Tax me if you can – The taxation of digital businesses in the EU


The proper taxation of digital companies has become a top policy issue in 2017 and is expected to gather even more steam throughout 2018. In this blog post, Paul-Jasper Dittrich and Pola Schneemelcher take a step back and assess the European debate and the main challenges ahead: The current discussion might fall short as it is mainly lead in the context of tax evasion and regulating American tech giants. Since digitalisation infuses each and every aspect of society, numerous of today’s business models will be or are already digitalized. Thus, a ring-fenced approach to answer the challenges that digitalisation puts on current tax rules is not enough.

1 The political debate is gaining momentum

In August 2016, the European Commission decided that tax deals negotiated by the Irish government with American tech giant Apple amounted to illegal state aid, and that the company had to refund €13 billion in unpaid taxes to the Irish authorities. The decision, together with recent incidents such as the search of Google’s offices in Paris, demonstrates a new European resoluteness: National governments and the European Commission alike are increasingly determined to press their claims against global technology companies over what they perceive as illegitimate practices of tax evasion and profit shifting.

However, a much deeper technical and political challenge underlies the headline-grabbing raids and political quarrels: The current international tax system is ill-prepared to deal with the changes in value creation brought about by the digital transformation. As a result, the taxation of businesses in the Single Market has become unbalanced. According to the Digital Tax Index 2017, the effective tax rates of traditional domestic business models (20.9%) in the EU exceeded the tax rate of digital domestic business models (8.5%) by far. In this blog post, we take a closer look at the most pressing technical challenges related to taxing digital business models in the Single Market. We focus specifically on the unsolved problem of non-existing Permanent Establishment (PE) and on how current tax rules do not fully account for digital value creation. We conclude with an assessment of and some caveats on the current debate within the framework of the Single Market.

The policy issue of the proper taxation of digital companies gathered political steam throughout 2017. Both the French President Emmanuel Macron and European Commission President Jean-Claude Juncker have made taxation a top priority in their political reform agendas for the EU. The topic gained further political momentum with the Digital European Summit in Tallinn in September 2017 and will become even more prominent in the spring of 2018, when both the OECD and the EU Commission are expected to present new proposals on combating tax evasion and on the taxation of the digital economy. The two political and technical processes at inter- and supranational level are accompanied by proposals from Member States: In September 2017, four European finance ministers, under the leadership of France, put the Commission under pressure by asking it to explore the possibilities of establishing an “Equalization Tax” that would aim at revenue instead of profits as the base for taxation of digital companies.

2 The Challenges – Overview

While decision makers are already presenting ambitious solutions, the definition of the challenges posed by taxation in the Single Market remains fragmented. Policymakers and scholars find that current tax rules no longer fit the digitalized context and lead to decreased competitiveness, unfair taxation and declining state budget revenues across the EU. The current EU tax framework leaves it to the Member States to decide upon their own tax systems and individually design cross-border transactions through bi-national treaties. Bi-national tax treaties among EU countries generally follow the so-called Model Tax Convention drafted by the OECD. The Model Tax Convention form a non-binding blueprint for international tax treaties, thus actual treaties can depart from it.

One defining feature of digital business models (e.g. social networks, search engines or cloud hosting companies) is their heavy reliance on, and use of, intangibles (see also figure, “value creation”, right column). Intangibles in contrast to tangibles like machines or office buildings do not need a physical or geographically bound presence. Intangible assets are assets such as patents, brands or a piece of software for most digital companies. Intangible assets are often non-rivalrous and scalable, i.e. they can be used by many at the same time without their availability diminishing. An example would be a piece of software that can be used across the world, no matter how many other people are using it at the same time. Owing to these economic properties, intangible assets are protected by property rights (licenses, copyrights, patents, trademarks and others) to encourage investment and innovation. Intangible assets, as the main value drivers of digital business models, challenge the core paradigm of current corporate tax rules – profits should be taxed where the value is created – due to two crucial features:

  • In contrast to non-digital business models, digital businesses often do not have “sufficient physical presence” (so-called “nexus” or “permanent establishment”) in the country where they create value. Thus, permanent establishment (PE) rules, which traditionally contribute to the definition of a tax base, do not apply to them.
  • The profits of digital business models are difficult to define. Often, profit is not only derived from goods and services delivered. Digital business models also raise the question of how to value intangible assets such as software or data themselves, as well as their contribution to the value chain (value creation)

Simplified depiction of main differences in value creation and taxable nexus between predomi-nantly digital or non-digital business models.

These two significant features of digital business models therefore challenge (1) permanent establishment rules and (2) the international rules governing the internal transactions of companies across borders (transfer pricing rules), for example the licensing of software from the parent to a subsidiary in the EU. In order to identify effective tax rules, these challenges require further attention.

3 A deeper look at Permanent Establishment

The current rules of the international tax system rely on the source taxation of profits and are largely producer-oriented. They attribute the taxable nexus and therefore the right to collect corporate income tax to value creation, i.e. to tax jurisdictions where the physical development and production of a product or service takes place (see figure, “nexus”, left column) . These rules were established for the industrial age and further developed to account for global supply chains. However, they do not adequately factor in the rapid changes in the way value is created as a result of the ongoing digital transformation of production and delivery of goods and services.

Digital business activities are spread across different jurisdictions and value is not necessarily only created where a company is physically present (see figure, “nexus”, right column). The challenge of the permanent establishment (PE) paradigm, and hereby the determination of a taxable base, is certainly not solely attached to the digital economy, but rather to a globalised world in general. Non-digital enterprises such as Starbucks or digitals such as Google both conduct R&D in the U.S., finance their businesses with loans from Swiss banks, and shift profit through royalties taxed in Ireland. In times of digital business models, the diversification of sites where value is created is an exacerbating factor.

In the current debate, the OECD and EU institutions stress that the concept of PE will remain one of the essential principles of global allocation of taxing rights on profits. The PE concept dates back to the first wave of globalization, when business became cross-border and multiple states claimed their share of profits. PE principles laid down in bilateral tax treaties guaranteed a market country could tax a foreign person to the extent she participated in its national economy – and guaranteed the same income would not be taxed twice in two different countries (so-called “double taxation”).

As mentioned above, PE is defined exclusively as physical in the OECD Model Tax Convention, through “a fixed place of business”. The OECD made certain amendments to the principle in recent years. Overall, its amendments do not completely abandon the physical presence requirement and are thus only partly sufficient to address the main challenges of the digital economy. At EU level, the Commission and the ECOFIN are asking for alternative indicators and discussing the long-term solution of a “virtual permanent establishment”. However, at the current state, such indicators remain undefined.

The status quo of the debate proves that a new definition of taxable nexus is needed.

4 A deeper look at taxing value creation

As described above, digital business models derive their value mostly from the development and subsequent exploitation of intangible assets, for example from the algorithms of a search engine. Depending on the specific digital business model, this entails the use, analysis, licensing or sale of software code and algorithms (Microsoft, Google), data (SAP), remote computing power (Salesforce) or digitized content, such as eBooks (Rakuten) and videos (Netflix), to name but a few.

Under current rules, profits should be taxed where the value creation takes place. For digital business models based on intangible assets, that means it is of utmost importance where a piece of software is actually created and improved, in which country the property rights on it remain, and how it contributes to sales in a third country. If the patents, marketing, software development and main operational activities of a company, say Facebook, are all located in one country, for example the US, the bulk of the value creation (and the main taxable nexus) is located there under current rules. As a result, non-EU companies with digital business models often do not have a considerable taxable nexus in the EU because their local activities are considered routine and negligible under current rules. According to current OECD guidelines, local collection, storage and preparation of data for analysis (with data mining) is considered a routine task that does not constitute a taxable nexus and is not considered part of (taxable) value creation. Subsidiaries of digital companies in the EU hence often claim that they only play minor roles as assisting entities, executing minor local functions in their parent company’s operations, and therefore contributing very little to its overall value creation. Such claims, although formally correct, often seem to be inconsistent with the actual importance of the subsidiaries, for example in terms of their crucial role in collecting local consumer data.

Companies with a digital business model also benefit extraordinarily from tax optimization and profit-shifting strategies based on transfer pricing rules involving intangible assets such as licenses. In general, such strategies are not new: Sublicensing property rights (patents, software, brands) to a local subsidiary at unduly high licensing costs is a typical example of the use of transfer pricing between different entities of a global company for the purposes of tax optimization. Having paid royalties and other fees to the right-holding entity within the same company, the subsidiary declares close-to-zero profits to the local tax authorities. There are, of course, rules for such transactions: under current OECD guidelines, internal transactions have to be made at comparable market prices (“arm’s length principle” for transfer prices). However, digitally derived intangible assets such as algorithms are often unique structures (like Google search algorithms). It is very difficult to determine a comparable value for them and apply the arm’s length principle, which would otherwise prohibit the charging of unduly high royalties for licenses on the assets used within the same company.

How can digital value creation and the understandable insistence of Member States on higher tax revenues from digital companies be realigned? One possible way forward is to focus on the definition and functional analysis of data-driven value in each tax jurisdiction. As described above, value is often created by processing and analyzing locally sourced information (data mining) and turning it into ad revenue, for example. New rules, based on a different definition of value creation can lead to new solutions for the determination of the tax base.

5 Assessment of the current European debate on taxation of digital businesses

As the debate on the taxation of digital businesses in the EU will intensify in the coming months the following four points should be considered. Firstly, the current discussion seems to lack a joint working definition of the “digital economy”. This leads to a limited debate that often narrowly focuses on a small number of American tech giants. As increasingly more businesses from the “traditional” non-digital economy are in the process of partly or fully digitizing their business models and value creation, the “digital economy” will become ubiquitous, and tax challenges for the EU will only multiply further. Car manufacturers, for example, will gradually transform into software companies as more and more of their revenue will be derived from software applications, data collection and algorithms in connected cars.

Secondly, the debate is mainly led within the context of tax evasion and profit shifting (BEPS). Addressing its challenges in the scope of respective BEPS-policy measures like the CCCTB seems to be a tempting and simple solution. However, in order to assess the implications for the Single Market and conceive the policy consequences of the digital transformation accurately, it should be perceived as a process in its own right. While digital businesses have numerous possibilities to shift profits and minimize their tax base, they did not invent current tax optimization strategies. They merely use existing international guidelines to their maximum advantage.

Thirdly, it is important to apply comprehensive functional analyses of specific digital business models to establish convincing arguments for a taxable nexus that is based, for example, on data mining. A cloud operator has a business model and market exposure that are very different from those of an E-commerce platform or a social network. Value creation and tax implications may vary dramatically depending on the specific digital business model. An American online TV show can be considered as a simple import to the Single Market and be taxed with VAT. However, what if the operator (say, Netflix) uses its European customer data to improve its algorithms and its European offer on the base of a data mining operation with their data? The latter would be a clear case of data-driven value creation for which a taxable nexus should be found.

Finally, the EU should converge towards a common stance on the international level. Given the depth and complexity of the issue, it would be risky to jump to premature conclusions. Before putting forward proposals such as an “Equalization Tax” or amendments of existing rules, we need to take a step back and comprehensively analyze the different aspects of the debate. Against the backdrop of global markets, an international debate and subsequent solution under the OECD umbrella is urgently required. Nevertheless, implications for the Single Market are vast and Member States should adopt a coordinated position to have a greater impact at the global level. Last but not least, the taxations of the digital Single Market has the potential to assist the debate over a new EU budget. As the High Level Group around Mario Monti recommended in its report, revenue stemming from EU digital Single Market policies could also serve to complement a new budget.