Skip to main content


Going GloBal: the OECD’s consultation on pillar two

After the excitement surrounding the OECD’s pillar one consultation, some might view the pillar two consultation as something of an anti-climax. Unlike pillar one, the launch was not heralded by a fanfare of announcements and webcasts but was quietly released on 8 November. However, one should not underestimate the significance of this document which could have a fundamental impact on the way all multinationals are taxed, including the possible introduction of a global minimum rate of tax. The issues raised involve huge politico-economic questions for the 135 members of the Inclusive Framework. However, those looking for answers and a clear mapping of the road ahead may, for the moment, be disappointed.

The story so far

In January 2019, the Inclusive Framework issued a policy note on Addressing the tax challenges of the digitalisation of the economy, under which they proposed that the issues should be addressed under two pillars. Pillar one looks to a new allocation of taxing rights through new nexus and profit allocation rules and the OECD consultation on this was published on 9 October 2019, together with an OECD Secretariat proposal for a unified approach (as to which see our previous article). The period for comments on this closed on 12 November, with a public consultation held on 21 and 22 November, so there may well be more on this soon.

Pillar two, also known as the global anti-base erosion (or GloBE) proposal, is designed to address remaining BEPS (base erosion and profit shifting) risks that arise around companies shifting profits to jurisdictions where those profits are subject to no or low taxation. This was developed in the May 2019 programme of work which suggested that the GloBE proposal would operate as a top-up to an agreed fixed rate of tax.

The proposal is designed to impact the behaviour of both taxpayers and tax administrations. For taxpayers, by introducing a global minimum rate of tax, the incentives to shift profits to no or low tax jurisdictions are to be taken away. Although the programme of work noted that profit shifting was ‘particularly acute’ in the context of intangibles which are prevalent in the digital economy, this initiative is not limited to digital businesses and could potentially impact all types of multinationals (MNEs) across the world.

For governments, the proposals are said to prevent a ‘race to the bottom’. The concern appears to be that governments might seek to encourage investment through tax incentives at the cost of public spending on infrastructure and public services, or perhaps shifting taxes to fund such public spending onto workers and consumers. The programme of work claims that this measure is needed to protect tax sovereignty, although those jurisdictions that do not support a minimum tax rate might argue it has the opposite effect.

The pillar two consultation: the unanswered questions

The OECD published its latest consultation document on pillar two on 8 November (see Whilst the pillar one consultation focused on a unified approach to the big issues to help reach consensus with the details to be fleshed out later, the pillar two consultation almost takes the opposite approach, delving straight into technical design elements, with the big picture points to be decided later.

Components and priority rules

The GloBE proposal is split into four components, namely:

  1. an income inclusion rule;
  2. an undertaxed payments rule;
  3. a switch-over rule (‘switching off’ the exemption method for double taxation relief in favour of the credit method); and
  4. a subject to tax rule,

as illustrated by the OECD diagram reproduced below.

The subject to tax rule complements the undertaxed payments rule and, in a sense, the switch-over rule is a variant upon the income inclusion rule.

One of the most fundamental questions is how these four components will interact. Proper coordination between them will be crucial to ensure taxpayers are not subject to double or multiple taxation. Similarly, the interaction between the proposed rules and existing international and domestic rules will need to be carefully crafted to prevent multiple layers of taxation. Although the consultation notes that ‘progress has been made’ with respect to these issues and that Inclusive Framework members are ‘actively engaged and identifying design solutions to address these challenges’, the consultation document is conspicuously silent on the direction of travel.

Obviously the four components are not intended to apply at the same time, but which rule will take priority? It has previously been indicated that, conceptually, the scheme is intended to work in a similar way to the hybrid mismatch rules, i.e. if the income inclusion rule applies, then there is no need to apply the other rules but, if not, then you may need to deny deductions or apply withholding tax. What is less clear is which rule would be the primary rule here? Back in July, Achim Pross of the OECD indicated that, at that time, most countries seemed to think that the income inclusion rule should apply first. Although the consultation document does not explicitly comment on this, the focus of the document is on big questions that primarily relate to the income inclusion rule. The document acknowledges that further work needs to be done on the undertaxed payment rule and the subject to tax rule, and there may be a further consultation on these.

Minimum rate

An enormous elephant in the room is the level of the minimum rate that would apply under the GloBE proposal. This is one of the most important questions to be answered with respect to the workability of the pillar two proposals. Those hoping for an answer to this will be disappointed. The consultation document notes that the rate will be discussed once other key design elements are fully developed. In all likelihood, this is a deliberate move to keep as many countries round the table for as long as possible, for the level of the minimum rate could be the deal breaker for many jurisdictions. The lower the rate it is, the less it will achieve its objective of changing taxpayer and government behaviours. The higher it is, the harder it will be to convince some Inclusive Framework members to agree to it. While the consultation document does not officially give any indication of the rates that are being discussed, the use of 15% in the examples in the annex (despite being ‘simply for illustrative purposes only’) perhaps gives an unofficial hint at the level being contemplated. Based on the latest OECD statistics on effective tax rates (from 2017, see 19 of the 75 countries listed (i.e. just over a quarter) had effective tax rates below 15%, which gives some insight as to how challenging it will be to get consensus on this.

Technical aspects

The three technical aspects that the consultation document does explore are:

  • determining the tax base;
  • blending; and
  • carve-outs and thresholds.

Determining the tax base

Once the minimum rate of tax is settled upon, one of the remaining fundamental questions is determining the tax base. The programme of work had initially suggested that the tax base would be determined by reference to CFC rules (i.e. in line with the tax base of the parent jurisdiction). Sensibly, the latest consultation document acknowledges the inherent inequalities that could arise from such an approach, accepting that differences in the design of different jurisdictions’ tax bases could result in inconsistencies in the application of the rules in circumstances where the policy concerns do not justify that result.

Instead, the consultation document advocates using financial accounts to establish the tax base, subject to adjustments needed to align accounting income with tax income. Whilst it would be simpler from a compliance perspective to compute the tax base using the accounting standards already used by the various entities in the group, this could again result in distortions based on the use of different local accounting standards, which would move away from the level playing field that the OECD is trying to achieve. It would also be harder for tax administrations to monitor compliance where the tax base is determined according to local accounting standards in another jurisdiction. The compromise suggested is to use the accounting standard applicable to the ultimate parent, provided that it uses an acceptable set of accounting standards.

Whilst a single accounting standard, such as IFRS, being mandatory across the board would help eliminate inconsistencies, the OECD is mindful of the need to balance this against imposing too great a compliance burden on taxpayers. These issues will be familiar to those who have been following the EU’s CCCTB proposals. However, by limiting the number of acceptable accounting standards to IFRS and a handful of local GAAP (US and Japanese GAAP are mentioned), it is hoped that the distortive effects could be minimised. There is also a broader concern as to whether having tax rules that rely on the accounts could lead to the accounting profits being manipulated to minimise the tax base artificially. However, it is hard to conceive of another method of levelling the playing field that does not give rise to its own issues. Limiting the number of accounting standards in play may at least help to ease some of these concerns.

Assuming that the starting point for the tax base is the financial accounts, there are then further questions around what adjustments should be made to bring the accounting income more in line with taxable income. Some jurisdictions that rely closely on the accounts for their tax base may not require much adjustment, whereas for others, relying on the accounts figures could mean that there is a significant overstatement or understatement compared with the conventional basis of taxation. The paper looks at how to tackle both permanent differences between tax income and accounting income (for example, where there are differences in the types of income or expense allowed in calculating income for accounting and tax purposes) as well as temporary differences (i.e. those which will reverse out over time).

Tackling permanent differences would require the exclusion of certain categories of income or expense from the financial accounts. The consultation document is open to suggestions here, noting that the required approach will depend on the position reached on other points, such as blending (see below). For instance, in tax calculations, dividends received from foreign corporations may be excluded from taxable income (e.g. wholly or partially exempted under a participation exemption or similar) to prevent double taxation at the payer and payee level. There may therefore need to be some adjustment to the tax base for the company receiving the dividend to bring the accounting income into line with the tax income. However, where a worldwide blending approach is used based on consolidated accounts, this should eliminate the problem (by eliminating intra-group flows) without need for a specific rule to address this particular issue.

Temporary differences are perhaps more commonly encountered, for instance differences in depreciation methods used for tax and accounting purposes. Three methods are proposed for dealing with these:

  • allowing carry-forward of excess taxes and tax attributes;
  • using deferred tax accounting; or
  • averaging the effective tax rate over a number of years.

The document indicates that the end solution may involve a combination of elements of all three. To the extent that a temporary difference results in an entity being taxed below the minimum tax rate in one or more years but exceeding it in other year(s), such that it would lead to a bumpy GloBE tax profile, each of these methods operates to smooth out those bumps in one way or another. But, once one begins to peer into the detail, significant complexity is apparent and there are clearly challenging issues with each of the approaches. The carry-forward method can be most easily targeted at the specific differences that are intended to be eliminated, whilst the averaging approach is the most conceptually simple, but also the bluntest. For those already using deferred tax accounting (itself a type of smoothing mechanism), this looks like it might be the least burdensome in terms of additional compliance (though there are knotty details here too).

Policy decisions will need to be made for all methods; for instance, the consultation indicates that items with long deferral periods, or where the temporary difference does not reverse out until the business is sold, may not be appropriate for the GloBE proposal. Similarly, changes in tax rates could potentially cause headaches under each approach. What will happen when an entity moves out of or into a group? To what extent should GloBE tax attributes move with it? Will the rules be fluid enough to cope with changes to accounting standards that may require new departures from the accounts to accord with tax policy?


Blending rules control the extent to which MNEs can combine low-tax and high-tax income from different sources in determining their effective tax rate. (Remember the old-fashioned dividend tax credit ‘mixer’ technique?) The consultation considers three options for blending:

  • entity blending (taking each entity separately);
  • jurisdiction blending (the MNE would aggregate foreign income apportioned to that jurisdiction and be subject to the top-up tax to the extent the total tax on this was below the minimum rate); and
  • worldwide blending (taking the aggregate total foreign income and total foreign tax – perhaps also resonating with the EU’s CCCTB proposals?).

The broader the approach to blending, the lower the MNE’s potential GloBE tax liability. For MNEs, a worldwide blending approach may reduce compliance costs, e.g. if it can be based on consolidated financial statements that have already been prepared for accounting purposes (although further steps may be needed to separate out foreign from domestic tax and income). Such an approach also alleviates a number of other tricky design issues which might otherwise need to be individually addressed. For instance, there is no need to introduce rules in relation to intra-group transactions such as dividends as these should automatically be disregarded in the consolidated accounts. It also seems to provide a simple solution to dealing with taxes imposed in other jurisdictions (e.g. under CFC rules). Under worldwide blending, tax paid on foreign income could be creditable without having to determine whether that tax was paid at branch level, head office level or under CFC rules of a third jurisdiction. It also indirectly helps address some of the points raised above in terms of temporary differences between tax and accounting. Although not targeted at this, in practice temporary differences in group entities may offset each other, such that their impact is minimised and there is potentially less need to introduce specific measures to address these, removing a layer of complexity.

On the other hand, the broader the approach to blending, the less effective the proposals will be in achieving their aims of changing taxpayer and governmental behaviours. However, entity and jurisdictional blending have their own challenges. For instance, if entity blending does not factor in group reliefs, then the tax calculated for GloBE purposes on an individual entity basis would not reflect the entity’s actual tax liabilities. Both a jurisdictional and entity basis for blending would mean preparing further (unconsolidated) accounts to show income broken down by jurisdiction or entity. Furthermore, in dealing with CFC-type charges, there would potentially need to be a ‘credit-transfer’ mechanism to ensure the effective tax rate is not understated for the jurisdiction where the income arises, and not over-stated in the jurisdiction where the tax is paid under the CFC rule. Likewise, specific rules would be needed to deal with dividends (and other distributions) where such amounts are treated as income for the recipient for accounting purposes, but exempt for tax purposes, to prevent the recipient over-stating its effective tax rate. One suggestion is that dividends could be excluded from income under a jurisdictional or entity approach on the basis that the income has already been taxed when earned, but then corresponding provisions would also be needed to exclude withholding taxes on such dividends. Further questions also arise as to whether to treat withholding tax paid as additional tax on earnings in the entity paying the dividend or distribution. All of these questions disappear on a worldwide blending approach.

There are also tricky issues around branches and transparent entities: to where do you allocate the income and to which entity should any tax that is paid be attributed? Again, some of these issues are simplified the broader the approach is to blending.

Carve outs and thresholds

The final aspect the consultation considers is also one of the most controversial: what carve outs should there be from the rules? The consultation acknowledges that including certain carve outs (such as one for regimes compliant with BEPS Action 5 on harmful tax practices) would ‘undermine the policy intent and effectiveness of the proposal’, and yet this has not yet been ruled out. Other suggestions include a potential threshold based on turnover, carve outs for a return on tangible assets (think GILTI), de minimis rules and sectoral exclusions. In particular, the turnover threshold is obviously a crucial factor in determining which groups are within scope and it is one aspect to which all multinationals should be paying close attention. The fact that it has not yet been agreed whether there will be a threshold (never mind what level this should be set at) indicates an interesting challenge ahead in reaching agreement on this.

Again, there is an emphasis on keeping things simple: should any carve outs be objective, formulaic and simple to apply, or should they be more qualitative and fact-dependent, increasing fairness but also leading to greater uncertainty and increasing compliance and administration costs?

Where do we go from here?

Comments were invited on the consultation document by 2 December, and there was a public consultation meeting on 9 December.

Assuming consensus can be reached, there will need to be both changes to domestic law as well as tax treaties. As with pillar one, care will need to be taken to avoid double taxation and infringing existing international obligations. An MLI 2.0 seems increasingly likely. Given the controversial nature of some aspects of the proposals, could there be some optionality in the final version of the pillar two rules, with jurisdictions able to select the elements they wish to incorporate, using a similar mechanic to the original MLI?

There is a recurring theme throughout the document of favouring simplicity over complexity and seeking ways to ease the compliance burden for tax administrations as well as taxpayers. But there is clearly a balancing exercise to be done between having a simpler set of rules that is less burdensome to comply with and ensuring that the aims of the GloBE proposal are addressed to the satisfaction of those pushing for a minimum tax (such as France and Germany).

The US has previously endorsed the idea of a minimum tax, but what happens if it looks considerably different to their existing GILTI and BEAT rules and they are invited to enact further measures? Will that be a deal breaker for US cooperation? The US might argue that its rules, which increasingly are aimed at ensuring that the profits of international businesses – particularly profits attributable to intangibles – are subject to current tax somewhere at a minimum tax rate, already respond to the concerns behind the OECD proposals. Although the imposition of minimum tax under the BEAT is not coordinated with foreign tax rules, very recent proposed regulations look to coordinate the GILTI rules with foreign tax systems by allowing an elective exclusion from GILTI for income that bears foreign tax at a specified rate. On the other hand, if the existing GILTI and BEAT rules are accepted by the OECD as being pillar two compliant, will the proposals have the impact that its proponents are hoping for?

At a recent conference, John Peterson of the OECD, who is leading the pillar two consultation, expressed a hope that by having a broad minimum tax, this could open up the window for countries to go back and rethink some of the more complex and dense parts of their foreign taxation regimes. This is indeed a laudable aim, although those that hoped for simplification of the raft of UK anti-avoidance rules when the UK GAAR was introduced might be sceptical that this is a realistic outcome.

It is hardly surprising that the consultation raises more questions than it answers given that so many of the issues are inter-related. Finding a solution to one question depends heavily on the answers given elsewhere. The path towards consensus is precarious. One wrong step could lead to a flood of countries walking away from the discussions. For now, the biggest questions remain unanswered. But if the project is brought to fruition, it could completely re-shape the international tax framework. Multinationals will be following developments intensely.

This article was first published in Tax Journal