Six questions plus one about the proposed EU Directive on the taxation of a “significant digital presence”

Kluwer International Tax Blog
April 20, 2018

Please refer to this post as: , ‘Six questions plus one about the proposed EU Directive on the taxation of a “significant digital presence”’, Kluwer International Tax Blog, April 20 2018,

On 21 March, just a few days after the publication of the (frankly not exciting) OECD report on the tax challenges of digitalization, the European Commission presented its own “package” on the taxation of the digital economy.

This comprises: (a) a proposal for a Directive on the corporate taxation of a “significant digital presence”, or “SDP” (hereafter, the “Directive”), (b) a Recommendation to Member States for including corresponding rules on a significant digital presence and profit allocation in their double tax treaties with third countries, and (c) a proposal for a Directive for an “interim solution” in the form of a new “Digital Service Tax” (DST), which is meant to be applicable until the new provisions on SDPs enter into force.

In a nutshell, the Directive extends the concept of “permanent establishment” to encompass an SDP. The latter will be identified by reference to some thresholds: (a) revenues generated in a certain State, (b) number of users located in that State, and (c) number of business contracts entered into with users located in that State. This new “digital PE” will only apply in relation to a certain range of “digital services”, defined in the Directive.

In addition, the Directive sets out principles for attributing profits to a SDP, generally (and arguably a bit simplistically) based on a mere adaptation of the authorized OECD approach to the performance of digital services (rather than on a definition of a totally new approach, specifically designed to allocate profits in the digital economy, as advocated by some scholars). Profit split shall be the method to be used, unless the taxpayer proves that another method is more appropriate.

The new rules are meant to override double tax treaties between EU Member States and will also apply to a European SDP of a non-EU business, where there is no treaty in place between the Member State concerned and the jurisdiction in which the non-EU business is established.

Member State are supposed to introduce the implementing laws and regulation by the end of 2019, which would imply the application of the new provisions from 1 January 2020.

Despite the Commission’s good intentions, thoroughly depicted in the “Explanatory Memorandum”, the proposal raises a few question marks:

1. What is the actual purpose of the proposal? Is it to adapt the existing international tax principles to the new scenario or is it to create brand new ones in order to achieve a new allocation of the taxing power among States?

The proposal seems to be more political than technical; talking about a mere adaptation of the existing rules sounds as an understatement (and arguably a bit hypocritical). It is true that in the digital economy, business models have dramatically changed, but what has also changed is the situation of some States, which in the “old economy” were capital exporters and have discovered themselves to be capital importers in the digital economy scenario, and they now understandably want to be able to tax a meaningful portion of the profits made by the digital businesses in their markets.

2. Is it actually true that “taxation where added value is created” is a general principle on which the existing international corporate tax rules were built, as postulated in the “Explanatory Memorandum”?

This is a mantra constantly reiterated by the EU Commission and by the OECD. However, references to such principle can be hardly found in the history of the OECD model treaties.

On the contrary, the OECD works in the post-World War II scenario have been dominated by the theory of “economic allegiance” and the prevailing attribution of taxing rights to the State of residence. This was politically acceptable (for the OECD and EU countries, at least) when a different balance between countries capital exporters and capital importers existed, and probably when the income of MNEs was expected to be taxed at least by the resident State, and not be lost in tax havens or tax haven-friendly countries where intermediate entities of multinational groups could be located (the US tax reform could arguably alleviate this last issue).

3. Significant Digital Presence (SDP) or Significant Economic Presence (SEP)? Is it fair to allocate taxing power to an SDP, only for a very limited range of “digital services”, if at the same time an SEP – outside the context of the digital economy – is not (or not yet) considered enough to constitute a permanent establishment?

The EU Commission has chosen to go on with the “ring fencing” of new rules specifically designed for the digital economy (while the US is strongly opposing this approach). However, if user-generated value justifies income taxation of business profits, as stipulated in the Directive, this should be true in all economic sectors, and not only in those specific ones to which the new Directive should apply.

(by the way, the Italian tax legislator is pioneering in this respect: an SEP is now deemed a permanent establishment under the recently amended Italian domestic tax rules).

4. Is it realistic to conclude that the functionally separate entity approach can be simply adapted to the taxation of SDPs (Article 5 of the Directive)? What exactly are the “economically significant activities performed by” an SDP, referred to in Article 5(3) of the Directive? Can an SDP perform activities?

This is probably the most difficult issue with regard to the taxation of an SDP and much more work is needed. One thing is to ascertain that some added value is created by the interaction with the users and by the data generously provided by them, another thing is to quantify that value and to distinguish it from the added value created, for instance, by the strategic decisions made by the key people of the company or by the performance of R&D activities.

Some interesting studies (e.g. Hongler & Pistone, 2015, and Petruzzi & Buriak, 2018) were recently published about methods to be used in order to calculate the added value of the so-called network effects and the income attributable to an SDP (or an SEP), but further analysis seems to be needed.

The preference of the EU Commission for an adaptation of the AOA to the digital economy is understandable, but ultimately a braver shift to innovative formulary apportionment methods should probably be explored.

5. Will the Member States be effectively able to check the application of the rules related to the localization of users (Article 4, paragraphs 4-7 of the Directive), without breaching the EU privacy rules?

The EU Commission is confident that the information necessary for the practical application of the new rules (namely, the geo-localization of the users) will be collected and managed in a way that will allow to keep it “anonymized”. However, the compatibility of these mechanisms with the new GDPR will need to be further assessed.

6. Will the impossibility to overcome double tax treaties with non-EU countries lead to a flow of digital economy businesses (or subsidiaries of non-European OTP groups) from Dublin or Luxembourg to non-EU “SDP-friendly” jurisdictions?

As the EU Commission cannot force non-EU countries to accept new double tax treaties including the “virtual PE” definition, some of them could try to play opportunistically and attract digital groups, wanting to escape the European SDP rules (this could even be the case for post-Brexit London? Even though at present the UK seems to be inclined to introduce a similar principle of taxation where user-created value exists).

But, most importantly:

7. Will the proposal ever be implemented with the unanimous vote of the Member States?

Nobody expects this to happen, at least in the short term (the reference to an implementation by the end of 2019, indicated in the proposal, is more wishful thinking than an actual rule). Moreover, the hypothesis of an implementation by way of enhance cooperation (if at all conceivable in tax matters) would make little sense for the SDP (as it would mean that the definition of SDP would not apply to the double tax treaties entered into by the EU countries not implementing the Directive, which would likely be those where the subsidiaries of the big digital groups are located).

As a consequence, the “interim measure” represented by the DST, if ever itself implemented by the Member States, could end up being all but temporary. However, the DST and the new definition of SDP are not fungible solutions for several reasons (for instance, their economic impact on companies, and also on the customers, would be very different, as the DST is a tax on the turnover, which will likely be passed on to the customers, while the SDP is meant to affect the application of income taxes).

To make sure you do not miss out on posts from the Kluwer International Tax Blog, please subscribe to the blog here.