1. Introduction

The international corporate tax system has been strongly criticized for being open to large-scale tax avoidance by multi-national corporations (MNCs) [1, 2]. In this article, we focus on the avoidance of corporate taxation that takes place through MNCs’ use of legal means to assign profits to low-tax jurisdictions that are neither the location of their customers, nor the location their factors of production, nor the domicile of the MNCs’ parent company. A central such means is the transfer pricing system, which enables ostensibly national firms that are part of a large international conglomerate to charge prices to each other for goods and services in such a manner that the conglomerate’s profits are redirected towards low-tax jurisdictions. Such profit shifting appears to be extensive, with a recent study estimating that 36% of multinational profits are moved to tax havens globally [3]. This process is made possible by two kinds of actors: ‘sinks’ (low-tax jurisdictions, such as Bermuda) and ‘conduits’ (pass-through jurisdictions, which facilitate this movement of paper profits, such as the Netherlands) [4].

As a consequence of such tax avoidance, a large majority of countries find themselves facing a collective action problem. Each member of this majority is separately incentivized to act in such a way that if they all follow this incentive, they will each be worse off than if they had acted collectively to secure their common interests. The principal form that this collective action problem takes is the ‘race to the bottom’: the competition for paper profits gives each country an incentive to lower corporate tax rates. The end result is lower tax rates than these countries would have sought had they been able to agree—and ensure conformity with—a shared set of rules for tax rates and the tax base. It is notable that this collective action problem exists primarily for medium and large-sized economies, while some smaller economies with advanced legal systems can benefit from the race to the bottom [2]. As a consequence of this race, the average corporate tax rate in the OECD has dropped from over 40% in the 1980s, to 32.3% in 2000 and 22.9% in 2021 [5].

The central role of the transfer pricing system in this process has led a number of academics and civil society organisations to argue for an alternative approach known as ‘unitary taxation’ or ‘the membership principle’ [2]. That approach involves taking all the profits a multinational has made around the world, adding them together and allocating the profits proportionally to those countries in which the economic activity was located, with each country having the right to tax the assigned profit share however they wish. Tying taxation rights to the location of economic activity makes profit shifting away from locations where genuine economic activity takes place more difficult. But unitary taxation alone of course determines only the tax base, and not the tax rate. Tax justice campaigners have therefore argued it should be paired with a minimum effective corporate tax rate.

The OECD’s response to the problems of profit shifting and the race to the bottom has been to adopt two packages of new corporate tax rules and standards over the last decade. The negotiation of the first round of changes was initiated in 2013, when the OECD and G20 adopted a 15-point action plan to address Base Erosion and Profit Shifting (BEPS). In 2015, this resulted in the adoption of a package of standards and guidelines running to almost 2,000 pages – known as the BEPS package. However, rather than introducing fundamental reforms to the international corporate tax system, this package attempted to resolve the problem of corporate tax avoidance with a toolbox of anti-avoidance measures that were often highly technical, complex, and difficult to implement effectively. Not long after the adoption of the BEPS package, political pressure started building for more ambitious and fundamental changes to the system. As a consequence, the BEPS package was followed by the Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy, which was adopted in 2021. Roughly, the two pillars of the latter reforms are:

Pillar 1: A partial shift from the arms-length principle to a form of unitary taxation. This partial shift involves the allocation of a right to tax corporate profits to ‘market jurisdictions in proportion to the share of users or sales that take place within a jurisdiction.’

Pillar 2: A new minimum corporate tax rate of 15% [6].

This two-pillar proposal seemingly goes some way towards meeting campaigners’ demands. And when first announced, these reforms were hailed as a significant improvement in the fairness of the global tax system. For example, on 8 October 2021, The Guardian described it as follows:

Almost 140 countries have taken a decisive step towards forcing the world’s biggest companies to pay a fair share of tax, with plans for a global minimum corporate tax rate of 15% to be imposed by 2023.

[The OECD] described the landmark deal as a milestone towards ending decades of countries undercutting their neighbours on tax [7].

In this article, we argue for a more sceptical perspective on these reforms. We will evaluate these reforms in the light of three complementary principles which capture aspects of tax fairness and argue that the reforms are unlikely to lead to a substantial improvement in any of these aspects.

The first principle is ‘no free riding’. The underlying concern is that it is wrong if MNCs are able to freeride by using goods provided by governments—such as physical infrastructure, a well-functioning legal system, and an educated workforce—in the countries where their management, production, and sales are located without paying their fair share of taxes on their profits to finance these goods.

The second principle is respect for and promotion of what has been called a country’s ‘fiscal autonomy’, which we can understand as a government’s policy opportunity set: the money the government could raise, the redistribution and services it could undertake if it so chose [2]. The extent of a democratic country’s fiscal autonomy determines what its citizens can collectively decide to do in terms of revenue raising, transfers, and public expenditure. Hence, it is an important component of their opportunity to engage in meaningful political debate and substantial self-governance in social and economic matters. Respect for other countries’ fiscal autonomy requires non-interference with this aspect of their self-governance and letting them decide the extent to which they wish to tax the things they can rightfully tax. We submit that one element of such respect is that a country does not, without other countries’ consent, undermine these countries’ ability to tax those profits that they should, from the perspective of justice, be able to tax because these profits were very substantially generated from activity that took place in these other countries while making extensive use of those countries’ publicly financed goods. In other words, respect for fiscal autonomy bars countries from acting as ‘conduits’ and ‘sinks’ in a manner that enables the most egregious forms of profit shifting and free riding by MNCs. The principle also prescribes that countries should be willing to promote other countries’ fiscal autonomy where it is not too burdensome for them to do so, and especially that they should be ready to join with other willing countries to establish agreements that, in an equitable manner, solve collective action problems around taxation that limit the vast majority of countries’ fiscal policy opportunities, such as the aforementioned race to the bottom in corporate taxation.

The third principle concerns the limitation of economic and relational inequalities. Profits are a form of capital income, which is much more unequally distributed than labour income. For example, the Gini coefficient for inequality in gross (pre-tax) capital income in the EU is estimated at 0.9, whereas the Gini for gross labour income is estimated at 0.38 [8, 9, 10]. Now, as Michael Devereux and John Vella note in their contribution to this issue, to assess the impact of a tax on inequality between individuals, one needs to look past the agent on which the tax is levied, and assess who will, in the end, bear its cost—what they call its ‘effective incidence’ [11]. They also note that the extent to which holders of capital bear the final costs of taxes on corporate profits is a matter of dispute, and that this extent will depend on market conditions and country circumstances [11]. But insofar as capital owners pay the final cost of this tax, a drop in the corporate tax rate will likely increase inequality, especially as governments have tended to replace the lost revenue through the greater use of regressive taxes such as value added taxes. (Value added taxes have risen in the OECD from on average close to 16% in the 1980s to 19.3% in 2020 [12].)

Relational equality is a further, distinct dimension of egalitarian concern. Among co-nationals, it focuses on individuals’ ability to live as free citizens who are willing and able to participate in social life and contribute to public decision-making on equal terms [13]. But social egalitarianism can naturally be extended to how individuals of different nations relate to one another through national and global power structures—for example, whether they, or their democratic representatives, have an equal opportunity to shape the rules that govern their economic interactions or whether, instead, these rules are shaped by a powerful minority predominantly in their own interests while the less well-off are dominated or marginalized [14].

We now turn to assess the impact of the proposed reforms using these principles.

2. Free Riding

To assess whether the agreement can be expected ensure that MNCs contribute their fair share to the collective goods in the countries where they operate, it is important to note the limited scope of these Pillars’ application.

Pillar 1 applies only to MNCs with a global turnover above 20 billion euros and profitability above 10% [6], which is expected to be around 100 MNCs globally. Corporations below this threshold will not be covered by the new reallocation of profits under Pillar 1.

For Pillar 2, the threshold is a global turnover above 750 million euros [6]. Although it thereby has wider scope than Pillar 1, the threshold still means the new minimum tax rate defined under Pillar 2 is only expected to apply to the largest 10–15% of the world’s MNCs [15]. While these MNCs are estimated to account for over 90% of global corporate revenues, the turnover threshold leaves around 85–90% of the world’s MNCs outside of the scope of both Pillar 1 and Pillar 2.

Since the transfer pricing system was open to freeriding (in the form of profit shifting) before Pillar 1 and Pillar 2 were adopted, there is every reason to believe that freeriding will continue to be an option for the overwhelming majority of MNCs that fall outside the scope of the new mechanisms.

Next, let us consider whether Pillar 1 and Pillar 2 will stop freeriding for the in-scope MNCs. The specific rules and standards of the OECD deal are still being developed. There is a further question about whether the pillars will be implemented, with it looking very unlikely at this point that all countries will be implementing both pillars. This is particularly important for Pillar 1, which depends on a multilateral agreement being reached for it to take meaningful effect. If, for example, the U.S. Congress does not agree to the deal, then the effect of the rules will be dramatically reduced.

Even if Pillar 1 were to be implemented globally, it is worth noting that the limitations in scope do not only apply to the number of MNCs covered, but also to the share of their profits that is covered. For the MNCs that fall within its scope, the reallocation of the rights to tax profits will only apply to 20–30% of ‘residual’ profit – defined as profit in excess of 10% of revenue [6]. At the same time, Pillar 1 forbids countries to use digital services taxes (DSTs). DSTs as commonly designed are a quite blunt type of tax, which nonetheless in their simplicity can be attractive for some countries, not least developing countries with low levels of administrative capacity. Depending on the tax rate applied, some countries might have more to gain from a DST than from joining Pillar 1. This is a concern that has, for example, been expressed by the Kenyan Revenue Authority Commissioner, after Kenya decided not to support the OECD deal. It has also been raised by the Intergovernmental Group of 24, whose membership encompasses many emerging and lower and middle-income economies [16, 17].

To determine how much each country might gain under Pillar 1, the formula for reallocation of profits is central. The rules have been designed to award more taxing rights to ‘market jurisdictions’, meaning countries where consumers of goods and users of services are based [6]. As a result, the problem of freeriding MNCs could at least be somewhat reduced in these countries (although the cost of foregoing the use of a digital services tax must be kept in mind). However, market jurisdictions are not the only ones that have been suffering from freeriding. Countries where production takes place have also been subject to the problem, but the formula has not been designed to benefit such countries. This point is especially significant for poorer developing countries, which are not primarily a location for large amounts of consumers and users of services but which are, in some cases, locations where production is based.

Compared to Pillar 1, Pillar 2 is significantly further along in terms of the implementation rules being finalized and published. This makes it somewhat easier to get a picture of what the mechanisms will look like in practice. However, two factors still create uncertainty about the result: lack of public data about MNCs’ economic activities in each country (often referred to as public country by country reporting) and uncertainty about how countries, and in particular low-tax jurisdictions, will implement the deal.

The lack of data on MNCs activities is particularly important because Pillar 2 comes with so-called ‘carve-outs’, which allow MNCs to exclude a part of their income from the 15% minimum tax depending on the value of tangible assets and number of employees they have in a jurisdiction [18]. The effect of this will be that MNCs can pay less than the ostensible minimum of 15%. A detailed estimate of the implications of the carve-outs for the potential revenue gain of Pillar 2 for the 27 EU Member States concludes that in the short term, these carve-outs will reduce the total tax gain by €19 (from €83 billion to €64 billion) and that in 10 years’ time (at which point the carve-out rates will decrease) they will still result in €12 billion of forgone revenue annually [19].

Another key question is where the revenue will go, including the extent to which profit shifting to low-tax jurisdictions will continue. A key concern with the original OECD agreement from October 2021 was the fact that it gave priority to the home countries of MNCs in deciding which jurisdiction would be allowed to collect the minimum tax in cases where MNCs had effective tax rates below the minimum [6]. This put smaller and poorer developing countries at a disadvantage, as they are rarely home countries of MNCs but, as mentioned above, do commonly host MNC operations. An alternative approach could have been to allocate the minimum tax in proportion to an MNC’s genuine economic activity in a country, taking account of sales, capital, and workers in each country. (Sol Piciotto and co-authors formulate a proposal of this kind [20].)

At the end of 2021, a fundamental change was introduced to Pillar 2, which will likely strongly affect where the revenue ends up and what the impacts of the average corporate tax rates will be. In the model rules published by the OECD in December 2021, the option of a Qualified Domestic Minimum Top-up Tax (QDMTUT) was introduced [18]. This gives low-tax jurisdictions the option of introducing an alternative to a general increase in their corporate income tax rates. This alternative allows low-tax jurisdictions to introduce a type of tax that only applies in the cases where Pillar 2 would otherwise have taken effect, and where other countries would have been able to collect the difference between the effective tax rate and the minimum tax rate. A detailed analysis of the consequences of this new rule highlights that the competitiveness of source countries, including low-tax jurisdictions that apply the QDMTUT, would now not be reduced by introducing Pillar 2 [21]. Moreover, in relation to tax competition and the race to the bottom on corporate tax rates, this analysis concludes that ‘following the introduction of Pillar 2, countries [will] still have an incentive to reduce Corporation Tax liabilities (through a lower rate or higher allowances), even possibly down to zero’ [21 p8]. Unsurprisingly, the QDMTUT seems to be gaining traction among low-tax jurisdictions. For example, in a public consultation document, the government of Ireland has published comments on the draft EU Directive on Pillar 2, which follows the OECD model rules published in December 2021. The Irish government notes that:

the Directive provides that implementation of a [QDMTUT] will be recognised as sufficient to satisfy the requirements of Pillar Two, meaning that other Member States are assured sufficient tax has been paid by companies in the jurisdiction applying a [QDMTUT] and no further top-up would be required. In effect, each jurisdiction applying a [QDMTUT] becomes a “safe harbour” in tax terms. This is an EU rule which is expected to be adopted globally. As Ireland’s 12.5% trading rate of corporation tax is below the agreed 15% minimum effective rate, it is very likely that Ireland will introduce a QDMTUT as part of the Pillar Two implementation process [22].

And in its draft budget law for 2022/2023, the government of Mauritius states that:

The Income Tax Act will be amended to cater for any change that may be required in connection with the introduction of a domestic minimum top-up tax, applicable to companies resident in Mauritius forming part of multinational enterprise groups having a global annual revenue of 750 million euros or more, to ensure that they are taxed at the global minimum rate of 15% [23].

Meanwhile, Switzerland has put forward a proposal that goes one step further. In addition to introducing a domestic top-up tax, its proposal specifies that it is expected to be budget-neutral for the federal government, and that a significant share of the revenue will be allocated to promoting the attractiveness of Switzerland as a business location [24].

In sum, it seems that several of the traditional low-tax jurisdictions are responding rapidly to the risks Pillar 2 creates and picking up on the option of introducing a QDMTUT. Given this, it cannot be assumed that Pillar 2 will end profit shifting to low-tax jurisdictions by the MNCs that fall within its scope. Nor can it be assumed that it will help end the era of international tax competition. Moreover, rather than end freeriding by MNCs in countries of operation entirely, these recent developments suggest that in-scope MNCs will pay more tax, but possibly in low-tax jurisdictions rather than in the countries of operation. Furthermore, within the low-tax jurisdictions, the revenues might to some extent be returned in the form of measures to promote the attractiveness of the jurisdictions as business locations. Lastly, since the QDMTUT can be introduced as a separate tax, rather than as a general corporate income tax at a rate of 15%, there is a clear risk that the race to the bottom on corporate tax rates might continue.

3. Respect for and Promotion of Fiscal Autonomy

Historically, tax havens have undermined the fiscal autonomy of other countries by providing tax incentives for MNCs to shift paper profits out of their countries of operation and into tax havens. Furthermore, low-tax jurisdictions have contributed to a situation in which countries are under pressure to lower their corporate tax rates, with nations in need of the revenue taking the view that something is better than nothing. As outlined above, there is reason to believe that these processes will not be much attenuated under the OECD’s proposals.

National fiscal autonomy is also affected by whether the rules will be imposed on countries that have not signed up to them. This question is less about the content of the deal and more about how actors such as the EU will use the agreement. Following the adoption of the 2015 BEPS agreement, the EU introduced BEPS compliance as a criterium for inclusion on its list countries that it considered non-cooperative on tax matters – the “tax haven blacklist” [25]. This meant that countries that had not been part of negotiating the BEPS package now risked sanctions by the EU unless they complied with the rules, putting pressure on them to come into compliance.

While the EU itself has yet to reach consensus on the directive to implement Pillar 2, the European Commission is considering the option of making Pillar 2 compliance a criterion for a new version of the list of non-cooperative tax jurisdictions [26, p9]. In this context, it is worth noting that over a third of the world’s countries chose not to participate in the OECD-led negotiations. Within this group of countries, the vast majority are developing countries, and almost half fall into the category of least developed countries [27]. Furthermore, of the countries that did participate in the OECD-led negotiations, four countries (Kenya, Nigeria, Pakistan and Sri Lanka) did not support the outcome. For these countries, it would be a challenge to their fiscal autonomy if they face undue pressure to follow the OECD deal.

4. Distributive and Relational Equality

The fact that Pillar 1 and Pillar 2 are unlikely to have much of an effect on the race to the bottom in corporate taxation or on profit shifting to low-tax jurisdictions implies that they are unlikely to affect the inequality-enhancing effects of this race that we outlined above. Moreover, both pillars include components, such as the allocation of taxable profits based on where sales rather than where production takes place, that can be seen as disadvantageous smaller and poorer developing nations. They are therefore likely disadvantageous for the individuals that reside in those countries. Consequently, their implementation is unlikely to contribute to the reduction of cross-country income inequality.

Finally, relational equality continues to be a concern in relation to the global tax negotiations. The negotiation of Pillar 1 and Pillar 2 took place behind closed doors, and citizens were not able to observe what their governments were saying on their behalf. Negotiating texts were not public until after agreements had been adopted, making it difficult for a wide variety of stakeholders to discuss and feed relevant input into the process, as well as feeding suspicion that the interests of the powerful dominated proceedings.

5. Potential Longer Term Impact and Alternative Initiatives

Despite this criticism, there is some reason to hope that the agreement may have a positive impact, not because of the quality of the rules and standards themselves, but because of the fact that they represent a broad international consensus in favour of taxing multinational corporations as unitary entities (Pillar 1) and for introducing a global minimum effective corporate tax rate (Pillar 2). This recognition has moved the debate forward and has coincided with the emergence of proposals for alternatives to the OECD approach. For example, in recent years, the Africa Group at the UN has repeatedly called for a new tax convention to be negotiated under the United Nations [28]. This has been echoed by civil society organizations, and in March 2022, the European Network on Debt and Development and the Global Alliance for Tax Justice launched a discussion paper on what a new UN tax convention could look like [29]. For the taxation of MNCs, the proposal suggests that the entire transfer pricing system (and not merely a small part of it, for a circumscribed group of corporations) should be replaced by a system based on unitary taxation, with an allocation rule that counts the location of factors of production as well as of sales, supplemented by a minimum effective corporate tax rate.

6. Conclusion

We have argued that the OECD’s new tax deal, even if implemented, is unlikely to substantially improve the fairness of the global system of corporate taxation along the dimensions considered. Its scope is too narrow and there are too many loopholes for it to have a sizeable effect on profit shifting. Moreover, it is likely to do little for countries’ fiscal autonomy, as it will not substantially expand most countries’ ability to raise tax and indeed involves a limitation on this power, since signing on involves a commitment to forgoing the power to use digital sales taxes. In addition, there are reasons to believe that poorer countries may face undue pressure to sign on to the deal. Due to its limited fiscal effects and elements that favour high-income countries (such as the allocation of the rights to tax a share of MNCs’ profits exclusively by where its users and consumers are based and its sales take place), it will be unlikely to ameliorate income inequalities. And the exclusionary and non-transparent process of decision-making, which has favoured richer over poorer countries, means it does not align with principles of relational equality. Nonetheless, a global recognition of the ideas of unitary taxation and a minimum effective corporate tax rate is welcome; and one may hope that a different, more inclusive decision-making process may eventually result in global agreements that encode these ideas in a more effective and fairer form.