Pillar 2: Avoiding Pitfalls on the Road to Consensus

Background

The digitalization of the economy has created challenges for traditional tax and regulatory systems. In particular, the international tax system, which has historically relied on a consensus-based approach to developing new rules, has struggled with the pace of change that digitalization has brought about.

While the traditional deliberative process has proven effective in the past, it has not kept up with the political will among some countries to tax a greater share of profits than the current consensus would allow. This political will has led to a proliferation of unilateral tax measures intended to operate outside the consensus-based system, which threatens the system’s ability to serve its purposes of preventing double taxation and providing certainty.

We commend the OECD for its efforts to develop an updated international consensus. This work, which has been divided into two pillars, has made preventing the further proliferation of uncoordinated unilateral measures an explicit goal. Pillar 1 focuses on the allocation of taxing rights between market and residence countries, and it appears that progress is being made toward reaching a global consensus, though significant work remains to be done. Uber has appreciated the opportunity to engage with the OECD and members of the inclusive framework on the development of workable solutions under pillar 1.¹

Regarding pillar 2, however, the OECD’s recently released public consultation document on the global anti-base-erosion proposal² appears to raise more challenges than it addresses. The stated goal of pillar 2 is to comprehensively address remaining base erosion and profit-shifting challenges “by ensuring that the profits of internationally operating businesses are subject to a minimum rate of tax.”³ This approach is based on the proposition that global action is needed to stop a “harmful race to the bottom” on corporate taxes, and that without a coordinated, multilateral solution, there is a risk that countries would take unilateral action to attract more tax base or to protect their existing tax bases.

Problems With the Proposal

While the goal of preventing uncoordinated, unilateral action is laudable, the consultation document does not provide meaningful information about the details of the pillar 2 proposal, including the minimum tax rate; how the income inclusion and base-eroding payments rules are intended to interact with each other; how cascading application of the income inclusion rule will be avoided; and how disputes will be prevented and resolved. Instead, the consultation document asks detailed questions regarding three discrete design issues: (1) the feasibility of using financial accounts as a starting point to determine the tax base to which the proposal will apply; (2) the level of blending across jurisdictions or entities that will be permitted in determining an effective tax rate; and (3) the use of carveouts and thresholds. While it is understandable that the more difficult political questions are unresolved, the lack of clear information about the overall design makes the specific design questions impossible to answer.

More importantly, even leaving aside the broader unanswered questions, the specific design features identified in the consultation document appear to take the proposal in a direction that is unlikely to lead to a workable regime that could accomplish its stated goals.

The focus on developing a single agreed tax base starting from financial accounting principles — whether for the income inclusion rule, the base-eroding payment rule, or both — seems inconsistent with how existing controlled foreign corporation rules (as well as the United States’ global intangible low-taxed income regime) work. Achieving agreement on a global common consolidated corporate tax base has been elusive in the context of corporate taxation more generally, and it is not immediately clear that it would be easier in the context of a minimum tax.⁴

Also, it is not apparent why agreement on a single tax base is essential to accomplishing the stated purposes of the proposal. Even if such an approach were considered necessary, however, the compressed time frame of BEPS 2.0 makes it extremely unlikely that the international community could agree on how to reconcile differences among existing accounting standards and address differences between accounting and tax standards in a way that could gather international consensus. Establishing a tax base, which countries would be bound to adopt, through OECD processes that differ from the domestic tax laws of all the participating countries would also appear to create significant tension with countries’ sovereign ability to determine a tax base that serves their domestic purposes.

The proposal would limit countries’ flexibility to determine for themselves whether a payment should be subject to anti-base-erosion measures and would instead impose a set of mechanical rules to identify payments that would be considered to be insufficiently taxed. This approach would mean that payments to a jurisdiction that has a domestic tax base that may differ from the agreed tax base could be subject to source-country tax, even when both source and residence country would have considered the payment to be subject to adequate tax. Similarly, the mechanical approach would limit countries’ ability to establish effective tax incentives to encourage investment.

We propose an approach that avoids the unnecessary complexity outlined in the consultation document, making it significantly more administrable, less burdensome, and more consistent with the goal of arriving at a consensus-based solution that effectively addresses the concerns underlying pillar 2.

Principle-Based Minimum Standards

Uber proposes that the CFC inclusion rule outlined in both the OECD’s work program⁵ and the consultation document have priority over the anti-base-erosion taxes (ABET) outlined therein.⁶ This means that when a country adopts the proposal’s CFC inclusion rule, no other base erosion measures would apply to payments made to that country. (The work program and consultation document each contemplate two: a denial of deduction and a denial of treaty benefits.) This simple yet effective priority rule would eliminate the risk of double taxation on the same income potentially taxed under both rules.

Further, the CFC inclusion rule should be principle-based and not prescriptive. A CFC inclusion rule that does not permit countries to make decisions defining their own tax bases and the operation of a CFC inclusion rule would encroach too far upon the sovereign rights of the countries. It would face too many objections, likely becoming mired in fundamental disagreements over technical interpretive matters that would prevent consensus among inclusive framework members.

Instead, the OECD could provide minimum standards that would serve as guidelines for countries implementing CFC inclusion regimes, and if satisfied, would prevent others from applying the ABET to entities within their jurisdiction. These minimum standards could be specific enough to ensure their effectiveness in addressing profit-shifting concerns while remaining sufficiently broad to allow countries the ability to craft domestic legislation that satisfies the standards. For example, an acceptable minimum standard for a CFC inclusion rule would require each country adopting such a rule to affirmatively address the following design features in their domestic legislation (as set out below)⁷:

1. ‘Blending’ determined by the parent jurisdiction. The OECD would not be prescriptive regarding how this testing is done except to provide a consensus threshold tax rate for low-tax testing. Testing would be permitted to be done at the worldwide, country, or entity level, allowing countries to decide what level of blending that they may permit.

2. CFC income that is low-taxed can be included in the parent jurisdiction taxable income. The parent jurisdiction would be permitted to tax the identified income up to the threshold rate.

3. Testing and inclusion mechanics are left to countries to implement. Countries should be free to determine for themselves the effective tax rate for a CFC, and the two main components of an effective rate test, the tax base and the taxes paid. Specifically:

   • Tax base. Countries may use any reasonable determination of the base, including the parent jurisdiction concept of taxable income, for example, as with GILTI, local statutory income, or allocated consolidated financial statement income.⁸

   • Taxes paid. Countries may use any reasonable determination of taxes paid, including accrued taxes (for example, GILTI), deferred taxes under applicable generally accepted accounting principles, and cash taxes paid.⁹

4. Timing and permanence differences. Countries should also be permitted to adopt their own rules regarding permanence and timing differences between the tax base for testing a company’s low-tax status and the tax base for domestic taxable income, including ignoring such differences (for example, as is done under the GILTI regime), tracking attributes at the parent level, tracking attributes at the CFC level, and averaging over multiple years.

5. Existing CFC inclusion regimes are accepted. Some existing CFC inclusion regimes, such as GILTI, should be certified as compliant regardless of what is chosen as the final low-tax threshold rate. Without such a concession, we fear that consensus may be elusive.

The CFC inclusion rule would provide that regimes that satisfy the above criteria would be certified and included on an “angel list” that would be published.¹⁰ As noted, if an entity is subject to such a CFC inclusion regime, that would turn off any ABET with respect to payments to that entity.¹¹

Permissible Anti-Base-Erosion Taxes

Payments made to related companies not subject to a CFC inclusion regime on the angel list would be subject to the ABET.¹² Prescriptive rules regarding the determination of whether a payment to a related person is sufficiently taxed so it nullifies any ABET introduces significant complexity regarding testing for low-taxed status and determining how those rules would operate. Such mechanical tax rate testing presents too many hurdles to overcome and the same challenges to countries’ sovereignty discussed above. Further, we do not believe such prescriptive rules are necessary to design an effective and efficient solution for payments to entities or groups with income that is considered insufficiently taxed.

Countries should be permitted, as they sovereignly determine, to apply or not apply ABET to payments made to entities in jurisdictions that do not satisfy the requirements of the CFC inclusion rule. In fact, no OECD consensus or permission is required for this to happen under existing rules. Countries are free today to impose ABET. We believe the existing economic incentives — a delicate balance between a desire to attract investment and protect the tax base — and fear of retaliatory measures best serve to regulate ABET rather than prescriptive global rules. However, it may be beneficial for the OECD to provide a principle-based framework for developing ABET rules for countries that wish to implement such regimes.

For payments to entities in jurisdictions in which the CFC inclusion rule does not turn off ABET (for example, parent jurisdictions), a principle-based framework for whether ABET could be imposed could look to the statutory tax rate of the payment destination jurisdiction. For countries that have a statutory tax rate equal to or greater than the consensus low-tax threshold rate and are on the angel list (because they have adopted a qualifying CFC inclusion regime), ABET should not be imposed.

Countries with sufficiently high tax rates to which all resident entities are subject and that have adopted the CFC inclusion rule minimum standard should be considered countries not contributing to the base erosion problem, and payments made to taxpayers considered residents in those jurisdictions should not be subject to ABET. However, for countries wishing to go further by looking to some proxy for the effective (rather than statutory) tax rate, they could add a criterion: whether the payment benefits from tax incentives (or special tax regimes) that the payer country deems abusive. Countries could determine abusive regimes by reference to work already undertaken in this area.¹³

The above framework merely provides guiding principles for countries wishing to adopt ABET rules. It would neither require nor prohibit countries from adopting rules, as is the case today. However, it would provide guidance and help alleviate the proliferation of unilateral measures that are inconsistent and lacking common principles.

Concluding Thoughts

The digital transformation of the economy continues to disrupt traditional operating models, which in turn upends the typical historic distribution of multinational enterprises’ profits between residence and source jurisdictions. In light of that disruption, while the OECD’s goal of ensuring that MNEs are subject to a minimum rate of tax presents new challenges to the boundaries of national sovereignty, it is not an unreasonable goal.

However, the proposal as laid out in the consultation document has several features that will likely struggle to attract consensus. The OECD and the inclusive framework would therefore do well to more carefully tailor the pillar 2 work in an effort to make real progress toward that global consensus more achievable. Uber’s proposal eliminates many of the more controversial aspects of the consultation document and maintains national sovereignty while preserving the key design features that would prevent a race to the bottom of corporate tax rates. A proposal following the principles that we outlined would thus form a much better basis for building a global consensus.