EBIT comments on the draft EU Directives on significant digital presence and a digital services tax

Brussels, 16 May 2018

To: European Commission

Sent via upload

Dear Sir / Madam,

EBIT’s Members are grateful for the opportunity to provide comments on the European Commission’s requests for feedback on its proposed Digital Tax Package, and in particular on:

  • The Commission’s Proposal for a Council Directive laying down rules relating to the corporate taxation of a significant digital presence (hereinafter referred to as the Digital PE proposal); and
  • The Commission’s Proposal for a Council Directive on the common system of a digital services tax on revenues resulting from the provision of certain digital services (interim proposal, hereinafter referred to as the DST proposal).

Whilst we acknowledge that the Commission has requested feedback individually on each of the proposed Directives, we believe that they cannot and should not be viewed in isolation of each other. EBIT’s Members have therefore prepared a consolidated response for submission to both of the Commission’s requests for feedback.

We also refer to our earlier submissions to the Commission of December 2017:

The comments provided in the present EBIT position paper are in line with these December 2017 submissions by EBIT.

General Comments

1.1. General approach

  • Tackling the tax challenges of the increasingly digitalized world is best achieved by international consensus to ensure feasible and practical outcomes and sustainable progress in modernizing the international tax system. The EU itself has underlined the importance and leading role of the OECD’s current examination of value creation and profit generation processes by the digital economy with the aim to design adequate policy responses.
  • Businesses within Europe need a sensible, internationally agreed approach for highly digitalized business models. EBIT Members strongly believe that unilateral, interim actions, at EU or national EU Member State level, will only impose a further administrative burden and cost on businesses, create less legal and tax certainty and possibly lead to double and multiple business taxation. That scenario may also foster concerns about the EU as a growth-friendly region and its attractiveness as a destination for innovation and investment opportunities. It is therefore essential for any measures agreed amongst the EU Member States to be consistent with any proposals from the OECD. In this regard, we believe the best approach is for the EU to follow the OECD lead and not to implement any interim measure before the OECD report is delivered in 2020.
  • EBIT’s Members are particularly concerned about how the interim proposal is in line with the OECD’s recommendations on interim measures.
  • Moreover, unilateral actions within the EU will need to take into account the EU’s VAT Directive 2006/112/EC in particular and EU VAT rules in general, EU State Aid considerations and fundamental freedoms as well as compatibility issues with treaty and WTO GATT/GATS obligations. International consensus is much preferred to ensure agreement on what will be effective in light of such international frameworks and to avoid a proliferation of defensive measures. This should also ensure that the agreed measures are appropriately targeted at only those situations, which are viewed as distortive.

1.2. Scope and definitions

  • The definitions within the DST and Digital PE proposals are deliberately brief, which will allow EU Member States to interpret the scope of taxation very broadly, inevitably resulting in double taxation. For example, Article 3(1) of the DST refers to “revenues resulting from” the activities specified in paragraphs (a) to (c) as “taxable revenues”, with little guidance as to how directly those revenues result from the activity. A large part of the measures in the draft Directives are intended to address taxation value created under new business models, which are hard to measure under existing tax systems.
  • To be effective, but without unduly burdening groups that are not directly targeted by the Commission, EBIT Members recommend that the nature of activities within scope be defined much more carefully and explicitly. The interim solution in any event should be targeted to address the perceived challenges arising for highly digitalized businesses. Just because a business makes an “electronically supplied service” from an EU VAT perspective does not mean that the business is highly digitalized and exhibits features that need to be addressed.
  • Traditional groups are still likely to have some technology activity, involving local web pages, communication servers, etc. As the Commission explains, these are not the target of the legislation, but based on the broad drafting of the Digital PE proposal, it could result in multiple PE registrations for such organizations, but with little, if any, taxable profit allocable to those PEs. This would neither be in the interest of the taxpayer nor of the taxing authorities; there could be a significant compliance burden for minimal taxation. An enhanced definition of auxiliary and preparatory could perhaps help avoid non-targeted transactions falling within scope of the Digital PE proposal.

1.3. Value creation and substantive issues

  • In general, EBIT Members remark that it is on several occasions not possible to ring fence the addressed digital economy from other, more traditional types of business. For example, providing goods or services to a customer through a business internet portal should not be targeted under either the interim or long-term measures, as these are activities that could equally be managed remotely by a telephone sales team or even through written correspondence i.e. it is not a new business model, merely the enhancement of an existing model by applying better technology.
  • Similarly, digital content is inherently no different to paper content and we would argue that it is anomalous that for instance the sale of an eBook results in a digital PE when mailing a paper book to a customer does not. The only difference is that the former is delivered digitally (with environmental benefits). The same applies of course more generally to for instance journals and newspapers. For this reason we would question whether all digital content should be within scope. In our opinion, there is cross investment from digital to traditional and vice versa. At some point, all businesses will likely be subject to these regulations which is not the intent.
  • It should also be observed that the attribution of value to the collection of data in all cases is a significant and far-reaching change to the principles of taxation. It is not clear that there is value in data collection in all cases, or that there is a real or implied bargain between the data provider and the data collector that the data is being provided for value or in return for access to a service. EBIT Members are of the opinion that the two proposals largely ignore the fact that digital business models are heavily dependent upon strategic management decisions that are focused on building “network effects”. These “network effects” in turn increase and retain the user base and ultimately facilitate the matching of users within the network. These management decisions that create a digital strategy, the “network effects” and the underlying computer algorithms are, in our view, the true value creating and income producing activity within most digital business models. In the view of EBIT, it is therefore not the collection of the raw data that creates value but rather the analysis and insights that are gained from the data, which could create value and ultimately taxable profits - or equally, taxable losses if the wrong analysis is undertaken (see also related to this point, example 3 provided in section 1.5 below).
  • Adding to this, there seems to be a contradiction between the assumption that data, without manual intervention or any human presence, creates value and the BEPS Actions 8-10 revision of the Transfer Pricing Guidelines that discounts the allocation of value to a location where there is no physical substance. It is difficult to reconcile this approach with a specific attribution of value to accessing users within a country.
  • It is not clear to EBIT’s Members whether the case for action by the EU is an anti-BEPS driven issue that has not been addressed yet, or a separate challenge that arises from changes in technology and the global economy. The justification for EU action does refer to the OECD BEPS project’s conclusions that profits should be taxed where value is created, but the Commission’s proposals either assume that value is being created where digital interactions take place (whether or not these lead to commercial value being derived), or in the case of the DST, create a tax where there may be no profit generated in the enterprise in any location.
  • The Commission’s reference to levying DST on revenues from the “sale of data” (by social media/search engines) is peculiar. Such business models usually do not involve selling the data collected from users. The Commission may wish to clarify this reference as this could inadvertently bring some businesses within the scope of the DST.
  • Introducing a new tax requiring global data and allocations of revenue using apportionment keys not in use today will require significant changes to IT systems particularly in large, diversified MNEs, which frequently operate multiple ERP systems from different providers with bespoke interfaces, which combined with the necessary ongoing compliance activities might (more than) offset the tax collected. Further, there should be sufficient time between enactment of legislation by the member states and the commencement date, to allow amendment of complex IT systems to capture the new data on a consolidated basis, including building the necessary interfaces required for global reporting.

1.4. The Commission’s impact assessment and core assumptions

  • The case for action made by the Commission’s impact assessment and accompanying documents stress the lower level of taxation enjoyed by digital companies. In this respect, we strongly encourage and urge the Commission to take any relevant academic studies and commentary on effective tax rates and value creation duly into account in its policymaking. We wish to draw in particular attention to the following three examples to underline this important point:
    • In its Communication of September 2017, the Commission claimed that digital companies face a tax rate of just 9%, compared with 21% for traditional business models. The Commission refers to the joint ZEW-PwC study “Digital Tax Index 2017: Locational Tax Attractiveness for Digital Business Models” to argue that digital business models pay a 9% tax rate, while it refers to the 2016 ZEW study “Impact of Tax Planning on Forward-looking Effective Tax Rates” to argue that traditional companies pay a 21% effective rate.
    • However, importantly, the co-author of these two studies, Prof. Christoph Spengel, has actually distanced himself clearly from the Commission’s assessment that digital economy companies are not properly taxed: “it is not correct to state that the digital sector is undertaxed." Prof. Spengel has stated that the study the Commission relies on for the point that traditional business models pay a 21% rate “considers a completely different type of investment” than the index he authored for digital businesses that considered intangible assets, royalty income and research, all of which receives a different tax treatment. The study on which the Commission relies for the 21% rate modelled national tax rates for the manufacturing sector, and concluded that manufacturers typically paid tax at a 21% rate after accounting for investments such as buildings and machinery.
    • As a second example, we refer to an article by Dr. Matthias Brauer entitled “Digital Companies and Their Fair Share of Taxes: Myths and Misconceptions1. Dr. Brauer argues that the Commission (as well as some governments) assumes that digital companies are currently paying lower effective tax rates, however, no evidence has been presented to date to support that assumption. In particular, the Commission’s hypothetical estimates for the effective corporate tax rates of digital companies do not reflect the high effective corporate tax rates that most of these companies actually pay. Dr. Brauer analysed data from DEG members and concludes that the narrative that digital companies pay lower effective tax rates than non-digital companies is inaccurate. Dr. Brauer also found that while the Commission’s hypothetical average effective tax rate for companies with digital international B2B models is only 8.9%, the average is actually 26.8% to 29.4%, so digital companies appear to pay a higher effective tax rate than non-digital companies.
    • As a third and final example, “EU Digital Services Tax: A Populist and Flawed Proposal”, published by Dr. Johannes Becker and Dr. Joachim Englisch, both professors of tax law at the University of Münster, analyse the Commission’s digital tax proposals in detail. The authors argue that the Commission claims that the traditional rules of allocating taxing rights do not sufficiently reflect the value created by user participation. Becker and Englisch argue that while it is true conceptually speaking, that supply without demand cannot create value, the international tax system uses the term “value creation” in a fundamentally different way. The value of a good or service is set by its selling price, and the supplier creates the value by producing the good or service. Therefore, the locus of value creation is the location where production, not consumption, takes place. In an earlier draft of the DST proposal, the Commission mentioned Facebook and Twitter as examples of platforms where users contribute content, and assumes that the use of the platform contributes to value creation. However, posting content onto a platform does not create value and would be a different definition of “value creation”, which blurs the conceptual division between production and consumption.

1.5. Interaction between the two draft Directives

  • The interaction of the two draft Directives is not clear and they do not appear to require that the interim solution (DST) is terminated if/when the significant digital presence is implemented. The objectives of the two draft Directives appear to be quite different. The interim solution is very targeted in its application, whereas the Digital PE proposal appears to be overly broad in its scope. Furthermore, there appears to be a significant risk that the interim measure will become permanent (at least in some countries) or could be easily expanded in scope if the long-term measure cannot be agreed.
  • From our day-to-day experience, it is not practical to try to implement the Digital PE at the same time as the DST, with the former applying where treaties have been changed or with respect to intra-EU transactions. Transactions tend to involve more than just two jurisdictions, so it would be very difficult from a practical perspective to apply both. This would also entail a higher risk of double-taxation. In addition, it would be unclear which tax should apply to purely domestic situations (e.g. a platform that operates only within one EU Member State will not by definition have any relevant tax treaties to look at, so does that mean that the DST applies?).

1.6. Levelling the playing field

  • It is not clear how the Commission’s proposals actually advance the goal of levelling the playing field for EU and foreign competitors. The tax is obviously focused on deemed under-taxed foreign companies, but domestic businesses must pay it as well, for non-discriminatory reasons. As a consequence, the existing differences in tax burden will be largely unaffected by these proposals. Is there now potential discrimination between an MNE that organises its business as buy/sell (specifically excluded from the DST) and one that is organised as an agent facilitating supply by third parties? Is this also potentially discrimination between different sizes of companies?
  • The case for EU action also does not appear to take into account the changes to international corporate taxation that will take effect in 2018-2020 (e.g. US tax reform, the implementation of the OECD BEPS recommendations, the EU ATAD I and II implementation, etc.). The basis for concluding that “quick fix” action by the EU is needed now is premature and increases the risk of tax uncertainty without safeguarding a global level playing field (and the competitiveness of EU-based businesses) for all MNEs. In addition, the EU proposals bring to the table a political risk for the EU for being perceived as targeting US MNEs.

1.7. Carve-outs

  • It would also be helpful to have clear exclusions within the proposals for data gathered by traditional industries as part of routine business activities or as an input into traditional sale of products. The UK has proposed a similar exclusion in its equivalent of the DST proposal, although this is also limited in scope and should be expanded for the purposes of the DST. A broader exclusion on “the digital services of a preparatory or auxiliary character” would also be necessary, especially when defining the Digital PE. Indeed a lot of traditional industries could have supportive/auxiliary activities (e.g. hotline, online help desk, etc.), whose value should have been included either in the sales of goods or in service charges, hence these services would be already subject to taxation in the place where the customer belongs.
  • The DST proposal should consider excluding the use of electronic media to support maintenance contracts on products sold, leased, or otherwise maintained under a contract with the customer. The DST proposal should also consider excluding the gathering of consumer data via the internet or mobile applications that is used as an input into the design or delivery of product marketing campaigns.
  • EBIT Members acknowledge, however, that as traditional business models and digital business models continue to converge that such exclusions or ring fencing may be difficult to achieve. This convergence also supports our position that an international consensus, led by the OECD is the only viable route to resolve this debate.

2. Specific Comments

2.1. DST proposal

  • EBIT Members are concerned with the principle of resolving a political issue about the appropriate taxation of corporate profits through the creation of a new unilateral turnover tax that targets specific companies and avoids the application of tax treaties and long-standing international tax principles. EBIT urges the Commission, national policymakers and also the OECD to be diligent and commit to sufficient detailed thought and analysis of the issues identified, and to resist politically motivated quick fixes before a comprehensive global solution is agreed, which we recommend be reached in 2020.
  • EBIT considers 3% of revenues is too high: if you assume an average profit margin of 15% and an average corporate income tax rate of 20%, then the 3% of revenues equates to tax on 100% of the profits. We would question whether this would be appropriate and fair, given that the bulk of the value is created by the significant people functions involved in designing and improving the technology. If the DST is not creditable, the effective tax rate on digital services will significantly increase through double taxation.
  • The list of digital services in Article 3 that give rise to “Taxable Revenues” includes:
    “(c) the transmission of data collected about users and generated from users’ activities on digital interfaces”.
    This is a very broad description and would allow Member States to draft legislation to include many types of data transmission. The presumption for the other definitions of taxable revenues seems to be that there are interactions with individuals that deliver advertising or provide access to social media, so that the principal focus of the DST proposal is to tax interactions between individuals and service providers.
  • There are many commercial activities that involve the transmission of data: is it the intention of the draft Directive to create taxable revenue from those activities? An example would be the collection of data remotely from machines, engines or other physical assets and the transmission of that data to a central location for analysis that may or may not result in the provision of additional services or supplies to the owner or user of the asset. Where software are embedded within product and would not be able to function independently or is ancillary to the main product supply, we would not expect that such “embedded software” to be included in the definition. Within indirect taxes, significant case law has been built up on when such services or software should be viewed as separable and believe that this would need to be considered in the context of determining whether DST would apply.

2.3. Digital PE proposal

  • Annexes II and III list services that are, or are not to be treated as “digital services” for the purposes of the proposed Digital PE Directive. In our opinion, the scope of Annex II is overly broad and should be narrowed as many of the transactions are routine business transactions or could inadvertently tax existing digital transactions already undertaken by traditional businesses today.
  • Increasingly, there are data connections and exchange of data between physical assets, machinery, engines, etc. and remote installations that monitor and analyse data, that may or may not be used for commercial exploitation. These transactions should be excluded from the scope of the Digital PE proposal.
  • The lists in both Annexes do not appear to recognize the Internet of Things as either specifically included or excluded. There should be a conscious effort to address this lack of definition to avoid unintended domestic law drafting that includes, or excludes, this data exchange in creating a Digital PE.
  • Unless specifically excluded, the interconnectivity of most modern appliances, machines and engines is likely to result in the creation of a Digital PE for most manufacturers or suppliers of connected assets.
  • The thresholds look rather low as well as being difficult to interpret. For example, what does 100,000 users actually mean? If they are discrete users, how would a business demonstrate that in practice? Do they have to be “active” users or all users? The relevance of “users” to different business models varies significantly – how will that be reflected? EBIT would challenge whether IP addresses are really a good proxy for geographical origin. This is already a problem with the use of VPNs, and, likely, this will just encourage more “disguising” of IP addresses to subvert these rules. It should be noted that the EU VAT rules for electronically supplied services move away from using IP as a proxy for the place of supply and instead use two or three pieces of non-contradictory commercial evidence. The decision to use a different set of criteria as a means of proxy in this case would create significant burdens and result in odd anomalies in having to account for VAT in one jurisdiction but recognize a digital PE elsewhere. As a result, there is a strong recommendation to move to away from IP address as a proxy towards “usual or habitual residence” or “registered office” criteria for such situations.
  • We note that the scope of the DST and Digital PE proposals are set differently. As outlined above, EBIT’s Members cannot see a policy reason to include the provision of digital content within the scope for a Digital PE, where it is inherently no different to the provision of paper content. If the rationale is that active users drive value, then surely the scope of the Digital PE should be defined only to catch those models where active users are a necessary/significant part of the business model. “Passive” users are no different to any other customer – if they are included in scope, then the Commission’s proposals are simply defining a PE based on the presence of a market. Moving taxing rights to the market could then be extrapolated to the entire economy by other, larger markets such as China, the US and India, and the EU will lose out.

EBIT’s Members trust that the above comments are helpful and hope that they will be taken into account by the Commission in its important work towards the much-needed international consensus in this area. EBIT is always ready to discuss with the Commission and any other stakeholders.

Yours sincerely,

European Business Initiative on Taxation – May 2018

1 Published by the think tank European Centre for International Political Economy (“ECIPE”) ↩︎