OPINION OF MR ADVOCATE GENERAL

VERLOREN VAN THEMAAT

DELIVERED ON 21 JUNE 1983 ( 1 )

Mr President,

Members of the Court,

1. Introduction

1.1.

All the relevant facts in Case 322/81 are clearly summarized in the Report for the Hearing. In the introduction to my Opinion in this case I can therefore simply draw attention to one or two main points and for the many complicated details refer to the Report for the Hearing.

The application lodged by NV Nederlandsche Banden-Industrie Michelin (hereinafter referred to as “Michelin NV”), supported by the French Republic, is for a declaration that a Commission decision of 7 October 1981 is void. Article 1 of that decision declares that during the period 1975 to 1980 Michelin NV infringed Article 86 of the EEC Treaty on the market in new replacement tyres for lorries, buses and similar vehicles by:

(a)

binding tyre dealers in the Netherlands to itself through the grant of selective discounts on an individual basis conditional upon sales “targets” and discount percentages, which were not clearly confirmed in writing, and by applying to them dissimilar conditions in respect of equivalent transactions; and

(b)	granting an extra annual bonus in 1977 on purchases of tyres for lorries, buses and the like and on purchases of car tyres, which was conditional upon attainment of a “target” in respect of car tyre purchases.

Article 2 of the decision imposes on Michelin NV a fine of 680000 ECU or HFL 1833184.80.

The submissions advanced in support of the application for a declaration that the decision is void relate, on the one hand, to the facts and, on the other hand, to the legal evaluation of those facts. I shall deal with the submissions in the order in -which they are summarized in the Report for the Hearing. I shall therefore first deal with the submissions on the relevant market, then with those concerning the existence of a dominant position “within the common market or in a substantial part of it”, thirdly the submissions concerning the two abuses of that dominant position found in the decision, fourthly the effect of the conduct in question on trade between Member States, fifthly the alleged breach of the rights of the defence and sixthly the contested imposition of a fine.

1.2. General remarks

Before I consider in turn each of the submissions advanced it would be as well, I think, to point out that this separate examination of them must not obscure the link between the various elements of the prohibition contained in Article 86 of the EEC Treaty. Nor should the nature of that link be forgotten.

First of all, in this case that link is of some relevance to the relationship between the nature of the alleged abuse and the evidence of the existence of a dominant position. As is clear from the last sentence of paragraph 35 and the second subparagraph of paragraph 48 of the contested decision, the Commission considers the existence of a dominant position partly proven by the fact that Michelin NV adopted a course of conduct which an undertaking not having a dominant position could not allow itself to do. In principle I consider that such an argument is in fact admissible and sometimes of practical value. For instance, an undertaking's elimination of its competitors through aggressive price competition consisting in the selling of goods at a loss might in certain circumstances constitute evidence of that undertaking's dominant (financial) position. Whether the link in question is relevant in this case too is a matter to which I shall return.

A second important point which should be made in this case is that it follows from Article 86 that the prohibited abuse of a dominant position may well relate to a product market on which the undertaking concerned precisely does not have a dominant position. This is clear in particular from the prohibition in Article 86 (d) of “linked sales”. This point is important in the present case because the alleged abuses relate partly to light tyres (for cars and delivery vans) in respect of which the Commission found that Michelin NV had no dominant position. Finally, for an appreciation of the parties' long exchanges on the question of what is the relevant “product market” in this case, it is important to remember that Article 86 requires the relevant geographical market to be determined but not the relevant product market. Some general observations on the relative value of the latter concept for the application of Article 86 therefore seem necessary.

The concept of the relevant product market is an idea which has been developed especially in American antitrust case-law, not in connection with the concept of a dominant position but mainly in connection with Section 2 of the Sherman Act and Section 7 of the Clayton Act which prohibit the monopolization or attempted monopolization of a section of trade and also mergers leading to a substantial lessening of competition or the danger of the formation of a monopoly “in any line of commerce in any section of the country”. Owing to the nature of the danger to be averted, it is indeed essential in order to establish whether there is a danger of monopoly to define the product market in question. Then the relevant product market is of course an element of the conduct to be resisted. In the application of Article 86 to mergers the concept may fulfil a similar function by enabling a risk of monopoly to be identified. In this regard I refer to the Court's judgment in Case 6/72 Continental Can [1973] ECR 215.

In the EEC Treaty only Article 85 (3) (b), which also mentions the possibility of eliminating competition “in respect of a substantial pan of the products in question”, expressly requires the relevant product market to be determined. There too, therefore, the requirement does not concern the question whether the undertakings concerned have a dominant position but whether their conduct on the market tends to monopolization.

For the purposes of Article 86, however, determination of the relevant product market is not expressly required. Even so, its determination may, as I have said, be implicitly required by the nature of the alleged abuse, particularly where there is a danger of monopoly. The concept is in that respect an element of the concept of “abuse”. In addition, the concept may be a useful tool for determining the existence of a dominant position of one or more undertakings. In that function, however, it does not have the character of an independent element of prohibition which must be established every time Article 86 is applied. Rather, it follows from the definition of “dominant position” given in the Court's case-law from which I shall cite presently that the existence of a dominant position can in principle be proven otherwise than by defining the relevant product market. A comparative legal analysis also shows that the concept of the relevant product market is important only in a few comparable bodies of legislation on restrictive agreements and practices and even then, as I have said, not always for determining the existence of a dominant position.

Since Article 86 does not prohibit a dominant position per se but only an abuse of a dominant position, the concept of the relevant product market is also indirectly dependent on the alleged abuse. The link between the two emerges plainly in the general definitions given by the Court in Hoffmann-La Roche (Case 85/76 [1979] ECR 461) of the concepts of dominant position (last subparagraph of paragraph 38) and abuse (last subparagraph of paragraph 91). It might even be deduced from the close link between the two definitions that the Court defines a dominant position as the ability to commit the abuse alleged in a specific case. In the case of many types of abuse this will mean that a dominant position need be proven only in respect of one or more products which are rightly or wrongly considered indispensable for carrying on their business by the trade customers harmed by the abuse. If, in the present case, for example, tyre dealers consider Michelin's new replacement tyres for heavy vehicles, or only certain types of such tyres, indispensable for their range of products, it is immaterial for an assessment of the discount system at issue whether Michelin also has a dominant position in respect of retreads or light tyres for cars and delivery vans. In so far as Michelin's new replacement heavy-duty tyres are indispensable to dealers, a dominant position on its part in respect of those tyres also enables Michelin — to refer once more to the “abuse” formula used in paragraph 91 of the decision in Hoffman-La Roche — through recourse to methods different from those which condition normal competition on the basis of the transactions of traders, to hinder the competition still existing on the market in respect of other types of tyre as well. To that extent it is therefore of no great consequence in deciding this case that it was firmly established at the hearing that the discount system in question applied not just to heavy-duty tyres for lorries but also to tyres for delivery vans, in respect of which the Commission had found no dominant position on Michelin's part, and that in this respect the Commission made a serious error when ascertaining the facts.

2. The relevant market

2.1. The product market

The main complaint of Michelin NV against the Commission is that its concept of the market is artificial and arbitrary. In what it considers the relevant product market the Commission includes non-interchangeable types and sizes of tyre whilst disregarding retreads — the very product which competes with every type of new tyre. To separate the market in tyres for lorries, buses and similar vehicles from the tyre market in general is also illogical. Finally, the applicant contends that the Commission's reasoning is self-contradictory inasmuch as the Commission puts itself alternately in the shoes of the consumer and the dealer.

I do not think that those contentions are capable of depriving the Commission's arguments of their force. In saying that I assume that, as I said earlier, the concept of the relevant product market may indeed be of practical value in providing the necessary proof. However, as I have said, it is not an element in the prohibition which must be proven separately but is a functional concept which depends for its substance on its usefulness in establishing the existence in a given case of a dominant position which is being abused.

According to paragraph 31 of the contested decision, the Commission considers that the distinction made between heavy and light tyres (for replacing the original tyres) is justified because they differ in three respects, namely in their technical characteristics, the use for which they are intended and their price. This is not denied or scarcely denied by the applicant. The argument that delivery vans in particular are also used for business purposes is correct in itself but immaterial as regards the distinction made. The point is that changes in the supply of and demand for light tyres caused by those three factors have no, or hardly any, effect on the supply of and demand for heavy tyres. Although the applicant further argues that not all types of heavy tyres appear interchangeable to consumers or, for technical reasons relating to production, to manufacturers either, that argument cannot further the applicant's case either. First, the most which could be deduced from this js that Michelin does not have a market share of 57 to 65% as regards all tyres belonging to the heavy tyre market but this would then logically imply that it has an even higher market share as regards a substantial part of that tyre market. Secondly, it is clear from a comparative legal analysis that in practice legal authorities reject distinctions between different types and sizes of products intended for the same group of consumers (in this case haulage contractors). From the manufacturing point of view, the Commission is right to consider the extent to which the various products complement one other technically, and not their interchangeability, the decisive factor. Thirdly, as the Commission rightly observes in the final sentence of paragraph 31 and in paragraph 33 of its decision, it is the dealers' market position which is crucial in the present case. In this regard it is not disputed that Michelin NV itself makes no distinction between different types and sizes of heavy-duty tyres when fixing sales targets.

If paragraph 32 is read together with paragraphs 5 and 6 of the Commission decision it appears that the distinction made in the decision between the market in new tyres and the market in retreads or renovated tyres is entirely dissociated from the marketing stage. Dealers are in a special position on the market in retreads because that market is separate from the market in new tyres. It appears from the paragraphs cited from the decision that the distinction drawn by the Commission is primarily based on three arguments:

(1)	Rightly or wrongly most consumers do not consider the tyres to be equally safe. Michelin NV does not directly deny this; it simply argues that this opinion of consumers is based on a misapprehension and is irrelevant in determining the actual position on the market.

(2)	There is a price difference which, regard being had to the residual value of a used new tyre and the price per kilometre, was estimated by the Commission to be 40% and by Michelin NV at the hearing to be 15%.

(3)

There is a distinct channel of trade for retreads, in which connection the most important point is regarded as being that consumers generally have their own tyre carcasses provided with a fresh contact surface (according to Michelin NV by dealers and according to the Commission by retreading firms). Michelin NV contests only the extent of the practice.

Furthermore, it is rightly pointed out in paragraph 32 of the decision that the market in retreads is only a complementary market and by definition retreads can never entirely replace new tyres. Increasing use of retreads will certainly produce a fall in demand for new tyres but the possibility of switching from new tyres to retreads will inevitably be limited once used tyres suitable for retreading become, as a result, increasingly scarce. Consequently the competition represented by substitute products is quantitatively restricted in any case and one may also argue, as the Commission does, that as in Hugin (Case 22/78 [1979] ECR 1869, paragraph 6 at p. 1896) the market in retreads is in the last resort essentially a service market and not a market for the sale of a product, so that in this respect too there is a different market.

Finally, the contention that the Commission's line of argument is self-contradictory must also fail in my view. It is alleged that the Commission put itself in the shoes of the consumer and of the dealer alternatively. I have pointed out already that the Commission's argument regarding retreads is quite independent of the stage at which they are marketed, although they may indeed affect the tyre trade at the resale level. Therefore this argument is indirectly relevant for an appraisal of trade at the resale level. As regards trade in new replacement tyres, the Commission's argument is indeed directed to the resale level, as the last sentence of paragraph 31 of its decision makes clear. I have pointed out already that this is logical in relation to the alleged abuse. After all, dealers are the intermediaries between manufacturers and consumers (who ultimately determine demand for new tyres). Therefore it is not illogical or self-contradictory that the Commission when examining the relevant market at the resale level should also examine the characteristics of the demand which dealers meet with.

2.2. The geographical market

Although the requirement of defining the relevant geographical market is expressly referred to in Article 86, this concept is also one which must be construed in its context. In other words, it is not possible to determine in the abstract what is the relevant geographical market for deciding whether or not an undertaking has a dominant position without taking into account the requirements for the case in point. An undertaking may well operate on the world market without having a dominant position on that market and yet still have a dominant position on the common market or on the national market of one or more Member States. It is also conceivable that an undertaking may abuse its dominant position on its home market by charging inflated prices there, from which it may finance ruinous price competition designed to strengthen its position on another national market. Here again it is therefore true to say that the abuse need not necessarily take place on the same geographical market as that on which the undertaking concerned has a dominant position. Determination of the relevant geographical market thus depends in turn on the nature of the alleged abuse.

In the present case the abuse is alleged to have been committed on the Netherlands market, not by the Michelin group as a whole but by Michelin NV which is considered to have a dominant position in the Netherlands. Furthermore, it is established that Michelin NV operates mainly on the Netherlands market. It is therefore logical that in paragraph 34 of its decision the Commission should regard the Netherlands market as the relevant geographical market, and in view of the Court's case-law it rightly considers the Netherlands market a substantial part of the common market. The Commission's argument on this point is also logical in the sense quoted. (Michelin NV's operations are “confined to this area, and it is here that the practices at issue took place” (paragraph 34 of the decision.)

Michelin NV's reply is that the Commission's case is based on a number of facts and circumstances which are far wider than those of Michelin NV and concern Michelin as a whole. Furthermore, the manufacturers with which Michelin competes in the Netherlands are undertakings which operate on a worldwide basis.

The Commission in my view rightly observes that this criticism is concerned not so much with the definition of the market as with the determination of the existence of a dominant position. Furthermore the Commission in my view rightly points out — and Michelin NV does not deny this — that tyre dealers in the Netherlands cannot obtain sufficient continuous supplies of Michelin tyres outside the Netherlands (paragraph 34 of the decision) and that competition from foreign manufacturers in the Netherlands also comes from their local subsidiary companies whereas here again dealers cannot in practice obtain supplies from sources located abroad (Commission's defence). I would add that, even if dealers could obtain supplies direct from abroad, that does not necessarily mean that Michelin NV has no dominant position on the Netherlands market. As I have said, the relevant geographical market is not an independent concept but is connected with the dominant position whose existence must be proved and to that extent is a relative one. Depending on the case in point, the Commission is entitled to regard the relevant geographical market as being the market which in the circumstances of the case is the most useful in investigating the case at issue and especially in proving its allegation of the existence of a dominant position.

2.3 Summing-up

To sum up, then, the applicant's criticism regarding the definitions of the relevant product market and relevant geographical market contained in the decision cannot in my view be accepted.

3. The dominant position

On the question whether a dominant position ; exists, which the applicant denies, I can be quite brief. Michelin NV does not deny that from 1975 to 1980 it had a 57 to 65% share of the market for new Michelin lorry and bus tyres in the Netherlands which I also consider to be the relevant market (paragraph 35 of the decision). According to the judgments of the Court in Hoffmann-La Roche (cited above) and United Brands (Case 27/76 [1978] ECR 207), such a market share must certainly be regarded as sufficient evidence of the existence of a dominant position if at the same time there is a considerable difference from the market shares of the major competitors. This is the case in this instance. According to the figures given in paragraph 35 of the decision, which are not contested by Michelin NV, the market shares of its five major competitors ranged between some 4 and 8% only. This means, as is rightly added in the decision, that a dealer who does not stock Michelin tyres suffers a loss of commercial prestige and the survival of his business may be endangered. To quote from paragraph 39 of the decision in Hoffmann-La Roche:“Such a position does not preclude some competition... but enables the undertaking which profits by it, if not to determine, at least to have an appreciable influence on the conditions under which that competition will develop, and in any case to act largely in disregard of it...”. Furthermore, it is not in substance denied that like Hoffmann-La Roche Michelin NV has “a very extensive and highly specialized sales network” (paragraphs 42 of the judgment and 36 of the contested decision). The essential point here, and this is not denied by Michelin ŃV either, is that the absolute size of Michelin NV's sales network was substantially bigger than its competitors'. In view of all this, the additional argument advanced by the Commission to establish the existence of a dominant position is in fact no longer of decisive importance. This is also true of the Commission's other arguments based inter alia on the fact that Michelin NV's range of heavy tyres is wider than its competitors'.

Michelin NV denies that a dominant position exists first of all because the evaluation of the market share was in its view based on an incorrect definition of the relevant product market. However, this argument has already proved untenable.

The other arguments advanced by Michelin NV are, in view of the Court's previous decisions, not sufficient to rebut the existence of a dominant position. As the Commission rightly points out, the advent of new Japanese competitors has not affected Michelin NV's market share. Nor is the evidence — which Michelin NV seeks to adduce from price and trading margins, poor financial results, production capacity and its competitors' financial strength and wider production-spread, to prove that it has no dominant position — sufficient to shake the conclusive evidence, based upon the absolute and relative size of its market share, that it does have such a position. The arguments based upon price levels and margins and losses sustained can only be of some relevance for establishing certain abuses. They may show for instance that there is no question of monopoly profits being made or an abuse consisting in, to use the words of Article 86 (a), “directly or indirectly imposing unfair purchase or selling prices” but such an abuse is not alleged against Michelin NV. Besides, it is well known that an undertaking which is making losses may in such an emergency situation be quite capable of engaging in ruinous price competition, discriminatory practices or other conduct which, if a dominant position exists, amount to an abuse. Experience shows that the sustaining of losses is by no means incompatible with a dominant position. In this regard I refer to paragraph 126 of the Court's decision in the United Brands case [1978] ECR 207, at p. 284. Moreover, the competitors' production capacity, financial strength and wider production-spread might at most be regarded as diminishing Michelin NV's absolute and relative market share as evidence of its dominant position if it appeared that those factors had actually affected its market share, but this has not been contended, let alone proven, by Michelin NV. Finally, the fact that users of heavy tyres are experienced trade purchasers, as Michelin NV contends, is not sufficient by itself to accept that there can be any question of a counterforce undermining or at least neutralizing its dominant position. This might be so only if trade purchasers had formed themselves into a purchasing association, which Michelin NV has not contended, nor does it otherwise appear to be the case.

4. The conduct in question

4.1.

According to the decision cited in the introduction to my opinion, Michelin NV is accused of two practices classified as abuses. They are the application of the general discount system and the grant of the extra bonus in 1977. In both cases Michelin NV contests first some of the facts found by the Commission and secondly the classification of those facts as an abuse of a dominant position within the meaning of Article 86. On the issue whether the facts found constitute an abuse of a dominant position Michelin NV is supported by the French Government. So far as necessary I shall deal in turn with the findings of fact and the classification of the facts found as an abuse of a dominant position first with regard to the general discount system and then with regard to the special bonus in 1977.

4.2. The general discount system

A — The facts

The general discount system is described in general terms in paragraph 22 of the decision and for the purposes of the hearing was summarized by Michelin NV itself in a simplified but partly more comprehensible form. For that summary I refer to the last part of the Report for the Hearing.

The factual issues between the parties, which remained unresolved throughout the proceedings, relate not to the general system itself but to its application in practice.

However, before examining those issues I consider it useful to recall what the Commission eventually declared proven in Article 1 (a) of its decision. The Commission finally declared it proven as against Michelin NV that:

(a)	through the grant of selective discounts on an individual basis

(b)	conditional upon sales “targets” and discount percentages which were not clearly confirmed in writing

(c)	Michelin NV bound tyre dealers to itself and

(d)	applied to them dissimilar conditions in respect of equivalent transactions.

For purposes of the later classification of my argument I have made my own subdivision into lettered subparagraphs in order to indicate precisely which aspects of the conduct in issue may be considered proven and which may not. I shall begin by taking the summary of Michelin NV's arguments contained in the Report for the Hearing. I have numbered each point:

(1)

Michelin NV contends in substance that the discount consists in a fixed component which is the same for each dealer and a variable component which is determined each year in proportion to the purchases of Michelin products on the basis of an annually reviewed progressive scale which Michelin NV notified to the Commission at the beginning of the investigation in 1977. I would point out that this scale, which was based on purchases in the previous year, was withdrawn in 1978.

(2)	The variable component fluctuated from year to year by no more than 5%.

(3)

A fraction of this component (0.2 to 0.4%, according to Michelin NV) was linked to the attainment of a jointly agreed sales target which was subsequently incorporated into Michelin's NV's production and sales programmes. I would mention here that after 1978 sales targets were, according to Annex 20 to the application, still fixed in relation to purchases in the previous year.

(4)	With a few exceptions all dealers who purchased more than 3000 tyres per year are stated to have received the same maximum discount percentage during the period 1975 to 1979.

(5)

Despite the greater maximum variation in the variable component of the discounts granted to retailers under the system described by Michelin NV itself (10 to 22 in 1975 to 1977, 4 to 15 in 1978 and 0 to 5 in 1979 and 1980), the difference in the discounts granted to 54 customers selected by the Commission's inspectors is stated to have been only 2 to 2.5% whereas the number of purchasers of heavy Michelin tyres made by those same dealers might vary from 13000 to 200 per year.

(6)	The differences between the discount percentages were due to the application of a scale fixed according to the total number of purchases made by the dealer from Michelin NV during the previous year.

(7)	However, the discount system, which according to Michelin NV was in principle quantitative only, was not applied in a mechanical fashion, it states, because dealers were not prepared to accept lower discounts when sales declined.

(8)

The agreed sales target was never the decisive factor in the discount. It represented a service rendered in the objective sense by the dealer to the manufacturer inasmuch as the information which the manufacturer received when fixing his targets enabled him to programme his production better and reduce costs.

(9)	According to Michelin NV, not a single dealer lost all his annual discount for any reason whatever. Each dealer knew from experience that in such a case he would lose only a few hundredths of a per cent.

(10)	The Commission's contention that targets and discounts were not notified is unfounded as they were agreed at the beginning of the year between the dealers and Michelin NV's commercial representatives. Any dealer who requested it always received official written confirmation.

In its defence the Commission points out in the first place that the emphasis which Michelin NV places in its application on the scale used for the annual bonuses is not consistent with Michelin NV's statements in the administrative procedure to the effect that the scale and relevant instructions were merely guidelines for internal use which were applied in a flexible manner in order to take account of the individual situation of each dealer. In my view that conclusion is substantially confirmed by my summary of Michelin NV's arguments taken from the Report for the Hearing, especially the fifth and seventh points of that summary.

In its defence the Commission secondly observes that the customer files clearly show that the discount was linked to a purchase target for heavy tyres and that this target consisted of a precise number of tyres to be purchased during the course of the year. Although it emerged during the hearing that the discount also depended on purchases of tyres for delivery vans, the fact mentioned is not contested by Michelin NV and is indeed plain from the customer files. The fact that the discounts were also related to purchases of tyres for delivery vans is in itself, I think, irrelevant in any case as far as the assessment of the discount system is concerned. As I pointed out earlier, an alleged abuse need not concern — certainly not entirely — the same products as those in respect of which a dominant position has been found to exist.

According to the Commission, the differences established between the discounts amount to 2 to 5% of the annual bonus, which is not consistent with points 2 and 5 of the summarized argument of Michelin NV. Michelin NV has pointed out that only 0.2 to 0.4% of that difference was linked to the attainment of a jointly agreed sales target and most of the difference should therefore be attributed to the variable component of the discount system of which I spoke earlier. However, I would point out another uncertainty which dealers had to face resulting from the difference between the total of the monthly and subsequent four-monthly advances and the maximum discount granted upon the final settlement of accounts. As is clear from the system described by Michelin NV itself, the margin of uncertainty varied between 2 and nearly 4%, depending on the period considered.

To sum up the arguments of both parties, it seems to me that of the facts referred to in Article 1 (a) of the decision, to which I attached letters (a) to (d), the following may be considered established or in dispute.

Sub-division (a): It is established that Michelin NV granted a considerable part of the discounts on an individual basis and to some extent, as is shown by what is recognized as being the flexible application of its general guidelines (certainly after withdrawal of the scale in 1978) and the agreement on individual sales targets, on a selective basis as well (points (2) to (7) of Michelin NV's argument read together with paragraphs 20 to 22 and the first 14 words of paragraph 23 of the decision). The extent to which dealers' loyalty to Michelin played a part in the selective determination of discounts is disputed but this point is not mentioned in the operative part of the Commission's decision and in any case does not seem to me fundamentally important as far as the assessment of the discount system is concerned.

Sub-division (b): It is also established that in particular the “target bonus” of 0.2 to 0.4% was linked to the agreed sales targets (point (3) of Michelin NV's argument) and that the sales targets were not officially confirmed in writing as a matter of course but on request only (point 10). To what extent the other variable element of the discounts, was linked to the agreed sales targets (or the sales figures for the previous and current years) is disputed as well as the flexibility with which final accounts were in practice made dependent on the attainment of the agreed sales targets. However, dealers were uncertain, to say the least, about the greater part of the variable element of the discount. In this regard Michelin NV merely relies on the fact that dealers could be sure from “experience” that flexibility would be shown and that they would never be deprived of all their annual discount (point (9) of the summary). Therefore, as Michelin NV thus admits, there was unquestionably no legal certainty that the annual discounts themselves were not made conditional upon the attainment of the agreed sales targets. Moreover, the importance of the extra bonus of 0.2 to 0.4% fixed in each individual case must not be underestimated in my opinion. In order to neutralize the incentive effect of such an extra bonus competitors would have to offer, in view of their market shares which I mentioned earlier, an extra bonus of 2.8 to 6.4% and, in order to obtain a competitive lead over Michelin NV equal to that which the latter had gained over them, twice the amount of those extra discounts. The issue whether the greater part of the variable element of the discount was also linked to the agreed sales targets or to the sales figures for the previous year is in my view immaterial. As I pointed out with regard to the third point of its summarized argument, Michelin NV itself states that the sales targets were fixed in direct relation to purchases in the previous year. In those circumstances it is merely a question of words whether one speaks of there being a correspondence with the sales targets (and thus an indirect correspondence with the purchase figures for the previous year) or of a direct relation to those figures; it makes no difference as far as the economic effect of the system is concerned.

My sub-divisions (c) and (d) of the statement in Article 1 (a) of the facts declared proven are on the borderline between being findings of fact and their classification as an abuse of a dominant position. In (c) the Commission states that Michelin NV bound tyre dealers to itself through the discount system applied. In paragraph 38 of its decision the Commission states that the discount system was aimed at binding dealers to Michelin NV and then examines this charge in more detail. In order to assess the discount system with reference to Article 86 it is in my view especially important to establish that the calculation in principle of part of the prospective discounts on the basis of the agreed sales targets is bound to put pressure on the dealer to achieve the maximum sales target. In order to establish whether the system has that effect it does not matter, either, whether the Commission's assertion in paragraphs 24 and 38 of its decision that the sales targets were usually set higher than the figures for the previous year is correct or not. Nor does Michelin NV's argument, summarized on page 34 of the Report for the Hearing, not shake that conclusion. The fact that the differences between the discounts received by the various dealers were purely quantitative in nature and based on the customer's total turnover, as Michelin NV contends, does not lessen the incentive effect of higher discounts for higher turnover. Furthermore, in its arguments regarding the “target bonus” of 0.2 to 0.4% Michelin NV expressly admits that it is possible to speak of an incentive discount. At the hearing Michelin NV itself stated that in order to programme its production and sales it needs to be certain that the sale targets will be be achieved. It simply denies that such an incentive discount constitutes an abuse within the meaning of Article 86. Whether it does or not is a question which I shall examine later, however. I therefore consider the factual content of my subdivision (c) of the Commission's statement of findings in Article 1 (a) sufficiently proven.

Michelin NV's objection to subdivision (d) of the operative part of the decision (application of dissimilar conditions in respect of equivalent transactions) is basically to its description as an abuse of a dominant position. It maintains that the differences between the discount percentages are due to the application of a scale which is adjusted to the dealer's total purchases from Michelin NV in the previous year and that they cannot therefore be regarded as discriminatory; the other slight differences, which it admits elsewhere (points (3) and (7) of my earlier summary), cannot be regarded as discrimination prohibited by Article 86 either. The evidence which the Commission has adduced of the other alleged discount differences, which cannot be accounted for in the way indicated, is in my view so distorted by the error previously mentioned regarding the basis of the discounts (tyres for delivery vans as well as heavy-duty tyres) that it cannot be given any weight. If tyres for delivery vans are taken into account, the cases adduced as evidence are not in fact equivalent so that it has not been proven that dissimilar conditions were applied to equivalent transactions. It appears, however, from paragraphs 40 and 41 of the decision that the Commission considers the alleged discrimination a separate abuse, which is important for the rest of my argument.

B — The assessment of the facts as an abuse

In an assessment of the facts which may indeed be considered sufficiently established it must first be recognized that there is no question of loyalty bonuses such as those in issue in the Hoffmann-La Roche case. There is in fact no obligation upon dealers not to buy any other than Michelin tyres or even to limit such purchases. Michelin NV was right to point this out.

On the other hand, in its decision (paragraphs 37 to 40) the Commission did not state this.

For an assessment of the established facts it is, I think, of importance first of all that periodic bonuses and annual discounts fixed on an individual basis, hence not communicated uniformly to all the dealers in advance, by their nature contain an element of uncertainty and make discrimination between dealers possible. In view inter alia of the results of a study of comparative law I believe that it is justifiable to consider such conduct on the part of an undertaking holding a dominant position sufficient by itself to constitute an abuse within the meaning of Article 86.

More important however in my view is the fact that annual discounts and advance bonuses are — certainly in so far as they are partly dependent on sales targets jointly agreed for the year in question — by their nature wholly different from purely quantitative discounts based on the quantity of each separate order. Whereas the latter type of discount may be justified by clear cost savings, this is not necessarily the case with purely periodical quantitative discounts based on sales in a given period, since a particular dealer's orders may be spread over the year in such a way that the cost savings for the manufacturer are greater in the case of a few large orders accounting in total for a lesser annual turnover. What is more, by its nature any system of periodic quantitative discounts puts pressure on purchasers to purchase as much as possible from the manufacturer in question. If, as in this case, the manufacturer's market share is larger than all its competitors put together, the offer of more generous terms may hardly be sufficient for those competitors to counteract the suction effect of the periodic discount system. In fact, in view of their market shares which I gave earlier, they would have to offer proportionately higher discounts as against the acknowledged variable component of 2.5 to 5% of Michelin NV's discount system (not including the extra “target bonus”). ( 2 ) To that extent the restrictive effect on competition and the absence of any economic justification for such a system may be likened to that of a “Gesamtumsatzrabattkartell” [restrictive agreement involving discounts on total sales]. As previously noted in point 8 of the summary of its argument, Michelin NV considers that the extra bonus, which in its view is small, paid upon the attainment of the annually agreed sales targets, is a reward for enabling the manufacturer to programme his production better and operate more economically. It is precisely that explanation, however, which makes it clear that the agreed sales targets do have, “through recourse to methods different from those which condition normal competition” on the basis of the transactions of traders, “the effect of hindering the maintenance of the degree of competition still existing in the market or the growth of that competition”, to cite again the general definition of an abuse given in paragraph 91 of the decision in Hoffmann-La Roche. The opportunity for Michelin NV to free itself from the competition of other tyre suppliers is therefore increased by the method of calculating the discount on the basis of annual turnover and by the fixing of annual sales targets which if attained attract an extra bonus. Whether the effect of the agreed sales targets is really so slight as Michelin NV suggests is disputed. I think it may be assumed that the uncertainty about the legal consequences of not attaining at least the lowest sales targets causes them to have a greater effect on dealers' efforts than Michelin NV claims. Moreover, I have already calculated how much higher discounts competitors having a so much smaller market share would have to offer simply in order to neutralize the effect of the extra target bonus of 0.2 to 0.4%. However, I do not consider an exact quantification of the aforesaid effects of the sales targets crucial in the last resort since the Court has already stated in the last sentence of paragraphe 123 of its decision in Hoffmann-La Roche that in the case of conduct of “an undertaking occupying a dominant position on a market where”, as a direct result of its presence, “the structure of competition has already been weakened, ... any further weakening of the structure of competition may constitute an abuse of a dominant position”. Finally, as I pointed out, Michelin NV specifically stated at the hearing that by means of the sales targets agreed it sought to obtain some certainty in order to determine its production and sales programmes. However, where there is effective competition, such certainty can never be obtained.

To sum up, then, I consider that in view of my foregoing observations the Commission was right to find in Article 1 of its decision that during the period 1975 to 1980 Michelin NV infringed Article 86 of the EEC Treaty on the market in new replacement tyres for lorries, buses and similar vehicles by binding dealers in the Netherlands to itself through the grant of selective discounts on an individual basis conditional upon sales “targets” and discount percentages which were not clearly confirmed in writing. On the other hand I do not consider it sufficiently proven that Michelin NV did actually apply to them dissimilar conditions in respect of equivalent transactions. It is certainly not possible to infer such discrimination simply from the established fact that the sales targets and discounts were fixed individually. The last two and a half lines of Article 1 (a) of the decision should therefore in my view be declared void. As I pointed out before, this allegation concerns a different kind of conduct to which specific exception is taken, which is separately mentioned in Article 86 and which by its nature is an abuse different from that of binding dealers to the undertaking in the way described, a practice which mainly affects competitors, not dealers.

4.3. The extra bonus in 1977

In Article 1 (b) of the decision it is stated that Michelin NV also abused its dominant position by granting an extra annual bonus in 1977 on purchases of tyres for lorries, buses and the like and on purchases of car tyres, which was conditional upon attainment of a “target” in respect of car tyre purchases. Michelin NV denies first that the bonus was conditional upon a “special target” in respect of purchases of car tyres. It contends that as far as light tyres were concerned the bonus depended on no other sales target than that agreed at the beginning of the year and in each case was calculated on the basis of the total volume of purchases of all current categories of tyres. The latter claim is not contested by the Commission (in the part of the decision cited). According to Michelin NV, the extra bonus should also be regarded as compensation for the fact that in that year it was unable to supply enough heavy lorry tyres and this point is not contested by the Commission either. In such circumstances I feel that the Commission's argument that an abuse of a dominant position exists here is unconvincing. That the bonus in question should represent an extra bonus for purchases of heavy tyres, as the Commission contends in paragraph 50 of its decision, is in itself improbable if purchases of them are restricted precisely because Michelin NV's ability to supply them is limited. To say that the extra bonus on heavy tyres is subject to the “condition” that a given quantity of light tyres be purchased is also unconvincing in these circumstances. As I have said, Michelin NV denies that the “special sales target” for light tyres was higher than the sales figures for light tyres in the previous year and this point has not been clearly proven by the Commission. It would appear to be inevitable that if, for reasons attributable to the manufacturer, sales of heavy tyres are temporarily limited, a “special” extra bonus of a compensatory nature should be linked to the sales targets for light tyres. The Commission has not therefore sufficiently demonstrated that the bonus in question was not normal compensation for the loss of heavy tyre sales, as Michelin NV contends. Finally, I have already pointed out that even under the normal system discounts on heavy tyres depend partly on turnover of tyres for delivery vans so that there may at the most have been an extension of the normal system to all light tyres on a single occasion and not a quite separate abuse, as the Commission claims. My opinion accordingly is that Article 1 (b) of the Commission's decision should also be declared void as the statement of the reasons on which that part of the decision is based is in any case insufficient to justify treating the bonus in question as a separate abuse.

5. Effect on trade between Member States

As regards Michelin NV's objection that the conduct in question was confined to the Netherlands and could not therefore adversely affect trade between Member States I can be brief. As is clear from paragraph 15 of the decision, a substantial proportion of the tyres made by Michelin NV's competitors are manufactured in other Member States and imported thence into the Netherlands. The proven restriction on their opportunities for marketing as a result of the effect of Michelin NV's discount system therefore automatically means that their import opportunities, and thus trade between Member States, are restricted as well. Hence Michelin NV's objection does not in substance concern this part of the prohibition contained in Article 86 but amounts in fact to a repetition of its arguments contesting the finding that it has abused its dominant position. In so far as those latter arguments must fail, so must its arguments on the issue whether trade between Member States is affected. In any case the Court has repeatedly held with regard to the application of Article 85 that practices carried out wholly within the territory of one Member State may nevertheless affect trade between Member States. I need only refer to the Court's judgment in Case 8/72 Cementhandelaren [19721 ECR 977.

6. Breach of the rights of the defence

The applicant's submission alleging a breach of the rights of the defence is in two parts. The first is that the Commission did not allow it to inspect vitally important parts of the file (see the application, page 62), in particular the findings made by the Commission in the case of various dealers, consumers and competitors of Michelin NV. In justification of its refusal to let the applicant see the results of its investigation the Commission relies on Article 20 of Regulation No 17. In any case the investigation did not relate to consumers and Michelin NV's competitors but only to certain dealers. The Commission also claims that the investigation did not disclose any new facts of which it was not already aware from the information obtained from Michelin NV. The applicant has not demonstrated or even adduced evidence suggesting that the Commission obtained from its investigation fresh incriminating evidence of the kind discussed in paragraph 14 of the decision in Hoffmann-La Roche. This part of the submission should therefore be rejected.

As may be seen from pages 64 and 65 of the application, the second part of the submission is to the effect that the Commission did not clearly indicate to what extent it accepted or rejected the statements made by the witnesses and experts at the hearing and other facts emerging from the hearing. An obligation of that kind cannot, however, be deduced from Community law as hitherto interpreted by the Court. As regards the rights of the defence the only relevant point is whether regard has been shown to them with reference to what is set forth in the statement of objections. It is clear from the decision that ample attention was paid to the refutation of the applicant's arguments regarding the statement of objections and to the statements made by the witnesses and experts at the hearing. It is not necessary for statements made by witnesses or experts supporting the applicant's case to be separately and expressly examined in the decision if the relevant points of his case and of the statements in question are adequately dealt with in the decision. This part of the submission must therefore be rejected as well.

7. The fine

Michelin NV takes the view that it cannot be accused of acting negligently or intentionally since the abuse with which it is charged by the Commission involves a new interpretation of Article 86 which it could not have foreseen. In particular that interpretation diverges from the Court's decision in paragraph 89 of the Hoffmann-La Roche judgment regarding the question of fidelity rebates since all the characteristics of fidelity rebates are absent in this case. The applicant also considers it unreasonable to fix the duration of the infringement at five years, from 1975 to 1980, as the Commission itself could have shortened that period by taking action earlier and could have prevented the payment of the extra bonus of 0.5%, which was not made until the end of 1977.

With regard to the fixing of the fine, at the hearing Counsel for Michelin NV expressly recognized the Commission's right to change the views which it previously held regarding the concept of an abuse: “C'est son droit le plus strict.” However, the Commission must then refrain from imposing a fine as this would otherwise infringe the legitimate expectation of those concerned.

I think that the applicant's other arguments regarding the amount of the fine may be dealt with briefly. The exchange of views on the relationship between the fine and turnover seems irrelevant to my mind because the fine imposed remains below the alternative maximum limit of one million units of account. As the Commission rightly points out, it is the duty of the undertaking concerned to cooperate in the administrative procedure. The fact that the undertaking fulfils that duty cannot therefore be a ground for reducing the fine. Similarly, like the Commission, I consider it self-evident that the Commission cannot indicate in the administrative procedure itself what the fine or the criteria for its assessment will be since the degree of guilt or negligence can be determined only at the end of the administrative procedure. It is even possible, and does sometimes happen, that the conduct in question will be terminated at the end of the administrative procedure, which may in some circumstances be a reason for reducing the fine originally contemplated.

As regards the main allegation made against Michelin NV's conduct in Article 1 (a) of the decision, I think that Michelin knowingly and successfully tried, by agreeing sales targets with the dealers, to bind them to itself and to maintain its production and sales at least at the level of previous years. As I mentioned earlier, it again specifically said in its reply at the hearing that “sans incitation pour atteindre l'objectif, il n'était pas sûr que cet objectif soit atteint, et c'est cela précisément que rémunérait la remise d'objectifs” (emphasis added). It is therefore clear to me that, to refer again to paragraph 91 of the decision in Hoffmann-La Roche, Michelin NV knowingly sought through recourse to methods different from those which condition normal competition in products or services on the basis of the transactions of traders to ensure that its discount system should hinder “the maintenance of the degree of competition still existing on the market or the growth of that competition”. I also share the Commission's view that the discount system applied by Michelin NV may from the economic point of view be considered a variant of a“fidelity rebate” in the same way as a “Gesamtumsatzrabattkartell” formed by undertakings acting in concert and that Michelin NV did intend to bind dealers to itself and thereby restrict sales of its competitors' products. To that extent I agree with the Commission that at the least Michelin NV acted negligently and that the legal difference between its discount and the fidelity rebates at issue in the Hoffmann-La Roche case cannot diminish its degree of negligence. Personally I think that it may even be said that there was a clear intention on the part of Michelin NV to freeze the structure of the market for the purpose of maintaining its market share.

On the other hand account must be taken of the fact that two of the subsidiary findings in Article 1 of the decision must in my view be declared void. Taking into account the fact that those two findings are in economic terms of an altogether secondary nature and the seriousness of the conduct to which exception has correctly been taken I propose that the Court should reduce the fine imposed to 500000 ECU or the equivalent amount in guilders.

8. Conclusion

To sum up, then, I propose that:

(1)	The last two and a half lines of Article 1 (a) and Article 1 (b) of the decision should be declared void;

(2)	The fine imposed on the applicant should be reduced to 500000 ECU;

(3)	The applicant's other submissions should be dismissed;

(4)	Each of the parties, including the French Republic acting as intervener, should be ordered to pay their own costs pursuant to Article 69 (3) of the Rules of Procedure.

---

( 1 ) Translated from the Dutch.

( 2 ) I think that the character of the discount system is illustrated very well here by the fact that according to the fourth point of Michelin NV's summarized argument, dealers who sell more than 3000 tyres all obtain the maximum discount percentage (excluding the target bonus). In my view it is precisely this fact which shows that the variable element of the discount is mainly intended to tic dealers to Michelin. Furthermore this practice confirms that the effectiveness of this means of binding dealers to Michelin NV is directly related to Michelin NV's and its competitors' market shares. If Michelin NV's market share were smaller, it would have to give higher periodic quantitative discounts to secure customer loyalty. To that extent too the discount system used is in point of fact also evidence of the existence of a dominant position, although for the purposes of argument it is simpler in this case to demonstrate first that a dominant position exists.