The OECD Global Anti-Base Erosion (“GloBE”) proposal

In January 2020 a group of CBT researchers, led by Michael Devereux, published a wide-ranging report on one of the two proposals under discussion by the OECD/G20’s Inclusive Framework: the GloBE Proposal (Pillar 2). Its proponents viewed it as a potential game changer for international business taxation, addressing issues of tax competition and profit shifting. But the proposal left many policy and legal questions unanswered. There was also a dearth of empirical evidence on its possible impact. The report sought to analyse the policy issues and to provide empirical evidence to facilitate a better-informed debate. The report covered four main areas. 

1. Policy Analysis
At the time at least, the GloBE proposal appeared to have two main objectives: first, to extend the BEPS project to further combat profit shifting and, second, to reduce tax competition. It was not clear if both main objectives, or only the first, would be pursued. 

From a policy perspective, this proposal moves in a different direction to the other proposal being discussed by the Inclusive Framework. While the Unified Approach (Pillar 1) moves in the direction of taxing business profit on a destination basis (in market countries), the GloBE proposal (Pillar 2) moves in the direction of taxing business profit in the location of its parent company. If the proposal were to target both profit shifting and tax competition, it would also constitute a notable departure from the policy consensus agreed only a few years ago during the BEPS project, according to which “no or low taxation is not per se a cause of concern, but it becomes so when it is associated with practices that artificially segregate taxable income from the activities that generate it.” 

The report also notes several issues of practice and implementation. It appeared likely that the GloBE proposal would not achieve its two primary objectives unless (i) it was adopted by all, or most, countries; (ii) countries agreed to a detailed set of harmonised rules; and (iii) the harmonised rules incorporated a strong form of minimum tax, which does not include a substance based carve out or worldwide blending. 

Subsequent reports approved by the Inclusive Framework seem to have recognised this by dropping the aim of combating tax competition, in favour of a more general goal of ensuring that multinationals pay a minimum level of tax. 

But in any case, questions remain as to whether some mechanisms that are at the heart of the proposal – most prominently the calculation of the effective tax rate – could work from a technical perspective.

2. The Effect on Revenues
The report uses data from a number of sources to estimate the impact on tax revenues of the introduction of the income inclusion rule, assuming that the proposal is adopted universally, without a substance-based carve-out. The approach did not allow for any behavioural response to the introduction of the GloBE. And the results are subject to considerable uncertainty due to the lack of reliable data.

The report reaches a central estimate that the jurisdictional approach with a threshold effective tax rate of 10% would yield additional revenue worldwide of around $32 billion, or around 14% of the taxes currently paid by foreign-controlled entities. This represents less than 2% of total taxes currently levied worldwide on corporate profit, and approximately one third of 1% of total worldwide corporate profit. More speculatively, if these revenues were collected by the country of the ultimate parent company, then the largest beneficiaries in absolute terms would be the world’s largest economies, such as China and the United States. Relative to their own tax revenues, several Eastern EU member states would also gain substantially. However, in this case, there would likely be a significant impact on tax rates in low tax countries; and if this were to happen, they would become the main recipients of the additional revenue generated.

The report also examined the worldwide blending approach, but this has subsequently been dropped by the OECD.

3. The Effect on Incentives
The report used a simulation model to identify the effects of the proposal for the income inclusion rule on three types of behaviour of multinational companies:

the location of real economic activity
the scale of real economic activity, conditional on location, and
the extent of profit shifting.
Compared to the worldwide blending approach, the jurisdictional approach would generate:

a much stronger effect in mitigating profit shifting, but
a steeper rise in effective average tax rate (EATR), and
a steeper rise in cost of capital.
4. Compatibility with EU Law.
EU law imposes several constraints on the GloBE proposal. The proposal should comply with existing directives, such as the Interest and Royalties Directive, or, more likely, the directives will have to be amended to accommodate the proposal. The difficulty here is political rather than technical as it requires unanimous agreement among EU Member States. The proposal should also comply with primary EU law including treaty provisions concerning Fiscal State Aid and the fundamental freedoms. The report focuses on the latter.

The safest route for compliance with the case law of the Court of Justice of the European Union on the fundamental freedoms is the inclusion of a substance-based carve-out. At least two alternative routes may be available if no such carve-out is adopted, although compatibility with EU law is less certain under both. The first is to extend the proposal to domestic subsidiaries. The second is to present the primary objective of the proposal to be that of achieving a broader policy objective such as the equality of treatment of domestic and foreign investment, rather than addressing avoidance. The likelihood that the second route would be found to be compatible with EU law is increased if the proposal is unanimously agreed by Member States and adopted through a directive. However, there is less certainty around these routes than there is on the inclusion of a substance-based carve-out.