15 May 2014 on markets in financial instruments and amending Directive 2002/92/EC and Directive

2011/61/EU (recast), OJ L 173, 12.6.2014, p. 349–496.

89

Commission, supra note 7, Article 3(4) subparagraph (b) and (c).

90

Commission, supra note 7, Article 3, paragraph 1(c).

91

Commission, supra note 7, Recital 17.

92

WARC, an advertising research agency, estimated Google’s market share of online advertising at 44%. Fisher, R.; 25 per cent of global ad spent goes to Google or Facebook, The Statista Portal, 7 December 2017 [online] Available at:

https://www.statista.com/chart/12179/google-and-facebook-share-of-ad-revenue/ viewed 4th May 2018.

(22)

the tax. In this matter, the DST may have a distortive effect on business choices for companies that rely on advertising. Unwittingly, the DST may reinforce the already dominant position of certain online advertising players.

The DST’s definition of taxable services poses difficulties for business models that could be described as ‘mixed’. Indeed, some business models may fall under two categories of services. In particular, many digital economy companies will fall under the online advertising category as well as the social media platform category. For instance, Facebook provides its social media platform for free to users but also places advertising on its Facebook interface (the same can be said of Instagram and Twitter). Both these services create a tax liability under the DST proposal and may therefore result in double taxation of the same revenue as it is counted twice towards the two services. The Impact Assessment conceded that this outcome was ‘theoretically conceivable’ but in that case ‘one service should take precedence over the other’93

. However, the proposal does not include a rule of priority or hierarchy between the different services. As such, the DST may violate principles of equity including the ability to pay by producing double economic taxation of the same revenue.

Further, some other mixed business models may combine taxed and exempt services. The DST excludes from its scope digital interface services where the sole or main purpose of making the interface available is the supply of digital content to users. This brings to mind online streaming services of videos or music such as Netflix, YouTube or Spotify and Deezer. However, these same services may also fall under the category of online advertising services. In the case of Spotify, its free offer includes the playing of advertisement between songs and in the case of YouTube the display of advertising before playing a video. This begs the question whether the Commission intended to exclude or include these services and puts in question the consistency of the proposal’s approach.

We see here that the DST produces different results based on the remuneration model, rather than based on the service provided. For example, the ‘free’ streaming of music online that is remunerated by advertising will be caught by the DST whereas the subscription model of the same service will not. The DST is not neutral according to the remuneration model, even within digital services. This also creates compliance difficulties and costs for the taxpayer who will have to segregate these revenue streams according the method of payment.

The tax does not affect all similar services in the same manner, running counter to the principle of equity and neutrality identified in Chapter I. For example, the DST will tax online newspapers displaying advertisements on their interface whereas free paper newspapers, funded by advertising in their pages, distributed on public transport will not be subject to DST. In terms of neutrality, the DST creates a tax bias ‘in the

93

(23)

treatment of old business models and new business models’. 94

This is precisely what the OECD warned against when it recommended in 2015 that the digital economy should not be ring-fenced. 95 Indeed, this specific regime on digital services drives an ‘inefficient wedge between the digital and the non-digital’ 96

version of the same service. In addition, liability under the DST for trading systems will depend on whether the service is regulated by EU law or not, rather than on the nature of the service. This purely formal distinction is somewhat arbitrary if the objective is to address the digital economy’s key features.

The transmission of data appears deceptively simple. The proposal tells us that the targeted services are those using digital interfaces as a way to create user input which they monetise but that internal use of the data is excluded. 97 In addition, the data must be generated by users’ specific activity on digital interfaces rather than data which has been generated by sensors. It is unclear what monetise means in this context. For instance, if the data is used to improve search results on a search engine and to better tailor advertising services, does this count a monetising because the service provider can presumably charge more for his advertising service or does it count as internal use? It is submitted that in its current wording, the transmission of data is too imprecisely defined and requires clarification to be workable.

Turning now to the specific objectives of addressing the key features of the digital economy, it is argued that the objective scope of the DST provides an answer to the perceived problem of value creation by users. Indeed, the DST targets services which are more reliant on user-generated content such as intermediation services, social media platforms and data transmission services. The OECD Interim report confirmed that social media platforms are the most user participation intensive while on the other hand, intermediation services rely on an underlying transaction between users and transmission services involve the aggregation of user data. 98

However, the exclusion of certain digital content services is questionable in light of the proposal’s stated goal of taxing the value generated by data and user participation. The OECD took the view that ‘intangible goods e-commerce operators [OECD specifically cites Deezer and Spotify] could offer even more interaction opportunities and thus have higher participation intensity’ because ‘[u]sers can increase the customer base by sharing their playlists’. 99

The Commission however takes the view

94

Brauner, Y. & Pistone, P., ‘Adapting Current International Taxation to New Business Models: Two Proposals for the European Union’, Bulletin for International Taxation, vol. 71, no. 12 [online]. Available: https://online.ibfd.org/document/bit_2017_12_int_1

95

OECD, supra note 1.

96

Schön, W. ‘Ten Questions about Why and How to Tax the Digitalized Economy’, Bulletin of International Taxation, vol. 72, no. 4/5 [online]. Available at:

https://online.ibfd.org/document/bit_2018_04_int_1

97

98_{OECD, supra note 5, paragraph 156.} 99

(24)

that for digital content platforms ‘it is less clear that the user plays a central role in the creation of value for the company supplying the digital content’. 100

Similarly, recital 13 clarified that digital retail stores such as Amazon are outside the scope of the DST. The OECD reaches a slightly different conclusion when it considered tangible goods operators in which ‘browsing data, reviews and ratings are employed to stimulate sales, customise services and improve customer targeting’. 101

In sum, the targeting of certain services and modes of remuneration as well as the risk of double taxation leads to questionable results for the DST under the general principles of good taxation. However, on the more specific objectives, the targeted services do really more heavily than others on data and user generated content.

Section 2: The subjective scope

The DST applies to all entities meeting the revenue criteria, regardless of their place of establishment. 102 As such, non EU resident companies will be subject to the tax as well as EU resident companies.

Article 4 paragraph 1 provides for the definitions of taxable persons under the proposal. An entity will qualify as a taxable person if it meets two conditions: (a) the total amount of worldwide revenues reported by the entity for the relevant financial year exceeds EUR 750 000 000 and (b) the total amount of taxable revenues obtained by the entity within the Union during the relevant financial year exceeds EUR 50 000 000. Article 4(6) provides that if the entity in question belongs to a consolidated group for financial accounting purposes, that paragraph shall be applied instead to the worldwide revenues reported by, and taxable revenues obtained within the Union by, the group as a whole.

An entity is defined in Article 2(1) as any legal person or legal arrangement that carries on business through either a company or a structure that is transparent for tax purposes. Pursuant to Article 2(2) a consolidated group for financial accounting purposes means all entities that are fully included in consolidated financial statements drawn up in accordance with the International Financial Reporting Standards or a national financial reporting system.

2.1 Comments on subjective scope

The relatively high thresholds are intended to exclude small enterprises and start-ups which would be disproportionately hit by the DST. 103 This reflects the intended narrow scope of the DST.

100

101

102

(25)

Regarding the EUR 750 million worldwide turnover threshold, the Commission claims that only companies that rely heavily on user participation, large user traffic and extensive user networks can achieve these level of revenues. In addition, the Commission claims that these large companies ‘are responsible for the higher difference between where profits are taxed and the value is created’ and concentrate the risks of aggressive tax planning. 104 These assertions are contestable. Indeed, the OECD BEPS’s Project made clear that the opportunities for aggressive tax planning were not confined to the digital economy. Indeed, the Commission itself has targeted, in its state aid investigations, aggressive tax planning in case of traditional ‘brick and mortar’ companies such as Ikea 105

, Fiat 106, Starbucks 107 and GDF Suez. 108

It is worth noting that the EUR 750 million threshold is inspired by the Country by Country Reporting Standard and is the same as the CCCTB (Common Consolidated Corporate Tax Base) proposal. The second threshold (taxable revenues exceeding EUR 50 million) aims to function as a proxy for significant digital presence in the EU and ensure that only entities with significant activities in the EU are caught.

These thresholds may lead to distortions and violate the principle of equity and fairness. The clearly stated goal of the thresholds is to exclude small and medium size enterprises but this also results in enterprises engaging in the same activities not being taxed in the same manner. The threshold also treats enterprises of the same size differently; for instance, a standalone digital advertising company with EUR 50 million turnover will fall outside the DST whereas an identical entity owned by a multinational with a EUR 750 million worldwide turnover will be taxed. Two similar businesses, with the identical digital presence in the EU, will bear a different tax burden.

The distinction between small and large companies ‘makes it hard to establish principle-oriented solutions’109 and indicates that preventing tax avoidance is the primary motive of the DST rather than addressing a fundamental change in the way business activities occur. This distinction may also conflict with principles of equity

103

104

105

Commission decision of 18.12.2017, State aid SA.46470 (2017/NN) State aid SA.46470 (2017/NN) – Netherlands - Possible State aid in favour of Inter IKEA, C(2017) 8753 final, Brussels, 18.12.2017.

106

Commission Decision of 21.10.2015 on state aid SA.38375 (2014/C ex 2014/NN) which Luxembourg granted to Fiat, C(2015) 7152 final, Brussels 21.10.2015.

107

Commission Decision of 21.10.2015 on state aid SA.38374 (2014/C ex 2014/NN) implemented by the Netherlands to Starbucks, C(2015) 7143 final, Brussels 21.10.2015.

108

Commission Decision of 19.9.2016, State aid SA.44888 (NN/2016) (ex EO/2016) – Luxembourg Possible State aid in favour of GDF Suez, C(2016) 5612 final, Brussels 19.9.2016.

109

(26)

and with the constitutional orders of some Member States. 110 It is submitted that the thresholds do not contribute to addressing the specific objectives of the DST as digital businesses, small and larger, all present the ability to achieve scale without mass as well as rely on intangibles and data.

Section 3: The territorial scope

One of the more novel aspects of this proposal is how taxable revenues are allocated between Member States. In essence, Article 5 of the proposal (‘Place of taxation’) puts forward a proportional allocation of taxable revenues between Member States based on the location of users. In this sense, the DST introduces a destination principle for revenues generated by these digital services.

Article 5(1) provides ‘taxable revenues shall be treated (…) as obtained in a Member States (…) if users with respect to the taxable services are located in that Member States. It is irrelevant of whether such users have contributed in money to the generation of those revenues, to ensure that all value generated by users is captured.’ The proposal then considers how to define the location of a ‘user’ and how to determine the portion of an entity’s revenue obtained by a Member State, for each type of service defined in Article 3.

First, in the case of advertising services as defined in Article 3(1)(b) a user shall be deemed to be located in a Member State if ‘the advertising in question appears on the user’s device at the time when the device is being used in that Member States in that tax period to access a digital interface’. The Member State shall then tax, pursuant to the proposal, the advertising revenue in proportion to the number of times the advertisement has appeared on the user’s devices in that tax period.

For services involving multi-sided digital interfaces offering intermediation services (where there is an underlying transaction between users), the proposal provides that the user is deemed to be located in the Member State if the user uses a device in that Member Stated to access the digital interface and concludes an underlying transaction on that interface. The Member State shall then tax the taxable revenues in proportion to the number of users having concluded underlying transactions on the digital interface during the tax period.

For other kinds of multi-sided digital interfaces caught by Article 3(1)(b), the Commission assumes that these usually require ‘a periodic payment after having registered or opened an account on the digital interface’.111

For these, the user is

110

Schön, supra note 96, pointing out in particular Article 3 paragraph 1 of the German Basic Law.

111