1

The applicant, Österreichische Post AG, is a public limited company governed by Austrian law 52.8% of the capital of which is owned by Österreichische Industrieholding AG, which in turn is wholly owned by the Republic of Austria. It provides full postal services and related services, in particular, in the Republic of Austria. Under Austrian law, it has been designated as a universal service provider in Austria.

2

By letter of 30 September 2013, the applicant sent the European Commission a request pursuant to Article 30(5) of Directive 2004/17/EC of the European Parliament and of the Council of 31 March 2004 coordinating the procurement procedures of entities operating in the water, energy, transport and postal services sectors (OJ 2004 L 134, p. 1), which it accompanied with various experts’ reports. That request concerned certain postal services as well as certain other services provided by the applicant in the territory of Austria. The purpose of the request was to obtain a declaration from the Commission that, because the services in question provided in Austria were directly exposed to competition in markets to which access was not restricted, contracts intended to enable those services to be provided were not subject to the procurement procedures in the postal services sector provided for in Directive 2004/17.

3

The services to which the applicant’s request related were the following:

4

By letters of 18 October and 5 December 2013, the Commission informed the Republic of Austria about the request and called on the Austrian authorities to communicate all the relevant facts to it. The Austrian authorities replied by letter of 17 December 2013.

5

On 20 November 2013, by a Notice concerning a request in accordance with Article 30 of Directive 2004/17 (OJ 2013 C 339, p. 8), the Commission extended to 2 April 2014 the deadline for taking a decision on the applicant’s request.

6

After several letters had been exchanged and various meetings held between the Commission and the applicant, on 2 April 2014 the Commission adopted Implementing Decision 2014/184/EU exempting certain services in the postal sector in Austria from the application of Directive 2004/17 (OJ 2014 L 101, p. 4, ‘the contested decision’), addressed to the Republic of Austrian, in which it partly granted the applicant’s request.

7

Article 1 of the contested decision states that Directive 2004/17 does not apply to contracts awarded by contracting entities and intended to enable the following services to be carried out in Austria:

8

In so far as concerns the other services to which the applicant’s request related, mentioned in paragraph 3 above, the Commission stated, in recital 102 of the contested decision, that the condition of direct exposure to competition laid down in Article 30(1) of Directive 2004/17 had not been met in the territory of Austria. Those services therefore remained subject to the provisions of Directive 2004/17.

9

By application lodged at the Registry of the General Court on 24 June 2014, the applicant brought the present action for the partial annulment of the contested decision.

10

By letter registered at the Court Registry the same day, the applicant requested that the confidentiality of any documents sent to the Court which contained business secrets be preserved vis-à-vis the public.

11

On hearing the report of the Judge-Rapporteur, the Court (Fifth Chamber) decided to open the oral part of the procedure.

12

By way of measures of organisation of the procedure provided for in Article 89 of its Rules of Procedure, the General Court invited the applicant to clarify what information it regarded as business secrets and called on the Commission to produce a document. The parties complied with those requests within the time allowed.

13

The parties presented oral argument and answered the questions put to them by the Court at the hearing on 29 October 2015.

14

The applicant claims that the Court should:

15

The Commission contends that the Court should:

16

In support of its action, the applicant puts forward seven pleas in law. It argues, in substance, that the Commission applied Directive 2004/17 incorrectly in that it failed to conclude that the conditions for the application of Article 30(1) of that directive were fulfilled. According to the applicant, since the postal services in question were directly exposed to competition in markets to which access was not restricted, the Commission erred in concluding that contracts intended to enable those services to be provided continued to be subject to Directive 2004/17.

17

The first plea in law alleges incorrect application of the criteria and methods for defining the market laid down in Directive 2004/17 and an inadequate statement of reasons relating to the method chosen by the Commission. The second to sixth pleas in law allege incorrect application of Directive 2004/17 and an inadequate statement of reasons. Those pleas concern the question of whether the applicant was directly exposed to competition on the market for postal services for addressed letters B2X on a national level (second plea), on the market for postal services for addressed letters C2X on a national level (third plea), on the market for postal services for addressed international letters B2X and C2X (fourth plea), on the market for postal services for addressed advertising letters on a national and international level (fifth plea) and on the market for postal services for unaddressed advertising letters on a national and international level (sixth plea). The seventh plea alleges an incorrect statement of reasons and breach of the duty to state reasons in relation to the market for postal services for the standard delivery of addressed and unaddressed newspapers.

18

The applicant argues that the Commission breached its duty to state reasons in so far as concerns the method which it chose and infringed Directive 2004/17 by applying the wrong criteria and methods in order to conclude that the postal services in question were not directly exposed to competition.

19

First of all, in the context of its argument that the Commission breached its duty to state reasons, the applicant observes that the Commission merely stated, in recital 7 of the contested decision, that the criteria and methodology used to assess direct exposure to competition under Article 30 of Directive 2004/17 were not necessarily identical to those used to perform an assessment under Articles 101 TFEU or 102 TFEU or Council Regulation (EC) No 139/2004 of 20 January 2004 on the control of concentrations between undertakings (OJ 2004 L 24, p. 1). However, the Commission failed to state the reasons for which it chose criteria and methods other than those which must be used to perform an assessment under Article 101 TFEU or 102 TFEU and to explain what the criteria and methods which it applied in adopting the contested decision actually were.

20

It must be recalled that, according to settled case-law, the statement of reasons required by Article 296 TFEU must be appropriate to the act at issue and must disclose in a clear and unequivocal fashion the reasoning followed by the institution which adopted the measure in question in such a way as to enable the persons concerned to ascertain the reasons for the measure and to enable the competent Court to exercise its power of review. The requirements to be satisfied by the statement of reasons depend on the circumstances of each case, in particular the content of the measure in question, the nature of the reasons given and the interest which the addressees of the measure, or other parties to whom it is of direct and individual concern, may have in obtaining explanations. It is not necessary for the reasoning to go into all the relevant facts and points of law, since the question whether the statement of reasons meets the requirements of Article 296 TFEU must be assessed with regard not only to its wording but also to its context and to all the legal rules governing the matter in question (see the judgment of 2 April 1998 in Commission v Sytraval and Brink’s France, C‑367/95 P, ECR, EU:C:1998:154, paragraph 63 and the case-law cited).

21

It is true that, in recital 7 of the contested decision, the Commission stated that the decision was without prejudice to the application of the rules on competition and that, in particular, the criteria and the methodology used to assess direct exposure to competition under Article 30 of Directive 2004/17 were not necessarily identical to those used to perform an assessment under Article 101 TFEU or 102 TFEU or Regulation No 139/2004.

22

However, it must be observed that, in recital 8 of the contested decision, the Commission stated that the aim of the decision was to establish whether the services to which the applicant’s request related were exposed to a level of competition in markets to which access was not restricted, within the meaning of Article 30 of Directive 2004/17, that would ensure that, even in the absence of the discipline brought about by the detailed procurement rules set out in that directive, procurement for the pursuit of the activities concerned would be carried out in a transparent, non-discriminatory manner based on criteria that enabled purchasers to identify the solution which was the most economically advantageous overall. By referring to Directive 2004/17 in this way, the Commission gave, in the contested decision, a sufficient statement of the reasons for which it considered that, for the application of that directive, it was not enough simply to borrow the criteria and methods employed in EU competition law.

23

In so far as concerns the criteria and methods actually applied, it must be observed that the Commission referred, in particular, in recital 3 of the contested decision, to Article 30 of Directive 2004/17. Paragraph 2 of Article 30 provides that the question of whether an activity is directly exposed to competition is to be decided on the basis of criteria that are in conformity with the FEU Treaty provisions on competition, such as the characteristics of the goods or services concerned, the existence of alternative goods or services, the prices and the actual or potential presence of more than one supplier of the goods or services in question. It may also be observed that those criteria are addressed in Commission Decision 2005/15/EC of 7 January 2005 on the detailed rules for the application of the procedure provided for in Article 30 of Directive 2004/17 (OJ 2005 L 7, p. 7).

24

By setting out those considerations, the Commission gave sufficient reasons in law for its choice of method. Consequently, the applicant’s argument in relation to an alleged breach of the duty to state reasons must be rejected.

25

Secondly, the applicant argues that the Commission infringed Directive 2004/17 in that it failed to apply the criteria and methods laid down in the FEU Treaty provisions on competition. According to the applicant, the approach taken by the Commission was in contradiction to Article 30 of the directive, Decision 2005/15 and the Commission Notice on the definition of relevant market for the purposes of Community competition law (OJ 1997 C 372, p. 5, ‘the Notice on market definition’). The studies, surveys and tests relating to the definition of the relevant market which the applicant carried out were, it alleges, based on both that notice and Decision 2005/15 and represented, according to case-law, legal theory and the Commission’s usual practice, recognised methods for establishing the substitutability of products and thus for defining the market for the products in question. Nevertheless, the Commission did not apply the criteria and methods which the applicant used. Nor did it examine the evidence furnished by the applicant sufficiently or put forward counter evidence, in accordance with the Commission document entitled ‘Best practices for the submission of economic evidence and data collection in cases concerning the application of Articles 101 and 102 TFEU and in merger cases’.

26

It has already been observed (see paragraph 21 above) that, according to recital 7 of the contested decision, the Commission considered that the criteria and methodology used to assess direct exposure to competition under Article 30 of Directive 2004/17 were not necessarily identical to those used to perform an assessment under Article 101 TFEU or 102 TFEU or Regulation No 139/2004.

27

That approach is not vitiated by any error of law.

28

First of all, whilst it is true that, in accordance with Article 30(2) of Directive 2004/17, the question of whether an activity is directly exposed to competition is to be decided on the basis of criteria that are in conformity with the FEU Treaty provisions on competition, the wording of that provision does not require that those criteria should be exactly the same as those referred to in the provisions of EU competition law. As the Commission asserts, Directive 2004/17 does not form a part of EU competition law; its legal basis is Article 47(2) EC and Articles 55 EC and 95 EC. The principle objective of the EU rules on public procurement is the free movement of services and openness to undistorted competition in all the Member States (see, to that effect, the judgment of 11 January 2005 in Stadt Halle and RPL Lochau, C‑26/03, ECR, EU:C:2005:5, paragraph 44), as indeed is clear from recitals 2, 3 and 9 of Directive 2004/17. It must be borne in mind in this connection that Directive 2004/17 is intended to open up the contracts to which it applies to EU competition by promoting the widest possible expression of interest among contractors in the Member States (see the judgment of 23 April 2009 in Commission v Belgium, C‑287/07, EU:C:2009:245, paragraph 103 and the case-law cited). In recital 8 of the contested decision, the Commission rightly referred to the aim pursued by Directive 2004/17 in order to determine what level of effective competition must be attained to support a finding, under Article 30 of the directive, that a given activity is directly exposed to competition. Furthermore, there is no reference in Articles 101 TFEU and 102 TFEU or in Regulation No 139/2004 to the concept of activities that are ‘directly exposed to competition’ which appears in Article 30(2) of Directive 2004/17.

29

Secondly, in so far as concerns Decision 2005/15, it must be recalled that, according to recital 2 of that decision, the conditions laid down in Article 30(1) of Directive 2004/17 are to be evaluated solely for the purposes of that directive and must be without prejudice to the application of the rules on competition. Moreover, it is clear from recital 40 of the directive that Article 30 thereof should provide legal certainty for the entities concerned, as well as an appropriate decision-making process, ensuring, within short time limits, the uniform application of EU law in this area. This confirms that, in examining a request made pursuant to Article 30, the Commission is not required to apply the criteria and methods laid down in the FEU Treaty provisions on competition as such.

30

In so far as the applicant argues that, in section 3 of Annex I to Decision 2005/15, the Commission confirmed that the criteria of the provisions of competition law applied, it must be observed that the first sentence of section 3 of Annex I to Decision 2005/15, on the definition of the relevant product market, is admittedly identical to the definition of the relevant product market given in point 7 of the Notice on market definition, which concerns EU competition law. However, Decision 2005/15 does not mention that notice and, under Article 30(2) of Directive 2004/17, the criteria for determining whether an activity is directly exposed to competition must be in conformity with the Treaty provisions on competition. Moreover, the Commission does not dispute that the references to EU competition law play an essential role in the context of the procedure provided for in Article 30 of Directive 2004/17, as is clear, inter alia, from recital 17 of the contested decision, in which the Commission expressly referred to the EU rules relating to competition with reference to its analysis of the substitutability of the services.

31

Thirdly, as regards the Notice on market definition and the Commission document entitled ‘Best practices for the submission of economic evidence and data collection in cases concerning the application of Articles 101 and 102 TFEU and in merger cases’, it must be observed that those documents refer exclusively to EU competition law. Nowhere do they indicate that the Commission is required to apply the criteria and methodology mentioned in them in the context of procedures involving public procurement law.

32

Fourthly, the applicant argues that the Commission’s decision-making practice gave rise to a legitimate expectation on its part that the market would be defined in accordance with competition law methodologies or the Notice on market definition. It refers in this connection to Commission Decision 2007/169/EC of 16 March 2007 establishing that Article 30(1) of Directive 2004/17 applies to certain courier and parcel services in Denmark (OJ 2007 L 78, p. 28), Commission Implementing Decision 2011/875/EU of 16 December 2011 exempting certain financial services in the postal sector in Hungary from the application of Directive 2004/17 (OJ 2011 L 343, p. 77) and Commission Implementing Decision 2014/299/EU of 22 May 2014 exempting certain services in the postal sector in Hungary from the application of Directive 2014/25/EU of the European Parliament and of the Council on procurement by entities operating in the water, energy, transport and postal services sectors and repealing Directive 2004/17 (OJ 2014 L 156, p. 10).

33

That argument cannot be accepted. First of all, recital 6 of Decision 2007/169, recital 10 of Implementing Decision 2011/875 and recital 6 of Implementing Decision 2014/299 emphasise that the assessment of whether or not the services in question were directly exposed to competition was carried out solely with reference to the application of Directive 2004/17 and was wholly without prejudice to the application of the rules on competition. Moreover, the legal basis of Implementing Decision 2014/299 was not Directive 2004/17 but Directive 2014/25/EU of the European Parliament and of the Council of 26 February 2014 on procurement by entities operating in the water, energy, transport and postal services sectors and repealing Directive 2004/17 (OJ 2014 L 94, p. 243). Furthermore, the Commission asserts that, in the contested decision, it correctly applied the method for defining the market as laid down in the Notice on market definition. On this point, it is clear from recital 17 of the contested decision that, for the purposes of its analysis of the substitutability of the services in question, the Commission referred to the criteria given in the first sentence of section 3 of Annex I to Decision 2005/15, which corresponds to the definition of the relevant product market given in point 7 of the Notice on market definition. To the extent that the applicant argues that the Commission failed to apply that notice, its argument is therefore ineffective.

34

Consequently, the first plea in law must be rejected.

35

In so far as the applicant puts forward, in the context of the present plea, the general argument that the Commission wrongly rejected the studies, surveys and tests relating to the definition of the relevant market which it had carried out in accordance with the Notice on market definition, it must be observed that this argument relates not to the Commission’s overall approach to the criteria and methods which it applied but to its assessment of the question of whether the various postal services at issue were directly exposed to competition. This argument will therefore be considered in the context of the second to seventh pleas in law, which concern the Commission’s findings as regards the direct exposure to competition of those postal services.

36

The applicant argues that the Commission infringed Article 30 of Directive 2004/17 and vitiated its decision with an inadequate statement of reasons in that it found, in recitals 14 to 33 of the contested decision, that postal services for addressed letters B2X at national level were not directly exposed to competition in Austria. According to the applicant, the Commission erred in its definition of the relevant market. More precisely, the Commission erred in taking the view that electronic delivery and postal delivery did not belong to the same domestic B2X market.

37

It is clear from recitals 14 to 33 of the contested decision that the Commission concluded that postal services for addressed letters B2X on a national level were not directly exposed to competition in Austria and that, consequently, Article 30(1) of Directive 2004/17 did not apply to contracts intended to enable the pursuit of those activities in Austria. In reaching that conclusion, the Commission noted, in recitals 14 to 30 of the contested decision, that, since the applicant’s assertion that electronic delivery and postal delivery belonged to the same relevant market could not be accepted, the relevant product market was the market for postal services for addressed letters B2X. Further on, in recitals 31 and 32 of the contested decision, it found that the applicant had a very strong position in that market, with an estimated market share of [confidential]%, ( 1 ) that the postal market had been fully liberalised since January 2011 and that that liberalisation had resulted in competitors gaining an aggregate estimated market share of just [confidential]%, even in the most readily competitive segment, that is to say, B2X addressed letter services.

38

It should be borne in mind at the outset that it has consistently been held that, inasmuch as it involves complex economic assessments on the part of the Commission, the definition of the relevant market is subject to only limited review by the Courts of the European Union (see, by analogy, the judgments of 17 September 2007 in Microsoft v Commission, T‑201/04, ECR, EU:T:2007:289, paragraph 482, of 15 December 2010 in CEAHR v Commission, T‑427/08, ECR, EU:T:2010:517, paragraph 66, and of 24 May 2012 in MasterCard and Others v Commission, T‑111/08, ECR, EU:T:2012:260, paragraph 169).

39

However, this does not prevent the European Union judicature from examining the Commission’s assessment of economic data: it is required to decide whether the Commission has based its assessment on evidence that is accurate, reliable and consistent, that contains all the relevant information that must be taken into consideration in order to assess a complex situation and that is capable of substantiating the conclusions drawn from it (see, by analogy, the judgments of 8 December 2011 in Chalkor v Commission, C‑386/10 P, ECR, EU:C:2011:815, paragraph 54, in Microsoft v Commission, cited in paragraph 38 above, EU:T:2007:289, paragraph 482, and of 9 September 2009 in Clearstream v Commission, T‑301/04, ECR, EU:T:2009:317, paragraph 47).

40

It must be observed that, in accordance with Article 6 thereof, Directive 2004/17 applies generally to activities relating to the provision of postal services. The procedure laid down in Article 30 of Directive 2004/17 provides for derogation from that rule where it is found, at the request of a Member State or a contracting entity, in accordance with paragraphs 4 and 5 of Article 30, that a given market for the provision of a service referred to in Article 3 to 7 of the directive is not subject to the directive. Where, as in the present case, a contracting entity has requested the Commission to establish that Article 30(1) of Directive 2004/17 applies to a given activity, it is incumbent on the Member State concerned, in accordance with the second subparagraph of Article 30(5) of the directive, to inform the Commission of all relevant facts, and in particular of any law, regulation, administrative provision or agreement concerning compliance with the conditions set out in Article 30(1) of the directive, where appropriate together with the position adopted by an independent national authority that is competent in the activity concerned. In accordance with the first subparagraph of Article 30(6) of the directive, for the adoption of a decision under Article 30, the Commission is allowed a period of three months commencing on the first working day following the date on which it receives notification of the request. That period may be extended once only, by a maximum of three months, in duly justified cases. It is clear from the fourth subparagraph of Article 30(5) of Directive 2004/17 that, if at the end of the period laid down in Article 30(6), the Commission has not adopted a decision concerning the applicability of Article 30(1) to a given activity, Article 30(1) is deemed to apply.

41

It follows from that procedure that the burden of proving that the conditions laid down in Article 30(1) of Directive 2004/17 are fulfilled rests on the applicant and the Member State concerned, inasmuch as, in this instance, the Commission has only limited powers in comparison with the very extensive investigatory powers that are conferred on it in the sphere of EU competition law by Council Regulation (EC) No 1/2003 of 16 December on the implementation of the rules on competition laid down in Articles [101 TFEU] and [102 TFEU] (OJ 2003 L 1, p. 1) and by Regulation No 139/2004. Moreover, the Commission is required to adopt a final decision within the period specified in Article 30(6) of the directive if it considers that the conditions laid down in Article 30(1) of the directive are not fulfilled.

42

The present plea comprises four parts. The first alleges that the Commission failed to consider the applicant’s arguments adequately and breached its duty to state reasons. The second to fourth parts concern certain errors of assessment which the Commission allegedly made in the contested decision. The second part alleges an error of assessment, in recitals 18 and 19 of the contested decision, of supposed hindrances to the substitutability of electronic and postal invoicing and of the situation in the market. By the third part of the plea, the applicant alleges that, in recitals 20 and 21 of the contested decision, the Commission misinterpreted the hypothetical monopolist test analysis (‘the HM test’) performed by the applicant and the data provided by it in relation to the evolution of quantities and prices resulting from the shock analysis, and breached its duty to state reasons. Lastly, the fourth part of the plea alleges an inadequate statement of reasons for the finding, in recital 24 of the contested decision, that the applicant would be able to pass cost increases on to its customers.

43

In order to substantiate its arguments and to explain, first, the study by the consultancy firm E., entitled ‘Austrian Communications Market’, of September 2013 (‘the ACM Study’), annexed to the applicant’s request for the application of Article 30 of Directive 2004/17, secondly, the shock analysis which it carried out and, thirdly, the regression analyses carried out by the Commission, the applicant requests that the author of the ACM Study and the shock analysis be heard as a witness.

44

It must be recalled, in this connection that, in the context of an action for annulment, the function of the European Union judicature is solely to determine whether the contested measure is vitiated by one of the grounds of illegality set out in Article 263 TFEU; it cannot substitute its own assessment of scientific or technical facts for that of the EU authorities (see, to that effect, the judgment of 29 June 2000 in DSG v Commission, T‑234/95, ECR, EU:T:2000:174, paragraphs 146 and 168 and the case-law cited). It is also necessary to bear in mind that, under Article 92(1) of the Rules of Procedure, the General Court is to prescribe the measures of inquiry that it considers appropriate. According to settled case-law, it is the sole judge of any need for the information available to it concerning the cases before it to be supplemented (see the judgment of 7 October 2004 in Mag Instrument v OHIM, C‑136/02 P, ECR, EU:C:2004:592, paragraph 76 and the case-law cited). It therefore falls to the General Court to assess the relevance of the request to the subject matter of the dispute and the need to examine the witnesses named (Order of 27 April 2006 in L v Commission, C‑230/05 P, ECR-SC, EU:C:2006:270, paragraph 47). Since the points to be clarified by hearing the witness relate to various different parts of the present plea in law, it is appropriate to determine whether it is necessary to hear the author of the ACM Study and the shock analysis after considering the arguments put forward by the applicant in the context of each of those parts.

45

The applicant argues that the Commission failed adequately to examine the studies and expert’s reports which it produced or the general trends and market definition which they indicate, namely the fact that the modes of postal delivery and electronic delivery belong to the same relevant market. By so doing, the Commission also breached its duty to state reasons. The Commission did no more than state, in recital 15 of the contested decision, that the assertion that the modes of postal delivery and electronic delivery belonged to the same relevant market was not in line with the assessments that it had made in previous decisions. More precisely, the applicant emphasises that the fact that the mode of delivery of business mail is technologically neutral has been taken into account by the EU legislature and consequently also by the Austrian legislature, in particular in the field of electronic invoicing, inasmuch as invoices delivered electronically are treated in the same way as invoices delivered by post. In a study carried out in 2008, Commission experts themselves noted that 57% of undertakings questioned stated that they sent invoices electronically. The great majority of those electronic invoices are sent as pdf attachments to emails. Moreover, various studies have demonstrated the close correlation between the significant development of broadband internet services and the replacement of postal communications with electronic communications. The high level of connection to broadband internet services in Austria thus also offers the technical means for substituting electronic communications for postal communications, in particular for the bulk of transactional mail, that is to say, invoices. That substitutability is also demonstrated by the correlation between the decline in postal items and the increase in electronic communications worldwide and in Austria, and also by the studies and the HM test carried out by the applicant. A number of studies have reached the same conclusion.

46

First of all, as regards the alleged breach of the duty to state reasons, it is necessary to bear in mind, in addition to the requirements to be satisfied by the statement of reasons mentioned in paragraph 20 above, the fact that, while the Commission is not required to discuss all the issues of fact and law and the considerations which have led it to reach a decision on the definition of the relevant market, it is nonetheless required under Article 296 TFEU to set out at least the facts and considerations having decisive importance in the context of its decision (see, to that effect, the judgment of 15 September 1998 in European Night Services and Others v Commission, T‑374/94, T‑375/94, T‑384/94 and T‑388/94, ECR, EU:T:1998:198, paragraph 95 and the case-law cited). The institution which adopted a measure is not required to define its position on matters which are plainly of secondary importance or to anticipate potential objections (see, to that effect, the judgment of 25 October 2005 in Germany and Denmark v Commission, C‑465/02 and C‑466/02, ECR, EU:C:2005:636, paragraph 106 and the case-law cited).

47

In this case, the applicant’s arguments do not demonstrate that the Commission breached its duty to state reasons. Admittedly, the Commission stated, in recital 15 of the contested decision, that the applicant’s assertion that the mode of delivery of business mail was technologically neutral — meaning that the modes of electronic delivery and postal delivery belonged to the same market — was not in line with the assessments that it had made in previous decisions. Nevertheless, it made other observations in recitals 17 to 26 of the contested decision.

48

In recital 17 of the contested decision, the Commission stated that, in accordance with EU competition rules, substitutability must be analysed, inter alia, on the basis of product characteristics, prices and intended use. In recital 18 of the contested decision, the Commission observed, in substance, that the characteristics and intended use of paper invoices and electronic invoices differed significantly in that, first of all, in order to be able to send or receive electronic invoices, some additional infrastructure might be required and, secondly, the use of electronic invoicing could be associated with a range of value added services and other benefits. In this connection, it stated in recital 19 of the contested decision that there were situations in which electronic invoices had become mandatory de jure and de facto and that in such cases the substitutability issue was not relevant, as senders did not have the option of choosing postal delivery. In recitals 20 to 22 of the contested decision the Commission disputed the results of the HM test performed by the applicant and the graphs illustrating changes in quantities and prices provided by the applicant and concluded that there was no robust and conclusive evidence that electronic and postal delivery were actual substitutes. According to recital 23 of the contested decision, customers, both business and private customers, who could not or who were not willing to switch to electronic communication would remain as a segmented, captured market that would presumably only be served by the applicant, which held a virtual monopoly in that market. In recital 24 of the contested decision, the Commission stated that the applicant would be able to pass cost increases on to customers who, given their intrinsic preference for postal delivery, would have no choice but to accept the price increase. According to recital 26 of the contested decision, although one of the main consequences of the increasing use of electronic means of communication was a reduction in the overall size of the letters market, it could not be concluded that electronic communication had introduced direct competition within the market for postal delivery.

49

Contrary to the applicant’s allegation, the Commission did not therefore confine itself to stating that the assertion that the modes of postal delivery and electronic delivery belonged to the same relevant market was not in line with the assessments that it had made in previous decisions. On the contrary, it gave in the contested decision an adequate statement of the reasons for which it considered that the relevant product market was the market for postal services for addressed letters B2X alone. It cannot be concluded that that statement of reasons does not enable the persons concerned to ascertain the reasons for the measure adopted or the competent Court to exercise its power of review.

50

It may also be observed that the degree of precision of the statement of the reasons for a decision must be weighed against practical realities and the time and technical facilities available for making the decision (see the judgment of 14 February 1990 in Delacre and Others v Commission, C‑350/88, ECR, EU:C:1990:71, paragraph 16 and the case-law cited). In the present case, it is clear from recital 16 of the contested decision that, having due regard to the legal and factual situation in Austria, the Austrian authorities were invited to express their views with regard to the substitution of electronic delivery for postal delivery and, more precisely, on the definition of the relevant product market, but were unable to provide additional information to support the applicant’s assertions. Moreover, the Commission had only limited powers and a short period of time in which to take a final decision (see paragraphs 40 and 41 above).

51

Secondly, the applicant argues that the Commission failed adequately to consider its arguments or the studies and expert’s reports which it produced in order to define the market.

52

In this connection, the applicant asserts, first of all, that the Commission failed to take into consideration the fact that, from a regulatory point of view, postal invoices and electronic invoices are treated in the same way. According to the applicant, the fact that the mode of delivery of business mail is technologically neutral has been taken into account by the EU legislature and consequently also by the Austrian legislature, in particular in the field of electronic invoicing, inasmuch as invoices delivered electronically are treated in the same way as invoices delivered by post.

53

Admittedly, Council Directive 2010/45/EU of 13 July 2010 amending Directive 2006/112/EC on the common system of value added tax as regards the rules on invoicing (OJ 2010 L 189, p. 1) introduced rules under which, as is clear from recital 8 of Directive 2010/45, paper invoices and electronic invoices are, in so far as concerns value added tax (VAT) requirements, to be treated equally.

54

Nevertheless, that does not demonstrate that the Commission made a manifest error of assessment by taking the view that the modes of postal and electronic delivery were not interchangeable. Indeed, the applicant does not dispute the fact that, as recital 19 of the contested decision indicates, as of January 2014 electronic invoicing became mandatory de jure at federal level in Austria in business to government relations. In that case, the substitutability issue was therefore not relevant, as the Commission rightly noted in recital 19 of the contested decision. Moreover, from a regulatory point of view, postal services, and not electronic services, are recognised as being services of general economic interest, within the meaning of Article 14 TFEU, as is clear from recital 3 of Directive 2008/6/EC of the European Parliament and of the Council of 20 February 2008 amending Directive 97/67/EC with regard to the full accomplishment of the internal market of Community postal services (OJ 2008 L 52, p. 3). As regards the applicant’s argument that the various methods of delivering letters are discussed exhaustively in Directive 2008/6, it must be observed that the applicant refers solely to recitals 14, 15, 19 and 22 of that directive, which mention modes of electronic delivery only very generally. In addition, under Article 3 of Directive 97/67/EC of the European Parliament and of the Council of 15 December 1997 on common rules for the development of the internal market of Community postal services and the improvement of quality of service, (OJ 1998 L 15, p. 14), it is the provision of postal services that is classified as a universal service. Contrary to the applicant’s allegation, it cannot therefore be concluded that, as a general rule, paper invoices and electronic invoices are treated in the same way from a regulatory point of view.

55

Secondly, the applicant argues that the Commission did not take sufficient account of the fact that the substitutability of electronic communications for postal communications, which has been made technically possible in Austria by the high level of connection to broadband internet services, has been demonstrated by macroeconomic data and trends which show that there is a correlation between the decline in postal items and the increase in electronic communications. According to the applicant, point 2.3.2 of the ACM Study shows that [confidential]% of media undertakings send electronic invoices and that [confidential]% of those undertakings receive electronic invoices. The changes in market conditions associated with electronic substitution give rise to sufficient competition, to which the applicant is exposed.

56

In this connection, it must be pointed out that section 3 of Annex I to Decision 2005/15 contains rules for defining the relevant product market. According to those rules, a relevant product market comprises all those products or services which are regarded as interchangeable or substitutable by the consumer, by reason of their characteristics, their prices and their intended use. Section 3 of Annex I to Decision 2005/15 also contains a non-exhaustive list of the factors which are normally considered to be relevant to the determination of a relevant product market and which are to be taken into account in the analysis. They are the degree of physical similarity between the relevant products or services, any differences in the end use to which the products are put, differences in price between two products, the cost of switching from one product to another in the case of two potentially competing products, established or entrenched consumer preferences for one type or category of product and product classifications (classifications maintained by trade associations, etc.).

57

Moreover, according to the case-law, the concept of the relevant market implies that there can be effective competition between the products which form part of it and this presupposes that there is a sufficient degree of interchangeability between all the products forming part of the same market in so far as a specific use of such products or services is concerned (see, by analogy, the judgments of 13 February 1979 in Hoffmann-La Roche v Commission, 85/76, ECR, EU:C:1979:36, paragraph 28, and of 30 January 2007 in France Télécom v Commission, T‑340/03, ECR, EU:T:2007:22, paragraph 80). Interchangeability or substitutability is not assessed solely in relation to the objective characteristics of the products or services at issue. The competitive conditions and the structure of supply and demand on the market must also be taken into consideration (see, by analogy, the judgment in CEAHR v Commission, cited in paragraph 38 above, EU:T:2010:517, paragraph 67 and the case-law cited).

58

Even supposing that there is, as the applicant alleges, a correlation between the decline in postal items, on the one hand, and the increase in electronic communications and a high level of connection to broadband internet services in Austria, on the other, those facts in no way demonstrate that, in the light of the criteria laid down in Decision 2005/15 and in the case-law cited in paragraph 57 above, the Commission made a manifest error of assessment by concluding that electronic services were not substitutable for the postal services in question. Indeed, it is clear from recital 17 of the contested decision that the Commission analysed substitutability, in accordance with EU competition rules, inter alia, on the basis of the product characteristics, their prices and their intended use. In recital 18 of the contested decision, the Commission observed, in response to the applicant’s assertion that broadband and internet penetration was very high in Austria, that it appeared that the characteristics and intended use of paper invoices and electronic invoices differed significantly. Those findings are not called into question by any possible correlation between the decline in postal items, on the one hand, and the increase in electronic communications and a high level of connection to broadband internet services, on the other.

59

As regards the argument that point 2.3.2 of the ACM Study shows that [confidential]% of media undertakings send electronic invoices and that [confidential]% of those undertakings receive electronic invoices, it must be observed that, according to point 1.2 of that study, only [confidential]% of all the operators questioned regarded electronic mail as a substitute for letters, while [confidential]% regarded letters and electronic mail both as substitutes and as complementary and [confidential]% regarded electronic mail as a complement to letters but not as a substitute for them. The applicant’s argument cannot therefore prove that the Commission made a manifest error of assessment in its definition of the relevant market. Moreover, in so far as the applicant argues that the Commission drew no distinction between the various types of invoice, it must be observed that, in recital 18 of the contested decision, the Commission considered various modes of electronic invoice and that the applicant has not proven that, during the administrative procedure, it itself drew such a distinction in order to establish the substitutability of electronic delivery for postal delivery. The illustration which it gave in the annex to its request, showing a break-down of mail in Austria, to which the applicant refers, is not sufficient in this regard.

60

Thirdly, in so far as the applicant observes that the Commission did not take sufficient account of the results of the HM test which it commissioned, it must be observed that the Commission considered those results in recital 20 of the contested decision and that it rejected them, taking the view that, even if the results implied that both paper and electronic mail belonged to the same relevant market, there were some technical aspects of the design of the survey that cast doubt on the validity of the results. The merits of the Commission’s decision to reject those results will be examined further on, in the context of the third part of the present plea, in which the applicant alleges misinterpretation of the MH test and of the data relating to the evolution of quantities and prices resulting from the shock analysis, and breach of the duty to state reasons.

61

Fourthly, the applicant argues that studies which the Commission itself commissioned conclude that the two modes of delivery, electronic and postal, are interchangeable. According to the applicant, the study carried out by the firm WIK-Consult, entitled ‘Main developments in the postal sector (2010-2013)’, of August 2013 (‘the 2013 WIK Study’) expressly states that there is substitution of electronic delivery for physical postal delivery and that, accordingly, competition exists. The study by Nikali, entitled ‘The substitution of letter mail in targeted communication’ and the study by Copenhagen Economics, entitled ‘Main developments in the postal sector (2008-2010)’, arrive at similar conclusions.

62

In this connection, it must first be observed, that, not having produced these studies to the Court, the applicant is unable to demonstrate that the Commission made a manifest error of assessment in its definition of the relevant market. The reference made in footnotes 34 and 39 to the request to websites relating to the Nikali and Copenhagen Economics studies cannot remedy that failure to produce the documents on which the applicant relies in support of its action, in accordance with Article 43(4) of the Rules of Procedure of the General Court of 2 May 1991. Secondly, inasmuch as an extract of the 2013 WIK Study has been produced by the Commission, it must be observed that the study sees in the possibility of electronic delivery a reason for a possible decline in demand in the postal services markets. However, the study also notes that, in so far as the competitive situation on the market for postal services is concerned, the intensity of competition remains weak and traditional service providers retain a dominant position. On that basis, it cannot be concluded that the Commission was mistaken to find that the postal services in question were not directly exposed to competition.

63

Fifth, inasmuch as the applicant argues in this connection that the Commission followed a decision of the French competition authority in which Voice over Internet Protocol (VoIP) was regarded as substitutable for fixed telephony and thus as belonging to the same market, it must be observed that it has furnished no evidence to show that that decision is comparable to the present case. Moreover, as regards the applicant’s argument that, in an advisory circular on competition dating from May 2014, the Commission confirmed that electronic substitution had direct effects on the postal sector, suffice it to observe that that circular expressly states that it is not binding on the Commission. Lastly, contrary to the applicant’s allegation, the fact that the Commission’s former Postal Services Unit has been renamed the Online and Postal Services Unit in no way proves that the Commission made a manifest error of assessment in its definition of the relevant market. The fact that a department within the Commission is named the Online and Postal Services Unit is not a material element in the analysis of the relevant market capable of establishing that those services are in direct competition.

64

Sixth, in the context of its argument that the Commission failed to take account of the fact that the adaptation of the universal service to actual market conditions has been under discussion, the applicant refers to a discussion document of the German regulatory authority dating from November 2014 and a discussion paper published by the European Regulators Group for Postal Services (ERGP) in September 2014. In this regard, it must be recalled that it is settled case-law that the legality of a contested measure must be assessed on the basis of the facts and the law as they stood at the time when the measure was adopted (see the judgments of 28 March 2000 in T. Port v Commission, T‑251/97, ECR, EU:T:2000:89, paragraph 38 and the case-law cited, and of 10 April 2008 in Deutsche Telekom v Commission, T‑271/03, ECR, EU:T:2008:101, paragraph 244 and the case-law cited). In the present case, it was therefore necessary in principle to take as a basis the regulatory framework as it stood under the relevant acts at the time when the contested decision was adopted. The discussion documents cited post-date the Commission’s adoption of the contested decision on 2 April 2014 and it cannot therefore be concluded that the Commission made a manifest error of assessment in its definition of the relevant market by not taking them into account. Furthermore, the applicant has not shown that, at the time when the contested decision was adopted, there were any proposals for the future amendment of the relevant regulatory framework in the European Union or in Austria.

65

In addition, in so far as the applicant emphasises that, according to recital 40 of Directive 2004/17, the effects of current or future opening up to competition should be taken into account, it must be observed that, under Article 30(1) of that directive, relevant markets are not subject to the directive if, in the Member State in which it is performed, the service in question is directly exposed to competition on markets to which access is not restricted. The assessment of direct exposure to competition must therefore be carried out on the basis of the information available at the time of adoption of the decision bringing the procedure provided for in Article 30 of Directive 2004/17 to a close, there being no bar on making a fresh request under that procedure at a future date.

66

The first part of the plea must therefore be dismissed.

67

The applicant argues, with reference to the ACM Study, that the Commission erred in finding, in recitals 18 and 19 of the contested decision, hindrances to the substitutability of electronic and postal invoicing, and that its assessment of the situation in the market is incorrect.

68

In recital 18 of the contested decision, the Commission observed, in substance, that the characteristics and intended use of paper invoices and electronic invoices differed significantly in that, first of all, in order to be able to send or receive electronic invoices which allow for automated processing, some additional infrastructure might be required, particularly in the case of B2B electronic invoices. Such infrastructure might imply the use of an external service provider or of a special internal application, for example, the use of an electronic signature. Secondly, the use of electronic invoicing could be associated with a range of value added services and other benefits, such as automated processing and the financing of payments by third parties. In recital 19 of the contested decision, the Commission stated that, as of January 2014, electronic invoicing had become mandatory de jure at federal level in business to government relations and that in that case the substitutability issue was not relevant, as senders did not have the option of choosing postal delivery. In addition, there were other similar situations in which electronic invoices had become mandatory de facto further to a large customer’s or a large supplier’s request.

69

First of all, as regards the Commission’s findings in recital 18 of the contested decision, the applicant argues that the Commission failed to take account of the fact that the great majority of invoices sent electronically are emails with pdf attachments, which have no characteristics or uses different from those of invoices sent by post in paper format.

70

It is apparent from point 2.3.2 of the ACM Study, to which the Commission referred in footnote 14 to the contested decision, inserted in recital 18 thereof, that there exist various models for electronic invoicing which differ according to their degree of automation. According to the study, there are both semi-automated internet processes, which use pdf invoices, web portals or service providers, and end-to-end automated electronic invoicing processes that are fully integrated from order to payment. It is apparent from the study that, while semi-automated internet processes which use pdf invoices are the preferred communication format in B2C relations, web portals have already been used by a significant number of undertakings that use electronic invoicing and services providers are used, in particular, for B2B electronic invoicing at European level. The ACM Study states that certain service providers secure the authenticity of documents with electronic signatures. Moreover, according to the ACM Study, in order to receive and send electronic invoices, undertakings may use solutions offered by an external service provider or specialised internal invoice handling applications.

71

In light of those observations in the ACM Study, which the applicant sent to the Commission, it cannot be concluded that the Commission erred in concluding, in recital 18 of the contested decision, that, in order to be able to send or receive electronic invoices, some additional infrastructure might be required and that the use of electronic invoicing could be associated with a range of value added services and other benefits.

72

Secondly, as regards the statements made by the Commission in recital 19 of the contested decision, the applicant points out that the share of the market taken by invoices sent at federal level in business to government relations stands at just [confidential]% and that it can therefore be left aside for the purposes of defining the relevant market. Nor is it appropriate or consistent with previous Commission decisions for account to be taken of situations in which electronic invoices may have become mandatory de facto. Furthermore, the Commission has said nothing of what that market share might be, or of the effects that might have on the definition of the relevant market.

73

First of all, it must be observed that, even though invoices sent at federal level in business to government relations may be so few in number as to be regarded as negligible, the fact remains that the Commission could, without making any error, take note of the existence of that type of invoicing, in relation to which the issue of substitutability was not relevant.

74

Secondly, in so far as the Commission stated that there were other similar situations in which electronic invoices had become mandatory de facto further to a large customer’s or a large supplier’s request, it must be observed that, in footnote 15 to the contested decision, inserted in recital 19 thereof, the Commission referred to point 2.3.2 of the ACM Study. It is apparent from that study that the choices which undertakings make with regard to modes of invoicing depend, inter alia, on relations between businesses, experience levels across the supply network and customer demand. According to the ACM Study, the economic reasons for which electronic invoicing is used depend primarily on external market forces, that is to say, the requests of large customers or suppliers, or government initiatives which render such invoicing mandatory.

75

Having regard to that reference to the ACM Study, the de facto situation to which the Commission referred is made clear, to the requisite legal standard, in the contested decision. As regards the market share which that situation represents, it must be recalled that, given the Commission’s limited powers of investigation under Article 30 of Directive 2004/17 and the fact that the burden of proof rests on the applicant (see paragraphs 40 and 41 above), the Commission cannot be required to go beyond the observations which the applicant has made in its request. It must also be observed that it is clear from recital 16 of the contested decision that the Austrian authorities were unable to provide additional information to support the applicant’s assertions (see paragraph 50 above). In so far as concerns the implications of that de facto situation for the definition of the relevant market, it is clear from recital 30 of the contested decision, which states that the Commission defined the relevant market on the basis of the information mentioned, inter alia, in recital 19 of the decision, that this aspect was taken into account by the Commission in its definition of the relevant market. Moreover, there is nothing to suggest that the criteria laid down for the definition of the relevant market (see paragraphs 56 and 57 above) exclude the taking into account of a de facto situation. Lastly, in so far as the applicant argues that taking such a situation into account is not consistent with previous Commission decisions, beside the fact that the applicant omits to indicate which Commission decisions might support its argument, it puts forward no argument capable of demonstrating that the Commission was bound, in this case, by its decision-making practice. This line of argument cannot therefore be accepted.

76

Consequently, the second part of the plea must be rejected.

77

This part of the plea comprises two complaints. The first concerns the assessment of the HM test which the applicant commissioned in support of its request. The second complaint alleges misinterpretation of the data provided by the applicant in relation to the evolution of quantities and prices resulting from the shock analysis.

78

The applicant argues that, in recital 20 of the contested decision, the Commission misinterpreted the MH test which it had commissioned on the subject of the market for addressed letters B2X on a national level, which it alleges shows that postal delivery and electronic delivery belong to the same market, and that the Commission breached its duty to state reasons.

79

It is clear from the case-file that the research comprising the MH test, which forms part of the ACM Study, was carried out among 451 undertakings and that it included two different types of questions. First, in a series of direct questions, the undertakings were asked about their practice in dealing with business mail. The central question was whether the undertakings would send their letters by electronic means if the cost of posting a letter were to rise from 62 to 65.1 cents. The undertakings had to choose one of seven answers: no change, a complete change, or a change in respect of 10%, 20%, 30%, 50% or 75% of their letters. Secondly, indirect questions were asked that did not directly address reactions to price increases. The undertakings had to state which of two different services, defined by a set of characteristics, they would choose.

80

In recital 20 of the contested decision, the Commission stated the following:

‘The applicant performed [an HM test] using survey data collected from 451 Austrian firms. The survey included a set of indirect questions aimed at eliciting the preferences for paper and electronic mail using conjoint analysis techniques. The results would imply that a 5% increase in postage price would decrease demand for sending letters by [confidential]%. Although those results would imply that both paper and electronic mail belong to the same relevant market, there are some technical aspects of the design of the survey that cast doubts on the validity of the results. For instance, the set of characteristics to define the products is chosen in a way that could induce some bias in favour of electronic means of communication. The controversial assumptions do not seem to be used in the relevant economic literature applying the conjoint analysis technique and it is not possible to measure their impact on the estimated decrease of [confidential]%.’

81

First, the applicant asserts that the Commission breached its duty to state reasons in that it failed to set out the reasons for which, in its view, there had been some bias in favour of electronic means of communication or what assumptions were controversial in the economic literature.

82

As regards the reasons given in the contested decision on the basis of which the Commission found some bias in favour of electronic means of communication, it must be pointed out that the Commission stated in recital 20 of the contested decision that it was, for instance, the way in which the set of characteristics to define the products had been chosen which led it to make such a finding. In so far as controversial assumptions are concerned, it is clear from recital 20 of the contested decision that these were the assumptions mentioned in the survey rather than any assumptions that might be controversial in the economic literature.

83

Admittedly, the lack of precision in the contested decision of the Commission’s criticisms of the results of the HM test may appear unfortunate. However, given the context in which the contested decision was adopted, marked, in particular, by the fixed time limit for the adoption of a final decision (see paragraph 40 above), the Commission did not, in the present case, breach its duty to provide an adequate statement of reasons for the contested decision, in accordance with the requirements laid down in the case-law mentioned in paragraphs 20 and 46 above. It would be unreasonable to require a detailed account of the criticisms relating to each of the characteristics chosen in the survey in question or of each controversial assumption mentioned in it. That is all the more so since the applicant was closely involved in the administrative procedure (see, to that effect, the judgment of 10 July 2008 in Bertelsmann and Sony Corporation of America v Impala, C‑413/06 P, ECR, EU:C:2008:392, paragraphs 179 and 180 and the case-law cited). Indeed, it is clear from the case-file that, after producing, on 2 December 2013, the questionnaires on which the survey was based, the expert engaged by the applicant to conduct the survey met with Commission staff on 6 March 2014. The following day, the Commission sent a draft of the decision to the expert and then discussed it with him at a meeting held on 28 March 2014. The draft decision that was considered at the meeting of 28 March 2014, which the Commission has produced in response to the measures of organisation of the procedure ordered by the General Court (see paragraph 12 above), already contained the wording in recital 20 of the contested decision.

84

It must also be observed that both the criticisms which the Commission made of the choice of characteristics to define the products and those of the controversial assumptions relate to the part of the survey involving indirect questions. The fact that the applicant asserted in its application that, in recital 20 of the contested decision, the Commission analysed only the indirect questions, and not the direct questions, is a further indication that it was aware of the Commission’s reasoning.

85

Secondly, the applicant complains that the Commission misinterpreted the HM test and that it did not submit any expert’s reports or surveys to counter its research. According to the applicant, although the HM test includes direct and indirect questions, the Commission did not analyse the direct questions asked of customers, which play an important role. Referring to the Commission’s practice and to the Notice on market definition, the applicant asserts that the Commission usually uses marketing studies made available to it by the undertakings concerned as evidence in its market definition, but did not do in this case. The results of the customer survey in relation to one direct question show, according to the applicant, that [confidential]% of the undertakings questioned would, in the event of a slight increase in prices, totally or partly replace their postal items by electronic mail and that [confidential]% of the undertakings questioned would continue to send items by post. Those results correspond to the results of another study concerning Austria carried out in 2012. Moreover, according to the applicant’s data, even if prices remained the same, [confidential]% of the undertakings questioned intended to send their invoices electronically in the future. The Commission failed to take account of the fact that a similar inclination to replace postal services with electronic services in the event of an increase in postal costs also existed in the United Kingdom and Germany, as is apparent from data collected by the United Kingdom postal services regulatory authority and the study by the firm WIK-Consult, entitled ‘Nachfrage nach Postdienstleistungen von Geschäftskunden’ (demand for postal service from business customers) of 2009 (‘the 2009 WIK Study’). As regards the Commission’s criticisms of the indirect questions, the applicant points out that those questions were asked, using a recognised survey technique, so that the situation could be analysed further. In accordance with the Notice on market definition, all the products perceived by customers as being interchangeable were included in the customer survey and there was therefore no bias in favour of electronic means of communication.

86

It must be observed as a preliminary point (see paragraphs 38 to 41 above) that, in defining the relevant market, the Commission enjoys a broad discretion that is subject to only limited review by the Courts of the European Union and that the burden of proof as regards definition of the relevant market rests on the applicant.

87

First, it must be observed that, contrary to the applicant’s allegation, the Commission does not dispute that an HM test may be of use in analysing the substitutability of products or services. Indeed, the Commission states that it applied the method for defining the market as laid down in the Notice on market definition (see paragraph 33 above). It is common ground that the HM test is, in principle, one of the methods contemplated in the notice. According to point 15 of the notice, one way of determining the range of products that are viewed as substitutes by the consumer is to conduct a speculative experiment, postulating a small but lasting change in relative prices and evaluating the likely reactions of customers to that increase. It is apparent from point 17 of the notice that the question to be answered is whether the parties’ customers would switch to readily available substitute products or to suppliers located elsewhere in response to a small (in the range 5% to 10%) but permanent relative price increase in the products being considered in the areas concerned. If substitution were enough to make the price increase unprofitable because of the resulting loss of sales, additional substitutes and areas are included in the relevant market. In accordance with points 39 to 41 of the notice, qualitative tests consisting of various econometric and statistical approaches (such as estimates of elasticities in the demand for a product, enquiries to gather the views of customers and competitors and market research studies submitted by the parties) may be relevant elements in the assessment of the demand substitutability of two products.

88

It is important to point out in this connection that footnote 7 to Decision 2005/15 states that, when defining the relevant market, the Commission considers only products which can be readily substituted for the products in question. Readily substitutable products are those to which consumers would switch in response to a modest but significant increase in the price of the relevant product, such as 5%. According to that footnote, this enables the Commission to assess the competitive situation in the context of a relevant market comprising all the products to which consumers of the products in question would readily switch. Again according to footnote 7, this does not however mean that the Commission does not take into account the constraints on the competitive behaviour of the entities concerned resulting from the existence of imperfect substitutes, that is to say, products to which a consumer would not switch in response to a modest but significant increase in the price of the relevant product, such as 5%. Those effects are taken into account once the market has been defined and market share quantified.

89

Secondly, as regards the applicant’s argument that the Commission did not submit any expert’s reports or surveys to counter its own research, suffice it to recall that the burden of proving that the conditions laid down in Article 30(1) of Directive 2004/17 are fulfilled rests on the applicant and the Member State concerned (see paragraph 41 above). The Commission was therefore not obliged to carry out its own studies.

90

Thirdly, the applicant’s argument that the Commission failed to analyse the direct questions must also be rejected. The fact that the Commission stated, in recital 20 of the contested decision, that the survey in question included a set of indirect questions and that the only examples it expressly gave were criticisms of the design of the indirect questions in the survey does not mean that it did not analyse the directive questions in the survey. The word ‘included’ does not imply that there were no questions in the survey other than the indirect questions, and indeed the Commission expressly asked the expert engaged by the applicant to conduct the survey for the questionnaires in the survey, which contained both direct and indirect questions. Moreover, in so far as the wording of recital 20 of the contested decision contains criticism of certain technical aspects of the design of the survey as such, it cannot be concluded that those criticisms relate only to the part of the survey comprising indirect questions.

91

Fourthly, the applicant argues that the Commission incorrectly assessed the results of the central direct question in the survey whereby the undertakings questioned were asked whether they would send their letters by electronic means if the cost of posting a letter were to rise from 62 to 65.1 cents.

92

As has already been observed (see paragraph 79 above), when answering that question the undertakings had to choose one of seven answers: no change, a complete change, or a change in respect of 10%, 20%, 30%, 50% or 75% of their letters. In recital 20 of the contested decision, the Commission confined itself to stating in this connection that, although those results of the survey implied that both paper and electronic mail belonged to the same relevant market, there were some technical aspects of the design of the survey that cast doubt on the validity of the results. In its defence pleading, the Commission stated that the fact that, if they envisaged switching from postal delivery to electronic delivery, the undertakings questioned had to indicate as a minimum 10% of their volume of mail, could have resulted in an overstatement of their degree of interest in switching in the event of a price increase. The Commission went on to state, in its defence, that the survey as a whole was quite misrepresentative, in terms of the size of the undertakings questioned, of the range of undertakings in the Austrian economy.

93

In response to the Commission’s criticism that the choice of a minimum 10% of the volume of mail might have resulted in overstatement, the applicant replies that the choice of the percentage in the survey was connected with the volume of mail, which was connected with market conditions. According to the applicant, customers that choose electronic delivery do not send only one letter electronically, but a given percentage of their volume of mail. In addition, it is clear from the results of the ACM Study that [confidential]% of the undertakings questioned send between 100 and 1000 items, including invoices, monthly. Thus, 5% of that volume of mail represents only between 5 and 50 items monthly. A volume of 10% represents between 10 and 100 items, which better reflects the conditions in the market. Furthermore, the results of other HM test studies carried out in the United Kingdom and Germany indicate that massive migration might be observed in the event of a price increase. According to the applicant, the choice of a threshold of 10% was therefore a prudent estimate. The applicant asserts that, if it had stipulated 5% as the minimum reconversion threshold, the propensity to switch to electronic delivery would have been even broader. Lastly, when analysed, the results of the survey were correlated not with the volume of mail, but with the number of undertakings questioned. If they had been correlated to the volume of mail, the decline in the event of a hypothetical increase of 5% would have been even more marked.

94

That line of argument does not, however, show that the Commission’s conclusion, in recital 20 of the contested decision, that some technical aspects of the design of the survey cast doubt on the validity of the results, was incorrect.

95

Indeed, as the Commission asserts, the reference to the number of letters sent by the undertakings questioned does not support the applicant’s allegation that a minimum percentage of 10% better reflected market realities, especially since it is clear from point 3.4.1 of the ACM Study that [confidential]% of the undertakings questioned sent 1000 letters or more monthly, with [confidential]% of them sending between 1000 and 2499 letters, [confidential]% of them between 2500 and 4999 letters, [confidential]% of them between 5000 and 9999 letters and [confidential]% of them sending more than 10000 letters monthly. As regards the results of the studies carried out in the United Kingdom and Germany, as the Commission states, leaving aside the fact that those studies do not concern the Austrian market and are therefore, contrary to the applicant’s allegation, not relevant to the definition of the relevant market, it is clear from point 3.4.1 of the ACM Study that the studies addressed the question of whether the undertakings in question were entirely or only somewhat inclined to switch to electronic mail in the event of a price increase. They did not, therefore, address the scale of such a switch. Moreover, as regards the study carried out in Germany, namely the 2009 WIK Study, the results of the survey carried out in that study were interpreted as indicating that, in the event of an increase in letters charges, undertakings would be more likely to switch service provider than to increase their use of electronic mail.

96

In so far as the applicant asserts that, if it had stipulated 5% as the minimum reconversion threshold, the propensity to switch to electronic delivery would have been even broader, it must be observed that that argument has in no way been substantiated and must therefore be rejected as unfounded. It must also be pointed out that, even if this argument were to be admitted, the case-file offers no indication of what degree of propensity might have resulted. The same applies to the argument that, if the results of the survey had been correlated with the volume of mail, the decline in postal delivery would have been even more marked.

97

In addition, in response to the Commission’s complaint that the survey as a whole was, in its view, quite misrepresentative, in terms of the size of the undertakings questioned, of the range of undertakings in the Austrian economy, the applicant replies that the undertakings questioned were representative of actual market conditions and of volumes of mail in the market for letter delivery. It emphasises that the ACM Study shows that it is above all the larger undertakings that send mail, with smaller undertakings sending fewer items. According to the applicant, had the survey been conducted with account being taken of the range of sizes of undertakings in the economy, it would have distorted market realities to a greater degree. There are 5641 customers in the applicant’s portfolio of business customers. 160 of them send [confidential]% of its total annual volume of mail. Having regard to actual market conditions, it is, according to the applicant, right to question those undertakings which generate an appropriate volume of mail, and not those that have no role to play on the market. 57 of the applicant’s 160 key professional customers took part in the survey, representing an estimated [confidential]% of the total volume of mail sent by all its professional customers combined. Moreover, the applicant included in the survey a proportion, 8.4%, of smaller undertakings that send fewer than 100 items monthly, representing only [confidential]% of the volume of mail, because its database does not include postal customers that take their mail to the post office themselves. The survey therefore gave a complete picture of market conditions. The applicant also states that it may be the case that larger undertakings are more inclined to switch to electronic mail. Nevertheless, even if only a small proportion of its large professional customers were to switch to electronic mail, that could generate network infrastructure costs that could have an impact on its other customers, potentially encouraging them too to switch to electronic mail.

98

That line of argument does not show that the Commission’s conclusion, in recital 20 of the contested decision, that some technical aspects of the design of the survey cast doubt on the validity of the results, was incorrect.

99

Indeed, since the applicant held a market share of approximately [confidential]% of the market for postal services for addressed letters B2X on a national level, a point which is not in dispute, then clearly almost all Austrian undertakings were its customers. As the Commission asserts, since the applicant itself regards it as plausible that undertakings of a certain size are more inclined to switch to electronic mail, the fact that the survey concentrated on larger undertakings entails a risk that it does not reflect the preferences of small and medium-sized undertakings appropriately. As regards the applicant’s allegation that, even if only a small proportion of its large professional customers were to switch to electronic mail, that could generate network infrastructure costs that could have an impact on its other customers, it must be rejected as unsubstantiated. Indeed, the applicant itself states that its prices are strictly regulated and that, consequently, an increase in prices is only possible with the authorisation of the regulatory authority.

100

Furthermore, as the Commission asserts, even supposing the volume of mail of the undertakings questioned is the right criterion, it does not appear that the research conducted among the undertakings questioned took account of relationships of size. Indeed, it is apparent from the Austrian statistics produced by the Commission that there were approximately 311000 undertakings in Austria in 2011. As has already been observed (see paragraph 99 above), almost all of those undertakings were customers of the applicant’s. In light of the foregoing, the applicant cannot assert, without offering any explanation in this regard, that taking into account the 5641 customers in its portfolio ensured a representative sample of undertakings questioned in terms of the volume of mail deliveries. On the contrary, choosing only the undertakings in the applicant’s portfolio of professional customers is likely to accord greater weight to undertakings having a large volume of mail, inasmuch as it may be assumed that undertakings having a smaller volume of mail will be less likely to be included in that portfolio. In addition, while it is clear from point 3.4.1 of the ACM Study that [confidential]% of the undertakings questioned had a volume of mail in excess of 10000 items per month, the applicant stated, in its reply, that only [confidential]% of its 5641 business customers had such a volume. Moreover, a strong bias was shown toward the larger undertakings among the 451 undertakings questioned in the survey inasmuch as, according to the applicant, 57 of the 160 large undertakings in its customer portfolio participated in the survey, while only 394 of the remaining 5481 undertakings participated in it. As the Commission asserts, the applicant has also failed to prove its allegation that smaller undertakings sending fewer than 100 letters per month were over-represented in the survey. Indeed, although, according to the applicant, [confidential]% of the 5461 customers in its portfolio were smaller undertakings, that segment represented only [confidential]% of the undertakings questioned in the survey.

101

In so far as the applicant argues in this connection that the results of the MH test correspond to the results of another study concerning Austria carried out in 2012, the applicant has not established that fact, having failed to produce the study in question, and this argument must therefore be rejected as unfounded.

102

Fifth, the applicant argues that the research carried out shows that [confidential]% of its customers expected to replace postal delivery of their letters with electronic delivery in the future, even if prices remained the same. That argument must be rejected. Indeed, that conclusion, expressed in very general terms, is vitiated by the same technical issues as those which vitiate the results of the central direct question in the MH test.

103

Sixth, the applicant points out that the indirect questions were asked, using a recognised survey technique, so that the situation could be analysed further. In accordance with the Notice on market definition, all the products perceived by customers as being interchangeable were included in the customer survey and there was therefore no bias in favour of electronic means of communication.

104

As has already been observed (see paragraph 79 above), it is clear from the case-file that the research comprising the MH test also included indirect questions. In those questions, which did not directly address their reaction to price increases, the undertakings were required to state which of two different services, defined by a set of characteristics, they would choose. In recital 20 of the contested decision, the Commission stated in this connection that, although the results of the survey implied that both paper and electronic mail belonged to the same relevant market, there were some technical aspects of the design of the survey that cast doubt on the validity of the results. According to the Commission, the set of characteristics to define the products, for instance, had been chosen in a way that might create some bias in favour of electronic means of communication. In its defence pleading, the Commission stated that the indirect questions asked of customers were arranged as ‘packages’ involving not only a choice between electronic and postal delivery, but also subjective aspects that called for consideration to be given, for example, to environmental protection, legal requirements and security. That method of questioning entailed the risk of constraining the undertaking questioned to express a given preference. Furthermore, according to the Commission, it is obvious that an entire series of those subjective aspects was formulated in such a way as to favour electronic delivery.

105

The applicant takes issues with those criticisms and argues that numerous studies mentioned in point 2.3.2 of the ACM Study establish that, in addition to price, senders also consider other preferences when making a decision. On the basis of the results of those studies, considerations relating to the environment, security and legal requirements are empirically proven preferences. It is normal, scientifically recognised practice to include such preferences in research. According to the applicant, the various preferences of senders were taken into account in the research in a balanced manner.

106

That line of argument does not demonstrate that the Commission’s assessment of the technical aspects of the design of the survey was incorrect. It does not appear that the Commission erred in taking the view that the arrangement of the indirect questions as ‘packages’ involving not only a choice between electronic and postal delivery, but also aspects that called for subjective assessment was likely to distort the results of the survey. Indeed, as the Commission asserts, having to consider those subjective aspects might have prevented the undertakings questioned from expressing their true preference for electric delivery or postal delivery. While it is true that other studies cited in point 2.3.2 of the ACM Study mention other preferences of undertakings among the factors in favour of electronic delivery, the fact remains that the ACM Study does not show that those other preferences have already been used in the same way in the relevant economic literature applying the conjoint analysis technique, as the Commission observed in recital 20 of the contested decision. In addition, by merely stating that those various preferences of senders were taken into account in the ACM Study in a balanced manner, the applicant has in no way established that that is indeed the case.

107

The first complaint must therefore be rejected.

108

The applicant alleges that the Commission was wrong to state, in recital 21 of the contested decision, that the data provided by it in relation to the evolution of quantities and prices resulting from the shock analysis did not confirm that electronic delivery and postal delivery were real substitutes. The shock analysis is an analysis of the decline in demand which followed the ‘shock’ consisting in the applicant’s price increases of May 2011, which were authorised by the regulator. According to the applicant, the demand elasticity of [confidential]% in the event of a 5% price increase indicated by the HM Study is of the same order of magnitude as the elasticity caused by the price increases of May 2011, which was [confidential]% or [confidential]%, depending on the service chosen.

109

In recital 21 of the contested decision, the Commission stated the following:

‘The applicant also provided graphs of the evolution of quantities and prices that suggested much smaller sensitivity of quantities to prices than the one reported in the survey analysis. In order to clarify this apparent contradiction, the applicant was asked to provide the real data used to produce the graphs and to quantify the sensitivity of demand. The sensitivity of demand calculated by the applicant using the real data ranged between [confidential]% and [confidential]%, depending on the technique used. Nevertheless, the applicant did not use the standard econometric analysis to produce those results. The preliminary results of a separate analysis conducted by our services using the same data set and standard econometric techniques for demand estimation in the economics literature suggest that the sensitivity of quantities to price changes could be as low as 3.1%.’

110

Footnote 19, inserted into the fourth sentence of recital 21 of the contested decision, states:

‘In demand analysis, the price is usually regarded as an endogenous variable. That endogeneity needs to be controlled for the associated sensitivity of quantities to prices to be considered an unbiased estimation of elasticity. The applicant did not perform that type of endogeneity check and did not properly justify the exogeneity of prices, assuming without justification that prices were exogenous.’

111

More specifically, the applicant argues that the Commission also uses shock analyses as evidence when defining markets. According to the applicant, there is no apparent contradiction in so far as concerns the results of the survey conducted in the context of the ACM Study. In the shock analysis, for practical reasons and because of technical considerations relating to turnover tax, it drew a distinction between standard letters left at a large post office, which fall within the scope of the universal service obligation, and letters that are taken directly to a sorting office, which do not fall within the scope of the universal service obligation, but are subject to VAT. For the month of May 2011, it calculated a demand elasticity of [confidential]% for a price increase of 1% for standard letters falling outside the scope of the universal service obligation and of [confidential]% for letters falling within the scope of the universal service obligation. A hypothetical price increase of 5% in an HM test would result into a decrease in demand for standard letter services of [confidential]% and [confidential]% respectively. The results of the HM test, which established a demand elasticity of [confidential]%, are therefore conclusive. According to the applicant, the shock analysis was confirmed by two plausibility checks, a simple econometric analysis and a complementary econometric analysis. Lastly, the applicant argues that the econometric analyses which the Commission performed on the basis of the data which it had provided were neither appropriate nor sufficient to discredit the reliability and clarity of its own analyses. While the results of the Commission’s first analysis indicated a demand elasticity of the same order of magnitude as that measured by the applicant, that is to say [confidential]%, its second analysis was incorrect, since it used unrelated variables and data for different time periods.

112

Those arguments do not demonstrate that the Commission’s assessment is vitiated by a manifest error.

113

First of all, as regards the argument that the Commission also uses shock analyses as evidence when defining markets, it must be observed that the Commission has not disputed that such analyses may be useful when defining a relevant market. The Commission has stated that it applied the method for defining the market as laid down in the Notice on market definition (see paragraph 33 above). According to point 38 of that notice, one of the factors which the Commission regards as relevant to its assessment of the demand substitutability of two products is any evidence of substitution in the recent past. According to point 38, in certain cases it is possible to analyse evidence relating to recent past events or shocks in the market that offer actual examples of substitution between two products and, when available, that sort of information will normally be fundamental for market definition. Again, according to point 38 of the Notice on market definition, if there have been changes in relative prices in the past, the reactions in terms of quantities demanded will (all things being equal) be decisive in establishing substitutability.

114

Secondly, as regards the argument that the Commission erred in finding there to be an apparent contradiction between the results of the survey carried out by the applicant and the results of the shock analysis which it performed, inasmuch as the graphs of the evolution of quantities and prices used in the shock analysis suggested that quantities were much less sensitive to prices than had been reported in the survey analysis, that too must be rejected.

115

Indeed, as the Commission asserts, it is clear from the graphs which the applicant gave it that demand had begun to decline in April 2011, before the applicant’s price increases of May 2011. While it is true that demand continued to decline in May, the fact remains that it cannot be ruled out that the exceptional causes which resulted in the decline in April — according to the applicant, the absence of alignment among some of its large customers — also contributed to some extent to the decline in May 2011, such that the Commission was entitled to conclude that quantities were much less sensitive to prices than had been reported in the survey analysis.

116

On the other hand, in so far as the Commission argues that the applicant should have taken into account the month of May 2011 alone, and not the period from April to June 2011, because the price increases under consideration entered into effect on 1 May 2011, that argument cannot be accepted. Indeed, while it is clear from the applicant’s advertisements dating from February 2011, produced by the Commission, that the price increases of 1 May 2011 had been brought to the public’s attention well in advance, it must be observed that that fact does not call into question the applicant’s assertion that the reactions of its customers to those price increases differed.

117

Thirdly, the applicant asserts that, in order to substantiate the results of the shock analysis, it carried out two plausibility checks, that is to say, a simple econometric analysis and a complementary econometric analysis. According to the applicant, the simple economic analysis of the data, which took the form of a regression analysis, substantiated the demand elasticities calculated using the standard formula. In that analysis the seasonal effects that may be observed in the postal sector were checked using a dummy variable. That analysis showed an elasticity of [confidential]% in the presence of a 5% price increase. The outcome of the complementary econometric analysis was an elasticity of [confidential]% in the presence of a 5% price increase. For the sake of completeness, the applicant submitted to the Commission a standard theoretical appraisal of demand elasticities in the postal sector, which also confirmed the elasticities measured by the applicant in the presence of a 1% price increase. According to the applicant, the Commission criticised the complementary econometric analysis simply because it was not a standard econometric analysis. In the complementary econometric analysis, the applicant had carried out a regression analysis using the volume of letters as an explained dependent variable and the price of letter delivery as an independent explanatory variable. In such a case, there is endogeneity, since the explanatory variable exerts an influence on the explained variable. However, contrary to the Commission’s inference, the recognised problem of endogeneity is accepted in scientific studies in the postal sector. The theory is that the price of letter delivery is exogenous, as is confirmed by the analyses produced by the applicant. The complementary econometric analysis is, according to the applicant, confirmed by the fact that it produced the same results as the applicant’s other analyses, including the simple econometric analysis, which, according to the applicant, was not affected by the problem of endogeneity since it did not include price as an explanatory variable. Contrary to the Commission’s allegation, the use of a standard econometric analysis is not therefore necessary.

118

First of all, in so far as concerns the argument that the Commission criticised only the complementary econometric analysis, because it was not a standard econometric analysis, and not the simple econometric analysis, it must be observed that recital 21 of the contested decision does not suggest that the Commission’s criticism relating to the absence of a standard econometric analysis concerned only the complementary economic analysis carried out by the applicant. Indeed, when it stated in that recital that, despite the fact that the sensitivity of demand calculated by the applicant using the real data ranged between [confidential]% and [confidential]%, depending on the technique used, the applicant had nevertheless not used the standard econometric analysis to produce those results, the Commission was in no way drawing a distinction between the simple and the complementary econometric analyses carried out by the applicant. On the contrary, it is clear from footnote 19 inserted into the fourth sentence of recital 21 of the contested decision (see paragraph 110 above) that the Commission rejected both of the applicant’s econometric analyses because the applicant had omitted to take the question of the endogeneity of prices into consideration and had not sufficiently established that prices were exogenous.

119

Secondly, as regard the applicant’s argument that the simple econometric analysis was not affected by the problem of price endogeneity since it did not include price as an explanatory variable and that it is therefore a valid analysis in any event, it must be observed that, as the Commission asserts, the applicant was seeking to prove by that analysis the sensitivity of demand to prices, or the reaction in terms of quantities to the shock caused by the price increases of 1 May 2011. Indeed, that analysis shows that, according to the applicant, the econometric modelling of demand for letter post made it possible to quantify the decrease in demand attributable to the price increases by isolating that effect from other factors. That analysis also included a result relating to the fall in volume attributable to the price increases. As the applicant stated at the hearing, the data which it examined in its simple econometric analysis related precisely to the period concerned by the price increases that came into effect on 1 May 2011. Consequently, it was necessary for the applicant’s simple econometric analysis to include the ‘price’ factor.

120

Thirdly, in so far as the applicant argues that the recognised problem of endogeneity is accepted in scientific studies in the postal sector and that the theory is that the price of letter delivery is exogenous, it does no more than refer to one study that supposedly supports its view, without however producing that study. The Commission did, however, produce that study, which deals with price elasticity in the market for postal services in the United States. As the Commission asserts, since the regulatory context is entirely different, the results of the study are of no use in so far as the Austrian market is concerned. Moreover, the study shows that, even in the American postal market, the theory that price is exogenous is deemed to be incorrect.

121

Similarly, in so far as the applicant states that it submitted to the Commission a standard theoretical appraisal of demand elasticities in the postal sector which also confirmed the elasticities which it measured in the presence of a 1% price increase, and merely refers to a list of scientific studies without however producing them, the applicant has in no way established that the Commission was wrong to reject its econometric analyses. Furthermore, it must be observed that none of those scientific studies concerns the Austrian market.

122

The applicant also argues that the results of its regression analyses may be evaluated using two quality criteria. The first is the t-test. According to the applicant, if the result of the t-test, that is to say the p (probability) level, is below 0.05, or 5%, the variable is regarded as statistically significant. In the present case, all the p values of the applicant’s two analyses are below 0.05. The second quality criterion is the coefficient of determination r2. This indicates the percentage fluctuation in the explained variable that may be explained by the regression equation and thus by the explanatory variables. The closer the r2 value is to 1, the greater is the explanatory force of the influence factor under consideration. The greater the explanatory force of the influence factor, the more accurate is the regression. According to the applicant, levels above 0.2 are acceptable. Levels above 0.4 indicate good estimates and levels above 0.5 indicate very good regression estimates. The applicant states that, in this case, its two analyses produced excellent levels: its simple econometric analysis returning 0.92 and its complementary economic analysis returning 0.88.

123

It must be observed in this connection that, in response to a question put to it by the General Court at the hearing, the applicant stated that p values and r2 levels cannot resolve the issue of endogeneity, Neither the t-test nor the coefficient r2 gives any indication of the causal nexus between the various factors used in regression analyses and the quantitative decline in question, which is the central question in any analysis of demand elasticity.

124

Fourthly, the applicant argues that the results of the two econometric analyses carried out by the Commission using the data which it had provided are incorrect. According to the applicant, the Commission did not state the reasons for which it had used the variables chosen. Moreover, the Commission’s regression analyses merely covered the period up to 2012, and the Commission used different observation periods and different time periods, something which is not done in any standard econometric analysis. In so far as concerns the Commission’s first regression analysis, the applicant asserts that the price of letter delivery explains the volume of letters, as in its own complementary econometric analysis. That first regression analysis also shows that the volume of letters decreased by 10.15%, which is a comparable result to those obtained by the applicant. In its second regression analysis, the Commission performed an estimate in instrumental variables and replaced the ‘postal costs’ variable and the ‘fixed broadband access lines/subscribers’ lines and propagation’ variable by the two variables ‘change in gross domestic product (GDP)’ and ‘price of electricity for industrial customers’. According to the applicant, in order to be instrumental, the ‘postal costs’ variable must have a statistical link with the ‘price of electricity for industrial customers’ variable, which, however, it does not. That lack of any long-term link has been confirmed by standard cointegration tests.

125

Those arguments do not demonstrate that the Commission’s finding, in recital 22 of the contested decision, that there was no robust and conclusive evidence that electronic and postal delivery are actual substitutes was incorrect. Indeed, it must be recalled that the burden of proving that the conditions laid down in Article 30(1) of Directive 2004/17 are fulfilled rests on the applicant and the Member State concerned, since the Commission has only limited powers (see paragraph 41 above). It follows that, in the present case, the Commission was not obliged to perform its own analyses.

126

In any event, it must be observed that the last sentence of recital 21 of the contested decision (see paragraph 109 above) expressly states that the analysis conducted by the Commission produced only preliminary results. In its letter to the applicant of 4 April 2014, the Commission stated in this connection that its analysis was not to be regarded as a definitive analysis of the postal services market in Austria and that more precise methods or better data might produce different results. It follows that the observations made in the last sentence of recital 21 of the contested decision about the analysis conducted by the Commission were supplemental to the statement of reasons given in the remainder of that recital. Consequently, even if that analysis were incorrect, that would not mean that the contested decision is vitiated by a manifest error of assessment.

127

In so far as the applicant alleged at the hearing that the Commission had infringed its right to be heard by not allowing it access to its data and econometric calculations before it adopted the contested decision, it must be observed that that argument is inadmissible since, under Article 48(2) of the Rules of Procedure of 2 May 1991, which corresponds to Article 84(1) of the Rules of Procedure, no new plea in law may be introduced in the course of proceedings unless it is based on matters of law or of fact which come to light in the course of the procedure, which is not the case in this instance. In any event, this argument is also unfounded, inasmuch as it is apparent from a letter which the expert appointed by the applicant sent the Commission on 1 April 2014 that, even though the Commission had not allowed access to all its data, it had nevertheless given the expert an opportunity to discuss its econometric analysis with it.

128

The second complaint, and consequently the third part of the second plea, must therefore be rejected.

129

The applicant argues that the contested decision is vitiated by an inadequate statement of reasons for the finding, in recital 24 thereof, that it would be able to pass cost increases on to its customers. The applicant states that its prices are strictly regulated and that it has produced extensive documentation to show that, when prices are increased, its customers find alternative means of delivery. Given the market dynamics, the Commission’s conclusion that it holds a dominant position is unfounded, nor is it adequately reasoned. Moreover, according to a customer satisfaction survey carried out in 2013, when questioned on the point, [confidential]% of its larger customers mentioned replacement by electronic means as the reason for the decline in the volume of mail that the applicant delivered for them. According to the applicant, the Commission has furnished no evidence to show that, contrary to the findings which it has produced, the overall reduction in the size of the market for postal delivery is unconnected with the increasing replacement of postal delivery by electronic means of delivery. Even though the Commission envisages an increase in paperless communications and mail taking the place of postal items, it denies any correlation between the collapse of postal deliveries of letters and the increase in electronic mail.

130

In recital 24 of the contested decision, the Commission stated the following:

‘Austrian Post would be able to pass the cost increases on to the customers who, given their intrinsic preference for postal delivery, would have no other choice than [to accept] the price increase. For the same reason, the clearly dominant position of Austrian Post does not guarantee that potential cost decreases will be passed on to customers.’

131

In footnote 21, inserted into the first sentence of recital 24 of the contested decision, the Commission stated that an increased tariff had been introduced in May 2011, following approval by the relevant national authorities of a request from the applicant.

132

In this connection, it must be borne in mind that the obligation to state reasons is an essential procedural requirement, as distinct from the question whether the reasons given are correct, which goes to the substantive legality of the contested measure (see the judgment of 22 March 2001 in France v Commission, C‑17/99, ECR, EU:C:2001:178, paragraph 35 and the case-law cited). It is apparent from the applicant’s arguments that, in the context of this part of the plea, it essentially alleges both breach of an essential procedural requirement and incorrect reasoning.

133

First of all, as regards the alleged breach of the duty to state reasons, it must be observed that the conclusions which the Commission set out in recital 24 of the contested decision satisfy the requirements of the duty to state reasons, as mentioned in paragraphs 20 and 46 above. As regards the conclusion that the applicant would be able to pass cost increases on to the customers who, given their intrinsic preference for postal delivery, would have no other choice than to accept the price increase, it is clear from footnote 21, inserted into the first sentence of recital 24 of the contested decision, which refers to the tariff increase of May 2011, following a request from the applicant, that the Commission considered that the applicant was able to pass on a cost increase, even though price increases must be approved by the competent national authorities. Next, as regards the reasons for the conclusion that the applicant holds a dominant position on the Austrian market, it must be observed that, in recital 23 of the contested decision, the Commission stated, with reference to an email of the applicant’s dated 2 December 2013 in which the applicant estimated its share of the market for national business mail at [confidential]%, that the applicant held a large majority of the market for paper mail.

134

Secondly, as regards the applicant’s calling into question the merits of the conclusions set out in recital 24 of the contested decision, that argument cannot be upheld either. Indeed, first of all, in so far as concerns the point that the applicant produced extensive documentation to show that, when prices are increased, its customers find alternative means of delivery, it must be observed that the applicant does not adduce any specific evidence in support of that assertion and that its argument relating to the MH test and its submission of graphs of the evolution of quantities and prices has already been rejected by the Court after examining the third part of the present plea. Secondly, as regards the argument that, given the market dynamics, the Commission’s conclusion that the applicant holds a dominant position was incorrect, it is clear from the Court’s examination of the second part of the present plea that the Commission did not err in its findings, in recitals 18 and 19 of the contested decision, regarding hindrances to the substitutability of electronic and postal invoicing and the situation in the market. Thirdly, in so far as concerns the assertion that, according to a customer satisfaction survey carried out in 2013, when questioned on the point, [confidential]% of its larger customers mentioned replacement by electronic means as the reason for the decline in the volume of mail that the applicant delivered for them, it must be observed that, by merely referring to the outcome of that customer satisfaction survey the applicant wholly fails to prove that the Commission made a manifest error of assessment in concluding that electronic and postal delivery were not actual substitutes. Fourthly, in so far as the applicant complains of the Commission’s failure to furnish evidence, the Court has already pointed out that the burden of proving that the conditions laid down in Article 30(1) of Directive 2004/17 are fulfilled rests on the applicant and the Member State concerned, since the Commission has only limited powers (see paragraph 41 above). Fifth, contrary to the applicant’s allegation, the Commission in no way denied that there was a correlation between the decrease in the volume of postal deliveries and the increase in electronic mail. Indeed, the Commission stated, in recital 26 of the contested decision, that, although one of the main consequences of the increasing use of electronic means of communication was a reduction in the overall size of the letters market, it could not conclude from that that electronic communication had introduced direct competition within the market for postal delivery.

135

It follows that the fourth part of the plea, and thus the second plea in its entirety, must be rejected.

136

As regards the request for the author of the ACM Study and the shock analysis to be heard as a witness (see paragraph 43 above), in light of the foregoing and on the basis of the documents before the Court and the answers given to the questions asked by the Court at the hearing, the Court considers that it has sufficient information, is able to understand all the economic issues involved and can therefore decide whether the Commission’s assessment is vitiated by a manifest error of assessment. Consequently, it is not appropriate for the Court to accede to that request in so far as the present plea is concerned.

137

The applicant argues that the Commission infringed Article 30 of Directive 2004/17 and gave an inadequate statement of reasons in that it found, in recitals 34 to 42 of the contested decision, that postal services for addressed letters C2X at national level were not directly exposed to competition in Austria. According to the applicant, the Commission erred in its definition of the relevant market. More precisely, the Commission erred in taking the view that electronic delivery and postal delivery did not belong to the same domestic C2X market. The applicant argues in this connection that the figures relating to internet use in Austria which the Commission took into account were incorrect. It also argues that the Commission failed to take account of the studies which it produced, of the constant developments and of omnipresent electronic substitution. In order to substantiate its arguments and to explain the ACM Study and its communications to the Commission of 8 November and 2 December 2013, the applicant requests that its expert, the author of the ACM Study, be heard as a witness.

138

It must be reiterated that the obligation to state reasons is an essential procedural requirement, as distinct from the question whether the reasons given are correct, which goes to the substantive legality of the contested measure (see paragraph 132 above). It is apparent from the applicant’s arguments that it is not essentially alleging breach of the duty to state reasons as a breach of an essential procedural requirement, but is instead calling into question the merits of the Commission’s reasoning in the contested decision, inasmuch as it relied on incorrect figures and failed to take account of all the relevant information that needed to be taken into consideration in order to assess the situation in Austria (see also paragraph 39 above).

139

It is clear from recitals 34 to 42 of the contested decision that the Commission concluded that postal services for addressed letters C2X at national level were not directly exposed to competition in Austria and that, consequently, Article 30(1) of Directive 2004/17 did not apply to contracts intended to enable the pursuit of those activities in Austria. In reaching that conclusion, the Commission noted, in recitals 39 and 40 of the contested decision, that the relevant product market was the market for postal services for addressed letters C2X, in which the applicant held a market share of approximately [confidential]%. According to recital 38 of the contested decision, electronic and postal delivery for addressed letters C2X do not belong to the same relevant market. The Commission stated in this connection, in recital 36 of the contested decision, that, according to the 2013 WIK Study, almost 30% of the population in Austria had never used the internet, around 55% of the population did not use internet banking and 75% did not fill in administrative forms online. According to recital 37 of the contested decision, the applicant has not provided further empirical evidence to support its assertions and to demonstrate substitutability.

140

First of all, the applicant argues that the figures on internet use in Austria quoted in recital 36 of the contested decision, which were taken from the 2013 WIK Study, are incorrect and irrelevant to the definition of the relevant market. According to the applicant, a census carried out by the Statistical Office of the European Union (Eurostat) confirms that, in 2012, only 17% of the population in Austria said that they had not yet used the internet. Moreover, it is appropriate to take households as a reference, more than 80% of which had internet access in 2013, according to the Eurostat data. Between 2002 and 2012, the proportion of households having internet access in Austria rose from 34% to 79% and between 2003 and 2012 the proportion of households having broadband internet access rose from 10% to 77%. In addition, in 2013, 75% of Austrians used the internet to send emails, the majority of them being users in the 25 to 64 age range. According to the applicant, as far as the definition of the relevant market is concerned, email use alone is significant, for one reason because it enables the transmission of invoices as pdf attachments. The applicant adds that these developments are in line with what may be observed in ordinary life, inasmuch as communication between consumers is via email, short messages (SMS) and other similar services and because letters that are to be sent by post are now written only in exceptional cases. Lastly, according to the applicant, it is significant for the definition of the market that, in 2012, internet availability for undertakings in Austria was close to 100%.

141

Those arguments do not demonstrate that the Commission’s assessment is vitiated by a manifest error in that it is based on incorrect evidence.

142

First of all, the burden of proving that the conditions laid down in Article 30(1) of Directive 2004/17 are fulfilled rests on the applicant and the Member State concerned, inasmuch as, in this instance, the Commission has only limited powers in comparison with the very extensive investigatory powers that are conferred on it in the sphere of EU competition law by Regulation No 1/2003 and Regulation No 139/2004 (see paragraph 41 above). Moreover, the lawfulness of a Commission decision under Article 30 of Directive 2004/17 must be assessed in the light of the information available to the Commission when it adopted the decision (see, by analogy, the judgments of 15 April 2008 in Nuova Agricast, C‑390/06, ECR, EU:C:2008:224, paragraph 54, of 12 October 2011 in Dimos Peramatos v Commission, T‑312/07, EU:T:2011:587, paragraph 95, and of 15 July 2014 in Italy v Commission, T‑463/07, EU:T:2014:665, paragraph 108).

143

In the present case it must be observed that, as is clear from recital 35 of the contested decision, on being invited by the Commission to express, in accordance with the second subparagraph of Article 30(5) of Directive 2004/17, their views with regard to the definition of the relevant product market, the Austrian authorities were unable to provide additional information to support the applicant’s assertions.

144

Secondly, as regards the argument that the figures on internet use in Austria taken from the 2013 WIK Study are incorrect and irrelevant, it must be observed that the applicant does not call into question the results of the 2013 WIK Study generally and, in particular, does not assert that the study is based on incorrect data. On the contrary, during the administrative procedure it was the applicant that submitted this study. Moreover, in support of its argument relating to direct exposure to competition in the B2X and C2X markets, the applicant referred to this study in order to substantiate its view that the decrease in postal deliveries was to be ascribed in particular to electronic substitution. Furthermore, it must be pointed out that the 2009 WIK Study, that is to say, the study preceding the 2013 WIK Study and produced by the same consulting firm, is cited on a number of occasions in the applicant’s ACM Study.

145

Thirdly, contrary to the applicant’s allegation, it does not appear to be incorrect, in the context of the present definition of the relevant market, to refer to the fact that around 55% of the population does not use internet banking and that 75% do not fill in administrative forms online. As the Commission asserts, those figures are relevant to the definition of the relevant market, since they relate to essential segments of the C2X market, which comprises mail between individuals and mail from individuals to businesses. It must also be pointed out in this connection that the applicant’s argument that email use alone is significant, for one reason because it enables the transmission of invoices as pdf attachments cannot be upheld in so far as the C2X market is concerned. Indeed, the applicant has in no way shown that the volume of invoices sent by individuals is other than negligible in this market.

146

Fourthly, as regards the applicant’s argument that communication between consumers is via email, SMS and other similar services and that letters that are to be sent by post are now written only in exceptional cases, it must be observed that it has not been substantiated by any evidence. In this connection, as regards the document of the German regulatory authority dating from November 2014, this was produced along with the reply, without any reason being given for its late production, and it must therefore be rejected as inadmissible, pursuant to Article 48(1) of the Rules of Procedure of 2 May 1991. Moreover, the Court has already pointed out that the legality of a contested measure must be assessed on the basis of the facts and the law as they stood at the time when the measure was adopted (see paragraph 64 above). Furthermore, the applicant does not even assert that the document of the German regulatory authority refers to the situation in Austria. Lastly, as the Commission asserts, it is clear from point 5.2.2.2 of the applicant’s request that, in 2012, it delivered approximately [confidential] letters from private individuals.

147

Fifth, as regards the applicant’s argument that it is significant for the definition of the market that, in 2012, internet availability for undertakings in Austria was close to 100%, assuming that to be established it would still be necessary to point out that, on the C2X market it is particularly important for the internet to be available to individuals, so that they can send electronic communications.

148

Secondly, the applicant argues that the Commission failed to take account of the studies which it produced, of the constant developments and of omnipresent electronic substitution, It refers, in this connection, to the general description of the development of the B2X market which it gave in its application, to the 2013 WIK Study, to the ACM Study, to paragraph 5 of its letter to the Commission of 8 November 2013 and to paragraphs 3 and 4 of its communication to the Commission of 2 December 2013. It alleges that the Commission incorrectly evaluated only the 2013 WIK Study, which it produced.

149

Those arguments equally fail to demonstrate that the Commission’s assessment is vitiated by a manifest error in that it is not based on all the relevant information that needed to be taken into consideration.

150

Indeed, first of all, in so far as the applicant refers to its arguments relating to the definition of the B2X market, it must be observed that the Commission took those arguments into account when it examined the B2X market, as is clear from recitals 14 to 33 of the contested decision, and that the Court has found, on examining the applicant’s second plea, that the Commission made no manifest error of assessment in that regard. In the context of the present plea, the applicant has added nothing new. It must also be observed that, as the Commission asserts, the applicant has in no way shown that the research conducted among undertakings and the studies into the structure of business demand also apply to the structure of the demand from private individuals and that they consequently constitute relevant information that the Commission was required to take into account when considering the definition of the C2X market.

151

Secondly, as regards the applicant’s general reference to the 2013 WIK Study, to the ACM Study, to paragraph 5 of its letter to the Commission of 8 November 2013 and to paragraphs 3 and 4 of its communication to the Commission of 2 December 2013, it must be recalled that, under Article 21 of the Statute of the Court of Justice and Article 44(1)(c) of the Rules of Procedure of 2 May 1991, every application is required to state the subject matter of the proceedings and a summary of the pleas in law on which the application is based. According to consistent case-law, while the body of the application may be supported and supplemented on specific points by references to extracts from documents annexed thereto, a general reference to other documents, even those annexed to the application, cannot make up for the absence of the essential arguments in law which must appear in the application. It is not for the Court to seek and identify in the annexes the pleas and arguments on which it may consider the action to be based, since the annexes have a purely evidential and instrumental function (see the judgments of 9 March 2015 in Deutsche Börse v Commission, T‑175/12, EU:T:2015:148, paragraph 354 and the case-law cited, and of 25 March 2015 in Slovenská pošta v Commission, T‑556/08, EU:T:2015:189, paragraph 434 and the case-law cited). Consequently, this general reference must be rejected as inadmissible.

152

The third plea in law must therefore be rejected.

153

As regards the request for the applicant’s expert to be heard as a witness (see paragraph 137 above), in light of the foregoing and on the basis of the documents before the Court and the answers given to the questions asked by the Court at the hearing, the Court considers that it has sufficient information, is able to understand all the economic issues involved and can therefore decide whether the Commission’s assessment is vitiated by a manifest error of assessment. Consequently, it is not appropriate for the Court to accede to that request in so far as the present plea is concerned.

154

The applicant argues that the Commission infringed Article 30 of Directive 2004/17 and vitiated its decision by an inadequate statement of reasons in that it found, in recitals 43 to 50 of the contested decision, that postal services for addressed letters B2X and C2X at international level were not directly exposed to competition in Austria. The applicant argues that, for the same reasons as it gave in relation to the B2X and C2X markets at national level, the Commission erred in its definition of the relevant market. The applicant also argues that, in any event, its share of the market for the postal deliveries in question was below [confidential]% and that, as a result of stiff competition, it fell continually between 2010 and 2012, a point which the Commission failed to take into consideration. According to the applicant, the Commission declined its request, wrongly and without sufficient explanation taking the view that it was necessary to split the relevant market into the subcategories B2X international and C2X international. In any event, it provided the Commission with estimates of market shares in the B2X international and C2X international markets. The Commission failed to explain why, in its view, those estimates were unreliable. At very least, following its own line of argument, the Commission should have exempted the B2X international market from the application of Directive 2004/17. In order to substantiate its arguments and to explain the ACM Study and its communication to the Commission of 2 December 2013, the applicant requests that its expert, the author of the ACM Study, be heard as a witness.

155

It is clear from recitals 43 to 50 of the contested decision that the Commission concluded that Article 30(1) of Directive 2004/17 did not apply to contracts intended to enable the provision of postal services for addressed letters B2X and C2X at international level in Austria. According to recital 43 of the contested decision, the Commission rejected the applicant’s assertion that the delivery mode of cross-border mail is technologically neutral and that electronic and postal delivery services belong to the same relevant market, essentially for the same reasons as for postal services for addressed letters B2X and C2X at national level. In recital 46 of the contested decision, the Commission stated that competition for cross-border letter post was very different for private individuals and for companies, that private individuals generally had no real choice but to send international mail with their national universal service provider and that the volumes sent by private individuals were generally too small to provide an incentive for new entrants into the market. According to recital 47 of the contested decision, the competitive situation also depended on the size and population of each city, because cross-border service providers did not maintain a nationwide access network, but generally collected mail directly at customers’ premises. In recitals 48 and 49 of the contested decision, the Commission stated that, in its previous practice, it had drawn a distinction between cross-border postal services for the market for addressed B2X letters and cross-border postal services for the market for addressed C2X letters, and there was no evidence that the situation was any different in Austria. Consequently, according to the Commission, two separate product markets had to be considered, namely cross-border postal services for outbound B2X addressed letters and cross-border postal services for outbound C2X addressed letters. According to recital 50 of the contested decision, the applicant had been unable to provide detailed information on its shares of each of the relevant markets, or on the market shares of its main competitors. The Commission concluded, in that recital, that, in the absence of information on the degree of competition in each of those markets, it was not possible to conclude that the conditions for granting an exemption under Article 30(1) of Directive 2004/17 to cross-border postal services for outbound B2X addressed letters and to cross-border postal services for outbound C2X addressed letters in Austria had been fulfilled.

156

As regards, first of all, the applicant’s argument that the Commission erred in its definition of the relevant market, for the same reasons as stated in relation to the B2X and C2X markets at national level, suffice it to observe that the Court has already rejected those reasons, on examining the second and third pleas in law concerning the B2X and C2X markets at national level.

157

The applicant’s second argument is that the Commission declined its request, wrongly and without sufficient explanation taking the view that it was necessary to split the relevant market into the subcategories B2X international and C2X international.

158

As regards, first, the alleged breach of the duty to state reasons, it must be observed that, contrary to the applicant’s allegation, the Commission did not confine itself to making a general reference, in recital 48 of the contested decision, to its previous practice. Indeed, it is clear from the footnote inserted into that recital, which refers to footnote 6 to the contested decision, that the Commission referred to a specific earlier case. It also set out, in recitals 46, 47 and 49 of the contested decision, the reasons for which it was necessary in its view to draw a distinction between cross-border postal services for the market for addressed B2X letters and cross-border postal services for the market for addressed C2X letters: competition was very different and account had to be taken of the fact that the competitive situation also depended on the size and population of each city. There was no evidence that the situation was any different in Austria. That statement of reasons, in the light of the requirements mentioned in paragraphs 20 and 46 above, was sufficient to enable the persons concerned to ascertain the reasons for the measure adopted and to enable the competent Court to exercise its power of review.

159

As regards, secondly, the argument that the Commission erred in taking the view that it was necessary to split the relevant market into the subcategories B2X international and C2X international, the applicant essentially confines itself to stating that it is impossible to give separate market share figures for those two subcategories. That argument does not, however, establish that the Commission made a manifest error of assessment in drawing a distinction between cross-border postal services for the market for addressed B2X letters and cross-border postal services for the market for addressed C2X letters. Indeed, it in no way calls into question the reasons which the Commission gave in recitals 46 to 49 of the contested decision for drawing such a distinction (see paragraph 155 above). It must be observed in this connection that the 2013 WIK Study, which was submitted to the Commission by the applicant, confirmed the conclusions which the Commission stated in recital 46 of the contested decision and that, while, in its request, the applicant drew a distinction between the B2X and C2X markets at national level, it gave no reason for not doing so with regard to the B2X and C2X markets at international level, either in its request or in its letter to the Commission of 14 January 2014. Indeed, it refused to make such a distinction, as the Commission has confirmed.

160

Thirdly, the applicant argues that it provided the Commission with estimates of market shares in the B2X international and C2X international addressed letters markets in paragraph 2.1 of its communication of 2 December 2013 and that the Commission failed to explain why, in its view, those estimates were not reliable. That argument cannot be accepted. Indeed, it is clear from paragraph 2.1.3 of that communication that the applicant confined itself to making estimates for a global B2X and C2X international market, without drawing a distinction between cross-border postal services for the market for addressed B2X letters and cross-border postal services for the market for addressed C2X letters. In so far as the applicant argues that, in paragraph 5 of its letter of 8 November 2013, it provided the estimates requested by the Commission, it must be observed that that paragraph states the proportion of C2X mail in the overall volume of the applicant’s letter deliveries, which does not, however, offer any conclusion as to the applicant’s share of the market for C2X mail at international level. The Commission was therefore entitled to state, in recital 50 of the contested decision, that the applicant had been unable to provide detailed information on its shares of each of the relevant markets, or on the market shares of its main competitors, nor did it breach its duty to state reasons by making that observation.

161

Fourthly, the applicant asserts that, at very least, following its own line of argument, the Commission should have exempted the B2X international market from the application of Directive 2004/17. According to the applicant, if account is taken of the point made in recital 46 of the contested decision that its services are not substitutable on the international C2X market, its share of the B2X international market would be considerably smaller than [confidential]%, which is also in line with the Commission’s finding in recital 47 of the contested decision that the applicant’s competitors are mainly located in urban areas.

162

That argument must be accepted. Indeed, the Commission has not disputed that the applicant’s share of the market for postal services for addressed letters B2X and C2X at international level was below [confidential]%, as indicated in point 5.2.2.2 of the applicant’s request. Moreover, as the applicant assesses, recital 46 states that the services in question are not substitutable on the international C2X market: according to the Commission, private individuals generally have no real choice but to send international mail with their national universal service provider. It follows that the applicant’s share of the market for postal services for addressed letters B2X at international level must be well below [confidential]%, a fact which the Commission failed to take into account when it concluded that those postal services were not directly exposed to competition. In light of the foregoing, it must be concluded that the Commission made a manifest error of assessment in not exempting postal services for addressed letters B2X at international level from the application of Directive 2004/17.

163

Consequently, the fourth plea in law must be upheld to the extent that it relates to postal services for addressed letters B2X at international level and dismissed in so far as it relates to postal services for addressed letters C2X at international level.

164

As regards the request for the applicant’s expert to be heard as a witness (see paragraph 154 above), in light of the foregoing and on the basis of the documents before the Court and the answers given to the questions asked by the Court at the hearing, the Court considers that it has sufficient information, is able to understand all the economic issues involved and can therefore decide whether the Commission’s assessment is vitiated by a manifest error of assessment. Consequently, it is not appropriate for the Court to accede to that request in so far as the present plea is concerned.

165

The applicant argues that the Commission infringed Article 30 of Directive 2004/17 and gave an inadequate statement of reasons in that it found, in recitals 51 to 56 of the contested decision, that postal services for addressed advertising letters on a national and international level were not directly exposed to competition in Austria. According to the applicant, following the complete liberalisation of postal services, there are no longer grounds for drawing a distinction between addressed letters B2X at national level and addressed advertising material. Consequently, for the same reasons as given in relation to the B2X market at national level, the Commission erred in its definition of the relevant market. Moreover, numerous studies have shown that the market for postal services for addressed advertising material has shrunken significantly as a result of electronic substitution. However, in the contested decision, the Commission failed to take account of the arguments and studies which it submitted, confining itself to referring to its earlier decisions, which relate to a different geographical market and do not take account of developmental trends in the communications market.

166

It is clear from recitals 51 to 56 of the contested decision that the Commission concluded that postal services for addressed advertising letters on a national and international level were not directly exposed to competition in Austria and that, consequently, Article 30(1) of Directive 2004/17 did not apply to contracts intended to enable the provision of those services in Austria. In reaching that conclusion, the Commission noted, in recital 51 of the contested decision, that addressed advertising mail was defined as mail consisting solely of advertising, marketing or publicity material and comprising an identical message. According to that recital, that type of advertising mail, which could be addressed to companies or to private individuals, had to bear the customer’s name and address and required the customer’s agreement to receive such information. In recital 52 of the contested decision, the Commission rejected the applicant’s assertion that the market for addressed advertising mail could be included in the market for B2X postal services for addressed letters, because the applicant had not, according to the Commission, provided any empirical evidence thereof and its assertion was not in line with Commission Decision 2007/564/EC of 6 August 2007 exempting certain services in the postal sector in Finland, excluding the Åland Islands, from the application of Directive 2004/17 (OJ 2007 L 215, p. 21). It therefore concluded, in recitals 53 and 54 of the contested decision, that the relevant product market was postal services for addressed advertising letters, of which the applicant held a market share of [confidential]%.

167

It must be observed that, even supposing the applicant’s allegation to be correct and that addressed letters B2X at national level and addressed advertising mail belong to the same market, the applicant’s argument would still be incapable of calling into question the merits of the Commission’s findings. Indeed, since the Commission was entitled to conclude that the relevant product market for postal services for addressed letters B2X at national level was the market for postal services for addressed letters B2X of which the applicant had an estimated market share of [confidential]% (see the second plea) and given that it is common ground that the applicant’s share of the market for postal services for addressed advertising letters on a national and international level was also [confidential]%, it must be held that the Commission did not make a manifest error of assessment in concluding that the applicant was not directly exposed to competition in Austria on that market.

168

In any event, the applicant’s arguments do not show that the Commission made a manifest error of assessment in finding, in recital 53 of the contested decision, that the relevant product market was the market for postal services for addressed advertising letters. In support of its argument that, following the complete liberalisation of postal services, there are no longer grounds for drawing a distinction between addressed letters B2X at national level and addressed advertising material, the applicant refers, in particular, to recital 17 of Directive 2008/6, which states that direct mail consisting solely of advertising, marketing or publicity material and comprising an identical message, except for the addressee’s name, address and identifying number, may be considered as an item of correspondence. However, as the Commission asserts, the fact that Directive 2008/6 treats addressed advertising mail as an item of correspondence from a regulatory point of view has no bearing on the correct definition of the relevant market in the context of the examination carried out under Article 30 of Directive 2004/17. Moreover, as the Commission states, the 2013 WIK Study, which the applicant submitted during the administrative procedure, similarly does not regard addressed letters B2X at national level and addressed advertising letters as belonging to the same market. Furthermore, it must be observed that the ACM Study also considers addressed advertising letters as a specific product and indicates that the applicant’s tariffs for addressed advertising letters and B2X letters are different.

169

The applicant also argues that numerous studies have shown that the market for postal services for addressed advertising material has shrunken significantly as a result of electronic substitution. However, in the contested decision, the Commission failed to take account of the arguments and studies which it submitted with its request, confining itself to referring to its earlier decisions, which related to a different geographical market and did not take account of developmental trends in the communications market.

170

It must be observed in this connection that, inasmuch as the applicant confines itself to referring to the arguments and studies which it submitted with its request in order to substantiate its assertion that the Commission failed to take into account all the relevant information when it considered whether electronic delivery had replaced postal delivery of addressed advertising letters at national and international level, the Court has already pointed out that a mere reference to annexes cannot make up for the absence of the essential arguments in law which must appear in the application, since it is not for the Court to seek and identify in the annexes the pleas and arguments on which it may consider the action to be based, since the annexes have a purely evidential and instrumental function (see paragraph 151 above). Consequently, this general reference must be rejected as inadmissible.

171

In any event, it is common ground that the applicant’s share of the market for addressed advertising letters in Austria was [confidential]%, as is stated in recital 54 of the contested decision, which refers to the applicant’s communication of 2 December 2013. Given that percentage, and the fact that the Commission noted, in recital 55 of the contested decision, that the liberalisation of the postal market for addressed advertising letters in January 2011 had resulted in competitors gaining a total aggregated estimated market share of just [confidential]% by the time the contested decision was adopted, it was entitled to conclude, without making any manifest error of assessment, that postal services for addressed advertising letters at national and international level were not directly exposed to competition in Austria. Indeed, by merely making a general reference to the declining market for postal services for addressed advertising letters, the applicant has not proven that there is any relevant information that the Commission should have taken into account when assessing the situation in the relevant market, having regard to the market shares mentioned above.

172

Lastly, as regards the argument that the Commission confined itself to referring to its earlier decisions, which related to a different geographical market and did not take account of developmental trends in the communications market, that too must also be rejected. Indeed, in recital 52 of the contested decision, in addition to referring to Decision 2007/564, the Commission also stated that the applicant had not provided any empirical evidence to substantiate its assertion that the market for addressed advertising mail could be included in the market for B2X postal services for addressed letters.

173

The fifth plea must therefore be dismissed.

174

The applicant argues that the Commission infringed Article 30 of Directive 2004/17 and gave an inadequate statement of reasons in that it found, in recitals 57 to 64 of the contested decision, that postal services for unaddressed advertising letters on a national and international level were not directly exposed to competition in Austria. According to the applicant, if the Commission had taken all the relevant facts into account it would have reached the conclusion that, in 2012, the applicant had an estimated market share of only [confidential]% of the market for unaddressed advertising material. In its reply, the applicant requests that its expert be heard as a witness.

175

It is clear from recitals 57 to 64 of the contested decision that the Commission concluded that postal services for unaddressed advertising letters on a national and international level were not directly exposed to competition in Austria and that, consequently, Article 30(1) of Directive 2004/17 did not apply to contracts intended to enable the provision of those services in Austria. In reaching that conclusion, the Commission noted, in recital 57 of the contested decision, that unaddressed advertising mail was characterised by the absence of a particular destination address and consisted in unsolicited advertising mail that met certain criteria such as uniform weight, format, contents and layout, for distribution to a group of recipients. In recital 58 of the contested decision, the Commission stated that the applicant had defined the market for unaddressed advertising material as including advertising in other media such as advertisements in daily newspapers and weekly regional newspapers. The starting point for the applicant’s reasoning had been, according to the Commission, a ruling of the Austrian cartel court of 2009 in which it was accepted that both direct advertising and unaddressed advertising material could be regarded as part of the same relevant market for free newspapers. That court, however, had recognised substitutability only for large customers and had drawn several distinctions in respect of the relevant geographical market. In recital 59 of the contested decision, the Commission stated that the applicant had extended that conclusion to include all newspapers and had implied that the delivery of free, unaddressed mail competed with advertising in all newspapers. The applicant had also performed an HM test the results of which the Commission had duly considered. According to the Commission, the applicant’s broad interpretation of the ruling of the Austrian court was not in line with its previous decisions, according to which different media types are complementary to one another, rather than interchangeable. According to recital 60 of the contested decision, on being invited by the Commission to express their views with regard to the proposed definition of the unaddressed advertising market, having due regard to the abovementioned ruling of the Austrian court and the current legal and factual situation in Austria, the Austrian authorities had been unable to provide additional information to support the applicant’s assertions. In recital 61 of the contested decision, the Commission therefore stated that the information currently available to it was not conclusive enough to support the market definition proposed by the applicant. Recitals 62 and 63 of the contested decision state that the relevant product market was defined as postal services for unaddressed advertising letters, of which the applicant had a market share of [confidential]%.

176

It must be observed as a preliminary point (see paragraphs 38 to 41 above) that, in defining the relevant market, the Commission enjoyed a broad discretion that is subject to only limited review by the Courts of the European Union and that the burden of proof as regards definition of the relevant market rests on the applicant. Contrary to the applicant’s allegation, the Commission was therefore not obliged to perform its own analysis, as the Court has already observed (see paragraph 125 above). Moreover, the Court has also already observed (see paragraphs 56 and 57 above) that a relevant product market comprises all those products or services which are regarded as interchangeable or substitutable by the consumer, by reason of their characteristics, their prices and their intended use.

177

First of all, the applicant argues, referring to a study by the consulting firm E. entitled ‘Austrian print advertising market’, of September 2013 (‘the APAM Study’), which it annexed to its request for the application of Article 30 of Directive 2004/17, that the Commission failed to take account of the fact that, in the eyes of users, unaddressed advertising material, direct marketing and advertisements in free newspapers all have the same characteristics and the same intended use and that the relevant market therefore includes unaddressed advertising material, direct marketing and advertisements in free newspapers. According to the applicant, from the demand point of view, direct marketing and advertisements in free newspapers are interchangeable as a form of advertising, since the cost of such forms of advertising and the access to potential customers which they offer are practically the same. As regards access to potential customers, the applicant considers that it has demonstrated that, beside the two major newspapers published in Austria, there has, since 2009, been cooperation among free weekly newspapers in Austria which together cover the majority of the country. According to the applicant, very extensive reach could already be achieved via free newspapers. In so far as price is concerned, the applicant asserts that the price of one thousand contacts for these forms of advertising ranges from EUR [confidential] for advertisements in daily newspaper supplements to EUR [confidential] for unaddressed advertising material and direct marketing. The difference in price between advertisements in daily newspapers (EUR [confidential]) and unaddressed advertising in the form of brochures (EUR [confidential]) is in the order of just four euros. According to the APAM Study, the only thing that matters is that the advertising information contained in the direct marketing material or the advertisement actually reaches households.

178

Those arguments do not demonstrate that the Commission made a manifest error of assessment in finding, in recital 61 of the contested decision, that it did not have sufficiently conclusive information to support the market definition proposed by the applicant. Indeed, as the Commission asserts, it is clear from the APAM Study that there are significant differences between the characteristics of unaddressed printed advertising material and of advertisements in newspapers.

179

First, contrary to the applicant’s allegation, it is not apparent from the APAM Study that the only thing that matters is that the advertising information contained in the direct marketing material or the advertisement actually reaches households. As the Commission asserts, according to point 2.2.2 of the study, the most important criteria in advertising are usefulness, credibility and whether the advertisement is informative. While it is true that, in that passage of the APAM Study, no comparison is made between unaddressed advertising mail, advertising supplements to newspapers and advertisements in newspapers, as the applicant asserts, the fact remains that those criteria relate to the characteristics of different forms of advertising. The study shows that it is precisely with regard to those characteristics that differences exist between postal services for unaddressed advertising mail and advertisements in newspapers. Accordingly, point 2.2.1 of the APAM Study indicates that the main problem with direct marketing is the lack of context or connection with a medium such as a newspaper, unaddressed advertising often being experienced as an ‘unwanted visitor’. According to point 2.2.1 of the study, in short, while [confidential]% of people receiving newspapers take their newspaper, including the advertising pages, home with them, only [confidential]% of recipients of unaddressed advertising do the same. The APAM Study concludes that it is therefore less likely that unaddressed advertising material will be read than advertisements in a newspaper. As the Commission asserts, the APAM Study also indicates that readers have greater confidence in advertisements appearing in newspapers and that confidence is a highly important factor in advertising.

180

Secondly, as regards the argument that the costs of direct marketing and of advertisements in free newspapers are practically the same, it must be observed that that argument is based on the findings relating to the cost per thousand contacts set out in the APAM Study. According to point 1.2 of the study, in the 4.3 million households in Austria, 2.3 people on average read advertising mail. Given that, as the Commission asserts, the population in Austria was approximately 8.51 million in January 2013 — a fact which the applicant did not dispute when asked by the Court at the hearing — it cannot be concluded that the applicant was entitled to assume that the 4.3 million households in Austria comprised 2.3 people on average. It must also be observed that, according to point 2.2.1 of the APAM Study, [confidential]% of recipients throw away unaddressed advertising mail.

181

In addition, the Commission was right to state that the applicant had based its calculation on the lowest of its three applicable tariffs, as is apparent from point 1.2 of the APAM Study. In so far as the applicant justifies the use of that tariff by saying that the majority of direct marketing material is delivered in areas of urban concentration, in which the tariff which it used applies, it must be observed that, according to point 2.2.1 of the APAM Study, the average recipient of direct marketing material lives in a rural area. It must also be observed that, as the Commission asserts, point 1.2 of the study shows that the applicant chose the price for the lowest weight category, without explaining why that was specifically relevant. Furthermore, as the Commission states, it is common ground that the tariffs which the applicant used do not include the cost of printing the advertising material, by contrast with the cost per thousand contacts of advertisements in newspapers.

182

In so far as the applicant argued at the hearing that, in Implementing Decision 2014/299, the Commission did not dispute the existence of a common market for unaddressed advertising mail and advertisements in free newspapers, it must be observed that that decision related to the postal sector in Hungary. While it is true that, in recital 16 of that decision, the Commission defined the relevant market as the market for services for the distribution of unaddressed advertising material provided by postal operators, the fact remains that that definition was specifically based on the information relating to the Hungarian market mentioned in recitals 11 to 15 of the decision. The applicant has neither asserted nor established that the situation in the Austrian market is comparable to the situation in the Hungarian market, which the Commission examined in Decision 2014/299. This argument must therefore be rejected.

183

Secondly, the applicant argues that, in accordance with the Notice on market definition, it studied demand substitutability and supply substitutability and, in the context of the APAM Study, performed an HM test, which confirmed the existence in Austria of a common market for unaddressed direct marketing materials and advertisements in or advertising supplements to free newspapers. The applicant asserts that, according to the HM research conducted among 248 undertakings, not individuals, as the contested decision suggests, a 5% price increase would lead to a decline in direct marketing of [confidential]%. That fall in demand would be so significant that a price increase of that kind would not be profitable for the applicant. The applicant asserts that, in the contested decision, the Commission failed to take sufficient account of its arguments or of the results of the APAM Study, in particular regarding the structure of the Austrian advertising market and the results of the HM test. The APAM Study shows, according to the applicant, that, on the demand side, direct marketing and advertisements in newspapers are interchangeable. The ruling of the Austrian court mentioned in recital 58 of the contested decision was merely the starting point for the applicant’s definition of the market and was not intended to replace the analyses and studies which it submitted. Moreover, by confining itself to stating, in recital 61 of the contested decision, that the information available to it was not sufficiently conclusive to support the market definition proposed by the applicant, the Commission breached its duty to state reasons. In addition, by failing to examine the arguments and evidence submitted by the applicant, the Commission also infringed the applicant’s right to be heard.

184

As regards, first, the applicant’s argument that the Commission’s assessment is vitiated by a manifest error in that an HM test showed that there existed in Austria a common market for unaddressed direct marketing material and advertisements in or advertising supplements to free newspapers, that argument cannot be upheld. Indeed, the Commission was entitled to find in recitals 59 and 61 of the contested decision, without making any manifest error of assessment, that after duly considering the results of that test, the questionnaires for which had been sent to it by the applicant in a communication of 2 December 2013, it did not have sufficiently conclusive information to support the market definition proposed by the applicant.

185

Having regard to the fact that the Austrian statistics produced by the Commission show that there were approximately 311000 undertakings in Austria in 2011, while just 248 undertakings were questioned, the Commission cannot be criticised for expressing doubt as to whether the survey carried out was representative. It must also be observed that table 52 of the APAM Study shows that the range of undertakings questioned, in terms of their size, was not representative of the range of undertakings in the Austrian economy. As regards the argument that it is first and foremost retail undertakings having more than 250 workers that use circulars as a form of advertising, it must be observed that the applicant has furnished no evidence to show that its selection was representative, especially since, according to the applicant, its selection of 248 undertakings from among its customers having the highest turnover was dictated solely by their availability.

186

Moreover, in so far as the applicant points out that, contrary to the Commission’s suggestion in recital 59 of the contested decision, undertakings were questioned, not individuals, suffice it to observe that the term ‘subjects’ includes both natural and legal persons. The use of the term ‘subjects’ does not, therefore, demonstrate any imprecision on the Commission’s part.

187

The applicant also argues that the Commission breached its duty to state reasons in that it confined itself to stating, in recital 61 of the contested decision, that the information available to it was not conclusive enough to support the market definition proposed by the applicant. According to the applicant, that finding was merely stated, without any indication being given of whether it was based solely on the research which it had carried out or also on the additional information gathered from the Austrian authorities. That argument must be rejected. It is clear from the word ‘consequently’, used in recital 61 of the contested decision, that the finding reached in that recital is a conclusion based on recitals 57 to 60 of the decision, which include an appraisal of both the applicant’s research and the reply from the Austrian authorities.

188

Thirdly, in so far as the applicant asserts that, by failing to examine the arguments and evidence which it submitted, the Commission infringed its right to be heard, which is guaranteed by Article 6 TFEU and Article 41 of the Charter of Fundamental Rights of the European Union, it must be recalled that Article 41 of the Charter, entitled ‘Right to good administration’, provides, in paragraph 1 thereof, that ‘every person has the right to have his or her affairs handled impartially, fairly and within a reasonable time by the institutions, bodies, offices and agencies of the Union’. Article 41(2)(a) of the Charter provides that that right includes the right of every person to be heard, before any individual measure which would affect him or her adversely is taken. According to the case-law on the principle of sound administration, where the institutions of the European Union have a power of appraisal, observance of the rights guaranteed by the legal order of the European Union in administrative procedures is of even more fundamental importance. Those guarantees include, in particular, the duty of the competent institution to examine carefully and impartially all the relevant aspects of the individual case (judgment of 21 November 1991 in Technische Universität München, C‑269/90, ECR, EU:C:1991:438, paragraph 14). It is clear from paragraphs 176 to 187 above that the Commission did not breach that duty.

189

The sixth plea must therefore be dismissed.

190

As regards the request for the applicant’s expert to be heard as a witness (see paragraph 174 above), the applicant considers that its expert should be heard so that he can explain the APAM Study and the criteria relating to accessibility, costs per thousand contacts and the research proving the substitutability of unaddressed advertising mail and advertisements in free newspapers. That request was made in the reply, without any reason being given for its lateness, and it must therefore be rejected as inadmissible, pursuant to Article 48(1) of the Rules of Procedure of 2 May 1991. In any event, in light of the foregoing and on the basis of the documents before the Court and the answers given to the questions asked by the Court at the hearing, the Court considers that it has sufficient information, is able to understand all the economic issues involved and can therefore decide whether the Commission’s assessment is vitiated by a manifest error of assessment. Consequently, it is not appropriate for the Court to accede to that request in so far as the present plea is concerned.

191

The applicant argues that the Commission gave an incorrect statement of reasons and breached its duty to state reasons in that it found, in recitals 65 to 69 of the contested decision, that postal services for the standard delivery of addressed and unaddressed newspapers were not directly exposed to competition in Austria. According to the applicant, if the Commission had taken all the relevant facts into account it would have reached the conclusion that there existed a common market for the standard delivery of daily, weekly and monthly titles and that the applicant was directly exposed to competition on that market. Consequently, the Commission had also breached its right to be heard. In order to substantiate its arguments and to explain the study by the consultancy firm E., entitled ‘Austrian Delivery Market for Newspapers’, of September 2013 (‘the ADMN Study’), the applicant requests that its expert, the author of the ADMN Study, be heard as a witness.

192

It is clear from recitals 65 to 69 of the contested decision that the Commission concluded that postal services for the standard delivery of addressed and unaddressed newspapers were not directly exposed to competition in Austria and that, consequently, Article 30(1) of Directive 2004/17 did not apply to contracts intended to enable to pursuit of those activities in Austria. In reaching that conclusion, the Commission noted, in recital 65 of the contested decision, that Decision 2007/564 established a distinction between early delivery of newspapers and standard delivery. According to recital 66 of the contested decision, the applicant was not active in the early delivery of newspapers, but was active in the standard delivery of newspapers. In recitals 67 and 68 of the contested decision, the Commission stated that the relevant product market was the market for postal services for the standard delivery for addressed and unaddressed newspapers and that the applicant had a [confidential]% share of that market. The main competitors were, according to the Commission, the national and regional networks organised by publishers, which deliver addressed and unaddressed newspapers to households. However, those competitors together had a combined market share of only [confidential]%.

193

First of all, as regards its argument that the Commission breached its duty to state reasons, the applicant submits that the general reference made in recital 65 of the contested decision to an earlier decision does not constitute an adequate statement of reasons. That earlier decision concerned the delivery market in Finland and the Commission has not explained why there should be a connection between the situation in the Finnish market and the situation in the particular market in Austria. Nor did it give any reason in that earlier decision for drawing a distinction between early delivery and standard delivery of newspapers. The applicant asserts that, since the structures of the markets in the various Member States differed significantly in some respects, the Commission should have demonstrated that there was a similarity between the Finnish and Austrian markets for newspaper delivery, so as to render its decision comprehensible. The Commission should also have analysed the specific situation in the Austrian market, and should have stated reasons for its decision accordingly.

194

It must be observed in this connection that, while it is true that the contested decision offers no substantial reason for the Commission’s decision to distinguish between early delivery and standard delivery of newspapers, the fact remains that the contested decision refers to the Commission’s practice as reflected in Decision 2007/564, in which it exempted certain services in the postal sector in Finland from the application of Directive 2004/17. It is clear from recitals 13 and 14 of Decision 2007/564 that, as a matter of practice, the Commission drew a distinction between the early delivery of newspapers and the standard delivery of newspapers. The Court has already noted that the applicant was closely involved in the administrative procedure and that the Commission even sent a draft of the decision to the expert engaged by the applicant, who had a discussion with the Commission at a meeting on 28 March 2014 (see paragraph 83 above). The draft of the decision considered at that meeting already contained the wording in recital 65 of the contested decision.

195

Moreover, it is clear from the Commission’s letter to the applicant of 28 November 2013 and its letter to the Republic of Austria of 5 December 2013 that it informed the applicant and that Member State of the fact that its previous practice had been to draw a distinction between early delivery of newspapers and standard delivery of newspapers. In its letter of 13 January 2014 to the expert engaged by the applicant, the Commission expressly mentioned that distinction.

196

Furthermore, it must be observed that, in point 3.1.6 of its request, concerning postal services for the standard delivery of addressed and unaddressed newspapers, the applicant confined itself, in substance, to asserting that the relevant product market included the distribution of addressed and unaddressed daily, weekly and monthly titles in Austria and to referring, generally, to the ADMN Study. It gave no indication of the reason for which the distinction which the Commission drew in practice between early and standard delivery of newspapers might be incorrect. In so far as the applicant argues that it established parameters for defining the market in a particular page of the ADMN Study, it must be observed that the passage to which the applicant refers simply contains a description of the Opinion of Advocate General Jacobs in Bronner (C‑7/97, ECR, EU:C:1998:264). However, it is clear from point 31 of that Opinion that the precise definition of the market was left open in that case. In its judgment of 26 November 1998 in Bronner (C‑7/97, ECR, EU:C:1998:569, paragraph 34), the Court of Justice left it to the national court to consider the question of market definition.

197

Consequently, since the applicant was informed of the Commission’s practice and, as the Commission also states, put forward no substantial reasons to warrant a different definition of the relevant market, and since the Commission was not obliged to anticipate potential objections (see paragraph 46 above), it cannot be concluded that the Commission failed to satisfy the requirements of the duty to state reasons, as mentioned in paragraphs 20 and 46 above (see, to that effect, the judgment of 1 July 2010 in AstraZeneca v Commission, T‑321/05, ECR, EU:T:2010:266, paragraph 81 and the case-law cited).

198

Secondly, in so far as the applicant alleges an incorrect statement of reasons and breach of its right to be heard (see paragraph 188 above), its position is essentially that the Commission failed to give sufficient consideration to its arguments and to the ADMN Study, which it produced during the administrative procedure. According to the applicant, in its request for the application of Article 30 of Directive 2004/17, it explained in detail that the relevant market included the delivery in Austria of addressed and unaddressed daily, weekly and monthly titles. Referring to the ADMN Study, it asserts that, in addition to itself, two networks also provided the nationwide delivery of newspapers and magazines. Also, in its request, it stated that any competitor could have free access to the delivery of newspapers and magazines and that, in the daily deliveries sector, its market share was [confidential]%. Even including the delivery of weekly and monthly newspapers and magazines, it would still, with a market share of [confidential]%, be directly exposed to competition from other distributors. According to the applicant, having regard to the ADMN Study, the Commission should in any event have defined the relevant market precisely, as it was required to do by point 41 of the Notice on market definition and as it failed to do in recital 6 of the contested decision. Moreover, the networks of its competitors already existed in Austria and those competitors might at any moment enter the market.

199

First of all, it must be observed that that line of argument does not demonstrate that the Commission made a manifest error of assessment by defining, in recital 67 of the contested decision, the relevant product market as the market for postal services for the standard delivery of addressed and unaddressed newspapers and by excluding the early delivery of newspapers. Indeed, the Court has already observed (see paragraph 196 above) that, in the arguments set out in its request and in the ADMN Study to which the applicant has referred in these proceedings, the applicant has in no way indicated why the distinction which the Commission draws in practice, and which it drew in the contested decision, between the early delivery of newspapers and the standard delivery of newspapers might be incorrect.

200

Moreover, as regards the general reference which the applicant makes to its request and to the ADMN Study, the Court has already pointed out that a mere reference to annexes cannot make up for the absence of the essential arguments in law which must appear in the application, since it is not for the Court to seek and identify in the annexes the pleas and arguments on which it may consider the action to be based, since the annexes have a purely evidential and instrumental function (see paragraph 151 above). Consequently, this general reference must be rejected as inadmissible. In any event, it must be observed that, contrary to the applicant’s allegation, the ADMN Study also contains passages which suggest that, for the purposes of defining the relevant product market, a distinction should be drawn between the early delivery of newspapers and the standard delivery of newspapers. Indeed, according to point 1 of the study, newspapers are delivered at different times of the day, and this is a quality factor. Point 1 also states that daily newspapers must arrive in the letter box before people the house, whereas weekly and regional titles, which are financed mainly from advertising, are delivered during the day, since they are not time-sensitive.

201

Secondly, it must be observed that the applicant has not disputed the Commission’s findings, in recitals 66 and 68 of the contested decision, that it was not active in the early delivery of newspapers and that it had a market share of [confidential]% of the market for postal services for the standard delivery of addressed and unaddressed newspapers.

202

Thirdly, the applicant argues that, having regard to the ADMN Study, the Commission should in any event have defined the market precisely in so far as concerns postal services for the delivery of addressed and unaddressed newspapers. The position which the Commission expressed in recital 6 of the contested decision, namely that the precise definition of the relevant market could be left open, is incorrect. According to the applicant, a precise definition of the market, in accordance with the methods prescribed in the case-law and in point 41 of the Notice on market definition, was indispensable.

203

That argument cannot be accepted. Indeed, in so far as the applicant refers to the ADMN Study for the purposes of defining the relevant product market, its arguments have already been rejected (see paragraph 200 above). Next, in so far as the applicant refers to recital 6 of the contested decision, it must be observed that, according to that recital, which appears in the part of the contested decision dealing with the legal framework of the decision, although a narrower market definition could be envisaged in certain cases, the precise definition of the relevant market could be left open for the purposes of the contested decision to the extent that the result of the analysis remained the same whether it was based on a narrow or a broad definition. In the present case, while the precise definition of the relevant market was left open by the Commission in so far as concerned, in particular, postal services for addressed letters B2X at national level, as is apparent from recital 29 of the decision, that was not the case for postal services for the standard delivery of addressed and unaddressed newspapers.

204

As regards the argument that the Commission should have defined the market precisely, in accordance with the methods prescribed in the case-law and in point 41 of the Notice on market definition, according to which consumer preferences are a relevant criterion in the assessment of the demand substitutability of two products, it must be recalled that the burden of proving that the conditions laid down in Article 30(1) of Directive 2004/17 are fulfilled rests on the applicant and the Member State concerned, inasmuch as, in this instance, the Commission has only limited powers in comparison with the very extensive investigatory powers that are conferred on it in the sphere of EU competition law by Regulation No 1/2003 and Regulation No 139/2004 (see paragraph 41 above). In the present case, it was therefore incumbent on the applicant to provide sufficient information to define the relevant product market.

205

Fourthly, in so far as concerns the argument that networks of competitors already existed in Austria and that those competitors might at any moment enter the market, since there are no barriers to entry, it must be observed that, given the very large market shares which the applicant enjoys on the market for the standard delivery of addressed and unaddressed newspapers, that is to say, [confidential]%, and on the markets for postal services for addressed letters B2X and C2X at national level, [confidential]% and [confidential]% respectively (see paragraphs 99 and 139 above), it cannot be held that the Commission made a manifest error of assessment by concluding that postal services for the standard delivery of addressed and unaddressed newspapers were not directly exposed to competition in Austria.

206

The seventh plea must therefore be dismissed.

207

As regards the request for the applicant’s expert to be heard as a witness (see paragraph 191 above), in light of the foregoing and on the basis of the documents before the Court and the answers given to the questions asked by the Court at the hearing, the Court considers that it has sufficient information, is able to understand all the economic issues involved and can therefore decide whether the Commission’s assessment is vitiated by a manifest error of assessment. Consequently, it is not appropriate for the Court to accede to that request in so far as the present plea is concerned.

208

In light of all the foregoing, the action must be partly upheld, to the extent that it relates to postal services for addressed letters B2X at international level (see paragraph 163 above). Consequently, the contested decision must be annulled in so far as it states that Directive 2004/17 continues to apply to the market for postal services for addressed letters B2X on an international level in Austria. The action must be dismissed as to the remainder.

209

Under Article 134(3) of the Rules of Procedure, where each party succeeds on some and fails on other heads, the Court may order each party to bear its own costs. However, if it appears justified in the circumstances of the case, the General Court may order that one party, in addition to bearing its own costs, pay a proportion of the costs of the other party.

210

In the present case, the Court must grant the form of order sought by the applicant in so far as it seeks the annulment of the contested decision in so far as it states that Directive 2004/17 continues to apply to the market for postal services for addressed letters B2X on an international level in Austria. On the other hand, it must dismiss the action to the extent that it relates to the other markets for postal services at issue. Accordingly, it is fair in the circumstances of the case to decide that the applicant is to bear its own costs and pay eight tenths of the Commission’s costs. The Commission must bear two tenths of own costs.

On those grounds,

THE GENERAL COURT (Fifth Chamber)

hereby: