1. 

In 2011, the Council of the European Union imposed an anti-dumping duty on imports of ceramic tiles originating in the People’s Republic of China (‘China’).

2. 

Because China is a non-market economy country without a reliable normal value on the domestic market, the Union institutions used a constructed normal value in a market-economy third country, namely the United States of America (‘USA’ or ‘US’), when determining the existence and margin of dumping. For that purpose, they used figures provided by a single US producer, making adjustments to allow for various differences but (for reasons of business confidentiality) without giving full details of the adjustments made. When determining the level of price undercutting, the institutions based their price calculations on samples of both Chinese and Union producers, though using different sampling methods in each case.

3. 

A German importer, objecting to aspects of those methods, has challenged the duty charged on the tiles it imported. The Finanzgericht München (Munich Finance Court, ‘the referring court’), hearing the case, seeks a preliminary ruling on the validity of the regulation imposing the duty.

4. 

Other issues concerning a different challenge to the same regulation have been referred to the Court in Case C‑569/13 Bricmate, in which I also deliver my Opinion today.

5.

The rules governing the imposition of anti-dumping duties are contained in Regulation No 1225/2009 (‘the basic regulation’). ( 2 )

6.

Article 1(1) of the basic regulation sets out the principle that an anti-dumping duty may be applied to any dumped product whose release for free circulation in the Union causes injury. Article 1(2) defines a dumped product as one whose export price to the Union is less than a comparable price for the like product, in the ordinary course of trade, as established for the exporting country.

7.

Article 2 lays down the principles and rules governing determination of dumping. Essentially, for a given product exported from a third country, a normal value based on the prices paid, in the ordinary course of trade, by independent customers in the exporting country, and an export price to the Union, are established. If a fair comparison of weighted averages shows that the normal value exceeds the export price, the amount by which it does so is the dumping margin.

8.

In that regard, Article 2(7)(a) provides:

‘In the case of imports from non-market economy countries …, normal value shall be determined on the basis of the price or constructed value in a market economy third country, or the price from such a third country to other countries, including the [Union], or where those are not possible, on any other reasonable basis, including the price actually paid or payable in the [Union] for the like product, duly adjusted if necessary to include a reasonable profit margin.

An appropriate market economy third country shall be selected in a not unreasonable manner, due account being taken of any reliable information made available at the time of selection. Account shall also be taken of time-limits; where appropriate, a market economy third country which is subject to the same investigation shall be used.

The parties to the investigation shall be informed shortly after its initiation of the market economy third country envisaged and shall be given 10 days to comment.’

9.

Article 2(10) provides:

‘A fair comparison shall be made between the export price and the normal value. This comparison shall be made at the same level of trade and in respect of sales made at, as closely as possible, the same time and with due account taken of other differences which affect price comparability. Where the normal value and the export price as established are not on such a comparable basis due allowance, in the form of adjustments, shall be made in each case, on its merits, for differences in factors which are claimed, and demonstrated, to affect prices and price comparability. …’

10.

Article 3 (‘Determination of injury’) provides, in particular:

‘...

2.   A determination of injury shall be based on positive evidence and shall involve an objective examination of both:

3.   ... With regard to the effect of the dumped imports on prices, consideration shall be given to whether there has been significant price undercutting by the dumped imports as compared with the price of a like product of the [Union] industry, or whether the effect of such imports is otherwise to depress prices to a significant degree or prevent price increases, which would otherwise have occurred, to a significant degree. ...’

11.

Article 6(9) provides: ‘… an investigation shall, whenever possible, be concluded within one year. In any event, such investigations shall in all cases be concluded within 15 months of initiation …’.

12.

Article 9(4) lays down, inter alia, the ‘lesser duty rule’:

‘Where the facts as finally established show that there is dumping and injury caused thereby, and the Union interest calls for intervention …, a definitive anti-dumping duty shall be imposed ... The amount of the anti-dumping duty shall not exceed the margin of dumping established but it should be less than the margin if such lesser duty would be adequate to remove the injury to the Union industry.’

13.

Article 17 concerns sampling. It provides, in particular:

‘1.   In cases where the number of complainants, exporters or importers, types of product or transactions is large, the investigation may be limited to a reasonable number of parties, products or transactions by using samples which are statistically valid on the basis of information available at the time of the selection, or to the largest representative volume of production, sales or exports which can reasonably be investigated within the time available.

2.   The final selection of parties, types of products or transactions made under these sampling provisions shall rest with the Commission, though preference shall be given to choosing a sample in consultation with, and with the consent of, the parties concerned …

…

4.   Where it is decided to sample and there is a degree of non- cooperation by some or all of the parties selected which is likely to materially affect the outcome of the investigation, a new sample may be selected. However, if a material degree of non-cooperation persists or there is insufficient time to select a new sample, the relevant provisions of Article 18 shall apply.’

14.

Article 18 concerns non-cooperation. It provides, inter alia:

‘1.   In cases in which any interested party refuses access to, or otherwise does not provide, necessary information within the time- limits provided in this Regulation, or significantly impedes the investigation, provisional or final findings, affirmative or negative, may be made on the basis of the facts available. …

…

5.   If determinations, including those regarding normal value, are based on the provisions of paragraph 1, including the information supplied in the complaint, it shall, where practicable and with due regard to the time-limits of the investigation, be checked by reference to information from other independent sources which may be available, such as published price lists, official import statistics and customs returns, or information obtained from other interested parties during the investigation.

Such information may include relevant data pertaining to the world market or other representative markets, where appropriate.’

15.

Finally, Article 19 of the basic regulation provides, inter alia:

‘1.   Any information which is by nature confidential (for example, because its disclosure would be of significant competitive advantage to a competitor or would have a significantly adverse effect upon a person supplying the information or upon a person from whom he has acquired the information) or which is provided on a confidential basis by parties to an investigation, shall, if good cause is shown, be treated as such by the authorities.

…

4.   This Article shall not preclude the disclosure of general information by the [Union] authorities and in particular of the reasons on which decisions taken pursuant to this regulation are based, or disclosure of the evidence relied on by the [Union] authorities in so far as is necessary to explain those reasons in court proceedings. Such disclosure must take into account the legitimate interests of the parties concerned that their business secrets should not be divulged.

…’

16.

The anti-dumping duty in issue in the main proceedings was first imposed by Regulation No 258/2011 (‘the provisional regulation’) ( 3 ) and then confirmed by Regulation No 917/2011 (‘the definitive regulation’ or ‘the contested regulation’). ( 4 )

17.

Following a complaint lodged by the European Ceramic Tile Manufacturers’ Federation (‘CET’), the Commission gave notice on 19 June 2010 of the initiation of an anti-dumping proceeding concerning imports of ceramic tiles originating in China. ( 5 )

18.

On 16 March 2011, the Commission adopted the provisional regulation, Article 1(1) of which imposed a provisional anti-dumping duty on imports of ‘glazed and unglazed ceramic flags and paving, hearth or wall tiles; glazed and unglazed ceramic mosaic cubes and the like, whether or not on a backing, currently falling within CN codes 6907 10 00, 6907 90 20, 6907 90 80, 6908 10 00, 6908 90 11, 6908 90 20, 6908 90 31, 6908 90 51, 6908 90 91, 6908 90 93 and 6908 90 99, and originating in [China]’. Under Article 1(2), the rate of duty applicable to such products manufactured by certain listed companies varied between 26.2% and 36.6%, with a rate of 73.0% being applied to those produced by all other companies.

19.

In that regulation, since China is a non-market economy, the Commission constructed normal value, pursuant to Article 2(7)(a) of the basic regulation, on the basis of data provided by a producer in an analogue market economy third country, namely the USA. In order to evaluate price undercutting, it also used figures from samples of both Chinese and Union producers, pursuant to Article 17(1) of the basic regulation, but selected the samples by two different methods.

20.

On 12 September 2011 the Council adopted the definitive regulation, Article 1(1) of which imposed a definitive anti-dumping duty on imports of the same products as the provisional regulation. Under Article 1(2), the rate of the definitive duty varied between 26.3% and 36.5% for products manufactured by the listed companies, with a rate of 69.7% being applied to those of all other companies.

21.

In the preamble to that regulation, the Council dismissed various criticisms as to the choice of the USA as analogue third country, the use of data provided by a single supplier in that country and the adjustments made to it, and the difference in sampling methods.

22.

I shall set out the detailed reasoning in those two regulations in the context of each aspect of the question referred to which it relates.

23.

In July 2011, Fliesen-Zentrum Deutschland GmbH (‘Fliesen-Zentrum’) imported ceramic tiles falling under CN code 6907 90 20, manufactured in China and declared them to the Hauptzollamt Regensburg (Principal Customs Office, Regensburg; ‘the HZA’). On 2 August 2011, the HZA issued a notice assessing the customs duty and import VAT payable and requiring Fliesen-Zentrum to provide security for provisional anti-dumping duty of 32.3%, amounting to EUR 9479.09, which Fliesen-Zentrum paid.

24.

On 5 August 2011, Fliesen-Zentrum lodged an objection to that notice, which the HZA dismissed on 19 October 2011.

25.

On 4 November 2011, the HZA made a definitive assessment of anti-dumping duty in the amount of EUR 9479.09, set off in full against the security already paid. Fliesen-Zentrum lodged a further objection against that assessment, which the HZA dismissed on 3 February 2012.

26.

Fliesen-Zentrum brought an action against the assessment before the referring court, calling into question the validity of the definitive regulation.

27.

The referring court took the view that four of the grounds advanced by Fliesen-Zentrum in that regard were plausible:

28.

After analysing those grounds, the referring court considered it necessary to seek a preliminary ruling on the question: ‘Is [the definitive regulation] valid?’

29.

The referring court’s question is simple and general. It does not identify any specific ground of possible invalidity on which the referring court seeks a ruling. However, that court’s concern is clearly with the four grounds outlined in the order for reference.

30.

Fliesen-Zentrum, the Council and the Commission, all of whom have submitted written observations, and all of whom presented oral argument at the hearing on 3 December 2014, have addressed those four grounds individually. I shall do the same.

31.

However, Fliesen-Zentrum has advanced two further grounds on which the contested regulation might be found invalid, neither of which is mentioned in the order for reference. The Commission, moreover, has anticipated and sought to refute certain of Fliesen-Zentrum’s arguments gleaned from the referring court’s file.

32.

Furthermore, while each of the four grounds of invalidity raised in the order for reference can be addressed independently, the first three all concern the way in which a constructed normal value was established and all raise issues concerning the adequacy of the statement of reasons in the provisional and definitive regulations. Those common features seem to me to raise general issues concerning both the scope and the standard of the Court’s review.

33.

I do not consider that the Court should examine the additional issues addressed by Fliesen-Zentrum and the Commission in the present proceedings.

34.

According to settled case-law, the procedure laid down in Article 267 TFEU is based on a clear separation of functions between the national courts and the Court of Justice. It is solely for the national court before which the dispute has been brought, and which must assume responsibility for the subsequent judicial decision, to determine in the light of the particular circumstances of the case the relevance of the questions which it submits to the Court. Moreover, Article 267 TFEU does not provide a means of redress to the parties to a case pending before a national court, so that the Court cannot be compelled to evaluate the validity of Union law on the sole ground that that question has been put before it by one of the parties in its written observations. It is therefore inappropriate to examine grounds of invalidity which the referring court has not referred to. ( 6 )

35.

As has been pointed out in both the order for reference and the observations submitted to the Court, it has consistently been held that the choice of analogue country falls within the discretion enjoyed by the Union institutions when analysing complex economic situations, although the exercise of that discretion is not exempt from judicial review and (pursuant to Article 2(7)(a) of the basic regulation) the choice must be made in a not unreasonable manner. ( 7 ) The extent of the Court’s review of a choice of analogue country, according to that case-law, is therefore relatively limited.

36.

In the present case, it is not only the choice of the USA as an analogue country for China, but also the reliance on data from a single producer within the USA, which is called into question. In the latter regard, the Court has accepted that, where necessary, a constructed normal value may be established on the basis of figures from a single producer. ( 8 )

37.

While the two aspects are separate, it might seem surprising that a normal value for the products of all ceramic tile producers in the whole of China should be constructed from data provided by a single producer in the USA, when that country has a significantly higher income per capita than China, ( 9 ) but produces less than 1/30 of the same quantity of tiles. ( 10 )

38.

In the presence of such a prima facie anomaly, I consider that the Court should not limit the extent of its review any further than is necessary in accordance with its previous case-law. In particular, it should verify carefully whether the institutions neglected to take account of essential factors for the purpose of establishing the appropriate nature of the country chosen and whether the information contained in the documents in the case was considered with all the care required. ( 11 )

39.

On the other hand, it is also necessary to bear in mind the precise extent of the institutions’ duty to state reasons in the particular context.

40.

The Court has regularly pointed out that anti-dumping regulations have a dual nature as acts of a legislative nature and acts liable to be of direct and individual concern to certain traders. ( 12 )

41.

In Petrotub, ( 13 ) a case in which an anti-dumping regulation was challenged by parties to whom it was of direct and individual concern, the Court stated: ‘the statement of reasons … must disclose in a clear and unequivocal fashion the reasoning followed … in such a way as to enable the persons concerned to ascertain the reasons for the measure and to enable the competent [Union] court to exercise its power of review. The requirements to be satisfied … 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. ...’

42.

However, the Court has also consistently held that the extent of the requirement to state reasons depends on the nature of the measure in question, and that in the case of measures intended to have general application the preamble may be limited to indicating the general situation which led to its adoption, on the one hand, and the general objectives which it is intended to achieve, on the other. ( 14 )

43.

Here, the contested regulation is, in principle, a regulatory act of general application. It is none the less of direct and individual concern to the exporters whose products are affected, and any linked importers, who have locus standi to challenge it directly before the General Court. ( 15 ) Others to whom it might be of direct concern in the absence of implementing measures could also challenge it. ( 16 ) But it is not of individual concern to Fliesen-Zentrum (which, as an independent importer not involved in the anti-dumping investigation, does not fall into any of those categories), and it does entail implementing measures (the imposition of anti-dumping duties on imports) — which are the object of Fliesen-Zentrum’s challenge before the national court. If Fliesen-Zentrum had attempted to bring an action before the General Court, it would therefore have been dismissed as inadmissible. ( 17 )

44.

In that regard, I do not consider relevant the assertion, made by Fliesen-Zentrum at the hearing, that Fliesen-Zentrum is part of a group of companies owned by members of the same family, another of which (Cera-Net, established in Luxembourg) took part in the anti-dumping proceeding. Such a circumstance (assuming it to be established) cannot confer on Fliesen-Zentrum the standing necessary to challenge the contested regulation directly before the General Court, as a person individually concerned. Indeed, Fliesen-Zentrum has not attempted to bring any such challenge.

45.

The question thus arises whether, when a party with locus standi challenges a regulation imposing an anti-dumping duty directly before the General Court, a stricter duty to state more detailed reasons should entail a higher standard of review whereas, where the Court of Justice is considering a request for a preliminary ruling in national proceedings brought by a party concerned, but not directly or individually, by the same measure the standard of review of the statement of reasons should be lower.

46.

I consider that there is a difference between the two situations and that, in the latter case, the Court of Justice should not treat the measure as a decision but rather as a regulatory act of general application. The higher standard of review to be applied by the General Court where a direct challenge is admissible might lead to annulment only vis-à-vis the applicant in the case before it. ( 18 ) By contrast, a declaration of invalidity by the Court of Justice in a preliminary ruling would be effective erga omnes, and the review of the institutions’ obligation to state reasons should be based on that premiss.

47.

I shall therefore now examine each of the four possible grounds of invalidity raised in the order for reference in turn, in the light of the considerations I have set out above.

48.

Recitals 46 to 52 explained the choice of the USA for the purposes of Article 2(7)(a) of the basic regulation:

‘(46)   In the notice of initiation, the Commission indicated its intention to use the USA as an appropriate analogue country for the purpose of establishing normal value for China, and invited interested parties to comment thereon.

(47)   A number of comments were received and several other countries were proposed to serve as an alternative, in particular Brazil, Turkey, Nigeria, Thailand, and finally Indonesia.

(48)   The Commission therefore decided to seek cooperation from known producers in these countries including the USA. However, only two producers of the product concerned in the USA replied to the questionnaires. A Thai producer also submitted an incomplete reply to the questionnaire; and in any case its product range was not fully comparable to the cooperating Chinese producers.

(49)   The investigation revealed that the USA was a competitive market for the product concerned. Several producers were active on the US domestic market and the import volumes were high. The investigation has further shown that the ceramic tiles originating in China and in the USA have basically the same physical characteristics and uses and that production processes were similar.

(50)   It was argued that since the US market is mainly characterised by imports, the ceramic tiles manufactured in the USA and those manufactured in China cover different segments of the market. Therefore, the domestically produced product types which would serve as a basis to establish normal value would not be comparable to the product types exported by China to the Union. However the investigation has shown that the US production covers a wide range of product types comparable to the ones produced in and exported from China, as mentioned above in recital 49.

(51)   It was also argued that the USA would be a relatively minor player in the worldwide ceramic tiles market. However, circa 600 million m2 [ ( 19 ) ] were produced domestically in 2009 which is considered significant. For comparison, China, the world’s major producer, manufactured 2000 million m2 in the same period.

(52)   One party argued that the USA had strict quality standards and effectively created non-tariff barriers for Chinese imports. However, the investigation revealed that as mentioned above import volumes from China in the USA were high and constituted the major share of the US domestic consumption. Therefore, the argument that non-tariff barriers in the USA affect imports and thus competition was rejected.’

49.

The same issue was dealt with in recitals 55 to 72:

…

…

…

50.

The referring court notes three groups of factors which might, in its view, lead to the conclusion that the choice of the USA was invalid: the US and Chinese tile markets differ significantly, with US manufacturers exporting very little and supplying only a small sector at the upper end of the domestic market, the lower three quarters being supplied essentially by imports; it is not clear that the Commission exhaustively investigated other potential analogue countries (in particular, Brazil, Turkey, Nigeria, Thailand and Indonesia); and there is no indication that it took account of other publicly available statistical information to guide its choice.

51.

Fliesen-Zentrum advances largely the same considerations and cites at length points 79, 97 and 103 to 119 of Advocate General Bot’s Opinion in GLS. ( 20 ) It concludes that the Commission had been given sufficient reasons, and had sufficient information at its disposal, to make it clear that the USA was not an appropriate analogue country for the purposes of its investigations, yet made no sufficient effort to determine an alternative country. That, it considers, was a manifest error of assessment and a breach of superior rules of law, namely Articles 2(7)(a) and 18(5) of the basic regulation.

52.

At the hearing, Fliesen-Zentrum stressed that dumping was not to be equated with social dumping, by which I understood it to mean that a product could not be considered to be dumped, for the purposes of the anti-dumping legislation, where its price was low simply because labour costs were low, and that sales prices of US production sold in the USA were necessarily formed to a significant extent on the basis of labour costs in the USA, which were higher than those in China.

53.

The Council considers that the preamble to both the provisional regulation and the definitive regulation provide sufficient information to dispel the referring court’s doubts.

54.

It stresses that a constructed normal value must be based on verifiable and verified figures from individual producers, not on publicly available macroeconomic statistics. The preamble to the provisional regulation makes it clear that the Commission endeavoured to obtain such figures from producers in all the countries mentioned but obtained responses only from one Thai and two US manufacturers, only that of one of the US manufacturers being usable. The Commission has no means at its disposal to elicit information from producers in third countries not subject to investigation who do not choose to cooperate. It did everything in its power to establish the facts and examine them with all due care, as required by GLS. ( 21 )

55.

In any event, some of the countries suggested (Turkey and the United Arab Emirates, for example) would be inappropriate for various reasons, such as availability of figures or comparability of markets. By contrast, the US market permitted comparability in terms of market, production and imports. Production exceeded the accepted threshold of 5% of exports from China to the Union. In addition, the types of product concerned were closely comparable, as verified by product control number (‘PCN’).

56.

The Commission begins by drawing three guiding principles from the relevant case-law. ( 22 ) First, the Union institutions must consider on their own initiative which market economy third countries present the appropriate characteristics and the Commission must make a sufficient effort to find producers in those countries willing to cooperate. Second, having found such producers, the institutions must select the appropriate country in a not unreasonable manner, taking account of any reliable information available at the time. That choice falls within the institutions’ broad margin of discretion in complex economic assessments, subject to only limited judicial review; in particular, it is not necessary to establish identical conditions in the analogue country in all respects, though comparability of production methods and availability of raw materials must be taken into account. Finally, the institutions are not required to take into consideration all the countries proposed by the parties in an anti-dumping proceeding, but must none the less examine the evidence put before them.

57.

The Commission then deals in turn with the three doubts raised by the referring court, while citing the preambles of the provisional regulation and the definitive regulation as providing, in its view, sufficient justification.

58.

As regards the differences between the Chinese and US markets, the point is not whether such differences exist but whether they cast doubt on the appropriateness of the USA as analogue country. ( 23 ) In this case, the high levels of competition and the wide range of products comparable to those exported from China make it a suitable choice. US producers are competitive and their production considerably exceeds the 5% of Chinese exports to the Union which the Court has held sufficient to establish comparability. ( 24 )

59.

Once the institutions have established that the USA is an appropriate analogue country, it is unnecessary to consider whether other countries might have been even more appropriate. The referring court’s further doubts therefore become irrelevant, but the Commission addresses them none the less.

60.

As regards the investigation of other potential analogue countries, the Commission did endeavour to contact producers in all the countries suggested by the parties to the investigation, in particular Brazil, but was unable to obtain cooperation, even with the help of the lawyer for (one of) those parties. There was no obligation on the Commission to search for further possible countries of its own initiative, on the basis of publicly available data; once it has found a suitable country, it is not obliged to examine other possibilities, at least unless they are proposed by the parties.

61.

As regards the use of publicly available data to determine normal value, the Commission stresses that only company-specific information can be used and that broader statistical data may not accurately reflect actual transactions. Such data may be used only if it is impossible to find a producer willing to cooperate.

62.

As the referring court points out, the Court holds the choice of analogue country to fall within the discretion of the institutions when analysing complex economic situations. The scope of any review by the General Court or the Court of Justice is therefore limited.

63.

However, there has been criticism of the tendency of the Union institutions to favour the USA as analogue country. ( 25 ) That criticism attaches in particular to the high proportion of cases in which the USA has been used as analogue country for China despite the difference in levels of economic development between the two countries ( 26 ) and to the likelihood that only those exporters in market economy countries with higher costs (higher normal value) than non-market economy exporters have an incentive to cooperate, since only their data will lead to higher dumping duties on the competing non-market economy exports. ( 27 )

64.

Criticism of that kind should in my view encourage the Court to verify thoroughly the aspects which are subject to its review. Those aspects are, in particular, whether the institutions took account of all the essential factors for the purpose of establishing the appropriate nature of the country chosen and whether they considered the information contained in the documents in the case with all the care required for the view to be taken that the normal value was determined in an appropriate and not unreasonable manner, bearing in mind that the objective of Article 2(7)(a) of the basic regulation is to find an analogue country where the price for a like product is formed in circumstances which are as similar as possible to those in the country of export. ( 28 )

65.

The structure of that provision is such that the institutions must seek to determine normal value first on the basis of data from a market economy country and second — and only if that is not possible — on any other reasonable basis. However, it is not suggested in the present case that it was impossible to determine normal value on the basis of data from a market economy country. The issue is rather whether the USA was an ‘appropriate’ market economy third country and whether it was selected in a ‘not unreasonable’ manner.

66.

It is clear from the preambles to both regulations, and it has not been challenged in the submissions to the Court, that the Commission did approach producers in Brazil, Turkey, Nigeria, Thailand and Indonesia, in addition to the USA. Only one Thai and two US producers responded, of which only one US producer provided sufficient and reliable data concerning comparable products. It is also clear that the Commission has no means whatever of requiring cooperation in such matters from producers in any third country. In that connection I accept the institutions’ argument that their assessment must be made on the basis of specific, verifiable data relating to actual transactions, rather than publicly available general statistics, which may not provide a true picture. It seems, therefore, that Commission did seriously endeavour to obtain data from a range of market economy third countries, but with very limited success. Was that sufficient?

67.

The institutions argue that, having obtained reliable data (albeit from only a single producer) from one market economy country, the Commission was not required to look any further. However, Fliesen-Zentrum and the referring court mention other third countries (the United Arab Emirates, Egypt, Malaysia and Tunisia) which could have been investigated but were not, and the documents produced to the Court by the Commission itself mention that Russia had also been suggested during the investigation proceeding as a possible analogue country. Furthermore, it is clear from recital 144 in the provisional regulation that the Commission was aware that ‘it is possible for importers and users to switch to products sourced from third countries … since the product under investigation is manufactured in several countries … (… United Arab Emirates, Egypt, … countries of South-East Asia, and others)’.

68.

In addition, it must be considered whether any difference in the level of economic development between China and the USA means that the latter is not an ‘appropriate’ third country for establishing normal value or that it was chosen in a manner which was ‘unreasonable’, given the need to find an analogue country where the price for a like product is formed in circumstances which are as similar as possible to those in the country of export.

69.

As regards the latter aspect, the Commission made a number of points at the hearing. It acknowledged that such a criterion had recently been accepted in the anti-dumping legislation of a number of countries (it cited India, South Africa and the USA) but not (yet) in the European Union, where the criteria are still those originally based on what were at the time (the Commission referred to GATT interpretative notes dating from 1955) those adopted by the USA. Those criteria place the emphasis not on comparable levels of economic development but on the need to use prices formed on a market in which a high level of competition is likely to exert a downward pressure. Although there is currently discussion between the Council and the Parliament as to whether a criterion of economic development should be introduced, that change has not yet taken place, and the institutions are bound by the law as it stands, not as it may stand in the future. The Commission also pointed out that a lower level of economic development, while it might in general imply lower labour costs keeping prices down, also usually correlated to lesser efficiency and a lower level of technological development having the contrary effect; consequently, it could not be concluded that data from a country with a higher level of economic development would necessarily provide a higher normal value than data from a country with a lower level of development. With specific regard to Russia, the Commission explained that, on the one hand, the country had been suggested too late in the investigation proceeding to allow data collection within the time-limit set and, on the other hand, the system of dual pricing for energy products operated there as between domestic consumption and export markets would have made comparisons more difficult.

70.

Against that background, it must be recognised, first, that there is currently no legal obligation on the Union institutions to take specific account of the level of economic development of a potential analogue country as compared to that of the exporting country under investigation. None the less, their choice of analogue country must be appropriate and not unreasonable. That means, in my view, that they have a duty to consider whether differing levels of economic development should be taken into account, but I accept that the consideration itself falls within the scope of their discretion when analysing complex economic situations.

71.

Next, it seems to me that the Commission’s explanations as to the roles of competition and technology in determining normal value are credible. It is almost certainly the case that a high level of competition will exert a downward force on prices, so that data from a highly competitive market will not necessarily lead to a higher dumping margin than data from a country with lower costs but also lower competition. ( 29 ) Likewise, it must be acknowledged that a higher level of technological development is likely to offset lower labour costs, so that higher labour costs cannot necessarily be taken as an indication of higher prices and higher normal value. ( 30 ) Clearly, these are matters which fall within the notion of analysis of complex economic situations.

72.

Finally, the specific reasons given by the Commission for not choosing Russia as an analogue country appear only partly convincing. The notice of initiation was published on 19 June 2010 and, in its response dated 6 August 2010, the CET suggested that Russia might be an appropriate analogue country, although only as an alternative to the USA which it favoured. It is not obvious how time constraints could have prevented extending inquiries to that country, given the fact that the provisional regulation was not adopted until 16 March 2011 and that the Commission was able to send questionnaires to producers in other countries suggested in response to the notice of initiation (the definitive regulation was adopted on 12 September 2011, within the 15-month deadline imposed by Article 6(9) of the basic regulation). However, the difficulties of comparing data from a country with a dual price system are no doubt real, and the Commission could justifiably consider that it was unnecessary to extend the investigation to a country suggested by the complainant as an alternative when it already covered the country considered by that complainant to be the most appropriate. Although it might have been preferable to mention the reasons for not sending questionnaires to Russian producers, that omission cannot, in my view, call the validity of either regulation into question.

73.

Those considerations lead me to the view that the choice of the USA as analogue country was not (or at least has not been shown to be) clearly inappropriate or unreasonable, in the absence of usable data from any other third country demonstrated to be more appropriate. Since normal value is to be determined ‘on any other reasonable basis, including the price actually paid or payable in the [Union] for the like product’ only where the use of data from a market economy third country is not possible, it therefore seems unnecessary to examine whether recourse to prices paid or payable in the Union might have been more appropriate.

74.

None the less, for the sake of thoroughness, I would make the following remarks. The dumping margins found in the definitive regulation ranged from 26.3% to 69.7% (recitals 88 to 93). The export prices and normal value from which those dumping margins were calculated are not in fact stated in either the provisional regulation or the definitive regulation but, in accordance with Article 2(12) of the basic regulation, the dumping margin is the amount by which the normal value exceeds the export price. In the provisional regulation, as confirmed by the definitive regulation, the average price of imports from China into the Union is stated to be 4.5 EUR/m2 during the investigation period, and the average price of Union sales within the Union to be 8.8 EUR/m2. In my Opinion in Case C‑569/13 Bricmate, also delivered today, I have calculated the correct figure for the average price of imports from China into the Union (taking into account a significant statistical error concerning the volume of imports into Spain in November 2009) to be 5.1 EUR/m2. Even on the latter basis, the amount by which prices for Union sales within the Union exceeded prices of imports from China would be, according to my calculation, 72.5% — higher than any of the dumping margins found on the basis of data from the USA. Consequently, if normal value had been taken to be the average price of Union sales within the Union, dumping margins, and thus anti-dumping duties, would have been higher than those actually found and imposed on the basis of the US data.

75.

I therefore reach the conclusion that the Union institutions committed no error in selecting the USA as analogue country, in the exercise of their discretion in analysing complex economic situations.

76.

With regard to the use of data from a single US producer, recital 53 stated:

‘The data submitted in their reply by the two cooperating US producers were verified on [the] spot. Only data from one producer visited was finally considered, as it was found to be reliable information on which a normal value could be based. The data from the second producer visited were found not to be reliable and had to be discarded, as this producer only reported part of its domestic sales and costs could not be fully reconciled with the accounts.’

77.

In the definitive regulation, the same issue was dealt with in recitals 61, 62, 64, 66, 69 and 74 to 77:

…

…

…

…

78.

The referring court notes that various provisions of the basic regulation, ( 32 ) by their use of plural forms, indicate that normal value — a crucial element in the determination of dumping — must be established on the basis of data from several producers. If that is true for market economy countries, it must apply also where non-market economy countries are concerned, as is confirmed by the fact that Article 2(7)(a) provides for reference to prices in an analogue country, not prices of an analogue producer. Objectivity and accuracy are considerably more difficult to achieve when reference is made only to a single producer and great care must be exercised, particularly when, as in the present case, there is a suggestion that the US producer in question is controlled by a Union producer and therefore not economically independent. In any event, data from a single producer will be conditioned by company-specific policy decisions. The referring court notes also that the Commission did not find it necessary to have recourse to any external expertise in this regard.

79.

Fliesen-Zentrum cites at length points 81 to 87 of the Opinion in GLS, ( 33 ) on the importance of the determination of normal value and the need for objectivity in that regard. Its arguments reflect in general the considerations expressed by the referring court, and it stresses that a failure to apply the same criterion (plurality of exporters or producers for the purposes of comparison) in the case of market and non-market economies would infringe the principle of equal treatment.

80.

The Council stresses that the Court has accepted that the fact that there is only one producer in a country does not preclude its use as analogue country. ( 34 ) The use of plural forms in the basic regulation must be seen as a general and indeterminate indication, not as necessarily requiring data from a plurality of exporters or producers; indeed some provisions alternate between singular and plural forms. There is, moreover, no need to refer to several producers if figures from a single producer are representative. The institutions’ practice has always been to contact all producers in the country in question, and to use the figures from all the answers received, whatever their number and even if only a single answer is forthcoming. Protection of confidential data from the producer is essential in order to ensure cooperation. There is no factual evidence to support the claim that the US producer’s figures might have been influenced by a Union parent company if such exists but, in any event, the Commission’s practice is to use only figures relating to sales to independent third parties.

81.

The Commission accepts that normal value can be more reliably constructed with figures from a number of producers in the analogue country. However, that is not always possible, and the Union institutions have no means of requiring cooperation. Producers in a third country not concerned by the investigation have no particular incentive to cooperate, unlike other producers in a country subject to investigation, who have every reason to provide their own figures if they are not to be assessed themselves, pursuant to Article 18(1) of the basic regulation, on the basis of the facts available.

82.

In the present case only one US producer provided usable figures. The Court accepts that the fact that there is only one producer in a third country does not preclude using that producer’s data for constructing normal value; the same should apply when there are several producers but only one cooperates. What matters is whether the prices in question are the result of competitive forces. The provisions cited by the referring court, which use plural forms, reflect the ideal situation, but not necessarily what is actually possible. Moreover, those provisions refer to possible alternative methods, as is clear from the auxiliary ‘may’; Article 2(6)(b) provides for use of figures for ‘the exporter or producer in question’, in the singular.

83.

Nor is there any question of inequality of treatment; determination of normal value in the country under investigation and in an analogue third country are two completely different situations, in which the same methods are not available.

84.

The danger that the prices from a single producer might be unduly influenced by that producer’s individual situation is minimal, whether the producer in question is controlled by a Union producer or not, ( 35 ) because the prices used must relate to a large number of transactions in a competitive marketplace and any prices for intra-group transactions must be adjusted accordingly.

85.

Finally, the fact that the Commission had no recourse to external expertise is irrelevant; the Commission never uses outside experts in such investigations, because of the confidential nature of the data.

86.

If it were to be decided that the choice of the USA as analogue country was vitiated to such an extent that the contested regulation should be annulled on that ground, it would be unnecessary to examine this issue.

87.

However, if the Court agrees with me that that is not the case, there are essentially three points to be considered: first, can a normal price be determined on the basis of data from a single producer; second, if so, can that still be the case even if the producer in question is linked to a Union producer; third, should the Commission have called in outside expertise?

88.

To take the third — and, in my view, simplest — point first, it seems to me that there are no cogent grounds for supposing that outside expertise should have been called in. Fliesen-Zentrum and the referring court appear to have attached what is, in my view, unwarranted importance to a standard subheading to be found in the explanatory memorandum to any Commission proposal for a Council regulation. In the absence of some specific reason to suppose otherwise (and no indication of any such reason has been given or is apparent in the present case), the institutions should in principle be regarded as capable of assessing the data adequately themselves.

89.

It is agreed by all the parties that it is always preferable and more reliable to base a determination of normal value on data from a broad range of producers. However, in Rotexchemie, the Court held that ‘the mere fact that there is only one producer in the reference country does not in itself preclude the prices there from being the result of genuine competition, since such competition may just as well result, in the absence of price controls, from the presence of significant imports from other countries’. ( 36 )

90.

The situation is to a large extent comparable less to that in Rotexchemie, where there was only one producer in the analogue country, as opposed to only one which cooperated usefully while others did not, than to that in Ferchimex, where the General Court stated: ‘… Potacan was the only Canadian producer which finally agreed to reply to the Commission’s questionnaire and to cooperate in the proceeding. In particular, the largest Canadian producer … refused to cooperate and merely provided certain information within the public domain which was inadequate for the purposes of determining normal value. … [T]he applicant does not deny that the Commission made every effort to obtain information relating to the Canadian market from sources other than Potacan. … Canada must therefore be regarded as an appropriate reference country, and it must be concluded that the Commission had no alternative but to use the information emanating from Potacan.’ ( 37 )

91.

The latter judgment, while not of course binding on this Court, was not appealed against and may be considered to have at least some persuasive authority.

92.

In the present case, it must be considered what alternatives might have been available to the Union institutions. I have already reached the view that there was little real likelihood that an obviously more appropriate analogue country, with a plurality of producers willing to cooperate in the investigation, could have been found. ( 38 ) Could more producers have been interrogated in the USA? From recital 48 in the provisional regulation, it may be inferred that more than two US producers were contacted, because it is stated that ‘only two producers of the product concerned in the USA replied to the questionnaires’, and there is no suggestion in the case-file that the number of producers contacted was insufficient. If no more appropriate analogue country could have been found, and no more producers in that country could be induced to reply to the questionnaires, then it seems that the only options were either to base the determination of normal value on the data supplied by the only US producer which did respond adequately or to take ‘any other reasonable basis’, of which none but the price paid or payable in the Union would have applied. As regards the second of those options, I have calculated above ( 39 ) that it might well have led to higher dumping margins.

93.

Finally, as regards the second point, the Court is faced with a lack of information. The institutions decline to say whether the US producer in question is or is not economically linked to a Union (allegedly Italian) producer and, while it appears to be suggested that Fliesen-Zentrum and/or the importers who took part in the investigation could identify the producer in question, no actual identification has been made in the papers or submissions before the Court.

94.

The Commission has offered to provide certain information (though not explicitly this particular information) to the Court, provided that it is treated confidentially vis-à-vis Fliesen-Zentrum. However, there does not appear to be any provision currently in the Rules of Procedure which would enable the Court to take such a course. ( 40 ) Confidential treatment of documents is referred to only vis-à-vis interveners in direct actions and appeals (Articles 131(2) and 190(1)). The Commission suggested at the hearing that this was an action for annulment cloaked as a preliminary reference and that, failing an explicit provision in its Rules of Procedure, the Court could apply an ad hoc interpretation of its judgment in Kadi II ( 41 ) in the light also of Article 339 TFEU (which imposes a duty of secrecy on members of institutions and officials of the Union) and Article 19(4) of the basic regulation.

95.

However, in the absence (for the time being) of any explicit provision in the Rules of Procedure allowing confidential treatment of information in the context of a preliminary ruling procedure, I do not consider that that would be a proper approach.

96.

If the information cannot be made available confidentially, it seems to me that the first step must be to take as a presumption that the US producer was indeed economically linked to a Union producer and to consider whether that fact could invalidate the determination of normal value. If not, there is no need to consider further whether access to confidential information is required.

97.

The institutions say that care was taken, as always, to use only data relating to sales to unrelated customers. In that regard, it seems to me unnecessary to require disclosure of information which the institutions would wish to keep confidential vis-à-vis Fliesen-Zentrum. Given the latter’s argument that the data used led to an artificially high normal price, it is implausible that sales to linked customers were taken into account. If anything, that might have been likely to bring down the normal value calculation, since it would seem more logical, in normal circumstances, for linked customers to benefit from favourable terms.

98.

A second question is: is it likely that the production costs of, or the prices charged by, a US subsidiary of a Union producer would be artificially inflated as compared to the costs or prices of an independent US producer? In fact, that seems rather unlikely, if the aim is to run a successful commercial operation in a competitive market (and, unless there is evidence to the contrary, that may be taken to be the US producer’s aim). The Commission verified the producer’s data on the spot, and it appears that the verification was thorough, since data from the other producer were discarded as unreliable. ( 42 ) Given the importance attached by the institutions to obtaining data from a highly competitive market, it seems unlikely that a non-competitive marketing strategy would have escaped such a level of scrutiny.

99.

In the present case, no specific reason has been advanced before the Court for supposing that a US producer linked to a Union producer (assuming that link to be established) would be likely to sell at higher prices on the US market than a producer without any such link.

100.

In those circumstances, it seems to me that the Court would not be justified in finding the contested regulation invalid on the ground that normal value was calculated on the basis of data from a single producer, regardless of whether that producer was linked in any way to a Union producer.

101.

The way in which the normal value was constructed on the basis of the data from the US producer is described in recitals 56 to 58 and 61:

…

102.

The calculation of constructed normal value was referred to in recitals 73, 77, 86 and 87:

…

…

103.

The referring court notes that, when constructing normal value from the US producer’s prices, adjustments were made to allow for differences affecting quality, marketing characteristics and cost discrepancies. In such circumstances, ‘it is incumbent upon the institutions, where they consider that they must make an adjustment …, to base their decision on direct evidence or at least on circumstantial evidence pointing to the existence of the factors for which the adjustment was made, and to determine its effect on price comparability’. ( 45 ) The reason for the adjustments must be demonstrable and the adjustments themselves must therefore be verifiable, but that is not possible here from the information in the provisional or the definitive regulation, or made available during the investigation. Even if it is necessary to protect the confidentiality of the data from the US producer, Article 296 TFEU requires adequate reasons to be stated for any legal act. Given that the establishment of normal value is an essential step in proving the existence of any dumping, the statement of reasons for an anti-dumping measure is of particular importance, and the rights of defence provided for in Articles 20 and 21(6) of the basic regulation are set at nought if interested parties have only vague information on how exactly the normal value was established and thus cannot submit factually well-founded observations.

104.

Fliesen-Zentrum’s submissions are again largely the same as the considerations expressed by the referring court, stressing the vagueness and unreliability of the data as expressed in the provisional and definitive regulations.

105.

The Council stresses the need for confidentiality vis-à-vis cooperating producers in analogue third countries. To disclose any of the details which Fliesen-Zentrum claims should have been disclosed would be a flagrant violation of that confidentiality. Moreover, the adaptations required during the determination of normal value fall within the institutions’ broad margin of appreciation in complex economic matters, and no grounds have been put forward to show that they abused their power of assessment. Finally, the adjustments made all favoured the producers whose products were subjected to an anti-dumping duty; where there are no adverse effects, the obligation to state reasons is less strict, and it was thus respected in this case.

106.

The Commission submits, first, that the statement of reasons in the definitive regulation meets the requirements in the case-law and, second, that Fliesen-Zentrum does not enjoy any ‘rights of defence’ but that, in any event, no such rights have been infringed.

107.

With regard to the statement of reasons, it cites the general requirement as expressed in Petrotub ( 46 ) before pointing out that in Beus ( 47 ) the Court found that the preamble to a regulation intended to have general application ‘may be confined to indicating the general situation which led to its adoption, on the one hand, and the general objectives which it is intended to achieve on the other. Consequently, it is not possible to require that it should set out the various facts, which are often very numerous and complex, on the basis of which the regulation was adopted, or a fortiori that it should provide a more or less complete evaluation of those facts’. In Petrotub ( 48 ) the Court found that, by merely stating that it had been ‘found that sales made using compensation were indeed made in the ordinary course of trade’, the Council had not provided an adequate statement of reasons; it had not given reasons for considering that prices charged in connection with those sales had not been affected by the relationship between the parties; it had not enabled those concerned to know whether those prices were correctly taken into consideration when calculating normal value, or whether there might be a flaw affecting the legality of the contested regulation; it had also prevented the Court from exercising its power of review and from determining whether there might have been a manifest error of assessment. However, the Commission submits that the obligation to state reasons must also take account of the obligation to respect confidentiality of business secrets as laid down in Article 339 TFEU and Article 19 of the basic regulation. In particular Article 19(4) of the latter distinguishes between disclosure of general information, which is not precluded, and disclosure of specific evidence, which is closely circumscribed. Finally, it is necessary to assess the statement of reasons in its context, which, in the present case, includes all the contents of the non-confidential file, disclosed to the parties.

108.

The Commission submits that the requirements set out above were all met in the present case, in particular by recitals 56 to 58 and 61 in the preamble to the provisional regulation, to which the definitive regulation referred. The mere fact that no specific or approximate figures are provided does not fail to meet the requirement to state reasons; the information is sufficient for the Court to request to see any confidential data necessary to verify the reasoning. Nor is the question whether the US producer is or is not linked to a Union producer in any way relevant for the validity of the definitive regulation, since the data used concerned prices for sales to unconnected parties. The figures were verified from the US producer’s records and adjusted as appropriate.

109.

With regard to the rights of defence, the Commission points out first of all that Fliesen-Zentrum took no part in the procedure which led to the adoption of the definitive regulation. It can therefore enjoy no rights of defence in that connection. Nor could it derive any such rights from the fact that it claims to be ‘related’ to Cera-Net Sàrl, a Luxembourgish company which did take part in the procedure, unless (perhaps) it could demonstrate that it actually controlled that company. In any event, Cera-Net itself enjoyed no such rights, since anti-dumping proceedings are directed only against third-country producers or exporters and, where applicable, their related importers, not against independent importers such as Cera-Net.

110.

In the alternative, the Commission submits that there was in any event no infringement of Cera-Net’s rights of defence. In Timex, ( 49 ) the Court considered that the traders or manufacturers concerned should be able to inspect the information gathered by the Commission during the investigation so that they could effectively put forward their points of view, although that right must where necessary be reconciled with the principle of confidentiality; the Commission should make every effort, as far as is compatible with the obligation not to disclose business secrets, to provide the applicant with information relevant to the defence of its interests, choosing, if necessary on its own initiative, the appropriate means of providing such information. In the present case, Cera-Net was provided with some of the information requested, and the hearing officer verified that the prices used had been adjusted where necessary. Actual disclosure would have been a breach of confidentiality, maintenance of which is essential in order to secure cooperation from third-country producers.

111.

This aspect of the case raises — like the previous aspect — an issue of confidentiality of information. Here, however, that issue must be considered first of all in the light of Fliesen-Zentrum’s procedural position vis-à-vis the anti-dumping investigation.

112.

It seems to me clear that Fliesen-Zentrum, not being linked to any of the Chinese exporters and not having participated in the investigation procedure, cannot itself claim any individual rights of defence, and that the Commission has put forward cogent reasons for concluding that it cannot derive any such rights from parties which did take part in the procedure.

113.

The question is therefore, in the context of the main proceedings and of the present request for a preliminary ruling, whether the statement of reasons was adequate for a regulatory act of general application. In other words, was it sufficient for those concerned (other than directly and individually) to understand the reasoning followed by the institutions and for the Court to exercise its power of review? For that purpose, was it necessary to set out details of the adjustments made to allow for differences affecting quality, marketing characteristics and cost discrepancies?

114.

All the information available to the Court, or to parties concerned in general by the anti-dumping duties, is contained in recital 61 in the preamble to the provisional regulation, to which recitals 77, 86 and 87 in the preamble to the definitive regulation refer, and reads as follows: ‘The normal value was adjusted for differences in characteristics — mainly due to OEM branding and for quality differences for certain types not produced by the analogue country producer — for the lower cost of non-porcelain tiles. Further adjustments were made, where appropriate, in respect of ocean freight, insurance, handling and ancillary costs, packing, credit, bank charges and commissions in all cases where they were found to be reasonable, accurate and supported by verified evidence.’

115.

It seems to me that such a statement must in principle constitute a sufficient explanation for an act of general application. It indicates the type of adjustment made and, implicitly at least, the direction in which the figures were adjusted — namely, downwards.

116.

Certainly, without actual figures, it is not possible for a party only indirectly concerned (such as an importer in Fliesen-Zentrum’s position) or for the Court to check the precise calculations. And, vis-à-vis Fliesen-Zentrum (and the ceramic tiles trade at large), it seems reasonable to accept that these figures should remain confidential.

117.

If there were particular evidence that the adjustments were not carried out correctly, the Court might envisage whether to ask for the figures to be produced and to examine them. Again, as the Rules of Procedure currently stand, I would counsel caution in that regard.

118.

However, there is no such specific evidence and, in its absence and in the light of Fliesen-Zentrum’s status vis-à-vis the anti-dumping procedure as a whole, there does not appear to be any reason for the Court to require further details of the adjustments. If Fliesen-Zentrum, like any other indirectly concerned party, must in principle be content with a statement of reasons which satisfies the requirements for a measure of general application, the fact that it brings proceeding before a national court which give rise to a request for a preliminary ruling by the Court of Justice cannot give it a right to demand that the Court examine, in that context, information which might have been available to the General Court in the different context of a direct action brought by a qualified party directly concerned by the anti-dumping measure.

119.

Consequently, while there may remain areas of uncertainty concerning the detailed accuracy of the calculation of normal value from the figures of the US producer in question, which might have been relevant in the context of a direct action brought by a qualified party before the General Court, it does not seem to me that those areas are such as to justify this Court in finding, in the specific procedural context in which the question is raised, that the contested regulation is invalid.

120.

Sampling of Chinese and Union producers in accordance with Article 17 of the basic regulation was explained in recitals 4 to 13 in the preamble. Recital 4 stated that all known exporting producers in China, importers and Union producers were asked to make themselves known to the Commission and to provide basic information on their activities related to the product concerned during the investigation period. The preamble continued:

‘2.1. Sampling of Chinese exporting producers

2.2. Sampling of Union producers

121.

Sampling was dealt with in recitals 9 to 33. Recitals 9 to 12 dismissed an objection concerning two Chinese exporting producers, which is of no relevance to the present case. Recital 13 concluded: ‘In the absence of any other comments, recitals 5 and 6 of the provisional regulation are hereby confirmed.’

122.

As regards issues relevant to the present case, recitals 16 to 33 stated:

...

123.

Recitals 125 and 126 are also relevant with regard to this issue:

124.

The referring court notes that, in order to determine the level of undercutting, the Commission used sampling, as permitted by Article 17(1) of the basic regulation ‘where the number of complainants, exporters or importers, types of product or transactions is large’. Two types of sampling may be used: ‘a reasonable number of parties, products or transactions by using samples which are statistically valid on the basis of information available at the time of the selection’ or ‘the largest representative volume of production, sales or exports which can reasonably be investigated within the time available’. It is clear that the first type was used in order to determine prices of Union producers, and the second type to determine those of Chinese producers. The referring court queries whether such an approach is valid, since it led to sampling essentially from a large number of smaller Union producers (representing only 7% of Union production) and a small number of larger Chinese producers (14.4% of Chinese production), even though both markets are highly fragmented and composed largely of SMEs. The samples might thus not be comparable, which would cast doubt on the objectivity of the determination. In addition, the Union producers sampled were all from higher-priced western Member States, with none from lower-priced Member States. It is therefore not clear that the levels of undercutting found (43.2% to 55.7% according to recital 113 in the definitive regulation) were accurately and objectively determined.

125.

Fliesen-Zentrum puts forward again largely the same considerations, though in greater detail, insisting on the need for the two samples to be constituted in a comparable manner. Moreover, the representativity of the Union sample should cover at least 25% of Union producers, in line with the minimum requirement for initiating an investigation. ( 50 ) The way in which the Commission failed to compare like with like infringes the principle of equal treatment, the requirement of objectivity in Article 3(2) of the basic regulation and the lesser duty rule in Article 9(4) of the same regulation. In addition, the use of samples representing only 7% and 14.4% of the respective markets does not meet the requisite degree of representativity.

126.

The Council submits first that the injury to the Union industry was evaluated on the basis of macro- and microeconomic data from, respectively, the whole industry and the Union producers sampled; it could not be affected by the composition of the Chinese sample. Second, the Union sample took account of the fragmented nature of the industry. Polish or Czech producers were not included because a Polish producer ceased to cooperate and Czech production is less that 3% of Union production, but the Commission ran a simulation including Polish prices and found it made little difference. Third, Fliesen-Zentrum made no observations during the procedure concerning the composition of the Chinese sample, and may not do so now in a preliminary ruling procedure. In any event, the use of a different method for the Chinese sample falls within the institutions’ discretion and, in this instance provided maximum representativity for each group. Finally, the anti-dumping duties imposed were, in any event lower than the dumping margins found.

127.

The Commission states that it is wrong to suppose that the purpose of the two samples is simply to establish the level of undercutting. The Union sample serves also to collect a wide range of microeconomic data which should cover all categories of Union producers in order to detect any variation according to geography or size, whereas the sample of exporters aims to account for as great a share as possible of Union imports in order to ensure that prices are as close as possible to the real average. The use of two methods is therefore fully justified, does not infringe the principle of equal treatment and does not call in question the objectivity of the analysis. As regards the absence of eastern European producers in the Union sample, the institutions have no means of requiring cooperation and the major Polish producers did not cooperate. In any event, western Europe accounts for 80% of Union production, and publicly available data from Poland and the Czech Republic showed that its inclusion would have made little difference.

128.

As regards the possibility that a lower duty might have been imposed if the undercutting margin had been established on a different basis, the Commission states that the contested regulation was drawn up in English and that the term ‘underselling margin’ (the difference between dumped prices and target prices for the Union industry — ‘Zielunterbietungsspanne’ in German and ‘marge de sous-cotation des prix indicatifs’ in French, equivalent to ‘injury margin’) was mistranslated ( 51 ) into German (and French) as if it had been ‘undercutting margin’ (the difference between dumped prices and actual prices of the Union industry, already depressed by the effects of the dumping — ‘Preisunterbietungsspanne’ in German and ‘marge de sous-cotation des prix’ in French).

129.

The principal issue here is whether, when the institutions found ’significant price undercutting by the dumped imports as compared with the price of a like product of the [Union] industry’ (Article 3(3) of the basic regulation), the comparison they had carried out was invalidated because one of the two methods of sampling permitted by Article 17(1) of the basic regulation was used to evaluate Chinese prices and the other was used to evaluate Union prices. Secondary issues are whether it was necessary to include, in the Union sample, one or more producers from reputedly lower-priced eastern European Member States, and whether the Union sample was sufficiently representative. Then, if any of those factors led to an overestimation of the level of price undercutting, there arises the question whether that fact should have led to the application of the lesser duty rule in Article 9(4) of the basic regulation.

130.

To take the secondary issues first, I find the reasons given for the absence of Polish and Czech producers to be reasonable and coherent, both in the provisional and definitive regulations and in the institutions’ observations. The choice made as to the selection of producers falls, it seems to me, well within the broad discretion available to the institutions when assessing a complex economic situation, and they cannot be blamed for the Polish producer’s withdrawal of cooperation during the course of the investigation, or for the lack of cooperation from other producers. Moreover, checks were carried out to verify that the results were not significantly affected by the absence of data from eastern European (specifically, Polish) producers.

131.

As regards representativity, Fliesen-Zentrum’s claim that the Union sample must include at least 25% of Union production is in my view unfounded. Article 5(4) of the basic regulation does not concern the representativity of samples, and none of the cases it cites supports its contention.

132.

Turning now to the principal issue, I note first of all that Article 17(1) of the basic regulation contains no explicit requirement that different samples should be constituted on an exactly comparable basis. Nor, it seems to me, can it be read as containing an implicit requirement of that kind, covering as it does ‘cases where the number of complainants, exporters or importers, types of product or transactions is large’ and limitation to ‘a reasonable number of parties, products or transactions’. The categories involved cannot themselves always be comparable, so there will often be no scope for comparability of sampling methods.

133.

Moreover, it is clear from Article 17(2) of the basic regulation that the final selection of parties, types of products or transactions rests with the Commission. The institutions have provided, in my view, adequate reasons for proceeding in the way they did. As regards the prices of Chinese exports, the paramount concern was to obtain the best possible approximation to an accurate average price, and it was logical for that purpose to select exporters who accounted for the greatest volume of sales. Sampling of the Union industry, however, had to be suitable for a wider variety of purposes and therefore could not be confined to the largest producers.

134.

That said, it is true that the determination of injury must involve an ‘objective examination’ of the volume of dumped imports, their effect on prices in the Union market and their impact on the Union industry (Article 3(2) of the basic regulation) and that the samples must be ’statistically valid’ (Article 17(1)).

135.

In that regard, it seems admittedly at least plausible that the method used for China (largest volume, and thus largest producers) could have produced a lower price than if a larger number of smaller producers had been included in the sample; larger producers may be in a position to benefit from economies of scale and thus to sell at lower prices. Conversely, and by the same token, the method used for the Union might have led to a higher price. In that event the level of price undercutting found might have been greater than if both samples had been selected according to the same method. (The same would apply to underselling, which was based on data from the same sources, so that the Commission’s submissions concerning terminology and translation in that regard do not appear particularly relevant.)

136.

However, those possibilities are by no means certainties, and no detailed figures or calculations have been produced to the Court to render them specifically credible. Another equally plausible possibility is that Chinese export prices were determined more by competitive conditions on the importing market than by the size or capacity of the producer, so that a different method of sampling the Chinese exporters would not have changed the result. ( 52 ) Without substantiation, it is therefore not possible to assume that the level of price undercutting or underselling would have been significantly different if the same method had been used for both samples.

137.

Consequently, I see no reason to consider that the assessment of price undercutting in the contested regulation was in any way invalidated by the choice of two different sampling methods, or that there was any incompatibility with the lesser duty rule in Article 9(4) of the basic regulation.

138.

In the light of all the above considerations, I am of the opinion that the Court should answer the Finanzgericht München to the effect that Council Implementing Regulation (EU) No 917/2011 of 12 September 2011 imposing a definitive anti-dumping duty and collecting definitively the provisional duty imposed on imports of ceramic tiles originating in the People’s Republic of China is not invalid on any of the grounds raised in its request for a preliminary ruling.