OECD/G20 - understanding the proposed Inclusive Framework & the tax challenges of the digital economy

​Pillar 1 - unpacking the confusion!

On 31 January 2020, the OECD/G20 Inclusive Framework¹ released a statement which sets out its approach to tackling the tax challenges of the digital economy.  Following on from public discussions on the approaches outlined in its discussion documents on both pillar one and pillar two, the statement affirms the commitment of the members of the inclusive framework to reach a consensus-based solution on both pillars by the end of 2020.
Although the proposed inclusive framework is likely to face some significant stumbling blocks in its implementation (which we set out below) - what is clear is that the traditional approach to international tax and transfer pricing is changing and likely to change fast.

In respect of Pillar 1, the inclusive framework has adopted the "unified approach" on the basis it believes this is the least complex and will create greater tax certainty.  In respect of Pillar 2, the inclusive framework acknowledges that more work is needed to batten down the nexus rules.

What does Pillar 1 propose in a nutshell?

The traditional approach to international taxation is to tax based on a source or residence basis.  Residence in this context means taxation of worldwide income in the country of residence, and source means taxation based on economic nexus created by some degree of physical presence in a particular country.  In order to apply a source basis of taxation, the operations must create a permanent establishment in terms of both domestic law and Article 5 of the relevant double tax treaty.  In a digital world this is where it starts to get muddy.  One of the largest value drivers for the digital economy is revenue created through the exploitation of millions of digital users or consumers across the globe, many in jurisdictions where the company providing the digital services has no physical presence.  This leads to so called "super profits" which are not aligned to the typical activities considered for reward under a traditional transfer pricing analyses.

Pillar 1 proposes an approach that seeks to ascertain these super profits and allocate them to the jurisdictions which have the users responsible for their generation.  Thus, Pillar 1 seeks to expand the taxing rights of these user or "market jurisdictions" and allocate a portion of the super profits.  

The approach identifies three ways of achieving this:

• [A] - A share of the residual profit generated through the exploitation of the digital business using a formula.  This would apply irrespective of whether a physical presence exists in that jurisdiction and would be determined based on the participation of users in the digital business.
• [B] - A fixed allocation of profit to the market jurisdiction using the arm's length principle to arrive at a baseline allocation according to the distribution and marketing functions which are undertaken in that jurisdiction.
• [C] - Additional profit over and above a baseline allocation where the activities undertaken exceed that determined under the fixed allocation method and an effective dispute resolution mechanism to address any risk of double taxation.

B and C above involve allocating a portion of the super profit to an existing entity in the market jurisdiction and therefore applies transfer pricing principles we are familiar with.  The key change is the approach envisaged under [A] which seeks to create a completely new taxing right not linked in any way to a physical presence in the taxing jurisdiction and potentially allocates a portion of profits to an entity present in that jurisdiction, over and above the arm's length amount.

The inclusive framework agrees that this method would only apply to large multinational groups and would require the creation of a new nexus test.

Which businesses could be affected by this change?

Whilst there is no doubt the Inclusive Framework intends these new rules to have a wide application, the key target is the digital economy.  Businesses which supply global digital goods and services to a large global customer base will be impacted. These will typically include businesses providing remote services which generate substantial value without the need for local infrastructure. Examples being Facebook, Amazon, Apple, Microsoft, and Google (Alphabet) which have a combined market capitalisation of over US$4 trillion.

Businesses listed by the OECD potentially subject to [A] include:

• Businesses which generate revenue from the provision of automated digital services provided on a standardised basis to a large population across multiple jurisdictions, e.g. social media platforms, digital content streaming etc.; and
• Consumer facing businesses which generate revenue from the sale of goods and services of a type commonly sold to consumers, e.g. personal computing products, branded foods and refreshments, luxury goods and franchise models.

Shipping, airline and extractive industries and sellers of commodities are excluded as are sales with commercial customers within the financial services sector, including insurance.  Retail transactions within the financial sector may also be excluded where the industry is regulated.

In defining what is a business with a large customer base, consideration is being given to imposing a threshold aligned to the current country by country reporting threshold applied to multinationals for transfer pricing reporting (ZAR10bn / EUR750m).  Additional carve-outs and thresholds are also being considered.

How will this be implemented?

In order to apply [A], a new nexus rule will be required to create a taxing right in the market jurisdiction.  It is proposed that this new rule will be designed to limit filing requirements in each of the market jurisdictions which would be cumbersome, possibly restricting the filing to the ultimate parent jurisdiction.

Determining the arm's length amounts under [B] and [C]

[B] is intended to reward the marketing entities within the group with a fixed return for baseline marketing and distribution activities.  To simplify the approach, a fixed distribution return is proposed to reward the distributor for its baseline marketing and distribution activities.  The definition of what constitutes baseline marketing and distribution activities is yet to be determined, as is the fixed amount which will inevitably change based on functions, markets and industries.  We could be lead to assume that baseline equates to routine when considering functionality, however this still needs to be finalised.  The fixed amount will also still need to adhere to the arm's length principle in order to ensure the existing treaty measures to prevent double taxation can be relied on.  What is clear is that there is still considerable work required to arrive at consensus on how [B] will be determined.
[C] represents additional profit to be allocated to the marketing and distribution entity where its activities exceed the baseline activities.  How this excess will be determined remains unclear and as [C] also focusses on the need for improved dispute resolution, we can anticipate that this amount will be subject to disputes.

How is the amount under [A] determined?

This is where the proposals deviate from traditional understanding of transfer pricing and specifically the arm's length test.  The approach envisages an amount allocated over and above that which would be allocated under a traditional transfer pricing analysis to a MultiNational Enterprise group member which has a physical presence in a jurisdiction.  This approach provides for a completely different basis of profit allocation based on a globally agreed formula.

It is proposed that the consolidated group accounts will be used to derive that portion of the profit to be allocated to the market jurisdiction.  Adjustments will be needed to account for differences in accounting periods across jurisdictions.  The profit line to be allocated will likely be the profit before tax.  The rules will apply equally to profits and losses as in any profit split analysis. 
A formula will be used to determine the portion of the residual profit to be allocated to the market jurisdiction and will be allocated to each market jurisdiction according to an allocation key.  The preferred allocation key will be sales generated in each market jurisdiction.

To align this approach with existing transfer pricing rules, [A] will operate as an overlay to the existing profit determined under the arm's length principle.  This should seek to eliminate the risk of double taxation, at least in theory.

To avoid risk of double counting [A] is determined through a three step system.  This means the amount to be rewarded to the entities with a physical presence within the multinational group would first be established using normal transfer pricing principles, including any adjustments made for [B] and [C].  The amount [A] would then be determined and allocated to the market jurisdictions, reducing the profit already allocated to those entities within the multinational group under the first step.

[A] seeks to allocate a portion of the residual profit (over and above the fixed return) to the market jurisdictions, and consequently there should be no significant overlap.  However as indicated above, [C] is designed to reward an additional portion of the residual profit to the marketing entities in the jurisdiction so there could conceivably be a significant overlap in allocating a portion of this residual profit to established marketing and distribution entities under [C] as well as to the market jurisdictions under [A].  Additional measures are therefore needed to resolve this.

Managing the risk of double taxation

The OECD/G20 Inclusive Framework recognises that the proposals run the risk of increased incidence of double taxation and the need for robust measures to address this.  The inevitable increase in the number of cases being referred to the Mutual Agreement Procedure is recognised and the recommendation is that a clear, administrable and binding process is developed to prevent disputes.  In addition consideration is being given to developing a mandatory binding dispute mechanism.

The reality of implementation

The implementation of the proposals will not only require changes to domestic rules, but also significant changes to bilateral tax treaties.  One mechanism to address this could be yet another Multilateral Instrument (MLI).
Any implementation will also depend not only on agreeing many issues still unresolved under Pillar 1, but also getting to a consensus for the implementation.  The statement lists eleven work streams which need to address the many outstanding issues associated with the implementation.  The aim of these work streams is to reach consensus on the outstanding items and finalise the proposals under Pillar 1.  The timeline for this is the end of the year.  To say the program is ambitious in its timing is an understatement.  In reality this approach could end up being a theoretical approach which gets limited traction in practice, hindered by a sluggish response to changes in domestic law (some of which have already been unilaterally devised to address the digital economy); and a lack of appetite to enter into yet another MLI.