I. Introduction
Professors like to joke on the first day of international tax class that “there's no such thing as international tax.” Sadly, like most tax jokes, understanding it requires technical explanation and, even then, it is not funny. The joke relies on the observation that tax systems are creatures of national law. So, when we study “international tax” in U.S. law schools, we study how the United States taxes the foreign income of Americans, and we study how the United States taxes foreigners when they earn income in the United States. The implicit premise of the joke is that each state makes its tax law independently of other states. Increasingly, this joke is not only not funny; it is also not true.
This Article discusses the G20/Organisation for Economic Co-operation and Development (OECD) Base Erosion and Profit Shifting (BEPS) Project, a multilateral effort to combat corporate tax avoidance. Academics have harshly criticized BEPS, conceiving of it as a mere technical project to close tax loopholes.Footnote 1 Disputing these conclusions, this Article argues that by focusing narrowly on the technical recommendations arising from BEPS, commentators have overlooked its more profound implications. In particular, BEPS reflected—and to a significant extent operationalized—major changes in the participants, agenda, institutions, norms, and legal instruments of international tax.
First, BEPS expanded the actors who participate in international tax policymaking. For most of the twentieth century, international tax policy was made by a small club of mostly rich countries under the auspices of the OECD. But BEPS opened international tax to the G20, bringing a much-needed emerging-economy perspective to international tax. Second, BEPS significantly expanded the agenda of international tax policymaking, thereby commensurately expanding the importance of the role of the OECD. Third, for most of the twentieth century, the strongest norm in international tax was that companies should not pay tax twice, but BEPS both reflected and effectuated what might be called “full taxation,” a norm that income should not escape tax.Footnote 2Fourth, BEPS introduced novel forms of law, including an innovative multilateral treaty to update participating countries’ bilateral tax treaties and new forms of soft law and institutional structures uniquely suited to both preserving tax sovereignty and defeating state defection. Among these instruments are what this Article calls “fiscal fail-safes,” devices designed to achieve full taxation by assuring that if one state does not tax all of a company's income, another state will. Fifth, BEPS intensified old debates about the distribution among countries of the entitlement to tax income from international commerce.
The fracturing of consensus over a “century-old, mostly European gentlemen's agreement”Footnote 3—and in particular, a sharpening dispute between the United States and Europe over the taxation of digital giants, such as Facebook and Amazon—has prompted ongoing negotiations over the global tax split. The newly more inclusive institutions of international tax may not be fit for the task of mediating these disputes. BEPS thus leaves international tax at an inflection point.Footnote 4 For a hundred years, international tax has consisted of a collection of isolated national tax regimes, connected on a piecemeal basis by bilateral tax treaties that follow a model drafted by a small set of OECD member countries. The future could bring at least three possibilities—reversion to bilateralism as the dominant form of international tax policymaking, partial rejection of the status quo bilateral tax-treaty network paired with greater unilateralism and the reemergence of double taxation, or expansion and deepening of multilateralism.
Part II provides background on the taxation of multinationals, and it explains how legal regimes and economic developments contributed to corporate tax avoidance in the twentieth century. In the twenty-first century, revenue pressures created by the 2008 financial crisis combined with public backlash against corporate tax dodging to generate the political impetus needed to embark upon the multilateral BEPS Project. Part III argues that although the most obvious outcomes of this cooperative effort—the BEPS “deliverables”—may seem modest, that modesty obscures potentially transformative developments in international tax. By expanding the ambition of international tax cooperation and increasing its revenue rewards, the BEPS Project also turned up the heat on long-simmering disputes over how to divide the tax spoils of globalization. To date, those disputes remain unresolved. Part IV takes stock of the changes wrought by BEPS from the perspectives of revenue, inclusivity, legitimacy and accountability, policy innovation, and durability. It then considers potential outcomes of the current allocation dispute.
II. Tax Policy and the 2008 Crisis
Throughout the twentieth century, states faced a collective action problem, stemming from pressures to compete for mobile business and investment, that prevented them from cooperating to combat corporate tax avoidance. This reluctance to cooperate changed in 2008 when public investigations, a series of leaks, and the global financial crisis provided the political impetus for meaningful multilateral reform, which resulted in the BEPS Project.Footnote 5
A. The Basics of the International Tax System
The current international tax system—which consists of domestic tax regimes and an extensive network of bilateral tax treaties to connect them—relies on the concepts of source and residence. Corporate tax residence is clearly, albeit arbitrarily, defined as a company's place of incorporation or its place of management and control.Footnote 6 Under tax treaties and international custom, a state may tax a nonresident corporation only on income “sourced” in its territory.Footnote 7 Unlike source states, the company's state of residence may tax all of the company's worldwide income.Footnote 8 Treaties require residence states to relieve any resulting double taxation either by exempting income taxed at source or by crediting the source tax against residence tax due.Footnote 9 In the absence of tax treaties, residence states typically relieve double tax unilaterally. In addition to assuring such double-tax relief, tax treaties shift tax revenue from source to residence by constraining source tax entitlements; tax treaties also coordinate other tax administrative functions, including information sharing.Footnote 10
Although the OECD has only thirty-seven members, most countries follow the OECD Model tax treaty,Footnote 11 which traces its origins to the 1920s.Footnote 12 Under this model, states agree to two limits on source taxation of business profits. First, a state may tax nonresident companies only if the company has a “permanent establishment” there, meaning a physical presence or dependent agent.Footnote 13 Second, if a nonresident has a permanent establishment in the source state, then that state may tax only the income “attributable” to that permanent establishment.Footnote 14 To determine the attributable income, the source state imagines that the permanent establishment is its own entity, independent of its head office, and then imputes an “arm's-length” return to the permanent establishment. Widely followed throughout the world, the OECD Transfer Pricing Guidelines and other OECD guidance explicate this attribution process both for permanent establishments and separate legal entities in a multilateral corporate group.Footnote 15 Before BEPS, the OECD Model tax treaty, Commentary on that model, and the Transfer Pricing Guidelines represented the most important modes of multilateral standard-setting in corporate taxation.Footnote 16 Despite being soft law, adherence to these standards is widespread among both OECD and non-OECD countries. For example, the OECD Model serves as the basis for both the U.S. and UN model tax treaties.Footnote 17
B. Corporate Tax Avoidance
For most of the twentieth century, policymakers saw corporate tax avoidance as unproblematic, or regarded the costs of curbing it as too high. This section points to legal and economic factors that intensified and continue to intensify corporate tax avoidance; it explains that unilateral efforts to curb corporate avoidance in the last century were largely unsuccessful, and it identifies barriers to multilateral cooperation to curb corporate tax avoidance.
1. Intensifiers
Several factors intensify corporate tax avoidance, including legal, economic, and technological factors; the growing obsolescence of tax treaties; and differences in tax rules across states. These same factors also intensify competition among states both for paper profits, which generate tax revenue, and for real factors of production, which generate wealth and jobs for national residents.
Changes in law, economy, and technology. Liberalization of trade, commercial, and financial flows increase pressures on states to make their tax regimes appeal to highly mobile business and investment.Footnote 18 As this section explains, the rise of the internet and other communications technologies enables new forms of business, and the dominance of the world economy by highly integrated, intellectual-property-intensive global firms has increased taxpayers’ access to tax planning techniques that depend on exploiting differences in national tax laws.Footnote 19
Obsolescence of technical provisions. The obsolescence of technical provisions in tax treaties also contributes to corporate tax avoidance. The requirement for a physical presence or a dependent agent in the source state to satisfy legal nexus for tax purposes was well-suited to a bricks-and-mortar economy. But the replacement of offices, factories, and warehouses with immaterial websites and cheap third-party contracts for manufacturing, warehousing, and delivery has enabled multinationals to avoid tax nexus in source states. Companies’ lack of physical presence in a jurisdiction does not necessarily mean lack of engagement in the local economy, or lack of associated profits, but increasingly, states cannot reach those profits because companies do not satisfy a tax nexus requirement designed for an earlier era.
National fragmentation. Even when multinationals cannot avoid having a taxable presence in a state, they bring the considerable creativity (and chutzpah) of the tax bar to the task of minimizing their tax bills via brazen and inventive tax planning techniques that exploit differences in national legal regimes. Fragmentation of tax law into national systems facilitates both tax competition among states and aggressive tax planning. Following common practice, this Article uses “profit shifting” to refer to any strategy for avoiding tax by manipulating where a company is deemed for tax purposes to earn income. In a world where states set their tax rates independently from each other, a company that can choose where to declare its income also can choose its tax rate.
Multinational groups shift profits by, for example, moving valuable intellectual property to low-tax jurisdictions and then charging artificially high licensing fees to related companies in high-tax jurisdictions. The related group member in the high-tax state gets a large deduction, and, by design, the recipient of the fee is taxable in a low-tax state. The group's overall profit remains the same, but it saves tax due to the rate differential. This kind of payment between related companies is called a “transfer price.”
Other inventive tax-avoidance techniques take advantage of differences in legal regimes or definitions between different states. For example, Apple, the U.S.-headquartered technology giant, famously exploited a difference between the U.S. and Irish definitions of tax residence.Footnote 20 Ireland considered a company to be a tax resident only if it was managed and controlled in Ireland. In contrast, the United States determined residence by place of incorporation. By incorporating subsidiaries in Ireland, but managing and controlling them from the United States, Apple created companies that resided nowhere for tax purposes.Footnote 21 By shifting income to these stateless subsidiaries, Apple moved a large portion of its global profits to nowhere, thereby escaping tax.Footnote 22 Other companies were less audacious than Apple, but they were able to achieve the same functional result by using tax havens with no or very low corporate taxes.Footnote 23
2. Unilateral Measures
Although governments knew about these tax-avoidance techniques, before BEPS their unilateral responses did not combat them effectively. Competitive pressures and the influence of private business help to explain this lack of effectiveness. On the competitive front, source states that allowed a lot of profit shifting presumably calculated that they preferred inbound investment and business activity to the additional revenue that strict tax enforcement would have generated. An influx of jobs and investment might warrant ignoring excessive outbound deductible payments made to related parties. Moreover, powerful private interests opposed measures to counter tax avoidance.
In principle, the growing inability (or reluctance) of source states to tax multinationals should have enlarged residence states’ tax entitlements. But likewise catering to business interests and driven by competitive pressures, residence states also often did not tax.Footnote 24 As states competed with each other, headline corporate tax rates fell.Footnote 25 Although they typically could not afford to offer the extremely low across-the-board rates found in tax havens, otherwise high- or middle-tax states lowered their tax rates and introduced preferential tax regimes to attract mobile activities.Footnote 26
As the economy globalized, states’ losses from failure to curb tax avoidance mounted. Mere paper profit shifting generated revenue losses, but factors of production also moved as companies responded to tax incentives.Footnote 27 Unilateral efforts to counter tax avoidance were not very successful. For example, in the late 1960s, the United States passed so-called controlled-foreign-corporation (CFC) rules.Footnote 28 Under the CFC rules, shifting profits to or among foreign subsidiaries triggered tax for the U.S. parent on the shifted income, at high pre-2017 U.S. corporate tax rates.Footnote 29 But in the late 1990s, the Treasury Department gutted the CFC rules when it implemented the “check-the-box” regulations. Weary of fighting with taxpayers about how entities should be characterized for tax purposes, Treasury decided that, within certain limits, U.S. taxpayers could elect whether their business entities would be taxed separately or as pass-throughs.Footnote 30 The only thing that companies—including foreign companiesFootnote 31—had to do was file a form and check the box for their desired treatment.Footnote 32 A tax-planner's dream, but a tax official's nightmare, check-the-box provided an easy way to manufacture hybrid entities that were taxed as corporations in one jurisdiction but passthroughs or branches in another. Without check-the-box, shifting income among foreign affiliates would potentially generate tax to the U.S. parent under the CFC rules. By checking the box to regard a foreign entity as a branch of its parent, however, the U.S. parent could make profit-shifting payments disappear—for U.S., but not foreign—tax purposes.Footnote 33 Check-the-box defanged the CFC rules, enabling U.S.-parented companies to shift profits between foreign affiliates with impunity.
In part to counter profit-shifting, the United States also developed—and convinced the rest of the world to accept—the arm's-length standard.Footnote 34 Memorialized in hundreds of pages of OECD Transfer Pricing Guidelines, as well as various national rules and regulations, the arm's-length standard limits profit shifting by requiring each member of a corporate group to report a market, or arm's-length, transfer price on transactions.Footnote 35 Determining the arm's-length price requires estimating what the price would have been if the members of the group were unrelated.Footnote 36 Analysis of comparable transactions among unrelated parties helps taxpayers and tax administrators determine this figure,Footnote 37 but the arm's-length standard has faced significant and warranted criticism because arm's-length comparables are often unavailable and other methods for estimating appropriate transfer prices are unreliable or subjective.Footnote 38 Fuzzy standards and methods provide taxpayers significant discretion over where to report their income. Even when governments contest transfer prices, they have found it difficult to persuade courts that their estimations of arm's-length prices ought to prevail over taxpayers’ estimations.Footnote 39 Indeed, the arm's-length standard is said to produce a “range” of correct answers, such that the choice of a point within that range is arbitrary.Footnote 40 Courts are understandably reluctant to impose unfavorable results on taxpayers based on subjective, unclear, or arbitrary standards and regulations.Footnote 41 The subjectivity of the standard and taxpayers’ successes in fending off adjustments by governments have emboldened taxpayers to aggressively manipulate transfer pricing in order to shift income.Footnote 42
Similar to the United States, other countries passed limited anti-abuse measures meant to stem the leakage from their own tax systems,Footnote 43 and OECD countries worked together to strengthen and revise the OECD Transfer Pricing Guidelines.Footnote 44 But on the whole, and motivated by competitive pressures and pressure from corporate special interests, most states took an indulgent attitude toward corporate tax avoidance. The history of check-the-box serves as an instructive example. One consequence of check-the-box was that it enabled U.S. companies to pay less foreign tax, without suffering any U.S. tax penalty. U.S. lawmakers became convinced that such foreign tax savings could help U.S. companies compete against companies headquartered elsewhere. In addition, lower foreign taxes for U.S. companies itself lowered the U.S. obligation to credit foreign taxes. This gave the United States both competitive and revenue reasons to turn a blind eye to U.S. multinationals’ foreign profit shifting. Other countries never retaliated for check-the-box, and efforts by U.S. tax officials to claw back the benefit faced major opposition from multinationals.Footnote 45
3. Barriers to Cooperation
Tax competition and corporate tax avoidance represented collective-action problems that no one state could tackle unilaterally without driving out valuable productive factors, such as jobs and capital. Yet numerous barriers prevented states from working together cooperatively; these barriers included heterogeneity of state interests, monitoring problems, reluctance to cede national tax sovereignty, lack of clear ideas about how to close even notorious tax gaps, inability to grasp the scope of the problem, interest-group capture, and views that tax competition was normatively desirable.
State heterogeneity and monitoring costs. Although states made some efforts to curb tax competition and corporate tax avoidance, the unequal distribution of the potential benefits from cooperation (which would mostly redound to high-tax states) constituted a major impediment to collective action. Small states with few natural resources used taxes to compete for inbound investment that they could not otherwise win; some even become tax havens.Footnote 46 Such countries could not be persuaded to cooperate by the promise that cooperation would yield more tax revenue. Instead, overcoming their incentives to defect required either strict monitoring and penalties or side payments. Side payments were politically unpalatable, as were penalties, and monitoring posed special challenges in taxation because competition takes place on a variety of fields, some nontransparent. For example, while headline rates are relatively transparent, states can also attract taxpayers through lenient tax enforcement, favorable legal interpretations, and even secret tax deals.Footnote 47
Tax sovereignty. Harmonizing tax laws across countries would eliminate tax competition, but concerns about preserving national tax sovereignty scuttled multilateral efforts to combat tax avoidance through harmonization.Footnote 48 Tax sovereignty encompasses several concerns, some normatively desirable, some not.
First, ceding authority to a supranational body to determine important tax policies would reduce national legislators’ ability to respond to voter preferences, which differ across states. The existence of such differences in preferences represents the most powerful normative argument in favor of tax autonomy and, by extension, tax competition. Second, and less normatively desirable, national lawmakers may be rationally reluctant to shift decision-making power to supranational institutions because doing so reduces their own opportunities to deliver particular tax-policy outcomes to interest groups. While accommodating rent-seeking at the national level may not be normatively desirable, it is not clear than rent-seeking at the supranational level would be any better,Footnote 49 and in any case, the desire to retain control over rents at the national level impedes international cooperation.
Concerns regarding national tax sovereignty apply regardless of state size. Tax autonomy enables small states to defect from cooperative solutions that are unlikely to benefit them.Footnote 50 Tax autonomy likewise enables large, and especially rich, countries to pursue a divide-and-conquer strategy with respect to smaller and poorer states. This can be seen in tax treaties, which shift tax entitlements from poor to rich states as compared to the no-treaty situation.Footnote 51 Yet rich states’ retention of tax sovereignty—which enables them to exploit their power asymmetry in other tax areas—imposes a cost; it weakens their ability to respond cooperatively to tax competition.
Legality of tax gaps. The perceived legality of corporate-tax-avoidance strategies also served as a barrier to collective action. Unlike tax evasion, which typically includes fraud or other overt law-breaking, tax avoidance techniques do not violate the law of any state. Instead, they often merely exploit differences in different states’ laws.Footnote 52 Apple's stateless companies are a perfect example—Ireland defined tax residence one way, the United States another. When income fell through the gaps between the two states’ systems, neither state felt obliged to catch it.Footnote 53 The effects of check-the-box were similar—why should another country care how the United States characterized an entity for tax purposes? As long as neither state regarded itself as losing revenue owed to it, why should either be concerned about the taxpayer's activities or income elsewhere?Footnote 54 As David Rosenbloom put it, “The beauty of international tax arbitrage, when practiced most skillfully, is that none of the objections to aggressive or abusive tax planning should apply anywhere because, from the vantage point of any single country, there is neither aggressiveness nor abuse.”Footnote 55 There was also a question of how countries could close loopholes without harmonizing their tax systems, which they did not want to do.Footnote 56 Who but the Irish could tell Ireland how to define corporate tax residence? Every state possessed autonomy to create the tax system its voters found most suitable, and that autonomy led inevitably to gaps and overlaps in taxation. The very legality of aggressive tax planning thus made it difficult not only to conceive of solutions, but even to recognize that there was a problem that needed to be solved.
Hidden scope. Yet another reason for complacence in combatting tax avoidance was that officials did not understand the full scale of the problem. The scope of corporate tax avoidance was obscured not only by its legality, but also its elusiveness. Economists’ estimates of profit shifting varied widely, as did estimates of revenue losses, reflecting both gaps in data and problems with cross-country comparability data that did exist.Footnote 57 Because companies report taxable income at the national, not global, level, a particular state's tax authorities could not see the full income picture for any multinational.Footnote 58 So while economists observed indirect evidence of income shifting—for example that the foreign direct investment into tax havens like Bermuda and the Cayman Islands was orders of magnitude higher than would be expected given the features of those countries’ economiesFootnote 59—no one knew how much income was declared in those havens, how much income was declared nowhere, or how much income, if any, a particular state lost from corporate tax avoidance.Footnote 60
Regulatory capture. Powerful, well-organized corporate special interests also resisted effective responses to aggressive corporate tax planning, whether unilateral or cooperative, and there is ample evidence that companies play jurisdictions against each other to get the best tax deal.Footnote 61
Normative reluctance. The traditional argument in favor of cooperation to curb tax competition is that, without it, states will be unable to meet their domestic social welfare commitments.Footnote 62 But empiricists had trouble showing such impacts, and proponents of tax competition pointed to its normative benefits, including that it acts as a salutary check on big government, an efficient means to satisfy voter preferences that differ across jurisdictions, and an efficient response to the mobility of certain investments.Footnote 63
Conclusion. Due to these barriers, a project at the OECD in the late 1990s to curb tax competition made important, but only limited, progress.Footnote 64 Likewise, a contemporaneous project within the EU to use state-aid enforcement to curb tax competition petered out.Footnote 65 Moreover, when countries undertook cooperative efforts to combat tax competition, because they were guided by their own interests in securing inbound investment and advantages for their companies, they inevitably disagreed about what forms of competition ought to be prohibited.Footnote 66 Thus, while states cooperated to eliminate tax overlaps that led to double taxation, they left tax gaps unresolved. Several states even made a tidy living by deliberately creating tax gaps for multinationals to exploit.Footnote 67 And when other states inadvertently created loopholes, such as check-the-box, they found that interest groups sprang into action to prevent repeal.Footnote 68
C. Crisis and the Road to Cooperation
Given these barriers to cooperation to curb corporate tax avoidance, the goals of transnational tax governance in the twentieth century were modest: to continually and incrementally improve common standards, such as the OECD Model, Commentary, and the Transfer Pricing Guidelines. International tax was largely ignored by foreign ministers and even, to a surprising extent, by finance ministers. Tax bureaucrats tinkered with technical challenges, but they did not undertake fundamental reform of the taxation of multinationals.
The tranquility of the tax bureaucrats and the complacency of political leaders ended abruptly with the 2008 financial crisis, which triggered job losses, budget and monetary crises, and a new intolerance of corporate tax dodging.Footnote 69 Leaks and public investigations increased scrutiny of corporate taxation and galvanized decisive action by national leaders. Voters who had mostly ignored cross-border tax issues suddenly started paying attention. This section explains that these new factors overcame previous barriers to cooperation, paving the way for the BEPS Project, a major multilateral effort to curb corporate tax avoidance. Changes in preexisting institutions—including the G20 and the EU—also facilitated multilateral cooperation.
Investigations by the United States and United Kingdom of corporate tax avoidance galvanized popular attention, especially among European voters, who had experienced years of austerity budgets following the 2008 crisis. In September 2012, the Senate's Permanent Subcommittee on Investigations held hearings on corporate tax avoidance that included testimony from executives from Microsoft and Hewlett Packard.Footnote 70 Less than two months later, the Committee of Public Accounts of the British House of Commons conducted similar public hearings, calling executives from Amazon, Google, and Starbucks to explain why their companies reported such small profits in the United Kingdom, despite their apparent success there.Footnote 71
The explosive parliamentary hearings included testimony from a Starbucks executive that the company had received a “low-tax ruling” from the Netherlands, but that the Dutch tax authority had “asked [Starbucks] not to share [it] publicly.”Footnote 72 The executive also insisted that, despite a dramatic expansion in the number of Starbucks stores in the United Kingdom, the company had been unprofitable in the United Kingdom for fourteen out of fifteen years of operation.Footnote 73 The hearings made headline news around the world and led to boycotts of Starbucks by British consumers upset with the company's payment of “single-shot taxes on Venti-size sales.”Footnote 74
A second set of Senate hearings revealed Apple's stateless foreign subsidiaries; these subsidiaries filed full tax returns nowhere on Earth.Footnote 75 At the same hearings, an Apple executive also testified that Apple had negotiated a special tax rate with Ireland, although the company later retracted this claim.Footnote 76 Newspapers carried front-page stories on the tax-avoidance efforts of Apple and other prominent companies.Footnote 77 “Double Irish with a Dutch sandwich” entered the vernacular as readers pored over complex diagrams published in major newspapers around the world.Footnote 78
These revelations inflamed public sentiment against corporate tax dodging, as did disclosure by a PwC whistleblower of a trove of secret tax administrative rulings granted by the Luxembourg tax authority to clients of PwC.Footnote 79 Although many of the leaked tax rulings were ordinary administrative guidance that merely confirmed application of Luxembourg law to a particular taxpayer's facts,Footnote 80 others appeared to grant taxpayers secret sweetheart tax deals.Footnote 81 In the public eye, this material, known as LuxLeaks, seemed to confirm suspicions that companies were not paying their fair share of taxes and, crucially, that certain governments were complicit in helping companies to avoid tax.Footnote 82 Suddenly, international tax became not only dinner party conversation but also the subject of protests in the street. In light of this “mainstreaming”Footnote 83 of international tax, the G20 countries needed to do something—or at least appear to do something—about corporate tax avoidance. In 2012, the G20 leaders expressed strong support for international cooperation to combat tax avoidance.Footnote 84 Lacking a permanent staff, the G20 delegated to the OECD the task of coordinating multilateral efforts, initiating the BEPS Project.Footnote 85 Within months, the OECD issued an ambitious Action Plan identifying fifteen areas of corporate tax needing reform.Footnote 86 The goal of BEPS was to combat profit shifting, ostensibly by “align[ing] … returns with value creation.”Footnote 87 The improbable deadline for most of the project was just two years, during which the G20 and OECD countries envisioned agreeing to coordinated solutions.Footnote 88 Stakeholders predicted that the OECD would fail to meet the deadlines to deliver draft BEPS proposals, and that, even if the OECD did meet its deadlines, the project would not result in significant changes.Footnote 89 These were safe predictions.
In addition to the revenue pressure created by the global financial crisis and the popular backlash from new revelations of aggressive corporate tax planning, events leading up to BEPS facilitated states’ efforts to coordinate their responses to corporate tax abuse. For example, in 2011, to better respond to the financial crisis, the G20 had shifted from a meeting of finance ministers and central bankers to one of national leaders. The G20 closely followed and strongly backed progress on BEPS,Footnote 90 enabling OECD officials to expand their own authority by claiming they had a “mandate” from the G20.Footnote 91
Likewise, successes in multilateral tax cooperation at the OECD on tax information-sharing and enforcement in the immediate pre-BEPS era paved the way for cooperation on corporate tax avoidance by increasing confidence regarding compliance and monitoring.Footnote 92
The late twentieth and early twenty-first centuries also saw a dramatic expansion of the role of EU institutions in international tax. The EU promotes tax cooperation in several ways. First, the ability of the member states to pool their sovereignty and the role of the Commission in advocating pan-European interests has enabled the EU to serve as an important counterweight to the United States in international tax relations, including by forcing the United States to the bargaining table on important issues such as the taxation of digital companies.Footnote 93 Second, EU-level legislation provides a streamlined and simultaneous way to implement agreed reforms in all member states, reducing the risk of defection.Footnote 94 Third, EU-level tax policymaking increases opportunities for issue linkage—that is, it increases opportunities to bargain across policy areas and time periods. Fourth, by threatening to use its enforcement powers to undo coordinated—but malleable—standards, such as the arm's-length transfer-pricing method, the Commission may have encouraged EU member states to cooperate at the OECD to forestall deeper Commission involvement in tax administration.Footnote 95 At the same time, the threat that the Commission might undermine or replace the arm's-length standard also motivated non-EU member states—especially the United States—to devise cooperative solutions that could discourage the EU from taking a greater role in tax policy.Footnote 96
These shifts in the economic, political, and institutional landscape overcame barriers to cooperation. The technical ability and policy ambition of OECD staff, the political resolve and crisis-fueled revenue needs in the OECD and G20 states, and voter dissatisfaction with corporate tax dodging proved a potent combination. Over the next two years, tax officials from the OECD and G20 countries hammered out proposals to curb aggressive tax planning, close tax gaps, and increase states’ ability to monitor both multinationals and each other through various tax transparency measures.Footnote 97 Skillful leadership at the OECD and participating states’ willingness to dedicate significant technical and political resources to finding pragmatic solutions meant that the OECD met its deadlines, generating reports and deliverables on every action item in just two-and-a-half years.Footnote 98
III. The Transformation of International Tax
As noted in the introduction, many observers understand BEPS as merely a gap-filling project that patched technical rules while missing the opportunity for fundamental reform. This Part argues that this view fails to recognize the potentially profound effect of BEPS on the decisionmakers, agendas, institutions, norms, and legal forms of international tax.Footnote 99
A. New Participants in Multilateral Tax Policymaking
The BEPS Project brought the G20 and OECD countries together “on an equal footing”Footnote 100 to make international tax policy. The OECD—a group that only recently reached thirty-seven membersFootnote 101—long has been criticized for excluding other countries from international tax policy decisions.Footnote 102 The G20 added to the OECD eight additional countries—Argentina, Brazil, China, India, Indonesia, Russia, Saudi Arabia, and South Africa.Footnote 103 These countries collectively have a population of over 3.5 billion and they bring to the OECD a developing-country perspective that was sorely lacking.Footnote 104 Moreover, as negotiators finalized the BEPS recommendations, the G20 insisted that the OECD take an inclusive approach, inviting more countries into the post-BEPS monitoring and standard-setting process. The BEPS Inclusive Framework was established; its membership currently stands at 137 states.Footnote 105 The Inclusive Framework is open to countries that commit to implementing the BEPS minimum standards.Footnote 106
Nongovernmental organizations (NGOs) have also taken on an increasingly prominent role in tax policymaking, including in forums provided by the OECD, such as various public discussion sessions and open invitations for public commentary that were held regarding each action item.Footnote 107 Watchdog groups, such as the Tax Justice Network and the BEPS Monitoring Group, emerged to attempt to influence the outcome of BEPS.Footnote 108
Although they perhaps cannot properly be called participants in the policymaking process, journalists, including the International Consortium of Investigative Journalists that revealed the global scale of tax evasion and corporate-tax avoidance,Footnote 109 galvanized popular attention and helped make reform possible.
B. Expanded Role for Multilateralism and the OECD
1. International Tax Agenda
Dominated by the United States for the better part of a century, the work of the tax division of the OECD prior to BEPS was the stuff of technical experts—of geeks, not diplomats.Footnote 110 BEPS expanded the international tax policy agenda, including by making extensive recommendations for changes to domestic law.Footnote 111 This was true for some of the minimum standards that emerged from BEPS, such as administrative-ruling disclosure and country-by-country reporting,Footnote 112 as well as certain non-minimum standards, including the interest-deductibility and anti-hybrid rules.Footnote 113 As will be discussed later, the agenda of the Inclusive Framework has now moved well beyond BEPS to how to redistribute the entitlement to tax income from cross-border commerce.Footnote 114
2. Role of the OECD
One important aspect of international tax policymaking has not changed: the dominance of the OECD as a forum.Footnote 115 The role and budget of the organization have expanded along with the composition of the decisionmakers and the agenda of international tax.Footnote 116 BEPS enabled the OECD to leverage its technical tax expertise into a broad mandate to guide international tax reform efforts—not only for its own thirty-seven members, but for over a hundred nonmember countries. The OECD facilitates international cooperation by providing deep technical expertise, effective monitoring of state compliance with agreements, and a forum for iterated negotiations that also allows for issue linkage.
C. Acceptance of Full Taxation as a Norm
Before BEPS, cooperation in international tax aimed principally to lower transaction costs for cross-border economic actors by preventing double taxation, reducing compliance costs for companies, and providing legal mechanisms to resolve disputes between taxpayers and tax administrations.Footnote 117 States overcomplied with the no-double-tax norm, resulting in widespread nontaxation of income.Footnote 118 Today, however, states increasingly support not only a no-double-tax norm, but also what this Article will call a “full taxation” norm that dictates that all of a company's income should be taxed in places where it has real business activities.Footnote 119 Full taxation is a shorthand that encompasses other frequently used terms that aim at the same idea, including “prevention of profit shifting,” prevention of “aggressive” or “abusive” tax planning, closing of “loopholes” or “gaps,” and the particularly inelegant “prevention of double nontaxation.”Footnote 120 In the content and form of its recommendations, BEPS both confirmed and operationalized full taxation as a new international tax norm. This section explains how the BEPS proposals endorsed full taxation.
The BEPS Project resulted in two kinds of recommendations: discretionary reforms and minimum standards, the latter of which the states participating in the Inclusive Framework agreed to implement.Footnote 121 Nearly all of the BEPS recommendations—and three out of four of its minimum standards—aim to secure full taxation.Footnote 122 For example, one minimum standard promotes tax transparency through several related measures, including by requiring country-by-country reporting. Before BEPS, companies reported to a state only profits and activities that took place in that state. Tax administrators lacked a complete picture of a multinational's global activities, which limited their ability to identify, let alone combat, income shifting and stateless-income planning. Multinationals must now submit country-by-country reports, which provide governments a global and per-country overview of profits, sales, employees, income, and where companies declare income and pay taxes.Footnote 123 These reports aid tax administrators in assessing whether companies shift profits; the reports can therefore help target enforcement activity. By heightening audit, litigation, and reputational risks for companies showing large profits in havens (or nowhere), the reports presumably will discourage profit shifting.
Another aspect of this minimum standard requires BEPS states to automatically forward to each other summaries of certain previously secret unilateral administrative tax rulings, including rulings applying the notoriously flexible arm's-length standard.Footnote 124 Before BEPS, a taxpayer could secure a ruling from State A in which the state agreed not to tax certain income because that income was more properly allocated to State B. But nothing forced the taxpayer to declare the income in State B. Thus, taxpayers could take inconsistent positions in different states about the same income. The exchange of tax rulings give states notice of administrative decisions made elsewhere that may affect domestic tax liability; it also limits taxpayers’ planning options. By increasing transparency, the exchange of rulings also reduces incentives for states to collude with taxpayers trying to avoid foreign tax.Footnote 125
A second minimum standard pursues full taxation by enhancing efforts to prevent abuse of tax treaties. For example, states agreed to change the preamble of tax treaties to reflect that their purpose is not merely to avoid double taxation, but also to prevent tax evasion and avoidance.Footnote 126 Likewise, states agreed to include stronger anti-abuse rules in their treaties.Footnote 127 A third minimum standard seeks to curb harmful tax competition by demanding a closer connection between business activities and entitlements to tax benefits.Footnote 128 Taxpayers must have “substantial activities” in a jurisdiction before they can receive certain tax preferences there. This prevents income shifting by limiting divergence between where a business conducts its activities and where it declares its income, including by limiting special low-tax “patent boxes.”Footnote 129 Under a fourth minimum standard—the only one aimed primarily at preventing double taxation, rather than ensuring full taxation—states agreed to improve tax-treaty dispute resolution mechanisms.Footnote 130
Although not representing minimum standards, other BEPS reforms—including revised transfer-pricing rules—aim to ensure full taxation by reducing opportunities for profit shifting.Footnote 131 BEPS also closes stubborn loopholes. For example, its anti-hybrid rules help combat the U.S. check-the-box rules, which the United States retained. Countries also developed model interest-deductibility rules and a set of best practices for CFC rulesFootnote 132 that aim to reduce profit shifting.Footnote 133 Although not formally part of the BEPS Project, Ireland succumbed to pressure to change its tax-residence rule that facilitated Apple's stateless income planning.Footnote 134
Commentators analyzing individual BEPS recommendations may reasonably conclude that none is by itself transformative. Yariv Brauner, for example, criticized the OECD for failing to “include a clear articulation of double non-taxation as a fundamental principle of the international tax regime.”Footnote 135 Taken collectively, however, the BEPS recommendations reflect not only an extraordinary new commitment to coordinating domestic tax rules, but also an unprecedented commitment to full taxation.
Because states already faithfully adhered to the no-double-tax norm, growing acceptance of full taxation as a goal of international tax brings states closer to implementing Reuven Avi-Yonah's “single-tax principle,” which requires income to be taxed exactly once, no more and no less.Footnote 136 Unfortunately, like its foil “no double taxation,” the concept of “full taxation” is underspecified. While the notion that a company should be subject to “full” but not “double” taxation seems unobjectionable, because there is no global benchmark for defining taxable income and because there is no ideal tax rate, views diverge as to when these standards have been met.Footnote 137 Despite its intuitive appeal, the concept of full taxation raises a number of problems that will be explored further below.Footnote 138
D. Novel Legal Forms
BEPS required multilateral tax cooperation on a scale never seen before. But it did not—nor was it expected to—result in the genesis of new supranational tax institutions that could adjudicate tax disputes or legislate tax policy. Acknowledgments by public officials of states’ inability to “go it alone” in international tax became commonplace in the BEPS era,Footnote 139 but officials remained reluctant to breach the last bastion of national sovereignty. For example, at a recent press conference with U.S. Treasury Secretary Steven Mnuchin, French Finance Minister Bruno Le Maire declared, “Let's accelerate work at an international level.”Footnote 140 But when asked about France's recent adoption of unilateral digital taxes that arguably discriminate against U.S. companies, Le Maire retorted, “France is a sovereign nation that decides its own tax rules.”Footnote 141 These comments reflect a tension between the growing need for states to cooperate to achieve shared goals and their reluctance to relinquish exclusive national control over taxation.
To coordinate action in the areas where they agreed, while simultaneously protecting their core tax prerogatives, the BEPS states relied on innovative legal forms that avoided the “fashionable despair” of multilateral tax cooperation.Footnote 142 Some of these legal forms—such as peer review and soft law measures—are familiar to international law and have been used previously by OECD tax initiatives. But others are novel, including the Multilateral Instrument that modifies bilateral tax treaties. States also relied on what this Article calls “fiscal fail-safes,” in which the failure of one state to tax triggers tax by another state, thereby assuring full taxation. Although unilateral fiscal fail-safes were previously found in domestic law, coordinated implementation of harmonized fiscal fail-safes enabled BEPS countries to more effectively achieve full-tax goals by deterring both state defection and aggressive tax planning.
1. The Multilateral Instrument
Some BEPS recommendations required changes to bilateral tax treaties. To implement these changes expeditiously without renegotiating thousands of bilateral treaties, the BEPS countries invented a new international law instrument.Footnote 143 Signed in June 2017, the Multilateral Convention to Implement Tax Treaty Related Measures to Prevent BEPSFootnote 144 amends and updates the signatories’ bilateral tax treaties without supplanting them.Footnote 145 With eighty-eight signatories so far, the Multilateral Instrument will impact over 1,500 tax treaties.Footnote 146 It provides a way for participating countries to quickly implement certain BEPS minimum standardsFootnote 147 and other BEPS recommendations.Footnote 148 By leaving preexisting bilateral treaties intact, while updating them, the Multilateral Instrument embodies the pragmatic innovation and accommodation of sovereignty concerns that typified BEPS.Footnote 149 It also may represent a significant step toward, and a proving ground for, a more ambitious multilateral tax treaty. Although the Multilateral Instrument has detractors,Footnote 150 and the United States is not a signatory,Footnote 151 most recognize that it has, as Allison Christians put it, the “potential to permanently alter the architecture of international tax relations.”Footnote 152
2. Coordinated Unilateralism
BEPS, and international tax more generally, follows a long tradition of what might be called coordinated unilateralism.Footnote 153 Coordinated unilateralism includes nonbinding standards that reflect pooled technical expertise, are set cooperatively, and are implemented via domestic law or administrative regulation rather than a treaty. Many BEPS recommendations involve coordinated modifications to domestic law, including the model interest-deductibility, anti-hybrid, CFC, and various disclosure rules.Footnote 154 Other BEPS recommendations were implemented partly through treaties, and partly through domestic law. For example, country-by-country reporting, a BEPS minimum standard, required both domestic reporting infrastructure and new treaties or treaty provisions to exchange the reports.Footnote 155
As others have observed,Footnote 156 coordinating domestic rules may reduce or circumvent domestic opposition to international cooperation, by, among other things, lowering the stakes of cooperation.Footnote 157 This could be particularly important for states uncertain of whether they will be able to build capacity to comply with some of the more onerous BEPS standards.Footnote 158 Coordinated unilateralism also leverages the ability of tax administrations to implement policy without enlisting other parts of government. For example, the U.S. Treasury Department took the position that it had authority to implement country-by-country reporting without specific enabling legislation or a new treaty.Footnote 159 Such coordinated administrative action facilitates cooperation by enabling domestic legislators to avoid costly votes in favor of stricter tax standards that are likely to be unpopular with corporate campaign donors.
Coordinated unilateralism makes sense for policies that are known to be more effective if more states adopt them, including any policy that generates network effects. Tax treaties themselves exhibit network effects,Footnote 160 as do many of the BEPS recommendations, including tax-rulings exchange, country-by-country reporting, anti-abuse rules, CFC regimes, and more.Footnote 161 As network effects accrue, benefits from cooperation increase. By facilitating rolling implementation, coordinated unilateralism also facilitates later cooperation by countries for whom the benefits of joining only marginally exceed the costs.
3. Fiscal Fail-Safes
BEPS facilitated the widespread adoption of a particular type of coordinated unilateral provision, which this Article will call “fiscal fail-safes.” A fiscal fail-safe provides conditions under which if one country does not tax, another country fills the tax void. By automatically filling tax gaps—and thereby clawing back the benefit of tax planning and state-provided tax incentives—fiscal fail-safes discourage both aggressive tax planning and tax competition. The fiscal fail-safe concept unites a seemingly diverse and unrelated set of tax anti-abuse rules adopted both within and outside of BEPS. Fiscal fail-safes not only share a mechanism—linking triggers—but they also share a goal, namely, full taxation. Thus, the increasing prevalence of fiscal fail-safes both reflects and effectuates the growing acceptance of full taxation as a new global tax norm.Footnote 162
The BEPS anti-hybrid rules exemplify the use of fiscal fail-safes to neutralize state defection from full-tax goals. Among other tasks, the anti-hybrid rules help defeat debt-equity arbitrage.Footnote 163 Debt-equity arbitrage takes advantage of the fact that because debt and equity exist on a legal spectrum, states may take opposite views of the same financial instrument, with one state seeing debt where another sees equity.Footnote 164 Such hybrid instruments can generate deductible interest when paid out of one jurisdiction, but excludable dividends when received in the other.Footnote 165 When made between affiliates of the same corporate group, payments on such hybrid instruments generate deductions in one state with no corresponding inclusion in the other state. From a global perspective, hybrids thus make income disappear for tax purposes.
The BEPS solution to hybrids is a “primary rule” that directs the state from which the payment is made to deny an interest deduction unless the payment will be included in the recipient's state.Footnote 166 Thus, the rule expressly links the tax treatment in one state to the tax treatment in another state. But what if the payer state is a jurisdiction that refuses to adopt the hybrid rules because it wants to help taxpayers to avoid tax?Footnote 167 The BEPS recommendation includes a “defensive rule” to be applied by the recipient state, but only if the payer state does not—for whatever reason—shut down the hybrid. In that case, the defensive rule directs the recipient state to include the payment in the recipient's income.Footnote 168 Each of these rules functions as a fiscal fail-safe; it fills what would otherwise be a cross-border tax gap. As long as at least one of the payer's state or the recipient's state adopts the BEPS anti-hybrid regime, the loophole will be closed, and the payment will be subject to tax.Footnote 169 This kind of innovative, pragmatic problem-solving typified BEPS. Similar anti-hybrid rules counter hybrid entities, some of which arise from U.S. check-the-box rules.Footnote 170
By triggering special tax treatment in cases of tax gaps, the hybrid rules enable states to soak up tax entitlement not exercised by another state. Thus, the primary and defensive rules enforce full taxation more effectively than would either rule by itself, and they work automatically, requiring no additional coordination between the states. Most importantly, by formulating the hybrid rules as a fiscal fail-safe, the BEPS states averted the need for states to harmonize their definitions of debt and equity. Instead of finding agreement on the thorny debt-equity distinction that has vexed generations of tax policymakers, the states merely agreed on a result—full taxation—and a method of achieving it.
By averting the need to harmonize definitions of debt and equity, the hybrid rules also made tractable a previously intractable tax arbitrage problem. Debt-equity arbitrage arose from two different kinds of failures to cooperate. First, high-tax states were unwilling to defer to each other or to collective judgement about what distinguished debt from equity—even though doing so would curb arbitrage. These states were not using their definitions of debt and equity as a vector of tax competition; instead, those definitions addressed domestic tax concerns that swamped international tax concerns. These states’ refusal to coordinate opened an unintended tax gap.
The second type of gap was intentional. Other states essentially sold their sovereign ability to deem instruments to be either debt or equity for the specific purpose of facilitating tax abuse. By commercializing their sovereignty, these states encouraged taxpayers to locate hybrid instruments within their respective jurisdictions, where they generated at least some tax revenue.Footnote 171 In effect, these states competed for paper profits. The states that opened gaps unintentionally and those that did so intentionally could not be expected to agree on a single conception of the debt-equity distinction, but the motivation of each type of state differed. The states that created unintended gaps had a common interest in defeating intentional gaps opened by the second set of states. By agreeing to a fiscal fail-safe that would close gaps without requiring harmonization of debt-equity distinctions, the first set of states could eliminate the gains redounding to the states that intentionally created gaps. If structured properly, anti-hybrid rules allow states of the first type to share revenue that otherwise would have been lost to states of the second type. At the same time, anti-hybrid rules make defection by states in the second group ineffective.
Several characteristics typify fiscal fail-safes. First, they include a mechanism to link tax treatment across countries.Footnote 172 Second, fiscal fail-safes specify the conditions under which tax treatment in one state triggers a response in the other. Third, they specify special tax treatment, which constitutes a deviation from the implementing state's ordinary tax treatment of the relevant income (e.g., the state treats as taxable interest a payment that legally qualifies as a profit distribution). Fourth, the goal of the triggered treatment is normative; it aims to achieve full taxation or otherwise curb abuse.Footnote 173
CFC rules, described above, function as fiscal fail-safes. Disclosure of information about tax rulings and country-by-country reporting also may act as fiscal fail-safes by providing states an opportunity to link their own tax treatment of a transaction to another state's treatment of the same transaction, thereby reducing the ability of taxpayers to take inconsistent positions in two states regarding the same income.Footnote 174 But not all anti-abuse rules constitute fiscal fail-safes. Thin-capitalization rules, substance-over-form rules, limitations-on-benefits provisions, and many other anti-abuse rules fall outside the definition of fiscal fail-safes.Footnote 175 And there may be more effective methods than fiscal fail-safes to achieve full taxation. For example, adopting a system that allocates income completely with no gaps—such as formulary apportionment—might better achieve full-tax goals.Footnote 176 Given mismatched national regimes, however, fiscal fail-safes can be an effective way to bridge gaps.
Fiscal fail-safes may also take the form of administrative rules. For example, by building a regulatory fail-safe into the country-by-country reporting rules, the BEPS participants helped ensure speedy adoption, even by countries, such as the United States, that initially resisted the measure.Footnote 177 Under the country-by-country reporting regime, a multinational must file certain reports with its headquarters state, but if its headquarters state does not participate in country-by-country reporting, then the company must file its report in another state in which it has a subsidiary, thereby reducing the possibility that a multinational could entirely avoid compliance with BEPS reporting requirements. Several business groups urged Treasury to allow taxpayers to elect to file country-by-country reports with the United States earlier than the deadline provided in U.S. law to avoid triggering the requirement that they file their report with another state's tax administration.Footnote 178
One might object that fiscal fail-safes are not really new. Unilateral actions have always been available to respond to tax gaps and other strategic tax practices. Indeed, even before BEPS, many states adopted uncoordinated fiscal fail-safes that linked tax treatment in the residence state to tax treatment at source.Footnote 179 Worldwide taxation and CFC regimes can be viewed as unilateral fiscal fail-safes; they trigger additional tax at residence when there has been (in the residence state's view) insufficient tax at source. Recently, the United States proposed bilateral fiscal fail-safes in its model tax treaty.Footnote 180 But it is easier for multinationals to avoid a particular state's fiscal fail-safe than it is to avoid the coordinated fiscal fail-safes of a large number of countries. The relative effectiveness of coordinated fiscal fail-safes compared to unilateral fiscal fail-safes may explain why, since the 1960s, the United States has advocated a kind of “mutually assured taxation” by campaigning for other states to adopt CFC regimes on the U.S. model.Footnote 181
A key feature of fiscal fail-safes is that they facilitate coordination while neutralizing holdouts. The gains of BEPS were not distributed evenly, and neither were its costs. While the larger goal of BEPS was to reduce corporate tax avoidance and increase tax revenue for its participants, BEPS also created risks not only for tax havens, which stood to lose revenue, but also for relatively high-tax countries, such as the United States. The need to neutralize defection by tax havens is obvious, but the need to neutralize U.S. resistance to BEPS requires further explanation.
The United States broadly supported the BEPS Project, but it also had good reasons to be concerned about its outcomes. First, the United States was worried about its own companies. The world's largest multinational companies are disproportionately American.Footnote 182 Thus, efforts to prevent corporate tax avoidance by multinationals could be expected to disproportionately impact U.S. companies operating abroad and potentially harm their competitiveness. Second, the United States was worried about losing real business activity. The BEPS slogan was to “align tax with value creation.” If factors of production would constitute the measure of “value creation,” then aligning tax with value creation would motivate companies to move productive factors out of high-tax states. The United States had more reason to be concerned about real factor movement than other countries because at the beginning of the BEPS Project it had one of the highest statutory corporate tax rates in the world. Third was a pure revenue concern. Because the United States had a worldwide tax system in which it relieved double tax by crediting other states’ taxes, avoidance by U.S. companies of foreign taxes ultimately redounded to the U.S. fisc in the form of lesser claims to foreign tax credits. BEPS measures that succeeded in increasing U.S. companies’ foreign tax burden therefore would, in time, hit the United States in the pocketbook. The unique position of the United States as a high-tax home to an outsize share of large multinationals engaged in aggressive tax planning therefore created special risks for the United States compared to its BEPS partners.Footnote 183 Thus, BEPS participants rationally desired recommendations that could be effective even if the United States or tax havens did not implement them.
That fiscal fail-safes are effective for achieving full taxation or at curbing state defection is not enough to conclude that they are a good idea. The normative arguments for fiscal fail-safes resemble those against tax sparing.Footnote 184 Fiscal fail-safes—which impose additional taxes where foreign tax was never due—are the inverse of tax sparing, which grants credits for foreign taxes never paid.Footnote 185 Fiscal fail-safes may increase revenue,Footnote 186 promote the perceived fairness of tax regimes (by discouraging tax breaks for multinationals or bespoke tax deals for particular companies), reduce rent-seeking activity, and reduce both paper profit shifting and real locational distortions.Footnote 187 Against these virtues would be set the vices of fiscal fail-safes. Specifically, fiscal fail-safes impinge tax autonomy and inhibit tax competition by clawing back the benefit of low tax rates or tax holidays. Fiscal fail-safes thus may be criticized as paternalistic, and they may drive competition to other arenas, such as direct subsidies.Footnote 188
Putting aside these normative considerations, conceptualizing familiar and longstanding unilateral anti-abuse rules as fiscal fail-safes reveals incremental acceptance of the full-tax norm. Moreover, the BEPS Project adds a new dimension by standardizing and coordinating fiscal fail-safes.Footnote 189 In addition to being more effective at curbing abuse, coordinated fiscal fail-safes reduce and spread the risk that companies and investment will flee to other jurisdictions. Moreover, coordination reduces the risk that fiscal fail-safes could create new tax arbitrage opportunities. Finally, coordination may lower compliance costs and double-tax risks for taxpayers, thereby facilitating cross-border commerce. Consistent with this reasoning, the EU member states used EU-level legislation to adopt many of the BEPS recommendations—including fiscal fail-safes that were not minimum standards, such as anti-hybrid and CFC rules.Footnote 190
4. Peer Review
As in other areas of law, such as banking and tax-information exchange, state compliance with BEPS will be monitored with peer review.Footnote 191 The key to peer review is the existence of clear and uniformly applicable standards. Now that the OECD and G20 states have agreed in principle to certain standards, the Inclusive Framework will flesh out the detailed requirements that states must meet to comply with those standards.Footnote 192 In the peer-review process, multinational teams of tax administrators examine compliance according to fixed rubrics. Gradations of compliance allow nuanced distinctions among states as well as rewarding them for incremental implementation of the relevant standards.Footnote 193 Peer-review processes are phased and iterative—a noncompliant country has opportunities to come into compliance, and the monitoring process provides clear direction to the reviewed state for how to comply and improve its rating. Phased reviews sort easy-to-meet standards before more demanding standards to promote trust among reviewed states and to increase the chances that states will cooperate in the next phase of review. The peer-review process also can be expected to generate positive externalities for enforcement and dispute resolution as tax administrators who work together on peer reviews build trust that they carry over into competent authority disputes or multilateral audits.Footnote 194 In contrast with the OECD's earlier Harmful Tax Practices Project, all participating states, including powerful OECD states, are subject to the same standards.Footnote 195 An underappreciated aspect of the process is that engaging the Inclusive Framework in developing peer-review standards gives non-OECD, non-G20 countries an opportunity for meaningful participation in the post-BEPS work.
Together, the various BEPS reforms—including the Multilateral Instrument and other bilateral treaty changes, coordinated unilateral measures including fiscal fail-safes, enhanced disclosure and transparency initiatives and peer-review monitoring—substantially increase countries’ commitments to international tax cooperation.
IV. The Future of International Tax
So far, this Article has argued that the impact of BEPS must be measured not only by its list of deliverables, but also by how, compared to prior practice, it changed participants, institutions, agenda, norms, and forms of international tax cooperation. This Part evaluates the normative desirability and durability of the BEPS changes, including its impact on revenue, inclusivity, and accountability; it also considers how BEPS may constrain state tax autonomy and entrench even poor policy choices. This Part also raises some conceptual objections to full taxation as a norm and argues that states’ unwillingness to grapple with distributive issues created new risks of double taxation. The unaddressed distributive questions and the specter of reemergence of double taxation have triggered a new project at the OECD, known as BEPS 2.0, to reconsider and update 1920s-era tax-treaty rules, including the permanent establishment requirement. This Part concludes that because states lack shared views regarding how to efficiently and fairly divide entitlements to tax cross-border commerce, bargaining over national self-interests will likely determine the outcome of BEPS 2.0. However, loftier values of efficiency and justice and the more quotidian consideration of administrability will likely govern the formulation of the technical rules to implement the agreed outcome.
A. Revenue
It is too early to know the revenue effects of BEPS.Footnote 196 We have never fully understood the size of profit shifting, particularly for countries for which good data is unavailable.Footnote 197 For example, one recent paper argues that the method economists use to measure profit shifting results in systematic overestimation.Footnote 198 That claim has been disputed.Footnote 199 Notwithstanding data gaps, because nearly all the BEPS recommendations were designed to prevent corporate tax avoidance, it seems reasonable to assume that the net outcome of BEPS will be to raise revenue. One of the BEPS recommendations will generate more reliable data for measuring profit shifting.Footnote 200
B. Inclusivity
Among the things BEPS did not change was the significant participation in international tax by the business community. Post-BEPS, however, more stakeholders participate in international tax policymaking than ever before. Still, it is too soon to declare victory for pluralism. First, commentators disagree over the value of participation by NGOs. Shu-yi Oei and Diane Ring argue that NGOs wield substantial power untempered by politically accountability.Footnote 201 On the other hand, participation by these groups makes international tax law more salient, converting it from a seldom-traveled backwater, navigated only by those with technical tax training, to a more mainstream process.
Second, it is too soon to evaluate the impact of the Inclusive Framework. Although all Inclusive Framework members ostensibly participate in BEPS implementation questions on an “equal footing,” it remains to be seen how much influence smaller and poorer countries will have on important agenda-setting or policy decisions.Footnote 202 Commentators have argued that the participation of non-OECD, non-G20 countries in the Inclusive Framework represents an attempt by ambitious OECD bureaucrats to increase their own influence and to ward off competition from the UN.Footnote 203 Certainly, the interests of Inclusive Framework members are divided, and a desire by OECD members to retain control of the agenda and content of the international tax regime suggests that these states—perhaps joined by the rest of the G20—will find ways to exclude other countries, for example by making important post-BEPS decisions outside of the Inclusive Framework.
Relegating the Inclusive Framework primarily to peer review of implementation of BEPS 1.0 minimum standards would reinforce developed countries’ control over international tax policy, while maintaining the appearance of inclusivity. Even if developing countries formally had an equal voice at the OECD in agenda-setting and policy decisions, they have vastly less technical capacity than the OECD and G20 states.Footnote 204 In tax, as in other areas, control over the technical rules often equates to control over substantive policy. The most significant impact of BEPS in terms of inclusivity, therefore, may have been to add G20 countries that are not members of the OECD to the formal decision-making process.Footnote 205
C. Accountability
Democratic accountability concerns regarding international tax policymaking—such as claims that national representatives to the OECD are unelected bureaucrats and that OECD members deliberate in secret—predate BEPS.Footnote 206 By expanding the agenda of international tax, BEPS exacerbates these concerns.Footnote 207 Those uncomfortable with the normative implications of this descriptive claim could advocate for increasing the accountability of the OECD by, for example, making its deliberations and budget public.Footnote 208
D. Constrained Policy Outcomes
As tax policy becomes more international, it may become more constrained, homogenous, and entrenched. Indeed, tying domestic hands is part of the point of international law. With fiscal fail-safes, this effect is intentional; states often implement them with the goal of influencing other states’ rules. Other types of cooperation involve unintentional restraints. For example, in a phenomenon dubbed “the Luxembourg effect” in a nod to the seat of the Court of Justice of the European Union (CJEU), Lilian Faulhaber observed that EU law may limit international tax reform efforts.Footnote 209 When negotiating at the OECD, EU member states can only agree to reforms that are compatible with their obligations under EU law. Thus, on issues where a harmonized approach is needed or desirable, EU law effectively limits the options of non-EU members.
But that influence goes both ways—just as EU law sets limits on what can be accomplished at the OECD, OECD standards diffuse into EU law, generating a “Paris effect.”Footnote 210 For example, the EU Commission modified its own anti-hybrids proposal to match the one recommended as part of BEPS.Footnote 211 Likewise, the CJEU has repeatedly deferred to OECD standards,Footnote 212 even though not all the EU member states are also OECD members.Footnote 213 And in its state-aid investigations, the EU Commission relied in part on the OECD Transfer Pricing Guidelines to determine whether member states made illegal sweetheart tax deals with multinationals.Footnote 214 Nor is the Paris effect limited to the EU. OECD standards diffuse into the laws of the many non-OECD states that rely on the OECD Model and other OECD guidance.
As the number and interests of countries making international tax policy expands, consensus will be harder to reach, which may result in compromise positions that then become entrenched. OECD policies became entrenched even in the pre-BEPS era when fewer states were involved in multilateral standard setting. For example, the widespread acceptance of the OECD Transfer Pricing Guidelines undoubtedly entrenched the arm's-length standard.Footnote 215 The recent outcry over the possibility that the European Commission might introduce an arm's-length standard that differed from the OECD arm's-length standard, and intervention by the United States to foreclose that possibility, highlights the constraining effect of internationalizing tax.Footnote 216 Entrenchment of international standards results from the sunk costs of developing and building consensus around standards. It also arises from clientele effects—once companies and tax practitioners learn how to exploit a standard, they advocate against its reform.Footnote 217 Entrenchment can be deepened by tying other benefits to agreed standards. For example, countries must commit to the arm's-length standard as a condition to OECD membership, entrenching it as part of what might be called the OECD acquis.Footnote 218 With BEPS, the OECD acquis expanded to nonmember states; to join the Inclusive Framework, states “must commit to the comprehensive BEPS package.”Footnote 219
That international regimes interact is not new. But as regimes proliferate, so do the cross-constraints they impose. As international obligations grow, they will continue to close off viable paths for reform and innovation. For example, the strictures of tax treaties may stand in the way of domestic and multilateral tax reform.Footnote 220 To the extent that such reforms are normatively desirable, the blocking effect is unfortunate. On the other hand, tax treaties also may prevent protectionist or unwise national taxes.
E. The Indeterminism of Full Taxation
While intuitively attractive, when not paired with a clear distributive rule that dictates which state or states should tax cross-border income, the concepts of full taxation and double taxation are indeterminate—there is no way to specify the tax base or rate that would satisfy them.Footnote 221 Although commentators using terms like “double non-taxation” may have in mind an approximate tax base (perhaps financial accounting income with adjustments for tax) and an approximate tax rate (something like the OECD average rate), such vague notions are too imprecise to generate clear policy prescriptions. Because tax systems differ across jurisdictions, determining how much income a company has, and the rate at which that income should be taxed, requires a rule for deciding which state's rules for calculating income and tax rates will apply. Thus, states first must decide how to split a multinational's income among themselves before they can determine whether the multinational has been subject to too much tax (“double tax”) or too little tax (less than “full tax” or “double nontax”). Without clear consensus rules to allocate tax entitlements, a generalized norm of full taxation can lead to double taxation in exactly the same way that a generalized norm of no-double taxation can lead to nontaxation.Footnote 222
Perhaps no issue better demonstrates the problems that arise from lack of consensus over allocation than recent digital tax debates. Even though the BEPS participating states identified misallocation of income from “the digital economy,” as a serious issue at the outset of the Project,Footnote 223 the United States, keenly aware of its own interests, convinced other states that the digital economy could not be “ring-fenced” or cleaved off from the rest of the economy and taxed separately.Footnote 224 In addition to sidelining discussions about the digital economy, negotiators also did not fundamentally rethink the 1920s-era tax-treaty distributive framework that is based on a source-and-residence paradigm.Footnote 225 But failure of that preexisting framework to keep pace with changes in the global economy, and its consequent inability to generate acceptable tax outcomes for source countries, has resulted in states' declining fidelity to it.
The idea behind BEPS was to end nontaxation of corporate income, and instead to substitute full taxation. Because states did not want to tackle the distributive question, however, they either arbitrarily allocated the new revenue that arose from closing tax gaps, or they relied on preexisting source-and-residence rules.Footnote 226 Consider the anti-hybrid rules that close tax gaps that arise from different states’ definitions of debt and equity. The primary anti-hybrid rule fills the gap by requiring the payer state to deny an interest deduction, thus allocating revenue to the payer state. If the payer state does not adopt the anti-hybrid rule, then the defensive rule requires the recipient's state to include the payment, thus allocating revenue to the recipient state. What BEPS did not do, however, was resolve which of the two states was more entitled to the revenue and why, nor did the BEPS states determine whether some third state might have a legitimate claim to the revenue.
Similarly, under the BEPS recommendation for CFC regimes, income that had not been subject to enoughFootnote 227 tax at source would be subject to additional tax at residence in order to achieve full taxation.Footnote 228 By clawing back the benefit of low-source taxes, CFC rules encourage source states to tax more. Like the anti-hybrid rules, CFC and other types of minimum-tax rules seem more concerned with whether companies pay tax than where they pay tax. Such BEPS recommendations reflect a new commitment to full taxation, but they also reflect ambivalence or reluctance to confront the underlying distributive issue. This ambivalence only went so far, however. Although income typically taxed at source might be taxed at residence, and vice versa,Footnote 229 states did not use the BEPS Project to fundamentally rethink which states should be entitled to the labels “source” and “residence.” Likewise, although the BEPS countries purported to align taxation with value creation, because they never adequately defined what that alignment entailed, advocates saw in it what they wanted to see.Footnote 230
The concept of full taxation says that income should not go untaxed, but it does not specify where it should be taxed—whether at source, residence, or in some third state that lacks a tax entitlement under current rules. Having accepted full taxation as a norm, countries now face a new challenge: to avert a kind of free-for-all in which many states try to fill the same tax void. Consider the following examples.
First, Leopoldo Parada has extensively explored the double taxation that may arise from application of the BEPS anti-hybrid rules.Footnote 231 Second, to the extent that country-by-country reports reveal untaxed income, they may invite taxation by multiple states according to different theories of “value creation.”Footnote 232 Third, double tax may emerge as a result of disclosure of previously secret tax rulings. The European Commission's state-aid investigations attest to the possibility that tax-ruling disclosure may lead to double-tax, or at least to conflicting claims to tax income. Analysis by the Commission of member state tax rulings revealed that U.S. multinationals were shifting their profits out of EU states and into tax havens or to nowhere.Footnote 233 But the Commission's response prized full taxation over clear thinking about where income should be taxed.Footnote 234 For example, as a result of its investigation of Ireland for issuing sweetheart rulings to Apple that the Commission determined aided Apple in its stateless income planning, the Commission ordered Apple to pay $14 billion in taxes to Ireland. Even though most people might agree that a substantial portion of Apple's income should not be stateless, that it should be taxed by Ireland is far from obvious. The result of the Commission's state aid investigation seems not only arbitrary, but generative of serious disputes with other countries—including the United States and other EU states—that would seem to have superior claims to Ireland to tax Apple's stateless income.Footnote 235 Assigning Apple's stateless income to Ireland ensured full taxation, but not in a way that genuinely reflected where “value was created.”
In sum, despite academic commentary that justifiably criticized BEPS for failing to squarely tackle the allocation question,Footnote 236 perhaps the most important implication of BEPS was that by building consensus for full taxation, it heated up discussions about which state should step in to fill tax voids. As the next section explains, as states scramble to fill tax vacuums, they generate renewed risks of double taxation, prevention of which may drive further multilateral cooperation.
F. Reallocation: BEPS 2.0
After a century of uneasy postponements, governments are now confronting the allocation question. The 137 states in the BEPS Inclusive Framework have set the end of 2020 as the deadline for devising “a fairer and more stable” allocation of tax entitlements among states.Footnote 237 These BEPS 2.0 negotiations involve three major issues: (1) how to redefine the permanent establishment threshold for the modern economy; (2) how to attribute income to the permanent establishment, as redefined; and (3) a proposal, not discussed here, to adopt minimum tax rates as a fiscal fail-safe that would also help curb tax competition.Footnote 238 Different proposals would modify the permanent-establishment threshold in various ways (for example, by introducing the concept of digital presence) and allocate more income to the states of users and consumers (collectively, market states). States have also floated new methods to attribute profits to the permanent establishment; these methods would depart from the sacred arm's-length standard by, for example, relying on formulas.
Although we will not know the results of BEPS 2.0 for some time, this section considers factors that may influence potential outcomes, and it concludes that national self-interest will be the most important driver of any resolution of the allocation question,Footnote 239 but efficiency, fairness, and administrability concerns will likely shape the formulation of any resulting rules.
1. Absence of Agreed Norms
Reformers insist that states must “align taxation with where value is created.”Footnote 240 Such assertions represent claims that the distributive impact of the current international tax regime—the allocation of tax entitlements—must change. But there is no ready answer to what distributive scheme should replace it. In the absence of shared conceptions of efficiency or distributive justice, policymakers will seek to promote national self-interest.
At the inception of tax treaties in the 1920s, policymakers asked economists how countries should tax cross-border income; then, as now, economists did not agree on a single answer.Footnote 241 Economics does not directly answer the question of where income arises.Footnote 242 That economics cannot answer the where question does not mean that economics cannot tell us whether some distributions are more efficient than others. But appeals to efficiency only get us so far. Neither economists nor policymakers agree about which rules generate the most efficient outcome or about whose welfare should be maximized.Footnote 243 Nor have clear prescriptions emerged from empirical study.Footnote 244
That national welfare considerations generate different policy prescriptions for different states seems obvious, but even pursuit of global welfare generates multiple incompatible policy prescriptions.Footnote 245 That different states may seek to maximize along different margins is well-known. It is enough to observe that state representatives do not agree, and are not likely to agree, about how to use their international tax laws to achieve efficiency goals. Contrast the lack of consensus in income taxation with the consensus on the goals of trade agreements. Widely accepted, the theory of comparative advantage guides trade policy to ostensibly correct results—fewer and fewer tariffs, freer and freer trade. In most cases, the ideal tariff rate is zero.Footnote 246 But income tax rates cannot be zero because states need revenue to fund public goods and to maintain social programs. If states are to ensure full tax while avoiding double tax, they must coordinate on the question of distribution—they must decide which state taxes which income. This does not mean that efficiency concerns have no place in the current allocation debate. But they are inconclusive on the most important question— namely, who gets what?
As with economic efficiency, there is also a lack of consensus on how to fairly allocate tax entitlements among states. Policymakers broadly adhere to the benefits principle, but that principle does not generate clear guidelines regarding the magnitude of each state's tax entitlement. Wolfgang Schön's trenchant observation about the concept of value creation—that it only tells us that income should not be allocated to tax havensFootnote 247—applies equally to the benefits principle. Although the benefits principle would largely reject allocation of tax entitlement to states where companies have no productive factors, it generates no clear guidelines regarding what proportions of income should be allocated to states with real factors of production (however generously defined).Footnote 248
Nor do concerns about interpersonal equity typically drive international tax negotiations.Footnote 249 Policymakers commonly phrase allocation questions in terms of efficiency, enforcement, and legal jurisdiction; distributive justice concerns are pushed down to the national level where they can be resolved through democratic processes.Footnote 250 Cooperative efforts, such as BEPS, that ensure that corporations pay their taxes can be seen as shoring up states’ ability to effectuate domestic conceptions of justice, but they do so without developing an international conception of distributive justice.Footnote 251
Commentators have offered many theories for how to divide income between source and residence, and for how to divide tax entitlements among states that may claim to be a source of a multinational company's income. Although splitting seems intuitively attractive, particularly if a company's productive factors—such as payroll, property, and sales—are present in more than one state, any particular split is hard to justify as a theoretical matter. Indeed, many current rules lack a convincing theoretical foundation—they represent arbitrary compromises.Footnote 252
Moreover, source and residence tax represent only the two most familiar bases for tax. The academy has functioned as a kind of Pandora's box for corporate-tax allocation theories; credible arguments—arguments that rely on economic theory, conceptions of distributive justice, and pragmaticism—support not only taxation at sourceFootnote 253 and residence,Footnote 254 but also by the company's shareholders’ state(s),Footnote 255 its customers’ states,Footnote 256 the state where its products or services are used,Footnote 257 and taxation according to various formulas that would take into account the presence in particular jurisdictions of the company's factors of production.Footnote 258
Nor is the problem of dividing tax entitlements new. As far back as 1917, shortly before the emergence of our current international tax regime, T.S. Adams formulated the problem this way:
Here is a corporation whose owners live in jurisdiction A, whose factory is in jurisdiction B, whose main offices are in jurisdiction C, and whose principal sales department is in jurisdiction D. It needs no discussion to prove that each of these jurisdictions will demand and in the long run will succeed in collecting some tax.Footnote 259
Globalization allows us to add more jurisdictions to Adams's example—the software engineers work in jurisdiction E, the company's financing subsidiary is in jurisdiction F, the intellectual property resides in jurisdiction G, and the company has customers, but no physical presence, in jurisdictions H through Z. But Adams's main conclusion—that every jurisdiction with a colorable claim to tax “will in the long run succeed in collecting some tax”—is as applicable today as in 1917. Unless multinationals will be subject to multiple taxation, the question remains how states will resolve their overlapping tax claims. As the U.S. Supreme Court observed, providing an answer resembles “slicing a shadow.”Footnote 260
In the absence of shared conceptions of efficiency or fairness, views regarding the proper division of income remain contested. The difficulty of resolving distributional questions is familiar from domestic law. Even the notion that the burden of paying taxes should be distributed according to ability to pay does not tell us what the tax rates should be. In the domestic context, we use democratic procedures to distribute the tax burden. But no global institution exists that would allow us to democratically resolve the question of how states should divide entitlements to tax international income. Thus, any future compromise by necessity will be constructed from interstate bargaining.
This is not to say that all possible divisions are normatively equivalent or that only politics matters.Footnote 261 But adherence to different theories about what is fair and of how to maximize welfare and whose welfare to maximize lead to different conclusions. Advocates garner support for their preferred outcome when they can point to convincing reasons based in efficiency, advancing justice, or other values.Footnote 262 While efficiency and fairness norms cannot generate precise allocation rules that cover all situations, widely held norms like the benefits principle generate shared intuitions and expectations regarding acceptable allocation bargains. Moreover, once states decide how tax entitlements should be allocated, efficiency, fairness, and administrability will inform how technocrats construct the rules to effectuate the tax bargain.
It is worth emphasizing that the permanent establishment is a political invention.Footnote 263 In the 1920s, a physical presence requirement for taxation made sense. Later, as intragroup transfer pricing ramped up, countries adopted the arm's-length method to counter profit shifting.Footnote 264 Arm's length seemed a reasonable response to a growing problem, and its defects were either unrecognized or regarded as better than the alternatives. Long use and widespread acceptance made our tax rules seem not only stable but well-grounded. Each new generation of tax lawyers, administrators, and policymakers learns its sacred texts—the OECD Model, the Commentary, the Transfer Pricing Guidelines—and tax experts confabulated post-hoc rationalizations for arbitrary allocation rules.Footnote 265 A hundred years later, technical experts decry the invasion of international tax forums by politics.Footnote 266 But questions about how to distribute the gains from globalization are as political today as they were in the 1920s, and they will be settled—if they are settled—by bargaining.
2. The Changing Politics of International Tax
If bargaining over national interests will play a starring role in determining the outcomes of BEPS 2.0, what do states want? This section can only provide a snapshot of what Hugh Ault characterized as a third type of tax competition—the competition to define the distributive rules in a way that maximizes national tax revenue.Footnote 267 Although high-tax developed countries generally share an interest in preserving an allocation system that disproportionately benefits them at the expense of developing countries, important cleavages in the interests of developed countries have emerged in the past two decades. Developed countries like France, whose benefits from the 1920s compromise have eroded, see an opportunity to gain revenue share relative to the United States, but they also face the risk that enlarged taxation by consumer or source states would cause them to lose revenue relative to the developing world, especially China and India.
This section briefly identifies four factors that seem especially important to the negotiations: the emergence of the BRICs (Brazil, Russia, India, China, and South Africa), the rupture in tax relations between the United States and Western Europe, the rise of tax unilateralism, and changes in the law and economy of the United States that affect its outlook on tax allocation.Footnote 268
Emergence of the BRICs. Major emerging economies—in particular, China and India—that never agreed to the 1920s compromise see an opportunity to gain revenue share by reducing the extent to which tax treaties shift revenue away from source countries and toward residence countries.Footnote 269 Criticisms of the distributive outcome of tax treaties are not new. Developing countries, which are disproportionately source countries, long have complained that tax treaties require them to cede too much tax to residence countries. Indeed, states developed the UN Model tax treaty as an alternative to the OECD Model to help counter this shift.Footnote 270 But developing countries have had little success in persuading their developed-county treaty partners to adopt the UN Model.Footnote 271
Rupture between the United States and Europe. Several recent events reflect important cleavages in the interests of European countries and the United States that may affect their ability to display a united front to developing countries seeking a larger share of tax revenue from international commerce.
The United States and European countries found themselves on opposite sides of many BEPS issues, as Bob Stack, the chief U.S. negotiator for BEPS, made clear.Footnote 272 This was perhaps inevitable. U.S. multinationals had stripped European tax bases for decades, a practice overtly facilitated by U.S. law.Footnote 273 New business conventions, such as third-party manufacturing, warehousing, and delivery enabled nonresident companies to reap significant profits in Europe while treading lightly or not at all on the soil of market states. An office, branch, factory, or store constitutes a permanent establishment, but a web presence does not. The obsolescence of the permanent establishment concept found extreme expression in the technology sector.Footnote 274 Dominance of this sector by U.S. multinationals has meant that Western European countries are increasingly on the losing side of tax treaties with the United States. Social media, internet search, and streaming companies—Facebook, Twitter, Google, Netflix, YouTube—extract huge profits from selling targeted advertisements and subscriptions, but user states cannot tax those profits because the companies avoid creating permanent establishments there. Likewise, companies that provide digital markets to link sellers and customers, such as Amazon, Airbnb and eBay, avoid creating permanent establishments in the states of the seller or the buyer.
Companies’ growing ability to avoid permanent establishments has exacerbated the revenue shift caused by tax treaties and widened the set of losing states.Footnote 275 Pascal Saint-Amans, the head of tax for the OECD, observed that “[t]he Europeans have experienced what it is like to be a developing country. They had clever people coming in, making money, not giving anything back and then leaving.”Footnote 276 That European countries increasingly felt the sting of the tax-treaty revenue shift divided the interests of the United States and Europe.
Other events reflected growing U.S.-European tensions. For example, the EU Commission's tax state-aid investigations focused on tax rulings that EU states issued to brand-name U.S. companies; these allegedly sweetheart deals allowed U.S. companies to avoid tax.Footnote 277 The EU state-aid rules—which forbid member states from subsidizing private enterprises unless they have permission from the Commission—had been understood to apply to tax subsidies since at least the late 1990s. But the Commission had never investigated member states for tax state-aid that they conferred to particular companies.Footnote 278 The novelty of the investigations combined with what seemed to be disproportionate targeting of U.S. companies through a nontransparent selection procedure elicited a sharp bipartisan rebuke from the Senate Finance CommitteeFootnote 279 and a tense exchange of letters between the U.S. Treasury secretary and the EU commissioner for competition.Footnote 280 In August 2016, the Commission ignited a firestorm when it assessed a staggering $14 billion recovery order against Apple, reasoning that all the income that Apple had shifted to the nowhere subsidiariesFootnote 281 should have been taxed in Ireland (of all places).Footnote 282 Although both Apple and Ireland have appealed the decision, if enforced, the recovery would represent both the largest tax deficiency in history and the largest ever state-aid penalty.Footnote 283 In a move the Wall Street Journal dubbed “a diplomatic broadside,”Footnote 284 the Treasury Department issued an extraordinary twenty-five-page white paper arguing that the Commission departed from prior EU law and its own practice in the recent state-aid cases involving U.S. companies.Footnote 285 In another unprecedented move, the United States tried to intervene in the Apple case, but the EU courts held that it lacked standing.Footnote 286 The state-aid cases sent a shock-wave through the international tax community, prompting respected academic Michael Graetz to write an op-ed arguing that to “a U.S. lawyer steeped in the requirements of due process and the rule of law, the process by which these fines were decided doesn't pass the smell test.”Footnote 287 Bob Stack confirmed that the state-aid investigations soured U.S.-EU relations at the OECD.Footnote 288
The Commission's unsuccessful proposal in 2018 for a digital-services tax further drove the wedge between the United States and Europe.Footnote 289 In response to popular dissatisfaction with corporate tax avoidance, and perhaps to consolidate its own power vis-à-vis the member states, the Commission proposed a new digital tax that was tailor-made for U.S. internet giants, including Airbnb, Amazon, Facebook, Google, and Twitter.Footnote 290 U.S. dominance in technology enabled the Commission to formulate a tax that would almost perfectly target U.S. companies, while exempting nearly all EU companies, even when those companies engaged in the same commercial activities as the targeted U.S. companies.Footnote 291 The EU proposal failed due to opposition by Ireland and other tech-friendly states,Footnote 292 but it elicited the condemnation of U.S. lawmakers, who argued that it was designed to discriminate against U.S. companies, would undermine the tax-treaty system, create trade barriers, lead to double taxation, and possibly violate trade law.Footnote 293
Although the EU Commission's digital tax proposal did not secure the requisite votes, several large European states, including France, Italy, Spain, and the United Kingdom, passed or proposed similar digital taxes.Footnote 294 In touting these taxes, politicians have not hidden their intent to discriminate against U.S. companies while exempting domestic companies.Footnote 295 The dispute between the United States and France became particularly heated when the U.S. trade representative published a withering analysis concluding that the French tax discriminated against U.S. companies in intention and effect and that the United States would be justified in retaliating with tariffs of up to 100 percent on French goods.Footnote 296 The United States responded with similar threats to other EU countries that adopted or threatened to adopt digital taxes.Footnote 297
Growing tax unilateralism. Digital taxes are only one of several recent unilateral moves. As consensus for the old allocation rules erodes, more countries depart from it.Footnote 298 Some of these departures—such as digital taxes or the Commission's state-aid investigations—seem to target U.S. companies, but others are broader. For example, in the midst of the BEPS negotiations in 2015, the United Kingdom enacted a “diverted profits tax” that penalizes nonresident companies that artificially avoid creating British permanent establishments.Footnote 299 India and Israel developed nexus rules that rely on “significant economic presence” rather than physical presence.Footnote 300 Tax treaties bar some, but arguably not all, of these rules.Footnote 301
Several observations are worth making about these unilateral measures. First, they reflect growing dissatisfaction—including among developed states—with the permanent establishment requirement in tax treaties. Second, because these proposals typically raise little revenue,Footnote 302 they must aim at other goals. One possibility is that they respond to voter displeasure with corporate tax avoidance, and in particular corporate tax avoidance by U.S. multinationals. Another possibility is that such taxes aim to convince the United States, which has opposed efforts to renegotiate tax entitlements, to take seriously complaints about the obsolescence of tax treaties. Thus, France promised to repeal its digital services tax if states reach a new allocation deal at the OECD;Footnote 303 France even promised to refund digital taxes to U.S. companies upon reaching a new deal.Footnote 304
By threatening to use new instruments outside the tax treaty framework to impose tax on U.S. companies, the EU and other countries have expressed their willingness to depart from international tax's most sacred norm—that companies should not pay tax twice. Such threats to depart from the international consensus have been at least partially effective in bringing the United States to the bargaining table.Footnote 305 OECD tax chief Pascal Saint-Amans seized on impending unilateralism to persuade countries to compromise. He recently warned, “There's no plan B. There is plan C, and C is for chaos if we don't reach agreement.”Footnote 306 Even representatives of the United States, the world's staunchest advocates for maintaining the status quo, envision the possibility of fundamental rupture. Chip Harter, deputy assistant secretary of the Treasury for international tax policy, recently observed that if countries cannot agree on how to allocate income, “we will be in a world that threatens chaos.”Footnote 307 Unilateral measures thus have been important because, in addition to motivating the United States to participate in allocation negotiations, they motivate multinationals to lobby their own governments to reach a compromise on a stable distribution that would avoid double tax.Footnote 308
Changes in the United States. Because the United States has a disproportionate share of multinationals, the proliferation of unilateral actions will affect U.S. companies disproportionately. As a result, the United States has a special interest in preventing unilateral actions that could result in double tax for its multinationals, thereby undermining their competitiveness. Additionally, any shift in revenue from residence to source may be expected to result in a loss for the United States, a major residence state. But significant domestic legal and economic changes have made the United States more amenable to international tax changes.
In the early twentieth century, the United States was a clear capital exporter.Footnote 309 This posture helps explain the 1920s tax treaty framework, which shifted tax revenue from source to residence. Over the course of the twentieth century and into the twenty-first, however, the United States became not only one of the largest capital importers, but also one of the world's largest and richest consumer markets.Footnote 310 These shifts are important for current negotiations. If the United States can minimize its revenue losses from an allocation shift because the new system reassigns tax share from residence states to consumer states, then the United States, which was a large winner from the 1920s compromise, should have fewer objections to the change. This is one reason why the United States insists that any solution negotiated at the OECD cannot be limited to the digital sector. That insistence also ensures that other traditional residence countries share the expected revenue losses from a change in the permanent establishment requirement. The U.S. argument for system-wide, rather than digital-only, reforms may find support from other countries with large markets, including China and India.Footnote 311
Major domestic legal reforms in 2017 have also changed international tax strategy for the United States. Under pre-2017 law, foreign tax avoidance by U.S. multinationals ultimately redounded to the U.S. fisc, although the tax was deferred until repatriation. The 2017 corporate-tax-rate reduction was funded, in part, by a partial elimination of deferral of tax on the foreign income of foreign subsidiaries of U.S. companies. In place of deferral, the United States substituted either permanent exemption or current U.S. taxation at about half the newly lowered U.S. rate.Footnote 312 Both lowering the corporate tax rate and exempting certain foreign-source income have the effect of reducing the revenue incentive for the United States to oppose source taxes by other countries.Footnote 313 The 2017 tax reform therefore had the unintended consequence of making the United States more willing to negotiate allocation changes.
G. Predictions
Despite the dramatic changes to international tax reflected in BEPS and detailed in Part III, in the Introduction to this Article, I described the end of BEPS 1.0 as an inflection point: the future could involve more multilateral cooperation, more unilateralism, or the persistence of a predominantly bilateral regime. While it is too soon to draw conclusions on the direction of international tax, this Subpart speculates as to the durability of the various BEPS reforms.
The thirty-seven OECD countries can no longer claim an exclusive role in international tax policymaking now that the rest of the G20 has become accustomed to having a voice. At the same time, however, BEPS firmly solidified the role of the OECD as an institution in tax policymaking—including by accommodating more involvement by the non-OECD G20 countries. The current difficult negotiations over allocation represent a risk for the institution, however. If states cannot agree on a more stable solution, the OECD as an institution may see its authority and role diminish.
Acceptance of full taxation as an international tax norm, although accelerated by BEPS, has been incremental and is likely durable. Although the Multilateral Instrument holds great promise, because it leaves bilateral treaties intact, it is cumbersome and raises complicated legal issues.Footnote 314 It seems likely that states will incorporate the Instrument's provisions into their bilateral treaties as they renegotiate them, over time obviating the need for the Multilateral Instrument. Thus, unless states adopt truly multilateral provisions that cannot be incorporated into bilateral treaties, such as rules to resolve source-source conflicts, we might expect a reversion to nearly exclusive bilateralism in tax treaties.Footnote 315 At the same time, however, certain BEPS benefits—coordinated fiscal fail-safes, other coordinated anti-abuse rules, and disclosure and information-sharing regimesFootnote 316—cannot be secured exclusively through bilateralism, so we should expect them to endure as multilateral coordinated measures.
As noted above, the fate of the Inclusive Framework is uncertain. It could develop into a highly effective organization that has only a limited scope, like the Global Forum on Tax Transparency, or it could deepen into a more meaningful policymaking body with a role in international tax policy.
It is worth repeating that, as with the no-double-tax norm, the full-tax norm cannot be fully specified without an agreed distributive rule that determines where income will be taxed. The growing acceptance of full taxation has thus led to a chaotic situation in which multiple states seek to fill perceived tax vacuums. The risk that double taxation could reemerge motivates cooperation among states.Footnote 317 The twin aims of full taxation and avoiding double taxation necessitate more international cooperation than did a no-double-tax norm alone. Avoidance of double taxation requires only that either the source state or the residence state recede. Full taxation, in contrast, requires not only that some states recede, but also a coordinated answer regarding which state or states will affirmatively tax.
Nevertheless, predicting the outcome of the current tax allocation dispute is hazardous at best. The allocation negotiation is zero-sum, or nearly so, its surplus arises from reducing risks of double taxation, compliance costs, and uncertainty. U.S. representatives to the OECD hope to keep the aspirations of the project as narrow as possible, and recently they have even argued that any new rules that give tax entitlement to market jurisdictions should function as safe harbors, rather than mandatory rules. Given the difficulties of the negotiation and the reluctance of other OECD members—particularly Western European states—to join source states’ call for wholesale reform (as opposed to reform that applies to digital companies only), possible outcomes include the modest reforms advocated by the United States, or no reform at all combined with tolerance for digital taxes. Such a compromise would allow the United States and Western Europe to reunite to fend off claims by source states to tax multinationals’ income in the absence of permanent establishments. The narrowness of digital taxes as a solution would help address popular demands in Western European countries to tax U.S. companies, while at the same time limiting to only a single sector any revenue shifts and opposition lobbying efforts. If the United States and the Western European countries cannot agree on a separate compromise, however, they may not be able to overcome claims by source countries to a larger slice of the pie.
V. Conclusion
Throughout the twentieth century, states oversatisfied the norm against double taxation, accommodating significant nontaxation of corporate income. If the international tax regime did not manage to assign an item of income to be taxed, that gap caused no special alarm. The twin pressures of financial crisis and voter discontent motivated states to embark on the BEPS Project to fix tax gaps and other corporate tax-avoidance issues. Notwithstanding the modest formal recommendations of the BEPS Project, and contrary to academic criticism, this Article argues that BEPS both reflected and achieved fundamental change in the decisionmakers, agenda, norms, and legal forms of international tax.
While confirming and effectuating seismic shifts in international tax, the BEPS Project shored up old and disputed allocation rules using innovative instruments of international law, including the Multilateral Instrument and fiscal fail-safes. The point of BEPS was to grow the revenue pie by shutting down corporate tax avoidance—not to re-slice the pie—but it was unrealistic to think that the states could pursue full taxation while remaining indifferent to which state received the tax. This Article thus argues that one of the most important—if underappreciated—outcomes of BEPS is that by increasing the salience of distributive issues and creating an inclusive forum for negotiating distributive outcomes, BEPS made it more, not less, likely that states would seriously reconsider longstanding distributive questions.
Now is a crucial moment for international tax. Over one hundred countries are locked in a complex negotiation to decide how states should divide the entitlement to tax income from cross-border economic activities. BEPS provided a forum for negotiation—the Inclusive Framework—and populated it with representatives of developing countries nurturing long-ignored grievances about the prior distribution. Although we will not know the results of BEPS 2.0 for some time, this Article considered factors that will influence potential outcomes, including a widening rift between the United States and Europe and recent changes in U.S. tax law. It concluded that national self-interest will be the most important driver of any resolution of the allocation question, but efficiency, fairness, and administrability concerns will shape resulting rules.
Despite dire warnings of impending chaos, however, deviations from the 1920s compromise have been modest so far, consisting of very narrow digital taxes, changes to domestic nexus rules that do not affect companies resident in tax-treaty partners, and stepped up anti-abuse enforcement. This modesty suggests that commitments to minimizing double taxation though coordinated solutions remain strong.
It is also worth remembering that, although we lack shared values that could guide us to a stable new distributive outcome, we never shared such values. The international tax regime has always represented a negotiated bargain among states over how to divide the spoils of globalization. Appeals to fairness, efficiency, and administrability informed this bargain, even though the architects of our current tax system disagreed about how to weigh them. Through incrementalism in the twentieth century and paroxysms in the twenty-first, countries have built consensus for a system in which all of company's income should be subject to tax once. Next, they will decide how to share that tax.
Target article
The Transformation of International Tax
Related commentaries (5)
A Positive Dialectic: Beps and The United States
Diverse Interests and International Legitimation: Public Choice Theory and the Politics of International Tax
Introduction to the Symposium on Ruth Mason, “The Transformation of International Tax”
Shifting Design Paradigms: Why Tomorrow's International Economic Law May Look More Like the Tax Regime than the WTO
What is Unilateralism in International Taxation?