The impact of digitalisation on international tax

The impact of digitalisation on international tax principles and domestic tax bases is a global issue, and it is important that the EU has a clear and unified position. We also believe the EU should be pushing for a globally agreed and consistent response.

Given the complexity of the issues, we have set out some wider considerations before considering the specific questions.

A wider political dimension

A first observation is that this topic has, to a certain extent, a wider political dimension which should not be ignored. The vast majority of successful companies in the new digital economy are companies from the US. This can lead to the perception that attempts to amend the current taxation of the digital economy are mainly targeted at US multinational enterprises (MNEs). A question arises, why is that the case? Most start-ups are not profitable in their first few years of operation, and therefore cannot benefit from special tax regimes or exemptions. Nor is the US known for giving special subsidies to start-ups. Therefore, the success of these US companies in the critical first five years of their existence cannot be explained by advantages in the US tax system. So, why is it that the US generates more than its ‘fair share’ of successful companies than the EU, especially when it concerns the new economy? Is this because the US – despite being a federation of states – is more one country and one market than the single market of the EU? Is it because the investment and business climate for (starting) companies is better in the US, than in the EU? Or is it simply due to a stronger focus on technology in US universities? The debate about taxation of the digital economy often talks about a level playing field between domestically operating old-economy enterprises in Europe and the multinational digital-economy companies from the US. Equally, in terms of a level playing field, a question could be asked whether the policies of the EU are providing sufficient support to business compared to, for example, the US. Given the high success rate of US companies in the digital economy, a concern is that if the EU (without an OECD agreement) unilaterally introduces special tax measures such as the “GAFA tax”, this may be perceived as an attack on US companies, which could exacerbate existing trade and tax tensions. In addition, care should be taken that taxes are not introduced that could end up penalising successful, outbound, EU-based digital businesses.

Tax jurisdictions

A second political issue is that a highly digitalised business may have significant revenues in a country but pay no (or little) corporation tax there under current rules – irrespective of whether or not tax is paid elsewhere. This is an important issue which clearly needs addressing.

There is also a concern that some countries believe their corporate tax base is reduced because deductible business payments (e.g. for advertising) are being made to foreign multinationals which then pay no tax in the source state. To a certain extent this concern is more about the impact of cross border trade, and we have seen no studies to suggest that digitally-enabled businesses do actually result in the reduction of some countries’ tax bases overall. It may be that by increasing efficiency and increasing economic activity, they in fact expand the tax base. So this is an area which the European Commission could usefully carry out further analysis.

The transformative issue before us is that an MNE may be doing business in another country, but not have a taxable physical presence there, which does not align with traditional tax norms. Put simply, with the emergence of the digitalised economy, it must be considered whether there should be agreement on some fundamental changes in international tax theory, shifting taxing rights towards the source state or market state and away from the residence state. There are arguments for and against such a change, and KPMG is not promoting one approach over the other – that is a political decision.

We do note, however, that the issues are too complex to deal with adequately in multiple choice questions. For example, many of the questions in the survey are necessarily rather generic, and others provide insufficient flexibility to note that many answers are currently unknown or unknowable. Our answers to the questions, therefore, need to be read in the light of this explanatory note.

The digital economy as part of the traditional economy

We also believe it is a mistake to talk about the “digital economy” as if it were separate from the traditional economy. The reality is that digitalisation is impacting all businesses and fiscal changes designed to tax a “pure digital” enterprise (e.g. that sells digital service into a country with no physical presence there) may well impact traditional businesses which use technology to manage their supply chain, create efficiencies and improve the customer experience. Well known examples include hotel room or flight reservations. Twenty years ago, they were all made through travel agencies physically present in the country of the consumer. Nowadays this is mainly done online with no physical presence required.

We believe, therefore, that the survey on its own does not sufficiently clarified the problem it seeks to address with respect to the “digital economy”. More clearly defining the problem will allow the European Commission, and the EU, to better tailor the scope of the proposed solutions in order to minimize unintended consequences and maximize effectiveness. Based on the information released by the Commission to date, we observe at least three possible problems that the Commission may be seeking to address:

(i) A concern with the underlying balance between residency and source based taxation in the existing international tax system - at least as it applies to certain digital businesses;

(ii) A concern with the existing permanent establishment rules – e.g. are they working properly to tax any physical presence in a source state such as the presence of a marketing workforce? or

(iii) A desire to prevent tax planning and / or tax avoidance schemes within the existing international tax system.

Clarifying which of these possibilities (or perhaps others) the European Commission intends to address with its digital economy policy will go a long way in enhancing the dialogue around these issues. We also note that the consultation only focuses on corporate tax. However, it is necessary to consider whether or not the issue of taxing cross-border sales is not better addressed by the use of VAT.

In our comments below, specifically through the examples and scenarios we provide in our comments to Question 4.1, we highlight some of the issues that arise depending upon on which problem is intended to be solved.

Legality of any new rules

A detailed analysis of the legality of specific proposals is not within the scope of our comments here. However, it should be noted that any proposals will have to comply with the fundamental freedoms, the VAT Directives, state aid rules, and the principles of proportionality and subsidiarity. They will also have to respect bilateral tax treaties – unless it is intended that they should all be amended. Finally, it may be necessary to consider the impact of The General Agreement on Tariffs and Trade (GATT) and The General Agreement on Trade in Services (GATS) under the World Trade Organisation rules.

Question 4.1: Are the existing rules working?

The answer to this question depends upon what the issue is perceived to be, and the desired outcome. We illustrate this by looking at the example of a company (USco) which is established in the US and sells its product worldwide through the internet. It does not matter for the example if it is selling purely digital services or physical goods.

i) The source versus residence issue

Assume the investment in USco has all come from US investors. In the start-up years, it made huge losses investing in the algorithms and technology needed to power its business. These losses are carried forward and used against US taxable profits in later years. The US economy has, therefore, carried all the economic risk of establishing the business model. The company continues to invest heavily in technology and its entire workforce is located in the US.

Once USco is established it is making significant revenues in the countries where it sells (e.g., Ireland). It has no need of a physical presence in Ireland to make the sales.

The heart of the “digital” debate is whether or not USco should pay corporation tax in Ireland in this scenario. Under current rules it does not. Allocating some of the taxing rights to where the customers are based would be a fundamental change, although arguments could be made for this under source and benefit principles. Conceivably this could be done by extending the notion of a permanent establishment (PE) to include a pure digital PE. However, even with the adoption of a digital PE standard, determination of the profit attributable to such a digital PE remains a challenge. Actions 8-10 of the BEPS project updated transfer pricing rules and focused on where risks and assets are managed. In the example we have used, the risks and assets are all managed in the US. Furthermore, even if some revenue is attributed to the PE, what proportion of the costs should be allocated? Would the US be able to attribute some of the carried forward tax losses from the start-up years to the PE? Without such (complex) cost allocations, could a situation arise where the US would still be giving relief for real losses, while source countries (where there were merely sales and no personnel) would be taxing some of the profits?

One of the arguments for taxing the digital presence is that the parent company is using data which is collected in the source state. The key question here is, what is the value of the data which is collected? Is there a difference between publicly available data as opposed to confidential data? Is there a different value to data depending upon the industry involved and its usefulness; or is it possible to ascribe a standard value to a certain quantity of data? In any case, the vast majority of the value is likely to come from the collection mechanisms and the way the data is analysed and used, which are all functions performed outside the source jurisdiction.

There is a risk that if a digital PE concept was introduced, it would create a significant compliance burden in terms of additional registrations and reporting, with very limited taxable profits in return. Alternatively, a digital PE may result in considerable disagreements between states about apportionment, and potentially result in double taxation. In the example given above, the US would resist any claim by Ireland to allocate significant profits to Ireland.

From this example, it can be seen that the fundamental, underlying issue is not so much about “fairness” or paying a “fair amount of tax”. Tax will be paid in the residence state under the existing rules (e.g. at 35% US Federal rates in the example above), or if changes are made, some of the profits will be allocated to, and taxed in, the source state (e.g., at 12.5% currently in Ireland in the example above). Thus, the issue is a political one about how to allocate taxing rights. If there is a “fairness” issue, it is between different states but the arguments are very complex and require a global solution.

ii) Taxing the “physical” parts of the business

Now assume that USco sets up a subsidiary in Ireland that employs many staff to carry out marketing functions in Ireland and Europe as a whole.

A concern is often raised that a multinational in this position does have a significant physical presence but still pays little corporation tax. This is clearly a valid concern but the issue is not really about the taxation of digital businesses as such. Rather, the issue is whether or not the method of calculating the profits of the marketing subsidiary are correct under traditional transfer pricing rules. It can be argued that there is an issue of “fairness” or “competition” in this case, as different business structures should not, in principle, give rise to different tax outcomes. However, as these rules were considered and updated under the BEPS project (Actions 8-10) it would be advisable to test whether or not the changes have been effective before making further ones.

There are also concerns that a multinational like USco can – or could – avoid having a traditional PE in the source state by fragmenting its physical operations (e.g. by putting marketing, after-sales services, logistics, etc., into different companies). Again, such concerns should have been addressed through BEPS Action 7.

In short, where concerns are about transfer pricing and business structuring, it should be possible to address them through existing rules without having to introduce new rules or taxes for digitalised companies.

iii) Concerns about tax planning or avoidance

Assume now that USco has transferred the right to use its intellectual property outside the US to a subsidiary (IPco) in a zero tax country. Assume further that it also sets up a European subsidiary (Opco), which makes all the sales to customers in Europe. Opco pays most of its profits to IPco and so pays little corporation tax in the EU.

The “political” issue now becomes more acute, as much of the profit that is made from sales in Europe is not taxed anywhere on an immediate basis. IPco’s profits would be taxed in the US if they were ever repatriated. Such structuring creates concerns about competition and fairness – as USco can reduce its effective rate of tax. However, in terms of taxing the digital presence in the countries where the sales are made, the same issues remain as in the original scenario in i) above – should there be a shift towards allocating tax to where customers are based? Furthermore, there are existing mechanisms to address concerns with this sort of structuring, such as:

(i) The US could tighten its anti-deferral rules so IPco’s profits would be taxed in the US (it is noted that the proposed changes to tax rules in the US at the date of writing may well address this issue to some extent);

(ii) The allocation of profit could be challenged through existing transfer pricing rules as applied to Opco; and

(iii) Anti-avoidance rules, such as withholding tax on payments to low or no tax jurisdictions, could be applied.

In conclusion, it is likely that existing rules can be used to address many of the perceived issues with taxing digitalised businesses. The key issue remains, however, should international tax principles be amended to allocate greater corporation tax taxing rights to where customers are based, and away from the location where it is considered that value is created under current rules.

Question 4.2

While the current state may in fact encourage some countries to seek to implement unilateral tax measures that contribute to the fragmentation of the European single market, we do not think this is the best approach for any single country. However, we similarly note that the problem with unilateral action is replicated on a global scale – with unilateral action by the EU on these matters threatening to destabilize the existing international tax system.

Question 4.3: Is there any need for action regarding taxation of the digital economy?

We have answered yes because we believe there needs to be global agreement on how to ensure the international tax system adequately taxes modern – and developing – business models. Clarity and certainty is needed for both governments, which require stable revenues, and for companies, which need to understand their rights and obligations and to avoid being subject to double taxation. That action could result in an agreement that the current rules are sufficient or it could result in new rules.

Question 4.10: What are the most important objectives that future legislation should pursue?

The “Other” should be ensuring tax is not distortionary between different types of enterprise and structure.

The third option refers to neutrality but goes on to talk about fairness which begs the question whether it is fair that, in a globalised world where tax remains a sovereign issue, a company located in one country can have a lower rate of tax than a company located in different state.

Question 5.1: Targeted temporary solution

We can understand the drivers from a political and revenue raising perspective, but counsel against temporary solutions. A temporary solution is not likely to be fully adequate and could lead to unforeseen consequences, such as distortion of the market and double taxation. There is also a risk that local temporary solutions become permanent – even where there is global agreement on the new rules.

Question 5.2: Specific temporary solutions

It is difficult to answer this question given the complexity of the issue and lack of information. The proposals might work to some extent if the perceived issue is simply a political one – that some digital companies are not paying tax in the local jurisdiction of the customers. The issues with all of them are:

(i) the proposals could create situations of double taxation, as there might be no relief in the country of residence where the primary taxing rights are traditionally held; and

(ii) how to actually collect the tax.

The digital transaction tax is particularly difficult to delineate. Is it a tax just on a quantity of data collected? How is that monitored? The value of the data collected will very much depend upon the industry and the use.

A tax on advertising revenue might appear more targeted – if the issue is that advertising payments are tax deductible, and therefore reducing a country’s tax base, but the income is not taxable. However, presumably advertisers use foreign digital companies for advertising for commercial reasons. This implies they cost less or drive greater revenues – either way, one would expect the profitability of the paying entity would increase, so as to drive up the local tax take, which compensates for the deductible payment. Has the EU done any studies which show that offshore digital services reduce the local tax take – or do they actually increase it via efficiencies?

Question 5.4: Long term solutions

We feel that the multiple choice format of this question is not suitable to address the complexities raised by these matters as there are too many unknowns.

The Common Consolidated Corporate Tax Base (CCCTB) could be used to address issues of digital selling if the aim is to shift tax to where the seller is based, but only within the EU. It would not solve the problem globally, especially as many of the largest technology companies are based in the US. Trying to resolve the issue with the CCCTB by the EU on a stand-alone basis may have a negative impact on the investment and business climate of the EU. Likewise, it may have an overall negative impact with adverse consequences if it involves imposing unilateral solutions which affect non-EU businesses.

A destination-based tax system, if adopted globally, would address the issue, but only if it is generally accepted that the taxable value which a company creates is not where the productive activities are carried out, but where the customers are based. This is contrary to existing practice and theory.

Unitary taxation has the same issue as a destination-based tax – but it is mitigated, in that the formula may apportion part of the profits to where the activities are carried out and where the assets are held.

Using a tax rate in the residence country which is a blend of all the taxes in the destination locations should solve the issue of MNEs looking to base their operations in low tax countries. However, it is very complex and does not address the underlying fundamental issue – that some countries see very large revenues without being able to collect any tax. In other words, it addresses the problem of country shopping and potential avoidance structures, but does not solve the fundamental issue of taxing rights and source versus residence.

Question 5.6: Impact on business of a digital tax

It is impossible to answer most of these questions without further detail on what the proposal would be. However, it is clear that digital taxes would increase compliance costs. Depending on the regime ultimately introduced, it may have an impact on the development of technology service in the EU.


The issues surrounding the digitalisation of business and taxation are highly complex, but clearly need addressing. The central issue is the fact that due to technology is it possible for a MNE to be doing business in another country (and on a very significant scale) without having a physical presence and therefore a taxable presence under existing rules. Proposals to shift the balance of taxing rights towards source-countries and away from the country of residence represent a fundamental shift in international tax policy and theory as agreed by countries over the past decades. The BEPS project represented a broad multilateral agreement on revised standards to minimize perceived abuses of the historic international tax policy, but expressly did not address the existing rules on the balance of taxing rights between source and residence states. The EU, and the European Commission, should be very clear in assessing the implications if the intention is to propose a shift in this underlying policy.

As stated above, KPMG is neutral on whether or not such a fundamental change should be made. This is a political decision. However, whatever the final view, KPMG believes that significant work remains to understand any unintended consequences.

Further, we believe that any changes should:

  • Be globally agreed as far as possible;
  • Be as simple as possible to apply for tax administrations and taxpayers;
  • Create certainty for all parties going forward;
  • Not create double taxation, non-taxation, or undue administrative burdens;
  • Not create a drag on the development and diffusion of digital technology;
  • Not distort commercial decisions;
  • Ensure in particular that they do not create a burden for small and start-up businesses.

We thank the European Commission for the opportunity to provide these comments and look forward to continuing the dialogue to help the EU to craft efficient and effective tax policy.