Tax Notes logo

The Trouble With ‘Pillars’ in International Tax Policy

Posted on July 6, 2020

Lucas de Lima Carvalho is a professor of international tax law at the Brazilian Institute for Tax Law and a professor of international law at the University of Fortaleza, Brazil. He is a PhD candidate in economic and financial law at the University of São Paulo.

In this article, the author discusses the OECD’s two-pillar approach to taxing the digital economy. He focuses on the narrative of “pillar-building” in international tax policy and the difference between building consensus and manufacturing consent.

In the 1989 movie Field of Dreams, Kevin Costner plays an Iowa corn farmer named Ray Kinsella (incidentally, Kinsella is the surname of W.P. Kinsella, the author of the book Shoeless Joe, which inspired the movie). While walking on his farm, Ray hears a cryptic voice that tells him repeatedly, “If you build it, he will come.” He interprets that phrase as an instruction to build a baseball field on his farm, which he does. The field attracts ghosts of players involved in the game-fixing scandal of the 1919 World Series, and that should in and of itself be a lesson for people wanting to build stuff in the middle of corn fields. The movie, however, has a happy ending, and the quote from the mysterious source became a sort of inspirational message, particularly for entrepreneurs.1

The problems caused by the COVID-19 pandemic have understandably shifted the focus of most debates about international taxation. However, one might recall that in the seemingly distant year of 2019, the OECD released a work program containing proposals to address issues that were either within the scope of action 1 of the base erosion and profit-shifting project (that were not resolved by that final report) or that were related to its concerns about the digital economy, but exceeded its scope.

Those proposals were packaged into two pillars: Pillar 1 addresses the allocation of taxing rights, and pillar 2 contains recommendations for what is essentially a global minimum tax on income, known as the global anti-base-erosion (GLOBE) proposal.2 Discussions of those two pillars have continued into 2020, and one could argue that they have become even more relevant in the wake of the coronavirus, because many of us are working from home, unable to travel overseas, and relying heavily on services and goods provided through digital channels.

It is perhaps because of that perception that we have witnessed the emergence of a “pillar-building” narrative in scholarly articles and commentary on international tax law. The general notion behind that narrative seems to be that if we build consensus around specific proposals, we might have a shot at resolving inherent problems of rules and principles regarding cross-border taxation — problems that may be worsened by the COVID-19 crisis. In that vein, Allison Christians and Tarcísio Magalhães have proposed a global excess profits tax, or GEP, as a possible pillar 3,3 and in a recent letter to the editor, Nathan Boidman proposed abolishing the corporate tax in a possible pillar 4 (which the author actually says is an antidote not only to pillar 3, but to pillars 1 and 2 as well).4 With that proliferation of pillars, in line with a comment made by Pasquale Pistone in a recent interview,5 we will soon have enough to build a Greek temple.

Pillars evoke the idea of strength or steadiness; one could argue that social, political, and legal systems rely on them for legitimacy. No pillar, however, should be used to rationalize the globalization of policy proposals that serve the interests of some countries to the detriment of others. To expand on that, this article focuses on the narrative of pillar-building in international tax policy, and particularly on the difference between building consensus and manufacturing consent.

A Lesson From BEPS

On paper, the BEPS action plan was a package of sensible and coherent ideas to address international tax avoidance strategies. Over 100 jurisdictions joined the BEPS inclusive framework to discuss the merits and challenges of those ideas. The progress of those discussions was closely monitored via public consultations with key stakeholders.

Also, it is fair to say that BEPS issues have been the main topic of academic studies about international tax for the better part of the last seven years. The field’s brightest minds provided critical assessments of the work carried out by the OECD time and time again, all in an effort to contribute to what was perceived as the most comprehensive reform of international tax law in history.

Although the OECD initiative was laudable, its results were suboptimal. Some of the BEPS actions failed to reach any meaningful consensus whatsoever, such as action 1 (digital economy) and action 3 (controlled foreign corporation rules). Other BEPS actions resulted in reinforced standards, such as action 7 (permanent establishments), and in minimum standards, such as action 6 (treaty abuse),6 but they have been implemented in different ways by different jurisdictions — the options of and reservations to the multilateral instrument, along with the refusal to sign the MLI by the United States, Brazil, and other countries, are a testament to that. Finally, unilateral measures to address BEPS opportunities have become widespread, and that goes against the whole ethos of the OECD initiative, which was to promote enhanced international tax coordination among jurisdictions worldwide.7

The lesson we learn from the BEPS action plan is that systems tend toward entropy. At a certain point in 2013, the system of international tax law may have seemed orderly (or at least poised to become orderly), with several jurisdictions joining forces in the inclusive framework to combat aggressive tax planning, but its natural progression is to a state of disorder. In fact, amid the COVID-19 crisis, countries like Ecuador,8 Chile,9 and Mexico10 have implemented unilateral VAT reforms to target digital services. Brazil11 and Kenya12 have either proposed or announced digital services taxes to pursue the same goal. On another front, while some jurisdictions have granted different types of financial aid to distressed companies in recent months, EU countries such as Denmark and Sweden have excluded from that aid companies registered in tax havens.13 Those are typically the jurisdictions featured in the EU blacklist,14 many of which are small, insular countries that struggle to abide by EU, OECD, and WTO standards when designing their own domestic tax policies.

Those examples demonstrate that seven years after a project that was meant to unite us in the fight against abusive tax practices, we are still as divided and competitive as ever. Unilateralism — even group unilateralism like what we see in the EU — is alive and well after BEPS.

Building Consensus vs. Manufacturing Consent

It could be said that the shortcomings of the BEPS project are indicative of the impossibility of building genuine global consensus on matters of international tax law. After all, if international tax law is a system, and if systems tend toward entropy, there is little the OECD, or any other organization for that matter, can do to try and harmonize government behavior from parties with similar goals but conflicting interests. In line with that view, it is perhaps preferable to just accept full-blown, no-holds-barred competition as the present and future of international taxation.

That is not my view. Asymmetries in the tax treatment of offshore profits, financial instruments, and digital services income are some of the reasons that led us to a project like BEPS in the first place. Examples of multinational enterprises taking advantage of those asymmetries have shown that international tax avoidance is real and that it causes significant shortfalls in tax revenues worldwide.15 Also, although the BEPS project could not prevent unilateral measures, it still produced tangible results: a global forum for the discussion of key international tax issues, a peer review mechanism for minimum standards, and a set of policy recommendations that have been instrumental in reshaping domestic and international tax law in many countries. Yes, countries are implementing those recommendations in different ways, but their efforts now seem at least cognizant of other countries’ efforts, which is a step forward from pure unilateralism. Further, together with country-by-country reporting, exchange of information for tax purposes has seen increasing adoption and coordination among tax authorities, which in my view shows that efforts like the BEPS project are still worth pursuing.

The challenge here, however, is to strive to build consensus and not settle for manufacturing consent. The notion of manufacturing consent was used by Noam Chomsky and Edward S. Herman to describe what they believe U.S. mass media does to its audience.16 In their view, mass media outlets marginalize dissent and push narratives that serve the interests of government authorities and institutional investors. They claim those narratives are conveyed as an objective assessment of reality, but that assessment is made within a framework designed to direct public opinion to a specific worldview. Evidently, manufactured consent presents itself as the result of a fair and unbiased intellectual exercise — it is genuine consensus to the naked eye. That parties reached consensus only under a predetermined context, however, may suggest that the consensus was actually manufactured consent.

On one hand, manufacturing consent in international tax law is not a simple endeavor. It requires a stronghold on the discourse about best practices and desirable goals for international taxation. However, because the OECD has developed that stronghold over time, manufacturing consent around its policy proposals seems far simpler than building global consensus. Pushing specific narratives as agreed-on standards of international tax law and leading countries toward sweeping reforms of their international tax policies is easier than mediating agreements that solve global problems for all those involved.

Notice that both alternatives are motivated by the same concrete and present threat of unilateralism, which causes harmful competition, tax planning opportunities, and revenue losses. However, while building consensus respects and champions the tax sovereignty of participating jurisdictions, manufacturing consent is tantamount to coercion — diplomatic, well meaning, and principled, but still a form of coercion, one that submits the will of national democracies to norms that were never designed to address their needs.

The Trouble With the Pillars We Have

The OECD’s pillars 1 and 2, plus pillar 3 (proposed by Christians and Magalhães) and pillar 4 (proposed by Boidman), are all based on defensible policy concerns. Pillar 1 is based on the recognition that traditional taxing rights (or income allocation rules) are unable to capture residual income earned by MNEs in the digital economy (and by MNEs exploring traditional business models sans physical presence). Pillar 2 is based on the perception that low or no income taxes force governments to a race to the bottom, which prevents them from collecting appropriate tax revenue and funding relevant public expenditures. Pillar 3 is based on the premise that some industries may have reaped “outsized financial rewards” during the coronavirus pandemic, and that a global solution should be pursued to counteract externalities to society and to the public budget generated by those excess profits.17 Finally, pillar 4 is based on the notion that corporate income tax is the most “destructive” form of taxation, and that its demise would stimulate growth and create jobs.18

While pillars 1, 2, and 4 are aimed at solving structural problems in domestic and international tax law, pillar 3 seems to solve an exceptional problem (excess profits during the pandemic). However, all four pillars require some form of global solution to address their individual concerns.19 The trouble, of course, comes with the solutions on the table for discussions among the United States, the EU, the rest of the G-20, and the nearly 100 jurisdictions of the inclusive framework that do not belong to any of those groups.

Pillar 1 proposes expanding the taxing rights of market jurisdictions via a unified approach (on a “without prejudice basis,” according to the OECD/G-20 January statement on the proposals). Parts of the unified approach are heavily based on formulaic computations that depart from the arm’s-length principle. Their scope is wider than just digital businesses (it also includes consumer-facing businesses), and their application relies on the widespread adoption of “early dispute resolution procedures” and possibly a brand-new multilateral instrument, the OECD says. Those are all political decisions to be made in the context of an agreed-on standard: the attribution of taxing rights to market jurisdictions. If a dissenting group takes the view that the scope of the unified approach is too wide, or that its formulaic computations end up taxing resident MNEs beyond their ability to pay, it is unclear whether the OECD would accommodate their reservations.20 It is also unclear whether their refusal to sign on to the consensus would lead to them being labeled noncooperative and perhaps subject to retaliatory tax measures by the consenting group.

The pillar 2 GLOBE proposal involves four distinct rules: an inclusion rule, a switchover rule, an undertaxed payment rule, and a subject-to-tax rule. Regarding manufactured consent, what is relevant about the proposal is that it allows one jurisdiction, typically a capital exporter, to counteract different forms of tax incentives from another jurisdiction, typically a capital importer. That is a clear initiative to transform a policy proposal that serves the interests of some countries (to the detriment of others) into an agreed-on standard. If a dissenting group thinks the proposal is a retaliation against its tax sovereignty and the use of tax incentives to attract foreign investment, or that the globally accepted minimum level of taxation exceeds what its local infrastructure, market, and institutional security may reasonably justify, there is little they can do about it. What is most peculiar about pillar 2 is that it is an attempt to globalize fundamentally unilateral measures. For instance, if State A has a domestic undertaxed payments rule, it can in principle apply its rule to payments made from State B regardless of whether B adopts the same rule. The genesis of the pillar 2 rules is retaliatory: They are designed to adjust the international tax system to a viewpoint that forcefully limits tax competition.

As said before, pillar 3 is different because it proposes an “emergency measure” in the form of a GEP. The same problems identified for the formation of a consensus under pillar 1 seem to affect pillar 3, especially in terms of the line-drawing exercise required to distinguish routine from residual profits, and how that could possibly be used to separate residual from abnormal or excess profits.21 If the GEP taxes the assets of companies (not their profits, given that most companies will be at a loss position during this time) in different countries, one challenge will be to convince government authorities that under their constitutional or legal standards, a tax along those lines would qualify as an income tax, or that it would be permissible under a stricter reading of local “ability to pay” standards.

Another obstacle to political consensus on pillar 3 would be the definition of the words “emergency” or “exceptionality,” which surely applies to COVID-19, but raises the question whether future crises would have to reach that magnitude to enable jurisdictions to respond with a tax like the GEP.

Finally, and this is a broader point of public policy, global agreement on pillar 3 could pressure local governments to counteract perceived externalities of COVID-19 (or future public health crises) by using a tax response, as opposed to an anti-corruption or antitrust response. One could argue that the choice of public policy response to those externalities should respect the features of the legal systems of individual jurisdictions and not be the subject of peer pressure at the OECD level.

If the OECD were to seriously consider pillar 4, it would have to give up on pillar 1, adjust the scope of pillar 2 to address only low or no individual income taxation, and never even start discussions on pillar 3. Pillar 4 proposes the abolition of the corporate income tax or, as a proxy, the tax exemption for reinvested business profits to all corporations (domestic or foreign) and all corporate taxes. The immediate merit of pillar 4 (over the other three pillars) is that it proposes abolishing the corporate income tax without replacing it with another tax based on an agreed-on standard. However, the viability of consensus around pillar 4 would really hinge on that elusive alternative, because a complete lack of taxation at the corporate level may be detrimental to developing economies that are unlikely to collect sizeable annual revenue from individual or other taxes.

Unlike the other three pillars, pillar 4 seems to require a consensus that is even broader than that of the others. One could say it would require unanimity, because any country that either abolished the corporate income tax or exempted from it all reinvested profits would attract foreign investment and pressure other countries to follow suit, much like tax havens do under zero-tax regimes.


Apart from perhaps pillar 4, which is revolutionary in itself, the pillars are similar to BEPS actions in the sense that they intend to patch the problems of the international tax system as we know it today. As should be evident to all members of the inclusive framework, building consensus on those reforms is probably the only way forward against harmful international tax competition. However, it is a political process; it is the pursuit of solutions that fit agreed-on standards that might not benefit all parties involved, with pillar 2 actually putting a package of unilateral measures on the table in an attempt to achieve global agreement on them.

Perhaps the ultimate proof that this is a political process is the realization that all pillars face a common enemy: unilateralism. Yet none of the pillars discussed to date explore the alternative of a consensus on switchover clauses or other retaliatory measures against unilateralism. None proposes that if a State A discriminates against non-treaty partner State B alleging that B encourages “abusive tax practices” (to quote the expression used by the EU to justify its own blacklist),22 B and all other jurisdictions will be entitled to impose retaliatory measures against A, not because of the merit of A’s concerns (which may be valid in principle), but because A adopted the policy prescription individually. If proposals to patch the international tax system never confront lone wolf initiatives that pressure other jurisdictions to modify their own domestic tax laws, those initiatives will likely always set the bases on which new proposals will be discussed.

After reading the last paragraph, some readers could ask whether this is a proposal for a pillar 5.

No, this is not a pillar 5, but my article also does not condemn recent pillar-building efforts per se. It is just a reminder that pillars should not be built on manufactured consent, because solutions to international tax problems should respect (and champion) the tax sovereignty of all jurisdictions involved. Building consensus should not be premised on agreed-on standards and proposals that fit the agenda of some jurisdictions to the detriment of others.

In short, the remedy to unilateralism in international tax policy should be the diligent, never-wavering pursuit of consensus solutions that reach unanimity or near-unanimity. If those solutions are unreachable, the OECD (or any other organization) should not yield to second-best options that are surely enticing for their revenue-generating potential but may prolongate and foment BEPS concerns to be solved under a BEPS action plan 3.0, 4.0, or 5.0 — or as many Greek temples as our imagination would allow.


1 See, e.g., “In Business, If You Build It, Will They Come?” Kansas City Business Journal, Sept. 10, 2018.

2 For prior analysis, see Lucas de Lima Carvalho, “GLOBE and the Supranational ‘Nudges’ Affecting Domestic Tax Policy,” Tax Notes Int’l, July 29, 2019, p. 421.

3 See Allison Christians and Tarcísio Magalhães, “It’s Time for Pillar 3: A Global Excess Profits Tax for COVID-19 and Beyond,” Tax Notes Int’l, May 4, 2020, p. 507.

4 See Nathan Boidman, “Boidman Offers Pillar 4: Abolish Corporate Taxes!Tax Notes Int’l, June 8, 2020, p. 1161.

7 See Pascal Saint-Amans, “Enhanced International Tax Cooperation Works,” Global Governance Project, June 6, 2018.

9 See Sebastian Valdenegro, “SII Informa que ya Comenzó a Regir el Nuevo IVA a los Servicios Digitales,” Diario Financiero, June 1, 2020 (in Spanish).

11 See Rafael Bitencourt, “Projeto na Câmara Propõe Cobrança Progressiva Sobre Faturamento Bruto,” Valor, May 5, 2020 (in Portuguese).

12 See Nana Ama Sarfo, “Developing Countries’ Search for a Post-Pandemic Tax Path,” Tax Notes Int’l, June 8, 2020, p. 1104.

13 See Yvette Lind, “Sweden and Denmark Incorporate Anti-Tax-Avoidance Rules Into Very Different COVID-19 Responses,” Tax Notes Int’l, June 8, 2020, p. 1127.

14 See EU, “Taxation: EU List of Non-Cooperative Jurisdictions” (updated Feb. 27, 2020).

15 See David Bradbury, Tibor Hanappi, and Anne Moore, “Estimating the Fiscal Effects of Base Erosion and Profit Shifting: Data Availability and Analytical Issues,” 25(2) Transnat’l Corp. 101 (2018).

16 Noam Chomsky and Edward S. Herman, Manufacturing Consent: The Political Economy of the Mass Media (2002).

17 Supra note 3, at 507.

18 Supra note 4. See also Boidman, “Is Corporate Tax Abolition Unrealistic?Tax Notes Int’l, Nov. 4, 2019, p. 433.

19 Boidman does not refer to a “global consensus” in his original piece; however, the abolition of corporate income taxes by Canada and the United States would pressure other governments to also consider abolishing theirs.

20 Recently, the OECD engaged in discussions with the Brazilian tax authorities to seek alignment between the OECD transfer pricing guidelines and the “fixed margin” approach used by Brazilian transfer pricing rules. To the surprise of Brazilian scholars, the OECD simply recommended full alignment of the Brazilian rules to its guidelines. See Sérgio André Rocha et al., “Brazil and OECD: Building a Future of Certainty and Equality — Public Statement Regarding the OECD-Brazil Transfer Pricing Project,” Kluwer International Tax Blog, July 30, 2019.

21 Supra note 4, at 509.

22 Supra note 14.


Copy RID