Clifford Chance

There is now immense pressure to reform the way multinationals are taxed. The status quo – unchanged in essence for a hundred years – is widely seen by the public and policymakers as inadequate for the modern world and the digital economy. Last year, the OECD proposed a radical proposal to reshape the international tax system, creating new rights for countries to tax multinationals.

We look at the current state of the OECD proposals and assess both their potential impact, and the likelihood they will be adopted. If the OECD process fails, many countries are likely to adopt unilateral measures and, with US companies the obvious targets, there is a real risk of retaliation by the US – even a trade war. Sadly that is in our view now a very plausible outcome.

Back in 2012, the G20 instructed the OECD to begin work on proposals to counter the arrangements that many multinationals were accused of using to artificially erode the taxable base in their customers’ jurisdictions and shift profits to tax havens. This became the Base Erosion and Profit Shifting Project (BEPS), with fifteen detailed proposals published in November 2015.

However many people, including apparently the OECD itself, regard BEPS as a failure. In part that is impatience: much of BEPS is only now being implemented, five years later. In part it’s disappointment that billions of new taxable revenue was not generated – probably because the scale of the avoidance was less than some newspaper headlines suggested. But ultimately there is a feeling in some quarters that BEPS was aimed at the wrong target. Plugging loopholes is not enough.

That led to the new OECD programme. It’s sometimes termed “BEPS 2.0”, but that gives a misleading impression that this is a mere evolution of the original BEPS proposals. That is very much not the case.

The programme of work

The OECD published a programme of work in June 2019 to find a long-term and consensus-based solution to the “tax challenges of the digitalising economy”. In its latest statement on 31 January 2020, the OECD re-affirmed its commitment to do so by end 2020 – which is an extremely ambitious timeframe, given that “critical policy differences” and “significant divergences” will have to be overcome.

Whilst the EU, UK and others have been proposing specific taxes on particular digital businesses, the OECD are proposing something much more radical, potentially the most significant change since the norms of international taxation were set almost a century ago.

The programme builds on “two pillars” of proposals, with the “key policy features” to be agreed by the 137 participating members in July 2020. There are now 11 technical workstreams, all interlinked and involving multidisciplinary teams. Working parties, task forces and reporting groups are striving to find consensus on all measures.

The trouble, of course, is that while the economics are difficult, the politics are nigh impossible.

Crucially we still do not have an economic analysis or impact assessment of the proposals.The latest statement alludes to data limitations and suggests that analysis of the investment and growth impacts will wait until March 2020.

This is all very uncontroversial if the answer is simply that digital and other businesses pay more tax everywhere. However the result is unlikely to be so straightforward. Two key questions are:

• What are the expected effects of the proposals on the level and distribution of tax revenues across jurisdictions? The almost inevitable consequence of the proposals is a redistribution of taxing rights from the home of large corporations (particularly the US) to the “market” jurisdictions where they make their sales (the rest of the world). It is not obvious why the US would agree to this. 
• More specifically: what economic impact will the various proposals have for different types of multi-national enterprises, sectors and economies (e.g. developing countries; resource-rich countries, etc.)? Many of the non-OECD countries involved in the process, and many NGOs, want to see a fundamental redistribution of taxing rights away from the developed world and towards the developing world. They, not unreasonably, see the BEPS focus on taxing “where the value is created” as in practice allocating taxing rights to rich countries. However this creates some difficulty for developed-world policymakers, whose populaces expect OECD initiatives to result in more tax being paid to their treasuries, not less. 

Without this analysis, it does not seem plausible for countries to give the “political steer” that the OECD requires to agree the key policy features by July 2020. How can countries agree with the policy before they understand its distributional effect, or how it impacts the relevant sectors of their economy?

The work will proceed in earnest but, as mentioned in the programme, the available data may not permit an analysis of the impact on particular sectors, industries or business models. The concepts are too novel, the methods untested, and the scope encompasses the whole economy.

Amid this uncertainty, the US has called for “safe harbour” protections to be considered. It has been reported that this would make the new regime voluntary, by allowing multinational companies to choose whether they will elect into the new regime. Naturally this is strongly resisted by the OECD and most of its members. However, the latest OECD statement confirms that this alternative will be considered in depth, and that a final decision on the matter will be taken only after the other elements of the proposal have been agreed upon.

Hopes rest on the reality that, without a long-term consensus, more discordant unilateral measures, such as the UK’s digital services tax, will proliferate. This may create the impetus for compromise. As the OECD warn publicly, their proposal is Plan A, there is no Plan B, and Plan C is chaos – the proliferation of unilateral tax measures and trade barriers.

But it is hard to see how the major actors, particularly in Europe and the US, can align by the middle of 2020 on core elements of proposal, while the economics remain uncertain. Any agreement to proceed may, therefore, be highly provisional, with the “real” political agreement having to be sought much later in the process, when the economic consequences of the proposals are more clear.

The work programme can be broadly summarised as follows:

Pillar One: Revised Nexus and profit allocation rules

Pillar One considers the reallocation of taxing rights. The aim is to allocate more profit to the jurisdiction where customers and users are located, known as the “market jurisdiction”. Previous proposals from the OECD, EU, UK and others focused on “user participation”, “marketing intangibles” and “significant economic presence” ideas. Those are now harmonised in a single model, called the “Unified Approach”.

The Unified Approach identifies three amounts of taxable profit that should be allocated to the market jurisdiction – referred to as Amounts A, B and C: 1.

• Amount A is a share of “non-routine” profit, to be allocated to the market jurisdiction using a formulaic approach at a group or business line level. There are four technical steps: (i) determine total profit; (ii) remove “routine profit”; (iii) determine what is in scope of the taxing right; and (iv) apply criteria (an “allocation key”) to split the profit between the entities or business lines, for example by reference to the relative value of their contribution. This measure would apply to “automated digital services” (e.g. search engines, social media platforms and online advertising services) and also “consumer-facing businesses” (including online retailers, mobile phone providers and the hotel sector). The boundaries of these categories are still unclear. A series of monetary thresholds must be satisfied for the measure to apply (with respect to global turnover, in-scope revenues, business line profitability and residual profit). The residual profit threshold may – counter-intuitively – exclude some of the digital giants that the proposal may want to target, if they have high revenue but low margins. Extractive industries and the financial services sector are likely to be exempted entirely, although the scope and practicalities of their exemption have not been decided. 
• Amount B is a fixed remuneration for distribution and marketing functions that would be allocated to the market jurisdiction, based on the arm’s length principle. It is logical, though awkward, that Amounts B and C are in fact applied before Amount A, to determine the initial allocation of profit between different jurisdictions, so that Amount A is then allocated as an overlay or partial override.
• Amount C covers any additional profit if in-country functions exceed the baseline activity compensated under Amount B.

There is also a workstream to design the “nexus rule” that would determine when a market jurisdiction has a taxing right over Amounts A to C. This includes, for example, a deemed permanent establishment if the group exhibits “a significant and sustained engagement” in the market jurisdiction. To avoid interpretive debates, this is likely to be determined by reference to a revenue threshold (subject to additional “plus” factors for consumer-facing businesses, such as the existence of a physical presence). Deciphering the country revenue in this context looks practically difficult – for example, revenues from online advertising services would be deemed to be sourced in the country where users (“eyeballs”) viewing the advertisements are physically located. That requires sophisticated tracking of users’ activities and could raise difficult privacy issues.

There is also an important technical point here. Changes to profit allocation rules can be made through consensus, by changing OECD guidelines. Individual countries may take different approaches and, whilst double taxation in some cases may arise, those different approaches should not prevent Pillar One from being more or less effective. However, changes to the nexus rule require changes to double taxation treaties. Otherwise any new domestic nexus rule introduced by, for example, France, would be overridden by the US/France double taxation treaty when a US company makes a digital sale to France.

That likely means another “multilateral instrument”, amending multiple tax treaties simultaneously. But more importantly, it means that corporations based in countries that do not ratify the multilateral instrument will escape the new nexus rule. Ratification delays are very common (one of the main reasons why the original BEPS proposals have taken so long to come into force). If the US doesn’t ratify the treaty implementing Pillar One, how effective will it be?

Pillar Two – Global antiBase erosion proposal

Now branded as GLOBE, the second pillar proposes to introduce a global variation of the US anti-abuse regime known as GILTI, introduced as part of the Tax Cuts and Jobs Act of 2017.

This would include an “income inclusion rule”, to ensure that income of a group is subject to tax at a minimum rate. The OECD prefers a fixed global rate as the simplest option. The proposal is seductively simple, but there is still considerable disagreement about its design, including the tax base, the rate, any carve-outs and co-ordination with other rules. It has been open for discussion for over a year now, without any sign of a growing consensus.

The other key aspect is an “undertaxed payments rule” that would deny a deduction for payments to related parties if the payment is not subject to the minimum rate in the recipient country (or in the country of the recipient’s parent under controlled foreign company rules). This could be complemented with a “subject to tax rule” by subjecting the payment to withholding or other taxes at source and denying treaty benefits. There are number of issues to be explored, including the benefits of a withholding tax over a deduction denial approach.

It will be noted that GILTI is a highly complex set of rules which, two years after its enactment, is still not completely understood by many affected businesses. In part this is because of the pre-existing complexity of the US controlled foreign company rules, which GILTI supplements but does not replace. GLOBE can, therefore, be considerably more streamlined than GILTI. However it will still be a complex set of rules, which poses a challenge in terms of both achieving consensus and achieving a ruleset that can be realistically applied by developing world tax authorities with limited resources.

The OECD’s statement includes a surprising acknowledgement that the actual rate to be applied under the GLOBE proposal has not yet been discussed. This could be due to ongoing disagreement about the basic policy objective. The statement suggests that some countries would prefer to focus on the design of the existing controlled foreign company rules, rather than invent a new systematic solution to ensure all businesses pay a minimum level of tax.

There is no specific timeframe to agree the policy design of Pillar Two, although the general expectation is that it should be easier than Pillar One. Any change to withholding tax rules in Pillar Two will require tax treaty amendments, raising the same practical difficulties identified above.

The road ahead

The OECD will progress each workstream, so that a recommendation on the core policy elements can be tabled for agreement at the beginning of July 2020. Some economic analysis will be prepared before the end of March 2020 to inform this decision, however as noted above, it is not clear how sophisticated or complete this will be.

In our assessment, due to the immense political and economic difficulties, we are likely to see one of the following outcomes prevail in the next few years:

• The OECD talks break down without agreement – resulting in the proliferation of digital services taxes and potential trade disputes; 
• The OECD proposal is generally agreed, but the US does not agree, or the US Senate does not ratify the new treaties. US corporations are then only partially affected, because they are not bound by the new treaties. Digital services taxes may still proliferate, with potential trade disputes; 
• The US opts out of the proposal but the other countries override the US treaties (or argue that the new rules are compatible with those treaties, because this is a new tax that is not covered by them). This may remove digital services taxes but not defuse the trade disputes; 
• The OECD proposal is agreed, including by the US, but it takes at least three or four years to draft, sign and ratify the new treaties (as was the case with the BEPS multilateral instrument). Digital services taxes and trade disputes may persist in that interim period; 
• The OECD proposal is agreed and ratified quickly, but it becomes clear that it has limited impact and digital services taxes still proliferate, with related trade disputes.

In addition to retaliatory tariffs, we might also see attempts to challenge the legitimacy of digital services taxes, which are seen to target US multinationals. Currently the countries implementing digital services taxes (and the EU) believe that their taxes can circumvent the limitations of tax treaties and not constitute a turnover tax in the nature of VAT.

The economic and political problem aside, it strikes us that there has been scant consideration of the legal drafting that will be vital to its agreement and implementation. There is, for example, reference to amending Articles 5 (Permanent Establishment) and 7 (Business Profits) of the OECD Model Convention, maybe also Article 9 (Associated Enterprises), perhaps in a supplemental or new multilateral treaty.

These will be hugely significant legal changes that require lengthy consideration. As and when the programme moves on from abstract theory to hard implementation, businesses will need to carefully assess the legal effect of the proposals.

For now, digital businesses will likely want to stay actively involved in the development of the proposals; others should at least maintain a watching brief. Those international businesses which have not yet familiarised themselves with GILTI may wish to do so, and to start a high level assessment of the likely impact on their group.