Skip to main content

简报

Tax 2.0 - Redesigning international tax for multinationals

Redesigning the international tax system for an increasingly digitalised world. 

On 31 May, the OECD published a “Programme of Work” designed to help achieve consensus on a solution to taxing an increasingly digitalised economy. Unlike the consultation document published in February 2019 (as to which, see our previous briefing), this is a consensus document – approved by the 129 members of the Inclusive Framework in recognition of (i) issues in the existing international framework which some feel have still not been resolved by the BEPS package of measures; and (ii) the need to limit the proliferation of uncoordinated and unilateral measures through which some jurisdictions are seeking to claim a bigger slice of the tax pie.

The concepts that are now being worked on represent a radical, even startling reshaping of the architecture of international tax for all multinational enterprises, not just the digital giants.

Despite not yet having many of the answers, the work programme is well-considered and carefully articulated. As with the February consultation, the latest proposals still rest on two “pillars” but through a subtle sleight of hand the pillars have now been recast.

Pillar one reformulated

The first pillar, whilst still seeking to re-allocate taxing rights to the jurisdiction of users and/or consumers (i.e. the “market jurisdiction”), has moved away from the three proposals set out in the consultation paper (around user participation, marketing intangibles and significant economic presence) and instead the question is formulated differently. It is now divided into three different blocks, which consider:

  1. how to quantify the profits that will be subject to the new taxing rights;
  2. designing a new nexus rule not constrained by physical presence; and
  3. how this “new taxing right” will be implemented and administered.

On quantification i.e. the amount of profit to be reallocated to the market jurisdiction, various options are proposed: a modified residual profit split or “MRPS” (splitting out routine profit, potentially using existing transfer pricing rules, and then allocating non-routine profits between relevant market jurisdictions – this will require the development of an allocation mechanism or “key”), fractional apportionment (determining profit, creating an allocation key to apportion it, and applying this in conjunction with existing rules in such a way that does not give rise to double taxation, or double non taxation), as well as considering whether there are other methods which might be simpler to apply.

These changes will require a departure from the arm’s length principle. Even the MRPS option contemplates “proxies” for finding the non-routine profit that will be subject to the new taxing right. It is significant that the fractional apportionment method is being considered as an option here, given the similarities to global formulary apportionment: an idea favoured by reformers and academics (and the EU in the Common Consolidated Corporate Tax Base) but long resisted by others as arbitrary and impractical. The work programme acknowledges that profits of a multinational can vary across business lines or regions and therefore one stream of work will consider whether it makes sense to quantify in-scope profits separately for different business lines/regions. Helpfully, time will also be spent exploring what the limits of the rules should be (such as safe harbours, or de minimis thresholds) as well as how to handle losses.

On nexus, there is little clue given as to what the new rule will look like, but given many jurisdictions expressed reservations about the three options in the February consultation document, and in particular the user contribution approach, it is perhaps unsurprising that the programme of work has distanced itself from specific options at this stage in an attempt to produce something all nations can agree on.

It is clear that there will be a new nexus rule not dependent on physical presence which will seek to tackle situations where a multinational has “remote but sustained and significant involvement in the economy of a market jurisdiction” (emphasis added). This may involve amendments to the existing treaty definition of PEs, or alternatively the introduction of a “standalone provision” (query what the legal basis for this would be) either through “a new taxable presence or a concept of source”. This in itself will be no mean feat, but also begs the question of how this would then interact with existing provisions.

The final building block in this re-constructed pillar is implementation of the new taxing right: how can double taxation be prevented; how will the new taxing rights be enforced and collected (where the taxpayer is not resident and has no physical presence in the taxing jurisdiction); how would taxing authorities (and taxpayers) get the information needed to administer (and comply) with the new rules?

The programme of work will also examine potential enhancements to dispute resolution procedures, including arbitration and multilateral competent authority agreements, as well as considering whether multilateral coordinated risk assessments could assist in applying the new rules under this pillar, potentially drawing on the International Compliance Assurance Programme (ICAP).

Pillar two recast

Pillar two, perhaps in recognition of the power of a catchy name in helping an idea gain traction (as happened with “BEPS”), has had a bit of clever branding applied to it. The newly styled “GloBE” proposal (as in Global anti-Base Erosion), as with the previous iteration of pillar two, still seeks to address remaining BEPS risks or profit shifting to entities that are subject to no or low taxation. The work programme explains that a key aim of this pillar is to stop countries introducing tax incentives to attract more tax base in a “harmful race to the bottom”. The aim is to create a level playing field across jurisdictions and, in particular, to shield developing countries from pressure to offer such incentives in an attempt to attract investment.

The rhetoric around the need for change here is politically charged and, at points, apocalyptic, with reference to “adverse consequences for all countries, large and small, developed and developing as well as taxpayers” if multilateral action is not taken, further noting that uncoordinated unilateral action risks “effectively undermining the tax sovereignty of nations and their elected legislators”.

As before, there are two aspects to this pillar:

  1. an income inclusion rule and
  2. a tax on base eroding payments.

The income inclusion rule would subject a multinational to tax on its global income at a minimum rate, with the aim of reducing incentives to shift profits to low tax jurisdictions. The basic idea is that where income is not taxed at least at the minimum level, the income inclusion rule would operate as a “top-up” to achieve the minimum rate of tax. The work programme indicates that the GloBE proposal could supplement CFC rules and that there is an intention for this to apply to foreign branches, potentially via a “switch-over” rule.

Another point that will no doubt be highly controversial is who will get that “top-up” payment – will it be the parent/shareholder jurisdiction potentially meaning that developing jurisdictions will only get a small share of the benefits of these new rules?

The agreement to explore such a measure in a work programme approved by 129 jurisdictions is without doubt dramatic. The outcome could fundamentally change the taxation of most or even all multinationals across the world, whether operating primarily in the digital space or not.

Whilst the exploration of the pillar two proposal is expressed to be on a “without prejudice basis”, the minimum tax and tax on base eroding payments are described as inter-related measures, rather than alternatives – the implication being that if the GloBE proposal is to go ahead, then the minimum tax would form part of it. And if the GloBE proposal does not go ahead, can the new construct survive balanced precariously on a single pillar?

There have been several calls for the development of a minimum tax in recent months, with France and Germany being key advocates of this proposal, and US Treasury Secretary Steven Mnuchin endorsing the idea (perhaps unsurprising given the existing US GILTI rules mean that US multinationals already have to grapple with a form of minimum tax). However, some jurisdictions see this as a challenge to their sovereignty over tax matters. This aspect of the GloBE proposals may take some of the pressure off “digital” tax issues, but only by broadening the canvas in a quite controversial way.

Even putting political factors to one side, from a design perspective, such a seemingly simple concept as a minimum tax rate has its own complexities – what rate should be used: a fixed percentage across all jurisdictions or a tax tied to the parent jurisdiction of the multinational (as is done in many CFC rules)? The work programme advocates the use of a fixed rate in the interests of simplicity and fairness, but will there be consensus on an appropriate rate?

The work programme suggests that the tax base will be calculated based on the rules of the shareholder jurisdiction. Could there be inherent unfairness and inequalities in the way different jurisdictions calculate their tax base? How will this work with complex ownership structures? How does this interact with other international tax rules: withholding taxes, CFC rules, not to mention transfer pricing and possible new pillar one rules?

In relation to the tax on base eroding payments, the work programme will explore both an undertaxed payments rule (denying deductions or imposing source-based taxation – including by way of withholding taxes) and a subject to tax rule (where treaty benefits may be denied) in each case if certain payments/income are not subject to tax at a minimum rate. Politically, there may be differences of opinion on when something should be considered to be “undertaxed” and what factors should be taken into account in making that determination.

There is acknowledgement that work will need to be done on co-ordinating these rules with existing rules, the interaction with EU fundamental freedoms (the Commission is keen to present a unified position on digital taxation at the G20 meeting in Japan this month) and other existing international obligations, as well as potential carve outs and simplifications to ensure that the compliance burden is not too great.

What happens next?

The work programme does not stop at the two pillars. Another, crucial, strand is analysing the economic impact of the proposals and assessing the revenue and behavioural implications they will produce. The final section of the work programme sets out specific working groups within the OECD that have been tasked with different aspects of the programme and this may be coupled with further public consultations.

The aim is still to reach a global consensus-based solution by the end of 2020, with the programme of work noting that this will require the outlines of the architecture to be agreed by January 2020, including which option is to be pursued under pillar one and the appropriate interaction between the two pillars.

Although there is significant emphasis throughout the document on the need for simplification, finding a solution that all Inclusive Framework members can agree on, negotiating the interaction with existing rules and determining how the new rules will be implemented and administered will be anything but simple. There is a great deal of work still to be done here but this is an organised and considered plan of attack. Notwithstanding this, we could be in for a bumpy ride.