During the 2013–15 base erosion and profit shifting project (BEPS), the Organization for Economic Cooperation and Development discovered the joys of what might be called “tax momentum.”
The principle is quite simple: set a timetable with very short time limits, and design a process that, whatever it may say in theory, can in reality only move forward. And for BEPS it worked very well, surprising many skeptics and producing an influential body of rules by late 2015.
However, the same momentum principle applied to the current Pillar One and Pillar Two project is not working out in quite the same way. Pillar One (about which I will not say more) is extending into the future—currently mid-2023 just for its signing, likely much later for its ratification and entry into force. And Pillar Two, which now looks likely to come in at the earliest at the beginning of 2024, is experiencing one of its periodic slowdowns, this one with a Hungarian flavor.
So what happened? Well … the pandemic to be sure. But more importantly, the difficulty of scaling up from 40 or so BEPS countries to 140 Inclusive Framework ones (with very disparate interests); and the difficulty of scaling up from a number of fairly discrete base erosion and profit shifting scenarios, to pretty much the whole international tax system requiring wide-ranging treaty and national law changes.
Pillar Two, still with some momentum, but perhaps with less clarity of purpose (is it a minimum tax? Tax harmonization?), moves forward … but to where exactly? I would argue that Pillar Two has rather lost its way, and that this “Hungarian hiatus”—until at least October 4 and the next ECOFIN meeting—is a good moment at which to take some actions to get Pillar Two back on track.
Why We Need Pillar Two to Work (Well)
The first thing I always say to the non-US governments that I speak to about this is that I am trying to make Pillar Two work, not trying to stop it. Why do I want to? Well, the US since 2017 has had a quite effective minimum tax regime—despite assertions about a continuing “race to the bottom” —so a more widely adopted minimum tax would actually level the playing field for US businesses.
However, what US businesses do not want is a Pillar Two regime that creates particular problems in the US. Nor do they want, more generally—and this is a widely-shared sentiment in the whole international business community—a regime that creates huge complexity, as well as bringing harsh and sometimes arbitrary results, especially through the operation of the modified deferred tax accounting rules. As things currently stand (and more and more evidence on this emerges every day) these are increasingly likely outcomes.
So what might help lessen the complexity and steer us away from future transatlantic tax wars and reduced (perhaps, significantly reduced) cross-border trade and investment flows? I would suggest action in three areas:
- dealing more rationally with the area of credits and incentives, which is currently creating significant political issues;
- sorting out some of the harsher elements of the hybrid “deferred tax accounting” system (and other accounting/tax base issues) which is the tip of the iceberg on technical problems;
- and allowing a long lead-in time for the rules, where the first few years are essentially a trial period. During this period results can be approximate, and penalties not apply, to give us time to deal with the first two actions, as well as allowing proper lead time for enterprise resource planning (ERP) system changes, tax authorities upskilling, etc.
But none of this is just “nice to have.” In fact, it is central to the success of Pillar Two. Because without these (and other) changes, if Pillar Two is fully brought in for, say, even just 40 (far less 140) countries on Jan. 1, 2024, an administrative and compliance disaster for both taxpayers and tax administrations could occur. We still have a little time to avert that result—so, let’s use it well.
The simplest and cleanest way to solve all of these issues is by reopening the OECD Pillar Two Model Rules and fixing the problems there. “Can’t be done,” we’re told. But there should be a way to do that, because the way these rules came about has significantly and detrimentally affected their operability.
The final Model Rules were introduced on Dec. 20, 2021, after almost no consultation with business, and with business being given no subsequent opportunity for input. That was ill-advised, because business could have provided useful advice on what worked and what didn’t, especially in relation to the deferred accounting structure, but much more broadly than that. If we are to move this forward, and get back on track, then the Model Rules—in a very limited number of circumstances—should be reopened.
Tax Credits and Incentives
As is now well known, qualified refundable credits and grants for activities such as research and development, receive a more favorable treatment under Pillar Two than do non-refundable tax credits supporting identical activities (see Articles 4.1 and 4.2). The result is that in one case the amount of the refundable credit/grant is added to Pillar Two income, thus relatively modestly reducing the Effective Tax Rate, while in the other case the non-refundable tax credit is stripped out of the covered tax amount thereby potentially reducing the ETR by a significantly greater amount as compared to the other treatment.
Why did this happen? Well, honestly, the real answer may be good negotiating by countries with pre-existing refundable tax credits or grants. The given reasons, however, are less convincing:
- One reason given is that “15% means 15%,” and nothing less. But that’s slightly meaningless because it presupposes that that is 15% of the “perfect” tax base. In fact, of course, many other decisions are taken in the Model Rules about what goes into the tax base, meaning that 15% is simply 15% of a defined (and heavily politically negotiated) tax base, not of some pristine, objective number.
- Another reason is a line of argument that credits are economically inefficient. The IMF has argued this in many cases, but they do agree that credits and incentives that correct for market failure can be appropriate (see, e.g., Economic Issues Paper No. 27 from the IMF: “Tax incentives can be justified if they address some form of market failure …”). As BIAC’s letter of March 11, 2022 pointed out, R&D credits, “social” credits (e.g., to incentivize low income housing), and renewables credits at a crucial time of energy transition, all correct for market failure in the sense that investment wouldn’t otherwise be made (or, at such scale) in high risk or low return activities.
- A further argument made is that treating (refundable) credits explicitly as public expenditure will reduce the number of such subsidies by forcing governments to make clear to taxpayers their support for (and the cost of) various activities. However, at least in the US this doesn’t hold true, as we’ve had many features of the tax system beyond refundable credits that have been treated as tax expenditures for decades without much visible effect on the reluctance of the government to use the tax system.
- Finally, and linked to this last point, it appears there is a lack of trust among some experts that politicians can be relied upon to do the right thing. Recently, a senior OECD tax official talked about this. He argued for “clear” incentives (i.e., not on the tax line) “rather than giving away the tax revenues of future generations, which current politicians don’t have to pay for, because it doesn’t show up in their budget expenditure … [Under the new rules] if you really want to incentivize companies, you have to do that through a cash incentive, rather than hide that in tax breaks that play out in the future.” This might be a little naïve, in that cash grants can be problematic. But it is also perhaps a little dismissive of the efforts of those “current politicians” who may well be trying to improve the welfare of their citizens.
So how to solve this problem—and also avert a tax war with the US if other countries levy the Undertaxed Profits rule on the US tax base, thereby nullifying US tax incentives? Conceptually it’s actually not too hard: you would treat certain types of “good” but non-refundable credits (e.g., for R&D, “social” and energy transition) in the same way as refundable credits. In other words, rather than non-refundable credits being a tax benefit that reduces covered taxes in Pillar Two (as such tax benefits would generally reduce tax expense under US GAAP and IFRS), you would instead include the benefit of the non-refundable credit in Pillar Two (“GloBE”) income—as happens with refundable credits and grants.
There would of course be definitional and boundary issues to be sorted out, and anti-abuse rules required—but there are many, many of those issues elsewhere in the project. However, by doing this you would solve a big problem, and not just in the US, but in several European countries. You would also have a tax system that did not fight against other important economic and societal priorities such as increasing innovative activity and supporting energy transition, for example, Fit for 55 in the EU—not to mention supporting wealth and job creation. And you would solve the business dilemma of being thrown into the UTPR solely by reason of taking advantage in your home jurisdiction of tax credits and incentives validly enacted by the legislature, often decades ago.
Problems Caused by Using Financial Accounts as the Pillar Two Tax Base
The second suggested area for action relates to the use of financial accounts, and particularly the new hybrid deferred tax accounting tax base, which has raised problems from the beginning with more issues emerging every day. Responding to business requests not to have a new system based on carryforwards, governments agreed to a modified deferred tax accounting base. But they also felt nervous—potentially in part because of a lack of understanding of how accounting works—so they created some harsh results which added to complexity and unfairness.
One example is Article 4.4.1 of the Model Rules, which restricts the benefit of a deferred tax attribute to 15%, regardless of whether the statutory rate in the jurisdiction is higher. Instead of the ETR smoothing that was promised in the October 2021 Blueprint, this will result in unjustified top-up tax under the Pillar Two rules, despite the fact that, subsequently, regular corporate income tax at a rate well above 15% will be paid in a jurisdiction. This inequitable treatment will not, to put it mildly, encourage an open and trusting relationship between taxpayers and tax authorities.
Another example of a harsh result occurs under Article 4.1.5, which can lead to Pillar Two tax being owed in a year even when there is no income—a rather counterintuitive result. This arises when a permanent difference creates or increases a GLoBE tax attribute (a loss) that could be used in future years. This result was deemed inappropriate by governments. But rather than allowing taxpayers to simply adjust the loss to remove the benefits of the permanent difference in the year the loss is monetized, the Model Rules instead levy tax in a year of no income, and when no benefit was taken. BIAC (Business at OECD) suggested this seemed unfair, and that under those circumstances the permanent difference could simply be held to create no benefit for Pillar Two purposes. This, in a simple way, would negate the perceived problem. But, “no,” because the Model Rules can’t be changed.
However, when it is in a government’s interest, it seems it may after all be possible to change the Model Rules. For example—and by analogy from a different accounting area of the Model Rules—in relation to Article 6.3, dealing with the transfer of assets and liabilities, a problem has emerged for some governments because of an overlooked difference (one of many that will emerge) between International Financial Reporting Standards and US Generally Accepted Accounting Principles. Under IFRS—upon which the Pillar Two rules appear to be based—gain can be recognized on an intra-group transfer (e.g., on a sale of intellectual property between separate entities within a consolidated financial statement group). However, under US GAAP, rules exist that would cause there to be no gain on the same transaction (the “common control rules”). Some countries want this gain to arise under GloBE, so it appears they are proposing, effectively, to achieve that result, by “reinterpreting” the Model Rules to turn off the US GAAP “common control” provisions, and instead deem an IFRS-style fair market value transfer in all cases.
But what’s sauce for the goose should be sauce for the gander, so let’s deal with the issues across Pillar Two in the most sensible way: by changing the Model Rules—rather than the hard, and possibly ineffective way of retrofitting the (non-binding) implementation guidance. Furthermore, it now seems clear that the Pillar Two accounting rules are going to be a source of growing complexity and concern (and of disputes and double taxation), and we will only be able to effectively deal with newly arising issues if there is ongoing flexibility to adjust the Model Rules themselves.
A Substantial Transitional Period
This final suggested action should follow pretty naturally from the last two. With the best will in the world there are many problems that have already emerged, are emerging, and will emerge, with the Pillar Two rules. And tax authorities and taxpayers are going to face insuperable difficulties with, respectively, administering and complying with these rules.
The OECD is working on safe harbors, which is good news, but there also needs to be a general rule that the full weight of Pillar Two (including penalties) will not apply for several years. Time is needed for these rules/commentary/implementation guidance to properly bed down, for teething problems to be worked through, and for safe harbors themselves to be beta tested and refined—because the alternative to that will be exactly the chaos we were told this project was meant to avoid.
While not everyone may agree, the Hungarian government (and the Polish one before them) have actually done us all a favor by allowing some additional time to get Pillar Two into a shape where it can work—and work in the same way across major jurisdictions. But there is much work to be done, and time is short. Let’s reopen the Model Rules in a limited way, solve some of the current problems, and put in place a system for more easily solving future ones. Then—but only then—Pillar Two may live up to its billing.
This article does not necessarily reflect the opinion of The Bureau of National Affairs, Inc., the publisher of Bloomberg Law and Bloomberg Tax, or its owners.
The views expressed in this article are personal and do not necessarily represent the views of either PwC or Business at OECD (BIAC).
Will Morris is PwC’s Deputy Global Tax Policy Leader and Chair Emeritus of the Taxation and Fiscal Committee of Business at OECD (BIAC).