Tax measures are selective when they constitute an exception or deviation from the normal or common system of taxation. In addition, the exception must be open only to a pre-defined category of undertakings.


Often, the decisive element in whether a tax measure constitutes State aid is the existence of selectivity. On 7 November 2014, the General Court ruled on two almost identical cases that focused on the concept of selectivity: T-399/11, Banco Santander v European Commission and T-219/10, Autogrill Espana v Commission. Banco Santander and Autogrill Espana had applied for annulment of Commission Decisions 2011/282 and 2011/5, respectively. The Decisions concerned the tax amortization of financial goodwill in case of foreign shareholding acquisitions, either outside the EU, in the case of Banco Santander, or inside the EU, in the case of Autogrill Espana. Given the very close similarity between the two Commission Decisions and the corresponding Court rulings, in this article I analyse the judgment in case T-399/11, Banco Santander[1] and Commission Decision 2011/282[2].

This is a landmark judgment. It interprets the concept of selectivity in a way that goes beyond formalistic comparison of companies that benefit from aid and companies that do not benefit from aid. Even if formally a measure is an exception from a more general tax rule, the measure does not necessarily become selective if it is open to any undertaking. The judgment raises the standard of proof for the Commission. In the future the Commission will have to prove not only that a tax measure deviates from the normal system of taxation, but it must also demonstrate that it is possible to identify ex ante a well-defined category of aid beneficiaries.


After a lengthy investigation of a tax advantage given by Spain to companies to amortize goodwill arising from the acquisition of shareholdings in foreign companies, the Commission, by Decision 2011/5 of 28 October 2009, closed the proceedings with regard to equity investments made within the European Union. The Commission found that the tax measure constituted State aid which was incompatible with the internal market.

However, the Commission continued the investigation with regard to equity investments made in companies outside the Union. On 12 January 2011, the Commission adopted Decision 2011/282.

The Spanish measure in question provided that in the case of acquisition by a taxable company in Spain of at least 5% of the shares of a foreign corporation and the shares were kept for at least a year, the goodwill arising from this investment could be deducted from the tax liability of the acquiring company. In simple terms, goodwill is the difference in the price paid for a company and the value of its underlying assets. This difference results from several factors such as brand recognition and commercial relations with customers. It is what makes companies valuable in the eyes of investors.

Bank Santander contested the Commission Decision on the grounds that the Spanish measure was not selective. It argued that the measure could apply to or benefit any undertaking in Spain. As is well known, a measure which is general in the sense that it applies to all undertakings in the same situation does not fall within the scope of Article 107(1) TFEU.

The judgment

The General Court began its analysis by recalling how the selectivity of tax measures is determined.

Article 107(1) requires assessment of whether, in the context of a given legal system, a national measure is likely to favour certain undertakings in relation to others which, in the light of the objective pursued by the system, are in a comparable factual and legal situation. [Paragraph 33]

The determination of the relevant system is especially important in the case of tax measures because the existence of an advantage can only be established with respect to the normal tax or the reference framework. [34-35] The normal tax or reference framework is the rule that would apply to an undertaking had it not been exempted from it.

However, when a given system appears to differentiate between firms, and it is thus a priori selective, in reality there is no selectivity when the differentiation arises from the nature or general scheme of the system [C-88/03, Portugal v Commission]. [36]

It follows that the qualification of a tax measure as selective is based on a three-step test:

First, the normal or reference tax system must be identified.

Second, it must be established whether certain firms are differentiated from others through, for example, an exemption. These firms must be in a comparable factual and legal situation in view of the defined objective of the tax system.

Third, it must be considered whether the measure is justified by the nature or general scheme of the system.

The Court pointed out that a measure can be selective even when it is not limited to a particular industry or sector of activity. [39] It usefully cited several examples: [40-43]

  • The measure affects many sectors [C-169/84, COFAZ v Commission; C-126/01, GEMO].
  • The measure applies to manufacturing undertakings across sectors [C-143/99, Adria-Wien Pipeline].
  • The measure covers all companies locatedin a geographic area [C-156/98, Germany v Commission].
  • The measure is limited to a particular period of time [even though it formally applies to all companies, given the short timeavailable tocompaniesto meet the qualifying requirements, the measure is in reality open only tocompanies that are already prepared] [T-211/05, Italy v Commission].

Then the Court added a significant clarification. When the category of aid recipients is particularly broad or diverse, the decisive element in determining the selectivity of a measure is not so much which company is included in it but which is excluded from it [T-55/99, CETM v Commission]. [44] This is correct. Often the selectivity of a measure is demonstrated not by showing that the set of beneficiaries is not the same as the set of companies which are in the same situation, but by showing that there is at least one company that is excluded from its scope.

Drawing on the above, the Court concluded that the identification of a particular category of undertakings which are the only ones which are favoured by the measure in question is an indispensable or necessary criterion for selectivity. [45] [Emphasis added]

The Court continued that even when a measure which departs or diverges from the common, normal or reference system of taxation is potentially applicable to all undertakings, there cannot be a distinction between undertakings that benefit from it and undertakings that do not benefit from it. [48]

Then the Court deduced that for selectivity to exist, it is necessary in every case to identify the favoured undertakings and that it cannot be inferred that a measure is selective by the mere fact that it is an exception from the common or normal tax regime. [49] [Emphasis added]

This statement is extremely important. I believe, it is the first instance in the case law that an EU court has made a distinction between a formal exception and identification of favoured companies in such a manner. In most if not all other cases, it is automatically presumed that the favoured companies are the exempted companies and that exception necessarily benefits only certain companies. Now the General Court appears to be saying that an exception may be so broad that in practice no company is excluded and therefore no sub-set of companies can be identified as the only beneficiaries.

A simple example can illustrate what the General Court may mean. The normal tax systems of many countries require employers to pay social insurance for their employees. However, a country, in order to encourage the employment of long-term unemployed, may allow companies to benefit from reduced social insurance rates in case they hire a person who was unemployed for longer than a year. The reduced rate is a deviation or an exception from the normal system and it is not justified by the logic of the social insurance system that employers [and persons in employment] contribute to the funds which are used to provide financial support in case a person loses his or her job or becomes unable to work. But in this case what appears to be an exception is in fact a general measure because it is open to any undertaking that hires a person who was previously unemployed for a long time. Although this may appear unusual, the Commission did in fact approve as a general measure a Belgian reduction of social insurance rates for employment of manual workers [Commission Decision N 132/1997, “Maribel Quarter”]. It should be noted that the earlier Maribel bis/tre measures which were limited to companies affected by international competition had been found to be incompatible State aid [see C-75/97, Belgium v Commission]. [A similar Danish measure was also found by the Commission to be general in nature and therefore not to constitute State aid].

The General Court then clarified that it is the responsibility of the Commission to prove that a measure is selective. It added, for example, that in C-279/08, Netherlands v Commission, the beneficiaries could be clearly differentiated form other companies as only a specific group of large industrial companies were given free emission allowances. Similarly, in C-106/09 P, Commission v Gibraltar & UK, only offshore companies were subject to a special tax system. [50-52]

Having established the general principles of how it would identify whether a measure was selective, the Court then turned its attention to how the Commission had attempted to prove the selectivity of the Spanish measure. The Commission had concluded that the selectivity of the measure was based primarily on the existence of an exception from the normal tax system or reference framework, as the measure applied only to share acquisitions abroad. The normal system or reference system was the tax treatment that applied to share acquisitions in Spain.

As mentioned earlier, Bank Santander argued that the measure did not favour any specific group of undertakings as it was open to all undertakings. The Court acknowledged that the measure applied to all acquisitions by at least 5% of foreign companies that were held without interruption for at least one year. The measure did not refer to a particular business or production category, but to a particular category of economic operations or transactions. [57]

Yet, the Court immediately recognised that in some cases, companies may in practice be excluded from the scope of a measure that at first glance appears to be a general measure [C-6/69, Commission v France; C-106/09 P, Commission v Gibraltar & UK]. [58]

In the present case, in order to take advantage of the measure, a company had to purchase shares in a foreign company. [59] According to the Court, this operation was purely financial and did not require the acquiring company to change its activity [60]. Moreover, the measure was independent of the nature of the business activity of the acquiring company [61]. It did not set any minimum amount that had to be invested in the minimum of 5% shareholding [unlike the monetary threshold set in case T-227/01, Alava Provincial Council v Commission] [62], nor did it exclude, a priori, any category of business [65].

The General Court also addressed the Commission argument that the measure was selective because it was limited to investments abroad. It rejected that argument on the grounds that the Commission did not show that the effects of the measure were such as to limit the advantage it conferred only to certain undertakings [67]. The Court recalled that whether a measure is selective has to be determined according to its effects [69]. In this connection, it observed that, in the judgment on 3M Italia [C-417/10], the Court of Justice held that it would not be sufficient to conclude that the measure in that case was selective merely because only taxpayers who satisfied the requirements of the measure could benefit from it [70].

Also, in the judgment in Commission v Gibraltar & the UK, the Court of Justice held that any tax differentiation did not necessarily imply the existence of State aid. It was also required that the measure was capable of pre-determining a particular category of beneficiaries according to their special features [their offshore nature]. [71]

The General Court believed that the approach proposed by the Commission could lead to a finding of selectivity for any tax measure whose application was subject to certain conditions, even though beneficiary firms did not have any specific characteristics that could distinguish them from other non-beneficiary companies. [72]

In response to the Commission view that the measure intended to promote the export of capital from Spain and to strengthen the position of Spanish companies abroad, the General Court recalled that state action that benefits equally all businesses located in a country does not constitute State aid. [74]

However, the General Court pointed out that although in certain cases [such as Commission v France], it was held that a benefit granted to exported products was State aid, crucially that finding was not in connection to the concept of selectivity but in connection to the affectation of trade. The distinction between domestic and foreign products is not taken into account in order to determine whether a measure is selective. [78] Indeed, this is true because all general measures in any given country still differ from similar measures in other countries. The Court added that it is clear that the finding of the selectivity of a measure is based on a difference in treatment between categories of companies within a single Member State and not a difference in treatment between companies in different Member States. [79]

Then it sought to reinforce the point it made in paragraph 78 by explaining the reasons why aid to the whole of exporting activities was found to be State aid. This was because aid was granted only to domestic products that were exported or because aid was in the form of repayment of interest only on export credits or because aid was in the form of deduction of tax only for companies with export activities. In those cases, it was possible to identify exporting companies as distinct from other companies not involved in exports. [82-85]

After this exhaustive analysis of the Spanish measure and the repeated conclusion that it did not pre-define a specific group of undertakings, it was inevitable that the General Court would annul the contested Commission Decision for failure to prove that the measure was selective.


The reasoning of the General Court may be surprising but it is sound. It can be summarised in a single sentence. Tax measures are selective when they constitute an exception or deviation from the normal tax system and, at the same time, the exception or deviation is limited only to certain undertakings. In other words, the selectivity of tax measures has two sides: a formal side [the exception] and a substantial side [benefits limited to a pre-defined group of undertakings].


[1] The text of the judgment can be accessed here:

[2] The text of the Decision can be accessed here:


[Photo by Mike Mozart from]


Click here to comment upon this article

State Aid Uncovered
Opinions by Phedon Nicolaides

Share this now

Follow by Email


Phedon Nicolaides

Professor at the College of Europe, Bruges, and at the University of Maastricht, and Academic Director at lexxion training

Search this Blog

Upcoming lexxion trainings with Phedon Nicolaides:

Inhouse Training on State Aid
Choose date & town

Recent Comments