Can Pillar 2 Be Leveraged to Save Pillar 1?

The tax community has been pessimistic about whether pillar 1 can be designed in a way that could be agreed to and implemented by the United States. This pessimism is largely driven by a view that pillar 1 would require tax treaty changes by the United States and that there is no reasonable likelihood of reaching the required Senate majority. In this article, I outline a proposal that could be acceptable on a bipartisan basis in the United States but nevertheless could be successfully implemented without treaty ratification. The full analysis is set out in a recent Oxford Centre for Business Taxation working paper.¹

In a nutshell, the proposal seeks to leverage key concepts from pillar 2 to create a payment mechanism and to avoid double taxation. It assumes that there is an agreeable mechanism for determining the amount A allocation to each market jurisdiction. The proposal addresses two difficult aspects of pillar 1. First, how to ensure that the market jurisdiction receives the relevant tax liability payment and, second, how to identify a “ceding jurisdiction” that gives up taxing rights to avoid double taxation.

Pillar 1

The high-level design contours of pillar 1 were released in the OECD’s October 8, 2021, statement.² Pillar 1 is expected to reallocate a portion of the residual profits of the largest and most profitable multinational enterprises to market jurisdictions. For in-scope entities with profitability over 10 percent (defined as profit before tax/revenue), 25 percent of residual profits will be allocated to market jurisdictions. This allocation is called amount A. A market jurisdiction can be entitled to an amount A allocation even in the absence of an entity or permanent establishment in the jurisdiction. However, its amount A allocation can be reduced (potentially to zero) if there are already residual profits taxed in the jurisdiction under the marketing and distribution safe harbor.

Amount A is an allocation of profit. The market jurisdiction can then impose a tax liability on that profit, which I will call an “amount A tax liability.” There are other key elements of pillar 1, including a mechanism to grant tax certainty to in-scope entities, and the amount B allocation. These aspects are beyond the scope of this article.

The Proposal

The proposal seeks to address two fundamental questions:

• Once the amount A allocation has been determined, how can the regime ensure that the relevant tax liability is paid to the market jurisdiction even in the absence of a resident entity or PE?

• How can the regime identify a ceding jurisdiction to prevent double taxation?

The design must either:

• be capable of achieving bipartisan support that would allow for Senate ratification of U.S. treaty changes; or

• produce acceptable outcomes even in the absence of changes to U.S. tax treaties.

Imposing Tax Liability

The proposal seeks to leverage the pillar 2 infrastructure of jurisdictional top-up taxes to ensure that the MNE has an incentive to pay the market jurisdiction. In essence, an amount A jurisdictional top-up tax could be added to the pillar 2 jurisdictional top-up tax if, and only if, it is not paid to the market jurisdiction.

Pillar 2 relies on the concept of a backup tax. If a sufficient level of tax is not paid in the source country, then an additional amount of tax will be imposed elsewhere on the MNE group. This removes the incentive for the MNE not to pay the minimum level of tax in the source country. This mechanism can also be used to ensure that the MNE pays the market jurisdiction any amount A tax liability. If the MNE does not pay the market jurisdiction, it will be subject to the same amount of additional amount A tax liability elsewhere in the group.

While the working paper considers two models, the simplest is below:

where:

• jurisdictional amount A top-up tax is the amount added to the jurisdictional top-up tax under pillar 2 and subject to the income inclusion rule or the undertaxed payments rule;

• amount A is the amount A allocation to the market jurisdiction;

• jurisdictional tax rate is the tax rate applied in the market jurisdiction to the amount A allocation; and

• amount A tax liability payments are payments from any entity of the MNE group to the market jurisdiction for the amount A tax liability (the jurisdiction’s amount A allocation multiplied by the jurisdictional tax rate).

Under this design, the top-up tax will simply be the amount A tax liability in the market jurisdiction. From the MNE’s perspective, it has no incentive not to pay the market jurisdiction because it will be subject to the same tax liability under the IIR or UTPR. This model makes the MNE financially indifferent about whether it pays the market jurisdiction or the IIR jurisdiction. Positive incentives for the MNE to pay the market jurisdiction could be added either through other elements of the pillar 1 design (for example, by limiting access to the tax certainty process for MNEs that do not pay the market jurisdiction) or by providing a financial incentive to pay the market jurisdiction. These options are further examined in the working paper.

Importantly, this mechanism does not identify a particular entity within the MNE to pay the relevant amount A tax liability. It is not necessary for the paying entity to be in the ceding jurisdiction. Double taxation is avoided through a separately identified ceding jurisdiction (see below). Accordingly, the jurisdiction of the paying entity should not grant a deduction or tax credit for the entity that makes the payment. If the paying entity receives a tax benefit for making the amount A tax liability payment, the MNE will get a double benefit (as there will also be a ceding jurisdiction that relieves the double taxation).

There may be questions about whether the market jurisdiction could receive this payment as a tax payment. To the extent that this was an issue under their domestic laws, each market jurisdiction could adopt a different approach to resolving what is fundamentally a domestic law problem. However, it is important that any solution be consistent with the pillar 1 regime. Several potential options are addressed in the working paper.

The above mechanism uses the pillar 2 top-up tax infrastructure to ensure that the MNE has an incentive to pay the market jurisdiction even if it does not have a tax resident entity or PE there. The mechanism is indifferent about which entity in the MNE group makes the payment but notes that no tax credit or deduction should be given in the paying country for paying the amount A tax liability. Double taxation is avoided by identifying the ceding jurisdiction.

Identifying the Ceding Jurisdiction

Pillar 1 reallocates taxing rights over amount A to the market jurisdiction. To reallocate the taxing rights, there must be an identified jurisdiction or jurisdictions that will cede their taxing rights over the amount A profits. This is the ceding jurisdiction.

The proposal leverages the definitions of excess profit and effective tax rate (ETR) from pillar 2. It is important to note the difference between residual profit under pillar 1 and excess profit under pillar 2. Residual profit is calculated on a global group basis (not assigned to individual jurisdictions) and asks whether profit before tax is greater than 10 percent of revenue. Excess profit is the profit booked in a jurisdiction beyond the amount covered by the substance-based income exclusion, which is a formulaic return on the payroll expenditure and carrying value of the tangible assets within the jurisdiction. The concept is that the exclusion reflects a standard return for the substance in the jurisdiction. The ETR is the tax rate being paid by the MNE in relation to the total profits booked in that jurisdiction. Under pillar 2, the MNE will generally be required to calculate the ETR and excess profit for each jurisdiction in which it operates. Detailed rules have already been designed and agreed to by the inclusive framework for the calculation of excess profit and jurisdictional ETR.

The proposal is for a “waterfall” allocation for which jurisdictions must give up taxing rights. If the total amount A allocation can be made from the first tier of the waterfall, there is no need to progress to the second and so on. The rule is that the ceding jurisdiction is the jurisdiction with the lowest ETR that has excess profit. The jurisdiction with the lowest ETR (as determined under pillar 2) will cede the total amount of its excess profit or the total sum of amount A reallocations (whichever is less). If the former (that is, if the total amount of excess profit is ceded), the jurisdiction with excess profit and the second lowest ETR will be subject to the same process. The waterfall will continue through the tiers until the total amount A allocation has been ceded by a jurisdiction. This is best demonstrated through an example.

An MNE has $400 million of residual profits. Under pillar 1, 25 percent of those profits are to be allocated to market jurisdictions. Five separate market jurisdictions are each entitled to an amount A allocation of $20 million. Accordingly, the total sum of amount A allocations is $100 million. The MNE group has a taxable presence in only four countries (A-D); that is, it books all its profits in these four jurisdictions on an arm’s-length basis. The profit, excess profit, and ETRs of each of those jurisdictions are listed in the table below, ordered by ETR. In this example, residual profits match excess profits. This is not necessary for the allocation mechanism to work.

An allocation is made to a country only to the extent of its excess profits.

While Country A has the lowest ETR, it has no excess profits. It is not a ceding jurisdiction. Country B has the second lowest ETR and has excess profits. While it has total profits of $55 million, an allocation is only made to the extent of its excess profits — $50 million is allocated. The same applies to Country C, up to its cap of $10 million. Finally, an allocation is made to Country D for the remaining $40 million.

The result is that the ceding jurisdiction will be the one with the lowest ETR but also with excess profit. A jurisdiction will never cede any taxing rights over amounts other than excess profits in its jurisdiction. To avoid unnecessary complexity, a de minimis exclusion would likely be appropriate. For instance, a jurisdiction would be ignored if it booked less than, say, 5 percent or 10 percent of the total excess profit of the group. The result would be that the ceding jurisdiction will be where the group has large amounts of excess profits subject to tax at low rates. A

Ceding Jurisdiction Allocations

‘Production’ Jurisdiction	Total Profit	Excess Profit	ETR	Allocation
Country A	$60 million	$0	0%	-
Country B	$55 million	$50 million	4%	$50 million
Country C	$50 million	$10 million	6%	$10 million
Country D	$350 million	$340 million	10%	$40 million
Total				$100 million

variation could also be considered that ensures that the ceding jurisdiction retains taxing rights over a certain portion of its excess profit. Importantly, there is no attempt under this mechanism to trace profit from a particular market to a particular source of excess profits within the group. The allocation is mechanical.

The rationale for this allocation is driven by the idea that at least 25 percent of the residual profit of an MNE group is likely to be mobile. MNEs structure to locate this excess profit into low-tax jurisdictions. Insofar as there is to be a reallocation of profit to a market jurisdiction, it is appropriate to take it from the jurisdictions into which it has been placed by the MNE.

Avoiding Double Taxation

Once the ceding jurisdiction has been identified, it is expected that it will avoid double taxation by either:

• exempting the amount of profit ceded from domestic taxation; or

• providing a tax credit for taxes paid on that income in the market jurisdiction.

Both possibilities remain consistent with the proposal, but a pro rata allocation of tax credits is required if the ceding jurisdiction adopts the credit approach. The working paper sets out how this would be done.

Nonparticipating Jurisdictions

The big question for any pillar 1 design is what happens if the identified ceding jurisdiction refuses to sign up or is unable to amend its tax treaties. The payment mechanism identified above would still apply. The market jurisdiction would still collect the relevant amount A tax liability unless the MNE could avoid the top-up tax mechanism by not operating in any country that had adopted the UTPR (with the pillar 1 extension under this proposal). While this ensures payment of the liability, an identified ceding jurisdiction could refuse to grant relief from double taxation.

Before stating that any possibility of double taxation renders the proposal untenable, it is important to consider how likely it is that double taxation would be imposed. The proposal’s allocation mechanism identifies the ceding jurisdictions as those with the excess profits and the lowest ETRs in the MNE group. These jurisdictions have generally attracted MNEs to book significant amounts of excess profits by (among other things) offering low tax rates. It would be surprising if jurisdictions with the lowest tax rates in the group chose to deliberately impose double taxation on MNEs operating in their jurisdictions. It is not clear why a low-tax jurisdiction would want to impose double taxation. To the extent it did, there is a real risk of encouraging MNEs to relocate their excess profits to another jurisdiction. It is important that the reallocated amount is only 25 percent of the residual profits of the group. In many cases, the low-tax jurisdiction will only be ceding taxing rights over a moderate proportion of the excess profits booked in its jurisdiction.

Finally, in the worst-case scenario, if double taxation is imposed because the ceding jurisdiction refuses to implement pillar 1, then at least double taxation will be imposed at the lowest ETR for any excess profit in the MNE group. While clearly undesirable, this is clearly preferable to two high-tax-rate countries imposing tax on the same profits.

The United States

The political reality is that a pillar 1 agreement will be almost impossible to achieve without the support of the United States. Ideally, the pillar 1 design could achieve bipartisan support, allowing it to be implemented through treaty ratification with a two-thirds majority in the Senate. The proposed design offers significant upsides to the United States with very limited downsides. This may be enough to achieve bipartisan support.

First and foremost, the United States is unlikely to cede taxing rights over much revenue. The amount A allocation is for 25 percent of the residual profit of the group. Under the proposed design, if more than 25 percent of residual profits are in jurisdictions with excess profits and an ETR below that of the United States, the United States will not be a ceding jurisdiction. That is, if there is any significant portion of excess profits booked in a foreign jurisdiction with a lower tax rate, that jurisdiction would be identified as the ceding jurisdiction ahead of the United States. This seems both likely and verifiable for pillar 1 in-scope MNEs. To the extent it is true, the United States would not lose taxing rights. The design results in a reallocation from low-tax countries with lots of excess profit to market jurisdictions — it does not reallocate from one high-tax country to another.

The United States would also have positive reasons to join the proposed pillar 1 design. First, it would be entitled to amount A allocations from other countries. Second, U.S. MNEs may gain significant tax certainty benefits from the tax processes agreed to as part of pillar 1. Third, the United States would avoid the outcome of potential digital services taxes imposed in lieu of pillar 1. Finally, an acceptable pillar 1 design may increase the chances of pillar 2 being adopted because some jurisdictions have argued that the pillars are a package deal, and they may resist implementing pillar 2 unless pillar 1 is accepted.

There is at least a prima facie case that the proposal could achieve bipartisan support. If it did, pillar 1 could be adopted by a treaty that passed the Senate ratification process.

A key benefit of this design is that even if the United States couldn’t achieve bipartisan consensus, the design could be adopted through domestic law.³ By leveraging the pillar 2 infrastructure, the proposed design creates an incentive to pay the market jurisdiction even when the ceding jurisdiction doesn’t agree to pillar 1. U.S. MNEs would still have incentives to pay the market jurisdiction even without treaty changes. It relies on the same mechanism (and legal reasoning) through which pillar 2 will achieve a minimum level of taxation despite unamended tax treaties.

The United States would need to implement this pillar 1 design through domestic legislation for at least three reasons. First, it would ensure that the United States would not impose double taxation under the (likely rare) circumstances that it is identified as the ceding jurisdiction. Second, domestic legislation would be required for the United States to participate fully in the regime (for example, in the tax certainty process). Third, it is important for its legitimacy that pillar 1 pass through the U.S. Congress.

Despite this, it would not be catastrophic if there were a delay. If the Biden administration agreed to a pillars deal that was adopted by the rest of the world, but domestic legislation stalled in Congress, the rest of the world could apply the regime on an interim basis. This may be important not only for pillar 1 but also to reach a political agreement to implement pillar 2. While a delay is clearly undesirable, it is important to recall that the alternative to a temporary pillar 1 may be a reversion to unilateral DSTs (either on a temporary basis or not). The proposed design may be considered to offer a more positive interim solution, even in the absence of U.S. legislation, than would otherwise arise.

Conclusion

I have attempted to set out the contours of a design for pillar 1 that could be supported and implemented by the entire inclusive framework, including the United States. It is not expected that this proposal will be a perfect solution for pillar 1. It is intended to advance the conversation and to provide a proposal for discussion that could be shaped by the inclusive framework to reach a multilaterally agreed outcome.