This is the second article in a series of three that reviews the Commission’s Notice on the Notion of State Aid.* Last week’s article focused on the concepts of economic activity and state resources.** Next week’s article will conclude with a presentation of the conditions under which trade is affected and competition distorted and an analysis of the application of State aid rules to public funding of infrastructure.
Section 4 of the Notice focuses on advantage. Paragraphs 66-72 explain that advantage arises from any intervention of the state that reduces the costs that undertakings should normally bear. These are the costs which are inherent in their operations. It is irrelevant that some of the costs are imposed by law or decisions of regulatory authorities. Also irrelevant is the fact that the rules prevailing in an industry or regulating transactions in one Member state may differ from those in other Member States with the consequence that firms in different Member States are liable for different costs. The existence of advantage is established by comparing the situation of an undertaking before and after state intervention in the context of a single Member State [72 (the numbers in brackets refer to the corresponding paragraphs in the Notice)]. It does not involve any comparison with undertakings in other Member States, which may derive other kind of advantages such as lower taxes, access to cheaper labour, etc. The irrelevance of differences between Member States has also been recently reiterated at length by the General Court in T-47/15, Germany v Commission concerning support of green electricity producers and energy intensive industries.
The Notice also deals with the question whether a public authority which raises its funds from taxation can claim to act as a market operator [please bear in mind that a market operator does not have powers of taxation]. The question has been answered definitively by the judgments in T-156/04, EDF v Commission and C-124/10 P, Commission v EDF. The form of the aid, including the source of the money, is also irrelevant. What matters is the effect of the aid. A rational investor would seek to maximise profit regardless of how cheap it was to raise the funds in the first place [presumably it is thought that governments can easily raise funds through taxation].
Section 4 pays particular attention to the market economy operator test [MEOT] which is the same as the better-known market economy investor principle [MEIP]. Both are based on the premise that the objective of a market actor is the pursuit of profit, either to maximise returns from a prospective investment or minimise losses from a past investment. The same premise applies to the private creditor test and the private vendor test. 
When a public authority claims to act as a market operator it must disregard all public policy aims and ignore any obligations it has as an organ of the state.  The proof that a public authority acts as a market operator must be derived from an ex ante assessment of the profitability of the investment or transaction. When the Commission happens to examine such an investment ex post, it must consider only information that was available at the time of the investment or transaction.  All possible impacts on the profitability of the investment or transaction must be taken into account. 
“A transaction’s compliance with market conditions can be directly established through transaction-specific market information in the following situations (i) where the transaction is carried out ‘pari passu’ by public entities and private operators; or (ii) where it concerns the sale and purchase of assets, goods and services (or other comparable transactions) carried out through a competitive, transparent non-discriminatory and unconditional tender procedure.” 
It is important to note that those are sufficient conditions to rule out the existence of aid but they are not necessary conditions. If a transaction is not carried out on pari passu terms or an asset is not sold through a competitive tender it cannot be concluded that State aid necessarily exists.  It may still be possible to show that the public authority concerned has acted in the same way that a market operator would have acted. The Notice explains that the behaviour of the hypothetical market operator can be established with the use of other criteria which are termed “benchmarking and other assessment methods”. 
A transaction is pari passu when:
- The public and private investors act at the same time.
- The terms and conditions are the same for both classes of investors.
- The participation of the private investor is not nominal.
- The starting positions of the public and private investors are comparable; i.e. either they both have or do not have prior investments in the same company or project. 
The sale of an asset is free of State aid when the sale is effected through an auction or bidding process which is open, transparent, non-discriminatory and unconditional. 
Benchmarking excludes the possibility that State aid is granted when the behaviour of the public authority matches that of an investor who is in a similar position as the authority. 
Other methods focus on direct measures of the profitability of the proposed investment or transaction such as the internal rate of return, the cost of capital and the net present value of the investment. 
Investment in a new company or project is treated in the same way as investment in an existing company or project. In either case the decisive element is the prospect of a return that matches that of comparable investments or at least covers the cost of capital.
Investments in troubled companies can still be economically rational if there is a reasonable prospect that the company can be turned around and made profitable again. When a public authority makes an additional investment in a troubled company in which it has an existing investment then it is also necessary to examine whether the authority would be better off by liquidating its initial investment [106-107]. This is because just like the hypothetical initial investor who always has a choice between investing and not investing, the hypothetical existing owner of shares or capital of a company always has the choice of either selling [or liquidating the company] or making the additional investment that can return the company to profitability.
The Notice also examines specific conditions under which loans and guarantees can be free of aid [108-114].
Section 4 concludes with a distinction between direct and indirect advantage. Aid that is given straight to an undertaking confers a direct advantage. Aid to the clients or suppliers of an undertaking can provide an indirect advantage. The Notice warns in paragraph 116 that “indirect advantages should be distinguished from mere secondary economic effects that are inherent in almost all State aid measures (for example through an increase of output). […] An indirect advantage is present if the measure is designed in such a way as to channel its secondary effects towards identifiable undertakings or groups of undertakings. This is the case, for example, if the direct aid is, de facto or de jure, made conditional on the purchase of goods or services produced by certain undertakings only (for example only undertakings established in certain areas). [Footnote 181 clarifies that “by contrast, a mere secondary economic effect in the form of increased output (which does not amount to indirect aid) can be found where the aid is simply channelled through an undertaking (for example a financial intermediary) which passes it on in full to the aid beneficiary.”] I think that this wording will confuse rather than enlighten many practitioners. For example, if travel vouchers are given to individuals, who are not undertakings, is there indirect aid to transport companies? The answer is yes. This is a typical Article 107(2)(a) case.
The Notice does not advance any new or controversial points of view on the MEOT. However, three observations are in order here. First, normally one can identify many different market rates and benchmarks that can correspond to market-based transactions. The Commission enjoys considerable discretion in choosing the relevant benchmark, indicator or rate. In addition, there can be several values for any benchmark or indicator. The Notice mentions that it is more appropriate to choose the “average” or “median” rate  or “range of possible values”.  But the average value can vary significantly depend on the indicator that is chosen and when some of the values give negative results, subjectivity creeps in the calculations. The big arguments surrounding the MEOT are not about the methods. The methods are economically sound and have all been endorsed by EU courts [see, for example, T-319/12, Spain v Commission, where the General Court agreed with the Commission’s use of the capital asset pricing model to calculate the expected return]. The big arguments are about the specific values that prove the rationality of the investment or the existence of aid.
Second, recent judgments of EU courts have shown that when the private creditor test is applied other issues become important such as legal and procedural aspects that affect the speed of liquidation of a company or the calculation of the likely revenue that can be obtained from forced sale of assets [see, for example, T‑103/14, Frucona Košice v Commission].
Third, the Commission in most of its decisions involving application of the MEOT employs more than one of the methods outlined above. This provides assurance that indeed public investments are economically rational. Relying on a single method is never prudent.
Section 5 of the Notice deals with the issue of selectivity, especially with regard to tax measures. The Notice makes a distinction between “material” selectivity and “regional” selectivity.
A measure is materially selective when it differentiates between undertakings in a de jure or de facto manner. An otherwise general measure is de facto selective when in practice it favours particular undertakings . A general measure whose application is subject to administrative discretion can become selective . However, the authorisation or confirmation of eligibility on the basis of objective criteria established ex ante is acceptable because it deprives the relevant authority from the possibility of exercising administrative discretion .
The Notice goes on to analyse the selectivity of measures which instead of conferring an advantage, they reduce a disadvantage. When the purpose of a measure is to “mitigate” or reduce charges or taxes imposed on undertakings, then it becomes necessary to consider whether such charges or taxes ought to be borne in the first place by the undertakings which benefit from the mitigation or reduction. For this purpose, EU courts have defined a three-step test:
- Definition of the reference system.
- Determination of whether the measure in question is a derogation from the reference system which should apply to all undertakings which are in the same factual or legal situation in light of the objectives of the system.
- Determination of whether the derogation can be justified by the nature or general scheme of the reference system. 
The definition of the reference system is normally an easy task. It is the charge or tax that is levied by a public authority [there must always be compulsion and therefore the force of the law or official authority behind a measure that imposes a disadvantage]. The Notice deals with the issue of the reference system in just two paragraphs. “The reference system is composed of a consistent set of rules that generally apply — on the basis of objective criteria — to all undertakings falling within its scope as defined by its objective. Typically, those rules define not only the scope of the system, but also the conditions under which the system applies, the rights and obligations of undertakings subject to it and the technicalities of the functioning of the system.”  “The reference system is, in principle, the levy itself.” 
I am not aware of a case where EU courts have defined the reference system as in paragraph 133 of the Notice. Normally, the courts simply equate the tax under consideration with the reference system. However, two cases have shown that sometimes it can be very difficult to define the reference system because the tax can be designed in such a way that it has a narrow field of application. A narrow system benefits those products which are not subject to it without the need to grant to them a formal exception or derogation. In the absence of a formal derogation, then it is not possible to proceed to the second step of the test or the outcome of the second test will be negative. In the British Aggregates case, a tax on certain aggregates [materials used in construction, which are mined or extracted from quarries] was too narrow because it did not apply to other environmentally harmful mined materials [see C-487/06 P, British Aggregates v Commission]. In the Gibraltar case, a newly designed corporation tax used criteria that seemingly applied to all companies. The primary criteria were office space and number of employees. However, the Court of Justice concluded that those criteria excluded offshore companies which did not have offices or employees in Gibraltar [see C-106/09 P, Commission v Gibraltar]. In both cases, the Court arrived at its conclusions by asking what the aim of the tax was. The scope of the tax was narrow because in practice it did not apply to all products and undertakings that would be in the set of products and undertakings as defined by the aims of the tax.
It is worth noting that after the Commission suffered another defeat before EU courts in the British Aggregates case [T-210/02 RENV, British Aggregates v Commission], it finally decided recently in Decision 2016/288 that the non-application of a tax on aggregates used as building materials was not selective in the case of slate because it was not used in construction, but it was selective in the case of shale because the latter was a substitute for sand and gravel which were used in construction.
With respect to the second step, a mitigation or reduction constitutes a derogation when it places in a more advantageous position products or undertakings “which are in a similar factual and legal situation, in the light of the intrinsic objective of the system of reference.” 
After establishing the existence of a derogation, the justification of the derogation in the third step of the test is only possible if it stems from principles or objectives which are “intrinsic” in the reference system such as avoidance of double taxation . External considerations such as regional development or promoting international competitiveness may be legitimate policy objectives but are not intrinsic to tax systems, nor do they stem from the nature of such systems.
The Notice also provides detailed explanation of “regional selectivity”. A central government measure which applies only to certain regions is normally selective. A sub-national government measure which applies only to the area of jurisdiction of that government is not selective if the criteria that the Court of Justice expounded in the Azores judgment apply [see C-88/03, Portugal v Commission]:
- Institutional autonomy [the sub-national government enjoys decision-making powers on taxation].
- Procedural autonomy [non-interference in the decision-making process by the central government].
- Economic autonomy [no mitigation of any revenue shortfall by the central government] [145-155]
The Notice ignores the recent judgment of the General Court in case T-461/12, Hansestadt Lübeck v Commission where Lübeck argued successfully that lower airport charges by its airport were not regionally selective because they applied to all airlines within its area of jurisdiction [i.e. all airlines that used its airport]. The judgment has been appealed by the Commission. If it is upheld by the Court of Justice it will corroborate the simple principle that a measure of a sub-national government that is applied to all undertakings in its jurisdiction is not selective regardless of whether the beneficiaries are not the universe of undertakings in a Member State.
Section 5.4 of the Notice deals specifically with corporate tax measures. It covers the tax treatment of the following types of undertakings or economic activities: cooperative societies, collective investments, beneficiaries of tax amnesties, beneficiaries of tax rulings, beneficiaries of tax settlements, depreciation and amortisation rules and anti-abuse rules.
In view of the negative decisions taken recently by the Commission in relation to Starbucks, Fiat and the Belgian Excess Profits Rulings and the opening of investigations on Amazon, Apple and McDonald’s, the most controversial part of the Notice is likely to be that on tax rulings and settlements. As explained in the Notice, “the function of a tax ruling is to establish in advance the application of the ordinary tax system to a particular case in view of its specific facts and circumstances. For reasons of legal certainty, many national tax authorities provide prior administrative rulings on how specific transactions will be treated fiscally. This may be done to establish in advance […] how ‘arm’s-length profits’ will be set for related party transactions where uncertainty justifies an advance ruling to ascertain whether certain intra-group transactions are priced at arm’s length.” 
Advance tax rulings are especially useful to multinational companies with subsidiaries which trade exclusively or almost exclusively with other companies that belong to the same group. Since the suppliers or clients of such companies are related companies, the parent company can determine the price at which goods or services are bought or sold. This means that a subsidiary may show zero profit and therefore pay zero tax or it may not even show any revenue at all. In order to prevent subsidiaries of integrated companies from avoiding to paying tax, tax authorities decide in advance how tax is to be calculated and how much tax should be paid. Both the companies and tax authorities avoid disputes later on.
There is selective advantage “where a tax ruling endorses a result that does not reflect in a reliable manner what would result from a normal application of the ordinary tax system”.  The problem with this statement is that, when prices and revenues can be set by the parent company, the normal application of the ordinary tax system may lead to zero or very low profits and therefore zero or low taxes. In the next paragraph, the Notice refers to the judgment of the Court of Justice in the Forum 187 case [C-182/03, Belgium v Commission]. In that case the relevant Belgian rules reduced artificially the taxable base by excluding without justification certain expenses. The Court implied, reasonably, that prices had to cover all costs. The application of the ordinary tax would result in higher taxes. In this instance it is possible to make the comparison that the Notice defines in paragraph 170. But the problem is that, where there is no artificial shrinkage of the tax base, the case law provides no guidance on how far above costs prices between related companies must be set. Hence it is difficult to understand what result can be obtained from the “normal” application of the tax system.
Nonetheless, the Notice asserts that “a tax ruling which endorses a transfer pricing methodology for determining a corporate group entity’s taxable profit that does not result in a reliable approximation of a market-based outcome in line with the arm’s length principle confers a selective advantage upon its recipient” . Please note that the “normal application of the ordinary tax system” is now equated with “an approximation of a market-based outcome”. This means that the profit of a company has to be estimated on the basis of a proxy. Indeed the Notice continues that “the search for a ‘reliable approximation of a market- based outcome’ means that any deviation from the best estimate of a market-based outcome must be limited and proportionate to the uncertainty inherent in the transfer pricing method chosen or the statistical tools employed for that approximation exercise” . The implication of this approach is that the tax liability of one company is determined according to the tax liabilities of other companies.
The Notice goes on to declare that “this arm’s length principle necessarily forms part of the Commission’s assessment of tax measures granted to group companies under Article 107(1) of the Treaty, independently of whether a Member State has incorporated this principle into its national legal system and in what form. It is used to establish whether the taxable profit of a group company for corporate income tax purposes has been determined on the basis of a methodology that produces a reliable approximation of a market-based outcome. A tax ruling endorsing such a methodology ensures that that company is not treated favourably under the ordinary rules of corporate taxation of profits in the Member State concerned as compared to standalone companies who are taxed on their accounting profit, which reflects prices determined on the market negotiated at arm’s length.” 
The three sentences in paragraph 172 raise three questions. First, are group companies in the same situation as independent companies? Many business economists would argue to the contrary. Second, has the arm’s length principle been applied to any other case before the recent decisions on Fiat, Starbucks and Excess Profit Rulings? It appears that the Commission tries to give a pedigree to a new approach. It is perhaps revealing that, apart from the Forum 187 judgment, no case law is cited in support of the statements in paragraphs 171 and 172. Third, is the Commission not transgressing its powers and encroaching into the fiscal prerogatives of Member States by using the arm’s length principle where Member States themselves do not use it?
The Notice concludes the analysis of tax rulings by restating that such rulings confer a selective advantage “[…]where the tax authority accepts a transfer pricing arrangement which is not at arm’s length because the methodology endorsed by that ruling produces an outcome that departs from a reliable approximation of a market-based outcome” . This conflates the procedure with its outcome. There can be different arm’s length arrangements, each resulting in a different outcome. However, inconsistent application of rulings to undertakings which are in a similar legal and factual situation certainly results in selective treatment.
In conclusion, the part of the Notice on tax rulings does not just identify the relevant case law and clarify current practice. It also lays down new principles and signals a new approach in the assessment of tax rulings.
* Commission Notice on the notion of State aid as referred to in Article 107(1) of the Treaty on the Functioning of the European Union, OJ C262, 19/7/2016, p.1. The text of the Notice can be accessed at: http://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52016XC0719(05)&from=EN.
** View last week’s article here: http://www.stateaidhub.eu/blogs/stateaiduncovered/post/6796.
 Commission Decision 2016/288, OJ L59, 4/3/2016.