COVID-19 and the future of Pillar one

By Oliver Treidler, CEO, TP&C GmbH, Berlin

In the opening weeks of 2020, most transfer pricing practitioners were fixated on the OECD reform proposal for taxing digital commerce.

The OECD seemed poised to re-write the rules of the game by implementing so-called Pillar 1. The reforms would fundamentally alter the allocation of taxing rights among jurisdictions (new nexus rules) and also change the approach to allocating profits among these jurisdictions (transfer pricing rules).

Facing intense political pressure to broker a global consensus by the end of 2020, the OECD advanced the reform proposal at a breakneck pace – especially when considering the fundamental nature of the reforms.

For many stakeholders, including myself, it was increasingly frustrating to realize that important conceptual considerations (e.g., how to define the concept of value creation for digital business models) were nonchalantly brushed aside for a pursuing the sensible but short-term political target of preventing the proliferation of “digital taxes” on the national level.

The message conveyed by the OECD was basically that “the train has left the station” and that Pillar 1 will be put on the table without a further conceptual or technical discussion (see my earlier MNE Tax article).

For transfer pricing practitioners, there seemed little left to do but to wait for how the negotiations within the so-called Inclusive Framework would shake-out.

It seemed increasingly unlikely that the myriad of open technical questions with respect to definitions (e.g., how to delineate a customer-facing business?) and thresholds (e.g., how to calculate the so-called Amount A?) would be resolved in a timely fashion. Thus, everyone seemed to brace and prepare for picking up the pieces caused by implementing a half-baked reform.

At the end of the day, however, transfer pricing practitioners tend to be a pragmatic lot, and somehow solutions will materialize. Best practices will be established sooner or later.

Now, by the end of April 2020, the focus of transfer pricing practitioners has changed. In the last weeks, we have all focused on COVID-19.

The immediate focus was on how to adjust our daily routines, with schools and kindergartens being closed and the economy on stand-by.

Frankly, at times, thinking about Pillar 1 (or other conceptual transfer pricing issues) seemed a bit inane during March and April. With an increasing yearning for normalcy, however, thoughts circled back to transfer pricing.

On LinkedIn, there emerged a plethora of articles and webcasts dealing with the impact of COVID-19 on transfer pricing.

Front and center of the recommendations was a review of the risk allocation among related-parties and the question of how to adjust transfer prices to cope with the effects of the extraordinary economic circumstances (target margins, interest rates, etc.).

With each MNE being affected in different ways and to varying extents, there are no one-size-fits-all solutions.

 It seems, however, that transfer pricing practitioners are generally confident that it will be feasible to determine adjustments that are commensurate with the arm’s length principle. It also seems reasonable to assume that managerial decisions coping with the crisis will focus on business necessities rather than tax planning.

From a transfer pricing perspective, it is thus rather easy to give the “green light” to implement the respective decisions. Provided that the rationale of these decisions (adjusting target margins) are documented (as will be recommended by any transfer pricing practitioner), there is a good chance that the arm’s length nature of such adjustments can be defended in future tax audits.

Meanwhile, the Pillar 1 reform is still lurking on the agenda, waiting to be addressed. The great question, however, is how the COVID-19 crisis will impact the upcoming negotiations.

When thinking about possible scenarios, the timeline eventually to be adopted by the OECD appears to be one of the most critical aspects. In broad strokes, three scenarios are conceivable.

Scenario one: Pillar 1 adopted by the end of 2020 (short-term orientation)

While the duration and the extent of the (inevitable) economic recession are uncertain, it must be anticipated that governments will feel pressure to raise tax revenues to finance the emergency measures.

While debt ceilings are currently lifted with comparative ease, there ultimately is a limit to the budgetary discretion of governments.

Politically, the discussion will likely revolve around the question of how much tax burden companies can and should shoulder without jeopardizing economic recovery. Most governments will face difficult choices, and establishing a sensible balance will be a moving target in the years to come.

In this situation, the taxation of MNEs – and especially the digital taxation – could, depending on how the discussion is framed, either be regarded as “sitting ducks” or a “drop in the bucket”.

From the perspective of national politicians taxing MNEs (digital commerce) might be viewed as an extremely appealing source of additional tax revenue (hence a “sitting duck”).

From the perspective of national politicians taxing MNEs (digital commerce) might be viewed as an extremely appealing source of additional tax revenue (hence a “sitting duck”).

By either introducing national digital taxes or negotiating for a composition of Pillar 1 that is heavily geared in favor of “market countries,” national politicians can extend the domestic tax base without (directly) putting a strain on domestic companies.

While this might be politically attractive for some, it is clear that such an expansion of the tax base is a zero-sum game and will come at the expense of other jurisdictions.

Introducing national digital taxes will (we have seen in the past) likely be met by retaliatory measures of the affected host jurisdiction.

Similarly, aggressive demands in the context of the Pillar 1 negotiations will likely be met with an unwillingness to surrender excessive chunks of the tax base.

A failure of the negotiations would not only jeopardize the prospect of reaching a consensus on the allocation of profits, it would also likely signal a nationalization of tax policies.

 As a consequence, the adoption of new nexus rules, which, in principle, arguably enjoy a reasonably strong consensus, could be deferred for a long time.

On the other hand, when assessing the tax revenue actually at stake, accounting for the limits imposed by a potential conflict, the potential for additional taxes might conceivably be perceived as hardly making a dent in the national budget.

This will naturally depend on the individual financial situation of each country; however, when calculating the potential additional taxable income as a percentage of the total budget, it seems possible that many politicians will attach a lower priority to finalizing the Pillar 1 reforms by the end of 2020.

Another factor is that the technical details are far from complete and that therefore implementation will not be trivial (requiring substantial resources from both tax authorities and MNEs).

The tradeoff between realizing limited additional taxes and having to implement extensive reforms might be perceived as less than appealing.

Admittedly, the above is highly speculative. Still, it is hard to envision a positive outcome for a short-term oriented approach.

For those transfer pricing practitioners that felt uncomfortable with the pace of the reform before COVID-19, an insistence on concluding the reforms by the end of 2020 must seem like one of the worst ideas of all time.

This could equally apply to practitioners that favor new profit allocation rules based on formulary elements and those (including myself) more critical of Pillar 1 and looking for approaches to sustain arm’s length transfer pricing).

In other words, clamoring to meet the 2020 timeline is likely to benefit no one.

Scenario two: adopting Pillar 1 by the end of 2021/2022 (medium-term orientation)

Agreeing to extend the timeline to the end of 2021 or 2022 would allow stakeholders to assess the situation without being distracted by the immediate crisis.

In the context of the sitting duck scenario, politicians, especially those from market jurisdictions, would not be forced to compromise their position in principle.

Extending the timeline for the negotiations would seem to greatly enhance the chances of actually reaching a mutually agreeable (sustainable) solution for Pillar 1. In the interest of full disclosure, however, politicians would have to acknowledge that reaching the consensus will still be difficult.

As cooler heads might prevail, it might be feasible for politicians to convince their constituents that prolonging the negotiations to find an international solution is worthwhile.

In the context of the drop in the bucket scenario, it would only be logical to extend the timeline and it should be an “easy sell” for politicians.

From the perspective of transfer pricing practitioners, extending the timeline would almost certainly be greeted with a collective sigh of relief.

While in this scenario, the OECD would also be expected to keep the essential components of Pillar 1 unchanged, there would be sufficient time to discuss, modify, and improve the technical and administrative issues.

While there is no reason to assume that Pillar 1 would cause a great deal of euphoria, the pain of having to align a somewhat alien (formulaic) concept with the existing regulations for international taxation and profit allocation (arm’s length) would be conceivably be mitigated when clear delineations (scope of application, definitions of key terms) are established.

Scenario three: back to the drawing board (long-term orientation)

Whether or not the Pillar 1 would eventually be adopted was not a foregone conclusion even at the beginning of 2020.

The OECD did not have a “plan B.” While this was probably a deliberate choice made under the influence of the prevailing political pressure, it also reflected a certain degree of stubbornness as well as lack of creativity.

To be sure, irrespective of the current crisis, the transfer pricing regulations must be modified to cope with digital commerce, and the window for reforms will remain small (political pressures will remain).

With the adoption of Pillar 1 by the end of 2020 in jeopardy, however, there may be an opportunity to reintroduce some of the conceptual questions into the discussion.

 Most importantly, we could address the underlying concept of value creation for transfer pricing in the 21^st century.

That discussion should ideally transcend Pillar 1. The OECD was (maybe too) quick to depart from the arm’s length principle in the context of Pillar 1.

The Pillar 1 proposal coincided with the largely unquestioned new wisdom that the classical (market-based and, perhaps, IP-centric) concept of value creation must be modified to give more weight to people functions (at times reminiscent of the labor theory of value) and to the value contributed by a pool of users-customers (value creation originating outside of the MNE).

The Pillar 1 proposal coincided with the largely unquestioned new wisdom that the classical (market-based and, perhaps, IP-centric) concept of value creation must be modified to give more weight to people functions (at times reminiscent of the labor theory of value) and to the value contributed by a pool of users-customers (value creation originating outside of the MNE).

A new coherent concept of value creation was, however, not formulated – let alone discussed.

There is intriguing new literature on the subject of transfer pricing value creation (mostly focused on the digital economy) that would provide fertile ground for this discussion. Establishing consensus on the concept of value-creation would be a solid basis for thinking about how we need to reform the existing regulatory framework.

Admittedly, such a discussion would take time. However, only by basing reforms on a coherent concept of value creation will it be feasible to create a sustainable regulatory framework for transfer pricing.

A blurred concept of value creation that reflects an unholy combination of market-based (arm’s length) elements and politically-based (formula apportionment-oriented on ambiguous notions of “fairness” and “equitable”) will not be sustainable.

If it is inevitable for political reasons to present more timely solutions to market jurisdictions, it seems preferable to agree to a temporary compromise that is not anchored in a reform of the existing regulatory framework.

Why not agree to allocate a  fraction of the taxes paid by MNEs pursuing highly digitalized business models (basically the FAANGs) that exceed certain thresholds (i.e., MNEs subject to country-by-country reporting (CbCR)) to market jurisdictions based on the proportional share of digital revenues realized in these jurisdictions?

 I can see several advantages of such a temporary compromise.

 First, the total taxes to be paid by the MNEs would remain unaffected, and negotiations among the countries would be limited to a single item, i.e., the share of taxes being designated for re-distribution (host nations could forfeit their taxable income on a pro-rata basis).

Moreover, the applicable amount of taxes for each MNE can be determined by comparative ease (CbCR).

An allocation-key should not be too complicated to compute (at least much easier compared to Pillar 1).

Also, the allocation of the taxes would be entirely politics-based and arbitrary (but opposed to Pillar 1 this would at least be explicitly acknowledged) and kept outside of the regulatory framework.

Perhaps most importantly,  finding a compromise would have an important symbolic connotation as it reflects that countries seek cooperation rather than nationalization.

There are few potential downsides of such a temporary compromise (sort of an Amount A -light).

Yes, market jurisdictions would not (yet) claim (permanent) new taxing rights, as the introduction of a new nexus rule is deferred.

On the other hand, the host countries would demonstrate their commitment to addressing the reforms.

The re-distribution of (an admittedly smaller) part of their tax revenues would hopefully signal that market jurisdictions will not be short-changed.

Compared to forcing an agreement on Pillar 1, the stakes, and potential conflicts, would be lower. The politicians could reasonably feel that they have reached a sensible solution considering the circumstances and alternatives – living to fight another day.

For transfer practitioners, a compromise would buy valuable time and signal an obligation to work on establishing the concept of value creation for transfer pricing in the 21^st century to prepare a reform proposal that the politicians can negotiate in good faith.

For transfer practitioners, a compromise would buy valuable time and signal an obligation to work on establishing the concept of value creation for transfer pricing in the 21^st century to prepare a reform proposal that the politicians can negotiate in good faith.

Why we need to fight against scapegoating the arm’s length principle

If forced to place a wager, my money would be governments deciding to adopt Pillar 1 by the end of 2021/2022”.

This would reflect a business-as-usual outcome and, while my preference would be scenario three, we could at least be happy to have avoided the worst.

One aspect I feel strongly about when talking about Pillar 1, as probably evident above, is the arm’s length principle.

One of the most detrimental aspects of Pillar 1, especially in the realm of politics, is that it is often being framed as being about curbing tax avoidance (a sort of BEPS 2.0).

That, however, is a misconception. The Pillar 1 reforms are first and foremost about creating new [!] taxing rights for “market countries,” without any conceptual link to curbing aggressive tax structuring.

It is rather simple, really; when a company does not establish a taxable nexus in a specific jurisdiction, it does not owe any taxes to this particular jurisdiction.

In such a case, there is no need for a company to engage in aggressive tax structuring or other shenanigans. To be sure, when looking at the digital economy, it is legitimate to question whether the rules that determine what activities of a company create a taxable nexus are adequate.

Without prejudice, one might present a coherent economic argument to support the general idea underlying the Pillar 1 proposal (extending nexus rules, broadening the concept of value creation).

Too often, however, policy proposals (and Pillar 1 is no exception) are being framed in a way that suggests that the arm’s length principle would, even post-BEPS, allow MNEs to “game” the system and thus facilitate (large scale) base erosion.

These detrimental effects, it is suggested, are exacerbated by the IP-based nature of the digital economy, which presents MNEs with all sorts of enticing levers for implementing aggressive tax and transfer pricing structures.

Scapegoating the arm’s length principle in such a way drastically increases the political pressure to implement fundamental reforms at a breakneck pace – without regard to the collateral damage caused by an incoherent regulatory framework (force-feeding formulary apportionment) and an increased nationalization of taxation (facilitating zero-sum game negotiations).

What is especially “toxic” in this context is the suggestion that solutions for taxing the digital economy depend on the elimination of the arm’s length principle, which tends to be portrayed as the root cause of the conflict on digital taxation.

A prominent recent example for such scapegoating of the arm’s length principle is an article on “How to End the Conflict Over Taxing Global Digital Commerce” by Arthur J. Cockfield (Berkeley Business Law Journal (2020)).

To contribute to a more rational discussion on the arm’s length principle and thus to somewhat diffuse the political pressure to implement fundamental reforms irrespective of conceptual merits and economic costs, I have compiled a detailed response to Cockfield’s article – “Debunking Claims Regarding the Shortcomings of the Arm’s Length Standard” – which is available on SSRN.

Comments are explicitly welcome.

The fight for the supremacy [as transfer pricing paradigm] between the arm’s length principle and formulary apportionment is a rather old fight.

There is quite a good reason for this, too; it is simply a draw. Neither method is conceptually superior.

Both provide a coherent methodological framework for allocating profits earned by MNEs between different jurisdictions. The arm’s length principle is based on simulating a “market process” for individual MNEs. In contrast, formulary apportionment is based on ensuring an “equitable” or “fair” distribution for the economy as a whole.

These paradigms, however, do not “mix” well, as a market-based process is based on exchange values that are not necessarily aligned with conceptions of “fairness” that are based on a concept of value which rather looks at inputs (e.g., labor).

These insights are, I believe, not a misrepresentation of similar statements made by Walter Hellerstein, who is one of the leading proponents of formulary apportionment.

Hellerstein opens his 2005 paper titled “The Case for Formulary Apportionment ” by discussing the question whether the “objective of income allocation is to determine the “true” geographic source of income [arm’s length principle] or to effectuate an equitable division of income based on considerations other than geographic income [formulary apportionment].”

Hellerstein provides a clear answer by stating, “In the end, I think resolution of that question may be more a matter of faith than of logic.”

In other words, political preferences and the respective concept of value matter.

Hellerstein, if I interpret him correctly, also acknowledges that the potential success of implementing a system based on formulary apportionment depends primarily on the homogeneity of political preferences (i.e., a common interpretation of “fairness”).

In contrast, the acceptance of a system based on the arm’s length principle depends on ensuring that the geographic allocation of value-added contributions within an MNE is determined based on sound economic analysis. In other words, we do not have to establish consensus on general notions such as “fairness”, but we have to continuously work on validating that profit allocation is aligned with value creation.

The BEPS reforms demonstrated (e.g,. by introducing the DEMPE concept for more accurately capturing value contributions relating to IP) that an evolutionary approach can indeed work quite well. Next, we need to find solutions for the digital economy – which might indeed imply adopting a broader concept of value creation. I believe finding a solution for this challenge is feasible – and it seems much more realistic than establishing a global consensus on “fairness”.

The BEPS reforms demonstrated (e.g,. by introducing the DEMPE concept for more accurately capturing value contributions relating to IP) that an evolutionary approach can indeed work quite well. Next, we need to find solutions for the digital economy – which might indeed imply adopting a broader concept of value creation. I believe finding a solution for this challenge is feasible – and it seems much more realistic than establishing a global consensus on “fairness”.

To be sure, I disagree with Hellerstein’s analysis and conclusion when it comes to whether the arm’s length principle can reliably approximate a “true” geographic source of income (value-adding contributions by separate legal entities).

Still, in terms of outlining the relevant decision-making framework for the politicians, he seems to be quite accurate.

As emphasized above, it seems imperative that we engage in an open and intense discussion regarding the appropriate concept of value creation for transfer pricing in the 21^st century.

 I am confident that we can do much better than Pillar 1. We should, however, not allow ourselves to be sidetracked by unsubstantiated allegations of shortcomings of the arm’s length principle and simplistic assumptions of formulary apportionment being the panacea.