Not a transfer pricing question, rather a State aid answer

The Commission of the European Union decided that the tax rulings, which accept advance price arrangements (APAs) of Fiat in Luxembourg, Starbucks in Netherlands, 35 multinationals in Belgium and Apple in Ireland, are prohibited State aid due to the transfer pricing accountabilities contained therein, because the questioned APAs did not delivered an arm’s length result. Although these cases are filled with transfer pricing issues, it is believed they could be solved only by analysing how the Commission assessed the State aid criteria contained in article 107, paragraph 1 of the Treaty on the Functioning of the EU. Therefore, this paper focuses solely on some of the most critical State aid questions contained within these cases.


The present article aims at raising some critical legal questions about the European Union (EU) Commission’s interpretation of the EU State aid criteria, prescribed in article 107, § 1 of the Treaty on the Functioning of the EU (TFEU) to the cases Starbucks, Fiat, Belgium and Apple.12 The questions raised below intend to spur a critical debate about the difficulties in assessing such criteria in tax measures, such as the tax rulings that accepted advance pricing arrangements (APAs) of multinationals, challenged by the Commission in these cases. Moreover, underneath this main objective, to question: how far is the Commission legitimized to interpret such rules and interfere with the sovereignty boundaries of the EU Member States (MS)?

Due to delimitation purposes, I chose to approach and question what I believe to be the most sensitive and critical Commission’s assessment and interpretation of the State aid criteria contained in the selected cases. Therefore, for delimitation purposes, this paper focuses on the EU State aid criteria discussion, rather than the transfer pricing (TP) methods one.

The following section 2 provides a brief background of the selected cases relevant for the EU State aid criteria interpretation and discussions raised in section 3. This article does not provide assertive answers to the possible outcome of these cases at the EU Courts. It does however consider some possible answers and legal arguments useful for the development of a convergent interpretation of the EU State aid rules, aiming at a higher level of legal certainty in this field. This article is summed up with some final comments for consideration.

EU Commission, case on State aid SA.38374 (2014/C ex 2014/NN) implemented by the Netherlands to Starbucks; EU Commission, case on State aid SA.38375 (2014/C ex 2014/NN) which Luxembourg granted to Fiat; EU Commission, on the Excess Profit Exemption State Aid Scheme SA.37667 (2015/C ex 2015/NN) implemented by Belgium; EU Commission, case State aid SA.38373 (2014/C) (ex 2014/NN) (ex 2014/CP) – Ireland Alleged aid to Apple. Decision against Ireland not published yet.

Article 07, §1, TFEU: Save as otherwise provided in the Treaties, any aid granted by a Member State or through State resources in any form whatsoever which distorts or threatens to distort competition by favouring certain undertakings or the production of certain goods shall, in so far as it affects trade between Member States, be incompatible with the internal market.


In 2013, the Commission initiated an informal period of State aid investigations, by requesting MS to provide information about APAs approved and ongoing to multinationals incorporated within their jurisdictions. Last August, the Commission announced it considered Ireland’s tax rulings that accepted Apple Sales International and Apple Operations Europe (Apple Ireland) APAs as illegal state aid. Since illegal State aid can only be recovered retroactively for period of 10 years, the Commission’s decision is limited to the years 2003 to 2014 and it amounts approximately €13 billions of recovery taxes plus interests.

In October 2015, before this decision, the Commission concluded that the TP schemes in practice by Starbucks in the Netherlands and FIAT in Luxembourg were also illegal State aid measures, because they did not deliver an arm’s length principle (ALP) outcome within the intra-group transactions. As a consequence, the tax rulings that accepted the APAs conferred an illegal State aid and they amount to approximately 20 to 30 million euros of recovery taxes plus interests each.

In the Starbucks and Netherlands case, the Commission understood that Starbucks Manufacturing Europe Middle-East and Africa BV (SMBV) taxation scheme contained in the questioned APA, which determined the level of royalties’ payments to Alki LP (UK) after the corporation income taxation accountability, constituted an illegal State aid.3 Similarly in FIAT and Luxembourg case, the Commission understood that the equity remuneration and taxable profit accountability of the company in Luxembourg also constituted an illegal State aid.4

In the beginning of 2016, in the same line of reasoning, the Commission understood that Belgium granted illegal State aid for 35 multinationals, which corresponds to approximately 700 million euros of recovery tax plus interests. Following the trend of these four cases, the same outcome is expected for Amazon and McDonalds (the next cases in line), because the Commission seems to be political motivated to tackle multinational’s aggressive tax planning and structures within the EU.5 Conversely, it also tackles the main purpose of using tax rulings to ensure legal certainty to the taxpayers about their tax burden within the jurisdiction concerned. Hence, these cases have the potential to change the level playing field within the EU market, if the EU Courts confirm the Commission’s decisions.

§§40–45 in Netherlands case (note 1).

§§43–51 in Luxembourg case (note 1).

EU Commission press release, available at: and (accessed 1 October 2016).


3.1 General

The EU State aid rule does not differentiate how the assistance is granted, because article 107, §1, of the Treaty says "any aid granted ... in any form whatsoever”, i.e. the Commission’s action to investigate MS tax rulings is absolutely justifiable by this provision.

The referred article prescribes four cumulative criteria that must be fulfilled in order to consider a measure prohibited.6 The reason why this paper approaches these criteria instead of TP issues regards to the fact that, these cases might have a different outcome at the EU Courts just by dealing primarily with State aid questions.

This writer considers that article 107, §1, TFEU contain four cumulative criteria: (i) by a Member State and through State resources; (ii) favourable tax treatment (or tax advantage); (iii) selective; (iv) distort or threatens to distort competition and trade between Member States.

3.2 By a Member State or through State resources

The EU State aid prohibitory rule does not require any particular form, simply that the aid is granted by a Member State or through State resources. The first part of this criterion, by a MS, is easily verified in the cases under analysis, since it was the MS’ tax authorities who recognised as valid the corporation’s APAs as valid. If such tax rulings supposedly conferred a tax advantage to the recipients of these decisions, they consequently legally bind the MS concerned, which is thus imputable to the aiding measure.

Regarding the second part of this first criterion, through State resources, according to the pacified case-law, taxation has the capacity of transferring directly or indirectly State resources. In this sense, any type of tax benefit can potentially make an indirect use of public resources, because the tax normally levied is not charged at its regular time.7 Consequently, the revenue normally raised never enters the public treasury, incurring an indirect loss of tax revenues or consumption of public resources. Following this rationale, the fulfilment of through State resources criterion requires that the recipient of the aid receives a tax relief of his or hers usual tax burden, which is called favourable tax treatment or tax advantage and represents the second State aid criterion.

Case C-379/98PreussenElektra AGv.Schleswag AG and other[2001] ECR I – 2159, paras 58–60.

3.3 Tax advantage, selective or selective advantage

In the Starbucks, Fiat, Belgium and Apple cases, the Commission concluded that the tax rulings investigated conferred a selective advantage to the corporations that benefited from the illegal State aid decisions. Although tax advantage and selectivity are in principle two different criteria, because it is possible to confer a tax advantage without being selective, the Commission carried out their assessment jointly.8 A clear distinction must be made between both criteria, since tax advantage refers to the tax burden of the recipient (non-horizontal benefit), whereas selectivity targets the recipient of such advantage.9

Notwithstanding the Commission’s joint approach choice, it is settled in the case-law that selectivity assessment in tax matters should follow three progressive steps.10 First, identify and examine the reference system which is also relevant for verifying the existence of a tax advantage and, consequently, the use of State resources. Second, verify within the reference regime, if the appointed tax advantage derogates from this regime by favouring some taxpayers and not others (called prima facie selective). Third, verify if this derogation can be justifiable by the structure, rationale and guiding principles of the tax system in reference. If the measure cannot be justifiable, it is thus selective, otherwise it is general and not State aid.11

Due to delimitation purposes, the selectivity and tax advantage analysis below focuses particularly in the Starbucks (Netherlands) and Fiat (Luxembourg) cases, although this discussion is still useful for the Belgium and Apple Ireland cases.

When appointing and identifying the normal or common reference system for the first step of selectivity assessment, the Netherlands argued that the Dutch Corporation Tax Act 1969 (CIT), specifically article 8b, §1 and, the Dutch Transfer Pricing Decree (Decree) constitute the reference regime. Netherlands argued that the reference system incorporates article 9 of the OECD Model Tax Convention on TP and is applicable only to corporation groups. Conflictingly, the Commission understood that the reference system is only the CIT, rejecting the application of the Decree as part of the system and, that the reference system is applicable to all companies, regardless their structure (standalone or groups).12

Similarly, in the Fiat case, Luxembourg appointed that the reference system is the Luxembourg Income Tax Code (LIR), specially article 164, §3, and is applied only to group of companies. The Commission divergently understood the reference system is Luxembourg Corporate Income Tax (IRC) and that it is applied to all companies, regardless their structure.13

Following the referred confronting arguments between the Commission and the MS, the first step of selectivity analysis and starting point question is: which is the proper common and normal reference system in Netherlands and Luxembourg cases? From this question two clear departing answers can be appointed. At one side, if the EU Court understands that the Commission’s decisions are based on the wrong or incomplete common and normal reference framework of the MS imputable for the tax ruling, it is safe to presume that these decisions will be overruled, since they would lack legality in the light of the proper reference regime. Contrary, if the EU Court understands that the Commission properly identified the reference system, the Court should than verify if the tax rulings questioned created a favourable tax treatment to the recipient of the measure. Therefore, the logic follow-up question and second step of selectivity analysis is: which are the taxpayers under the same factual and legal situation according to the reference regime?

This analysis draws a circle around all taxpayers in a comparable situation in the light of the reference regime and it is relevant to understand how the Commission concluded that, the MS’ tax rulings granted a selective advantage to Starbucks and Fiat in comparison to their competitors. First, the Commission pointed out that Starbucks and Fiat income taxation results changed after the tax rulings were granted.14 During the process of investigating the existence of a State aid measure within these administrative tax decisions, the MS, recipients of the rulings and third parties provided substantial information to the Commission. A brief distinction must be made, Fiat only carried out business within the group, whereas Starbucks carried out business within the group as well as with non-related parties.15 Finally, the Commission understood that the investigated tax rulings granted a selective tax advantage to both companies, mainly because of the arguments presented below.

First, the method used to deliver corporations groups and standalone companies income tax base was accounted differently.16 Therefore, is the reference system applicable to both standalone companies and corporation group of companies? If the Court understands that the answer is ’no’, following the MS position and contradicting the Commission’s argument, corporation group of companies and standalone companies are not in the same comparable position and cannot receive the same tax treatment. On the other hand, if the Court follows the Commission’s position, this understanding strengthens the selectivity criterion fulfilment and requires an analysis of other arguments.

Second, the Commission challenged that the accountability methods contained in the APAs and approved by the challenged tax rulings did not deliver an arm’s length result regarding: (i) the corporation income tax base; (ii) the level of royalties’ payments in Starbucks case; and, (iii) the return tax rate on equity payments in Fiat case.17 Hence the TP methods applied did not consider the normal market conditions.18 The Commission understood that the questioned tax rulings fulfilled tax advantage and selective criteria, with the following argument:

The arm’s length principle therefore necessarily forms part of the Commission’s assessment under Article 107(1) of the TFEU of tax measures granted to group companies, independently of whether a Member State has incorporated this principle into its national legal system. It is used to establish whether the taxable profits of a group company for corporate income tax purposes has been determined on the basis of a methodology that approximates market conditions, so that that company is not treated favourably under the general corporate income tax system as compared to non-integrated companies whose taxable profit is determined by the market. Thus, for any avoidance of doubt, the arm’s length principle that the Commission applies in its State aid assessment is not that derived from Article 9 of the OECD Model Tax Convention, which is a non-binding instrument, but is a general principle of equal treatment in taxation falling within the application of Article 107(1) of the TFEU, which binds the Member States and from whose scope the national tax rules are not excluded.19

The Commission legally based the cited paragraph with another paragraph of a ECJ case, which says "it should be pointed out, first, that rules relating to tax are not excluded from the scope of Article 87 EC.”20 It is unclear why the Commission cited this part of the referred case to fundament its position, since within the context of the referred ECJ case, the Court was answering to an argument that the Commission did not have stands to review MS tax matters, according to article 87 EC (article 107, §1, TFEU).

The Commission also cited other parts of the referred ECJ precedent (§§96–97) to imply that the arm’s length result is a prerequisite of the State aid criteria in tax measures. However, the ECJ understood in the light of the regime analysed in that case, that the TP accountability which did not include the expenditures costs relevant for the determination of the business’ taxable income, represented a selective tax advantage. Following the rationale that, once such costs are normally charged in a free market, but only 5% of the undertakings in the same factual and comparable situation (coordination centres in Belgium) benefited from this measure, such tax law represented an illegal State aid scheme.21 Consequently, the arm’s length result in this case was part of the logic of Belgium’s tax law regime and not a general rule of article 107, §1, TFEU, as sustained by the Commission.

In this sense, the key question to this second step analysis and for the future of the EU State aid rules interpretation in tax matters is: is the arm’s length principle a general principle of article 107, §1 of the Treaty, when it comes to ensuring an equal tax treatment amongst taxpayers?

If the Court’s understanding is similar to the Commission’s position, this will represent a landmark in State aid cases involving tax matters, since this position will legitimise the Commission implementation of the ALP as a general principle of the Treaty (State aid) rule. It seems very aggressive to adopt such understanding, since the letter of the referred article does not intrinsically imply such interpretation, rather the Commission does. Moreover, article 109, TFEU requires that the Council adopt secondary law (regulations) for the application of article 107, thus such interpretation of the referred State aid article could represent a breach of article 109.

Contrary, if the Court disagrees with the Commission’s position, the conclusion that the ALP is a general principle of article 107, §1, TFEU falls apart, which consequently maintain the MS sovereignty rights for this tax matter and some level of legal certainty, in the sense that the ALP can only be implemented through national law or EU secondary law. Following this rationale, the argument that the tax rulings in Starbucks and Fiat cases granted a selective advantage, because the tax base and tax rate applied in these companies’ taxation did not represent an arm’s length result, has to be based strictly in the national law scheme and not a Treaty interpretation.

Finally, in the third step of selectivity analysis, the Court should answer if the tax rulings questioned have in fact deviated from the reference system in an unjustifiably way, fulfilling the selectivity criterion with a positive answer. If these cases are not solved by the Court until this level of State aid analysis, TP methods and ALP will most likely become the main legal discussion, since the last criterion (distort or threatens to distort) of State aid are easily fulfilled with the presence of the other three criteria.

Section 9.2 in Netherlands case; Section 7.1 in Luxembourg case; Section 6.3.4 in Belgium case; §§48, 52, 55 and 70 in the investigation procedure against Ireland (note 1).

See Aldestam M, EC State aid rules applied to taxes – An analysis of the selectivity criterion (Iustus Förlag 2005), pp. 64–65.

Case 730/79Phillip MorrisECLI:EU:C:1980:209, para. 11; Joined Cases T-298/97, T- 312/97 etc.AlzettaECLI:EU:T:2000:151, para 80.

Case C-88/03 Portugal v. Commission [2006] ECR I-7145, §81.

§§237–238, 245, 249–250 in Netherlands case (note 1).

§§193–210 in Luxembourg case (note 1).

§§234–236 and Table 3 in Netherlands case; and, §§197–199 and Table 7 in Luxembourg case (note 1).

§§148–150, 202–212 in Netherlands case; Table 3, §66 in Luxembourg case (note 1).

§61 in Netherlands case; §197 in Luxembourg case (note 1).

§§ 235 and 360 in Netherlands case; §§304, 232–233 in Luxembourg case (note 1).

§§255–263 in Netherlands case; §§219–227 in Luxembourg case (note 1).

§228 in Luxembourg case; §264 in Netherlands case; highlight added (note 1)

Joined Cases C-182/03 and C-217/03 Belgium and Forum 187 v. Commission ASBL ECLI:EU:C:2006:416, § 81.

Idem, §§ 96–97, 104–105.


This paper raises some of the most critical and primarily EU State aid questions in the Starbucks and Fiat cases, according to my point of view, and with the intention to spur a legal debate about their possible answers at the EU Courts. Most of the questions raised if not all of them, could be also used to discuss the Belgium and Apple cases, as well as the other cases next in the Commission’s line of State aid investigations, which targets MS tax rulings that accepted multinationals’ APAs.

1) Which is the proper common and normal reference system?

2) Which are the taxpayers under the same factual and legal situation according to the reference regime?

3) Is the reference system applicable to both standalone companies and corporation group of companies?

4) Is the arm’s length principle (ALP) a general principle of article 107, §1 of the Treaty, when it comes to ensuring an equal tax treatment amongst taxpayers?

The answers considered and briefly explained in this paper, represents only a small parcel of the difficulties about interpreting the EU State aid rules in tax rulings cases. Each question has the potential to be analysed and discussed more thoroughly within the EU State aid field, as well as under national tax law perspective and international tax law – e.g. transfer pricing.

Regardless which choice of legal approach one may use to analyse these cases, it is undisputable that they are relevant and may change legal practices, businesses’ tax planning and national tax policies within the EU internal market. At this moment, it is only left for us to debate these cases in a constructive manner and wait for their final verdict at the European Court of Justice in many years to come.

Joana Pedroso is doctoral candidate in tax law at the School of Business, Economics and Law, Gothenburg University.