EU State Aid Rules as a Means of Promoting Equity within International Taxation - With a focus on inter-taxpayer equity in light of the Apple case

1. Introduction

We can safely assume that the person on the street regards an MNE′s effective tax rate of 0.005% on its European profits as unfair. 1 Ultimately, his or her effective tax rate is likely considerably higher. Neither will such a person be happy to hear that almost 40% of MNE profits have been shifted into “tax havens.” 2 That is a lot of money, the person will think, and it is I or my children who will have to fill the emerging holes in the budget.

To what extent this logic holds true in reality is hard to say. MNEs’ aggressive tax positions are often the tax benefits that states offer—out of pure self-interest. 3 So the person on the street on a small island in the Caribbean will have a different view about the matter from that of someone living in a European welfare state. Yet, in the aftermath of the financial crisis, the public outcry of the latter was loud 4 —and politicians reacted. 5 As stated in the final declaration of the 2013 G20 summit in St. Petersburg, “[i]n a context of severe fiscal consolidation and social hardship, in many countries ensuring that all taxpayers pay their fair share of taxes is more than ever a priority. Tax avoidance, harmful practices and aggressive tax planning have to be tackled.” 6

What started in 2013 with the OECD Base Erosion and Profit Shifting (BEPS) Project 7 is meant to peak in a “consensus-based multilateral solution to overhaul the international tax system.” 8 As part of these wider efforts, the commission has also launched various high-profile state aid proceedings 9 that should recover taxes from a number of MNEs that benefitted from “sweetheart-deals” with certain tax authorities; 10 among them is the above-mentioned MNE with an effective tax rate of 0.005%: Apple.

State aid is defined in Article 107(1) of the Treaty on the Functioning of the European Union (TFEU) as prohibiting “any aid granted by a Member State or through State resources in any form whatsoever which distorts or threatens to distort competition by favoring certain undertakings or the production of certain goods shall, in so far as it affects trade between member states.” The application of state aid rules on tax measures is not at all a new phenomenon but already happened in 1974. 11 Thus, the CJEU recognized rather early that a lower tax bill can be state aid as much as paying a subsidy. 12 Nevertheless, the field, often referred to as fiscal state aid, was regarded to have been relatively quiet. 13 With the start of the commission′s 2014 offensive, this definitely changed. The attention for this area grew starkly because of the almost absurdly high sums that are at stake in these proceedings. 14

The general objective of state aid rules is to promote fair competition between enterprises in the EU. As a result of the increased competition neutrality, the functioning of the internal market should be improved. Leaving this and dogmatic legal access points to the topic of fiscal state aid untouched, 15 this article aims to provide for a new perspective by asking whether state aid rules can also promote fairness in European taxation. In view of the statements of Commissioner Vestager, it can also be assumed that they are meant to do so. In 2016, the latter said:

For me, that issue of fairness is the most important message from our enforcement of state aid rules in tax cases and it is my duty as Commissioner for competition to make sure that the rules apply in a fair manner to any company that does business in the EU's single market-regardless of size, sector or nationality. ... Responsible taxpayers don’t have to pick up the bill left by companies that avoid tax. And the public gains confidence that there's a fair chance for everyone, not just for a few well connected companies 16

Inquiring into this issue is more than an academic hobby. While it takes quite some effort to build up an analytical framework against the background of which the analysis can be executed and admitting that the notion is still very hard to grasp, let alone, to measure, there are still important aspects that seem to make investigations such as this one worth conducting. In fact, our tax systems are built on voluntary compliance, and the extent of this, in part, depends on people's opinions on the fairness of taxation. Compliance was found to decrease when the perception that the tax system is unfair increases. 17 To see how much that matters, consider that according to some estimates, tax evasion may even amount to 850 billion to 1 trillion USD per year in the EU alone. 18

Section 2 will provide a minimalist discussion of the notion of fairness in corporate taxation. Thereafter, an analytical framework (section 3) will be set up against which the Apple case, as decided by General Court of the European Union (GC), will be tested. The aim is to see how close the case came to a stylized theoretical optimum of inter-taxpayer equity that can be reached in the EU with distinct tax systems. To account for the inevitable existence of trade-offs between different tax policy goals, we will also reflect on the reasons that explain why the GC has not always gone that far—and thus show the limits of the law, set by reality, that the person on the street needs to accept in the current tax landscape (sections 4–6). It is worth mentioning upfront that the article does not discuss Irish residency rules—despite their being an integral part of Apple′s tax planning strategy. Apart from the fact that these rules have already been changed, 19 they have—to the surprise of some 20—not been challenged by the commission in the Apple case.

2. Theory: Equity in (Corporate) Taxation 21

The debate on equity in taxation has been going on for more than 2,000 years. 22 In short, it is about fairness in sharing the tax burden among taxpayers (inter-taxpayer equity) 23 and sharing tax revenues among nations (internation equity). 24 This contribution will, as mentioned, focus on the first dimension.

Before we are able to do so, some issues require a brief preliminary discussion. First, while fairness underpins income taxation, 25 a consensus in “paraphrasing” 26 it is unachievable given the great importance of personal values involved in the discussion. 27 Second, and in a related way, the traditional way of assessing fairness between taxpayers by focusing on well-known notions of horizontal equity (i.e., equals should be treated alike) 28 and vertical equity (i.e., an appropriate distinction should be made in respect of the treatment of people who are not alike) 29 is hardly undisputed. 30 Nevertheless, it is fair to say that horizontal equity matters in practice. As argued by Lindsay, it “is a fairness norm useful for structuring compromises between people who disagree about ideal principles of tax justice or about the empirical consequences of tax policy”. 31 What is more, taxpayers tend to view more horizontally equitable tax policies as more legitimate. 32 As mentioned above, perceptions of a tax system's unfairness can even go in hand with taxpayers’ being less inclined to comply with it. 33 As a result—and in relative ignorance of powerful voices to the contrary—this article will deem the promotion of horizontal equity as increasing fairness. 34

The next challenges lie in how to measure when people are equal and when they are not. Two notions are traditionally relied on in this respect: The principle of ability to pay and the benefits principle. The former focuses on the economic capacity of a taxpayer and demands that taxation should reduce each taxpayer's welfare by an equal amount. 35 The latter, on the other hand, looks at the benefits a person has received from the public hand and calls for those who have received more to pay more in taxes. 36 Both principles have done their part in forming the international tax system. 37 For this article, the principle of ability to pay plays a more important role because this notion underlies the concept of income. 38

What is worth stressing in discussing the fairness of a corporate tax context is the fact that, ultimately, it is never the corporation itself that bears the tax. 39 Rather, the so-called “incidence” of the tax will fall on someone else who, at the end of the day, is an individual. 40 This may, for instance, be the corporation′s owners (through lower dividends), workers (through lower wages), consumers (through higher prices), or the suppliers (through lower prices). 41 It is not known who eventually pays the bill, and it is understood that economists also struggle to answer this question with sufficient precision. 42 However, in terms of fairness perceptions, these consideration at least do not appear to matter much. Instead, it is plausible to assume that people tend to focus on the taxation of the corporation—be it the “right” perspective or not. Maybe it is here also where the actual value of the discussion lies. 43 Consequently, it is broadly assumed that perceived fairness may be promoted if those corporations that earn as much pay as much income tax.

3. The Analytical Framework

The above sounds easier than it is. In fact, there are quite some methodological challenges that need to be overcome before entering into the analysis. For the sake of finding out whether and in as far the outcome of the Apple case can contribute to inter-taxpayer equity, some outside frame of reference is needed against which this question can be analyzed.

Based on what was discussed in section 2, this benchmark would have to be the horizontally equitable tax treatment of comparable economic activity in the whole EU. That being said, it is clear from the outset that such “absolute EU-wide inter-taxpayer equity” cannot be reached. Mere disparities between member states’ tax systems are, in principle, unproblematic under primary EU law. 44 They also widely exist with respect to both the tax base and the tax rate. Hence, comparable economic units in different member states are typically subject to a different tax treatment. State aid rules cannot and are not intended to change that. This stems from the delineation of competences in the EU whereby member states have largely retained their sovereignty in corporate taxation.

Therefore, it would not be helpful to measure the contribution of state aid rules to inter-taxpayer equity against this background. Rather, the authors think that more insight can be gained by accepting this reality and setting a somewhat lower optimum. If we keep in mind the underlying issue at stake in the Apple case, it is thought that the “second-best inter-taxpayer equity” result lies in every transaction made by an economic unit in the EU being taken into tax consideration in precisely the same manner throughout the whole EU. It is then up to the member state to decide whether the income or expense is reflected in the tax base and the rate at which the net result is taxed. This would mean that differences in tax treatment should be attributed to only the economic activity subject to tax in different member states but not to the organizational or legal form in which it is conducted in the EU. In more concrete terms, that would result in income from economic transactions being recognized, measured, and allocated in exactly the same way regardless if it was entered into by, say, a Portuguese independent entity, the German subsidiary of a Dutch MNE, or the Irish PE of a tax resident entity nowhere—while the attached tax consequences differ according to the member state in which the income is taxed. 45

Against this background, the authors will analyze how close the Apple case 46 comes to this stylized “optimum of EU-wide inter-taxpayer equity”. 47 In this context, it is considered that an inquiry into three different dimensions of state aid is necessary. The first one concerns the reference framework. If the tax measure in question does not deviate from what is considered “normal” taxation, there is no advantage and thus no state aid. In principle, the benchmark to assess whether a tax measure is general or selective is the national tax system of each member state. However, if we assume that the general objective of transfer pricing rules in all member states is to eliminate, for tax law purposes, the effects of the absence of market forces between transactions of integrated companies and thereby to provide equal treatment of independent and integrated companies, the frame of reference could be essentially uniform in all member states. Hence, for the sake of maximizing EU-wide horizontal equity, a broad frame of reference entailing the aim of an equal treatment of independent and integrated companies must prevail in all member states. This notion of a broad frame of reference would also be in line with the general objective of state aid rules to provide competitive neutrality between economic operators, i.e., independent and integrated companies. Second, the standard used in testing for the equal treatment of the taxpayers affected must be the same in the whole EU. This means that optimizing EU-wide inter-taxpayer equity calls for a separate and distinct EU arm's length principle. 48 Third, for the sake of maximizing EU-wide inter-taxpayer equity, inaccuracies that occur in applying the arm's length principle must not be tolerated.

As will be discussed in section 5, the Apple case does not always reach this optimum. At this stage, the narrow focus of the analysis will be relaxed and consideration will be made of reasons unrelated to inter-taxpayer equity that can support not having gone that far. After all, tax policy is full of trade-offs that often occur between ensuring inter-taxpayer equity, on the one hand, and the administrability of rules on the other hand. 49 Stated differently, the fairness conception of the person on the street is not the only thing lawmakers (can) think about.

4. Description of the Relevant Parts of the Apple Case

The Apple case concerned two rulings issued by Ireland in 1991 and 2007 to the two Irish subsidiaries, Apple Sales International (ASI) and Apple Operations Europe (AOE), of Apple Group. 50 Within the Apple Group, Apple Operations International is a fully owned subsidiary of Apple, Inc. Apple Operations International fully owns AOE which in turn fully owns ASI. ASI and AOE were incorporated in Ireland but were not tax residents in Ireland. 51 This was because the subsidiaries’ boards met in the United States, and under Irish law that applied at the time, they were nonresidents of Ireland for tax purposes. 52 Both ASI and AOE had an Irish branch where all of the employees were located and were responsible for, among other things, carrying out procurement, sales, and distribution activities associated with the sale of Apple-branded products in certain regions including Europe and for the manufacture and assembly of computer products in Ireland. 53

The rulings issued by the Irish tax authorities set a method for determining the amount of taxable profits in Ireland. For that purpose, the main issue was the allocation of assets, functions, and risks of ASI and AOE between the head offices and the Irish branches. Particularly, the allocation of Apple's intellectual property (IP) licences held by ASI and AOE were of crucial significance in the allocation of profits between the head offices and the Irish branches. The rulings issued by the Irish tax authorities allocated those licenses and profits therefrom outside of Ireland. 54

According to the commission's primary line of reasoning, the allocation of the IP licences held by ASI and AOE outside Ireland was incorrect since the head offices had no own employees who could control and manage the IP licenses. It held that the misallocation of the IP licenses outside Ireland resulted in a selective advantage in the form of tax reduction in Ireland because the profit allocation between the head office and the Irish branches did not correspond to a reliable approximation of a market-based outcome in line with the arm's length principle they see enshrined in Article 107(1) TFEU. 55

As a subsidiary line of reasoning, the commission contended that, even if the IP licences were correctly allocated outside Ireland, the rulings employed inappropriate methodological choices. It held that those choices were liable to amount to a selective advantage in that the outcome was not in line with the arm's length principle enshrined in the application of Article 107(1) TFEU. 56

Finally, the commission also put forward an alternative line of reasoning. In the first place, it claimed that the Irish tax authorities departed in the rulings from the national arm's length principle for the advantage of ASI and AOE. Alternatively, in the absence of an arm's length principle under Irish law, the commission alleged that the Irish tax authorities conferred a selective tax advantage for ASI and AOE on a discretionary basis. 57

The GC annulled the commission's decision on the aid measure implemented by Ireland in favor of Apple. The reason for the annulment was that the commission was unable to demonstrate the existence of a selective advantage for the purposes of Article 107(1) TFEU in favor of ASI and AOE in respect of the amount of taxable profits allocated to the Irish branches. The GC confirmed, similarly to its earlier decisions in the Starbucks and Fiat cases, that the commission may use the arm's length principle as a “tool” or “benchmark” for the purposes of examining whether the outcome accepted in the transfer pricing ruling sufficiently corresponds to a market-based outcome. 58 However, in doing so, the commission cannot set aside the national tax rules, including those of allocating profits between the head office and the branch of a nonresident company that must be used for determining the normal level of taxation in Ireland. 59

With respect to the commission's primary line of reasoning, the GC held that the commission misapplied the Irish tax rules on the allocation of profits to a branch of a nonresident company that formed part of a reference system for determining normal taxation in Ireland and denied the presence of a selective advantage. 60 More precisely, it held that under Irish law the commission was not entitled to allocate the income from IP licenses to the Irish branches of ASI and AOE solely because the head offices lacked any substance. 61 According to the GC, the commission's analysis of the functions performed by the Irish branches with respect to controlling and managing the Apple IP licences was defective and did not support the view that the income related to them should have been allocated to the Irish branches under Irish tax law. 62 Thus, the commission's attempt to allocate the profits by default to the branches because they cannot belong to the empty head office (“exclusion approach”) was dismissed. 63

With respect to the commission's subsidiary line of reasoning on methodological errors, the GC admitted that there were inconsistencies and lack of documentation in the tax rulings. However, in line with its argumentation in the Starbucks case, it held that methodological defects pointed out by the commission are not in themselves sufficient for establishing the presence of a selective advantage. In the same vein, the GC rejected the commission's alternative line of reasoning insofar as it was based on the use of broad discretion by the Irish tax authorities.

5. Promotion of Equity by Means of State Aid Rules

6. Conclusions and Outlook

In this article, the authors have examined to what extent the state aid rules contribute to inter-taxpayer equity in international corporate taxation between different legal organizations (independent companies and groups of companies) and legal forms (independent companies and branches) in which economic activities may be carried on. The examination was carried out in light of the GC's recent decision in the Apple case. The starting point was that in the absence of a full-scale harmonization of corporate taxation in the EU, no absolute EU-wide inter-taxpayer equity can be reached. However, the theoretical framework of a second-best alternative was created; a so-called “optimum of EU-wide inter taxpayer equity” that would pursue the objective of allocating income from all economic activities similarly irrespective of the organizational or legal form in which economic activities are conducted, except for the differences arising from the variances in tax basis and tax rates across the Member States.

At the outset, three different legal issues were defined related to the application of the state aid rules that would determine the extent to which they contribute to inter-taxpayer equity. These issues were analyzed based on the GC's decision in Apple. First, the authors found that the endorsement of a broad reference framework that includes all companies subject to tax in a member state is a remarkable contribution to EU-wide inter-taxpayer equity. Second, it was found that the GC did not endorse an autonomous EU arm's length principle that would be distinct from the national arm's length principles of member states. To that end, the GC's approach does not maximize the contribution of state aid rules to EU-wide inter taxpayer equity in that it allows differences between the national arm's length principles. Having noted this, nevertheless, the authors found the GC's approach sensible and well balanced in view of, among other issues, the current delineation of competencies between the EU and member states in corporate taxation and issues related to legal certainty and administration. Finally, it was concluded that, based on the GC's decision, the state aid rules impose only a certain minimum standard with respect to the application of the arm's length standard in a given member state in that they require the application to fall within a range of inaccuracies inherent in the application of the arm's length principle. Seen in isolation, this is not optimal from the perspective of EU-wide inter-taxpayer equity. Yet, it is sensible in view of the nature of the arm's length principle. In addition, potential concerns were raised in relation to the strictness of the burden of proof that the GC imposed on the commission that may be liable to have a negative impact on the enforcement of state aid rules.

The authors are most curious to observe the commission's future actions in this respect. After all, there is no doubt Apple engaged in profit shifting and ultimately widely avoided paying any corporate income tax on profits it derived from its sales in Europe, and the tax ruling agreed with Ireland was an important part of Apple's overall tax planning structure. 101 There are several options for the commission. On the one hand, it can appeal against the GC's decision, which it had already decided to do. In this regard, it may try to uphold its sui generis EU arm's length standard by means of which it could try to apply the exclusion approach. 102 Should the GC's sentence be upheld by the ECJ, the commission could start again and launch a new investigation whereby it could concentrate on different factors. Parts of the literature seem to see better odds for the commission should it challenge the Irish residency provisions that are said to have been geared toward allowing MNEs like Apple to create stateless income. 103

The authors further stress that although the commission lost the Apple case at the GC level, the fact that it won the question with respect to the reference framework is not to be underestimated. Apart from that, it seems that the commission has established itself as a serious player in the international tax arena. The launch of the state aid investigations at the very least played a role in Ireland in abolishing its residency rules. This means that such a tax structure could not exist today. 104 This is at least a small comfort for the person on the street.