Introduction

Factual Background

Contentions of the Belgian Government

I — Manifest error of appraisal of the facts

(i) The frontier price

(ii) The nonpublic and guide price aspect of tariff F

(iii) Gasunie's profit margin

(iv) Gasunie's costs

(v) Gasunie's export markets

(vi) Belgian ammonia imports

(vii) Comparability of markets

(viii) Alternative expon possibilities

(ix) The commercial justification of tariff F

II — Error of law in the interpretation of Article 92 of the Treaty

(i) The lower price of gas in other countries

(ii) Differing stages in the production process

(iii) Costs of substitution

(iv) Renunciation of profit

(v) The specific character of tariff F

III — Annulment for defective reasoning

Conclusion

1.

This is an action taken by the Kingdom of Belgium seeking the annulment of a Commission decision taken on 29 December 1992 terminating the procedure it had initiated pursuant to Article 93(2) of the European Community Treaty regarding a preferential tariff system applied in the Netherlands to supplies of natural gas to Dutch nitrate fertilizer producers. ( 1 )

2.

The relationship between the Dutch gas supplier, Gasunie, and the Dutch nitrate fertilizer manufacturing industry has occupied the attention of affected interests, Member States, the Commission, and the Court, on State aids grounds, for over ten years. Natural gas is the chief raw material and represents 90% of the cost in the manufacture of ammonia. Ammonia is, in turn, the chief constituent in the manufacture of nitrate fertilizers. Gas, in effect, represents 70% of the cost of production. ( 2 ) Up to the 1980s, Community nitrate fertilizer manufacturers largely produced their own ammonia. ( 3 ) Gasunie has a de facto monopoly of the supply of gas in the Netherlands, and of the export of Dutch gas. The Dutch State owns 50% of Gasunie's capital, directly (10%) or indirectly (40%, through the State-owned Energie Beheer Nederland), and appoints half the supervisory board whose powers include the fixing of the tariffs applied by Gasunie. Furthermore, prices set by Gasunie are subject to the approval of the Dutch Government. In Van der Kooy & Others v Commission, ( 4 ) the Court concluded from this ownership and control structure that the fixing of a Gasunie tariff (in that case, a preferential horticultural tariff for natural gas supplies) was the result of action by the Dutch State and fell to be considered as an ‘aid granted by a Member State or through State resources’ under Article 92 of the Treaty.

3.

The Commission first initiated the procedure under Article 93(2) of the Treaty against the Netherlands in October 1983. According to the Commission, Gasunie, under the aid scheme then applied, granted special rebates to Dutch ammonia producers by means of a two-tier tariff structure, which had the effect of reducing the cost of natural gas used by them as a raw material. ( 5 ) The ammonia industry was charged the standard industrial price for gas to be used in production for sale on the European Community market, and a substantially lower price in respect of production for export to third countries. In the course of the procedure, the Commission delivered on 13 March 1984 a reasoned opinion in which it found that this tariff structure constituted a State aid within the meaning of Article 92(1) of the Treaty and did not qualify for any of the derogations provided for in Article 92(3).

4.

The Dutch Government informed the Commission on 14 April 1984 that Gasunie had abolished the contested tariff, and had added to its industrial tariff structure, with retroactive effect to 1 November 1983, a new tariff known as tariff F, which is the subject of the present action. Tariff F was made available to very large industrial users established in the Netherlands, other than those in the energy industry (and essentially, as we shall see, to the ammonia sector), on condition that they:

The new tariff F was invoiced at the price applicable under tariff E, which applied to users with an annual gas consumption of between 50 million and 600 million m3, less 5 Netherlands cents/m3. It later transpired that the minimum annual consumption required of tariff F users was 500 million m3, and that the 5 cents rebate was a maximum, with the actual rebate descending to as low as 2 cents/m3 on occasion. ( 6 )

5.

The Commission continued its investigation, in the light of the new tariff, which, it concluded, secured for Gasunie economies in the cost of supply greater than the value of the 5 cents rebate, by reason principally of the combined effect of large volume and regularity of offtake. It also concluded that tariff F formed part of the general tariff structure for gas users in the Netherlands, did not discriminate between sectors, and contained no element of State aid. It, therefore, decided on 17 April 1984 to terminate the procedure against the Netherlands under Article 93(2) of the Treaty.

6.

This decision was challenged in annulment proceedings taken by a number of French competitors of the Dutch nitrate fertilizer manufacturing industry, in CdF Chimie Azf v Commission. ( 7 ) The Court commissioned a report from three expert consultants on the gas industry (hereinafter ‘the experts' report’), upon which it based the findings of fact in its judgment. The experts' report analysed the savings for Gasunie attributed by the Commission both to the individual alleged elements of economy of supply and, also, to their totality; in each case, they found that the Commission had committed a manifest error of appraisal:

The experts reported that the tariff F rebate must be attributable to other considerations. The Court decided that the Commission had committed a manifest error of appraisal and annulled the Commission's decision in its judgment of 12 July 1990.

7.

Although the economies of supply, then advanced in justification of tariff F, no longer figure in the arguments, and the Commission, in the decision now contested, has replaced them by quite different grounds, the fact remains that the Court has already, in CdF Chimie, considered tariff F. It is useful to recall two particular aspects of the judgment of the Court. On the one hand, the Court rejected the contention by the French applicants that tariff F was a special and secret arrangement, negotiated only with the Dutch nitrate fertilizer manufacturers on a confidential basis. In fact, the Court held that it was ‘a public tariff whose conditions of availability are public and perfectly open’, and that it was ‘available to all customers who fulfil the objective conditions prescribed for its application’. ( 8 ) On the other hand, the Court upheld the argument that tariff F was essentially intended to apply to the ammonia manufacturing industry and that its provision to a single undertaking outside that industry did not undermine its essentially sectoral nature. ( 9 )

8.

The Commission reopened the procedure under Article 93(2) of the Treaty in January 1992, publishing a notice which pointed out both the apparently neutral objective conditions for tariff F and the fact that the Dutch ammonia producers had been its chief beneficiaries. It also pointed out in this notice that the rebate was variable, with 5 cents/m3 representing merely its maximum value. ( 10 )

9.

The Commission adopted a decision (hereinafter ‘the decision’) on 29 December 1992 once more to terminate the procedure. ( 11 ) The Commission drew upon the experts' report, which had suggested that, while the alleged savings had not materialized, there might be other commercial justifications for tariff F, viz. ensuring that valued customers for natural gas were not charged unaffordable prices which would drive them either out of business or to sourcing their ammonia elsewhere. ( 12 )

The Commission stated:

‘[W] ith regard to Gasunie's tariff F, the aim was to resist competition on the nitrate fertilizer market from ammonia produced in other countries, notably non-Community countries. A nitrate fertilizer manufacturer ... can himself produce the ammonia needed to produce nitrate fertilizers or he can buy it from other producers and use it to manufacture his product If the price of the gas which he uses in order to manufacture the ammonia he requires is too high, he will decide to purchase the ammonia, if possible, elsewhere at a lower cost than he would have to pay if he produced it himself This was the situation in the Community ammonia industry in the 1980s, and if Gasunie had not granted the Dutch nitrate fertilizer producers special tariffs, they could perfectly well in the long run have shut down the ammonia producing factories, obtained their ammonia supplies from outside the Community and continued nevertheless to produce nitrate fertilizers.’ ( 13 )

10.

The main element of the decision was that Gasunie needed to protect very large and vulnerable gas customers:

The Commission concluded (i) that tariff F was commercially justified; (ii) that it gave no preference over other Member State producers and (iii) that the absence of any loss of revenue showed that the Dutch State had not acted differently from any ordinary shareholder. Conclusions (i) and (iii) are effectively the same point, differently stated. The question whether conclusions (i) and (ii) of the Commission (which I regard as the principal bases of the decision) are interdependent is of some significance, and is discussed further below.

11.

Gasunie abandoned tariff F in 1991 in favour of a distinct pricing system for gas when used as a raw material rather than as an energy source. The new system was approved by the Commission on the condition that its terms, and any subsequent modifications thereto, be extended to export markets. ( 20 ) The Commission later approved a decrease in this price (on Dutch and export markets) on grounds identical to those expressed in conclusion (i) of the impugned decision in the instant case, viz. the need in a period of heightened competition from non-Community ammonia producers to respond to the potential loss of a very significant market. ( 21 )

12.

The Kingdom of Belgium commenced the instant annulment action against the decision on 1 March 1993. By an order of 30 June 1993, the President of the Courtn permitted the Kingdom of the Netherlands to intervene in support of the Commission. By an order of 6 December 1993, the Court granted the applicant's request that the Commission make available copies of the experts' report and of a 1991 joint report by the Commission and the Association Européenne des Producteurs d'Engrais (EFMA), entitled ‘L'industrie des engrais dans la CEE: situation et perspectives’.

13.

The Belgian Government seeks the annulment of the decision, and the condemnation of the Commission to pay the costs of the action, both of which claims are resisted by the Commission. The Belgian Government founds its application on three contentions: that the Commission was guilty of manifest errors in its appraisal of the facts; that its interpretation of Article 92 of the Treaty was erroneous; and that the decision is defective in point of reasoning. Each of these contentions comprises a number of distinct arguments. Some of the arguments relevant to one contention are rearticulated as arguments relating to another. It seems better none the less to examine each of the Belgian contentions in turn, and to address any overlaps between the three as they arise.

14.

The Commission has relied upon Article 42(2) of the Rules of Procedure of the Court to argue that it need not respond to certain arguments which, it alleges, were not introduced by the Belgian Government (or were not introduced in coherent form) until the later stage of the written pleadings. This argument on the part of the Commission will be addressed where it arises in the discussion of the contentions of the Belgian Government. ( 22 )

15.

The Belgian Government advances nine distinct arguments under this rubric, which can, however, be categorized in accordance with their relevance to one or other of the two principal conclusions upon which the Commission based the decision: (i) that tariff F represented a normal commercial response on the part of Gasunie to the difficulties faced by its important and vulnerable customers in the Dutch ammonia industry and (ii) that tariff F gave no preference to the Dutch over other Member State ammonia producers. All but the first two arguments relate to conclusion (i).

16.

It is important to establish whether the impugned decision can survive if supported by only one of the Commission's two conclusions, in case the arguments of the Belgian Government should be accepted in part only. They appear on their face to concern distinct elements of the test in Article 92(1) of the Treaty for the identification of a State aid incompatible with the common market (although neither is stated to relate solely to one of these separate requirements).

17.

Article 92(1) of the Treaty defines prohibited State aids (subject to certain exceptions) by reference to a number of cumulative criteria. A measure must:

18.

Conclusion (i) of the impugned decision (the ‘commercial response’ ground) is concerned with whether tariff F can be classified as an aid at all. The Court has made clear in its consistent case-law that the grant by enterprises under State control of preferential prices favouring certain undertakings or sectors can constitute State aid, but only where profits that might otherwise have been earned are thereby foregone. ( 23 ) In such cases, the Court asks whether an ordinary commercial operator would have acted in the same fashion. ( 24 ) The Commission sought by reference to a number of considerations — the difficulties of the ammonia industry, its importance to Gasunie, the profits which continued to be earned and investments repaid, the lack of alternative gas markets, etc. — to establish this in the instant case, culminating in conclusion (i).

19.

Conclusion (ii) (the ‘frontier price’ ground) is primarily traceable to the separate requirement in Article 92(1) of the Treaty that aid favouring certain undertakings or fields of production must be such, in order to be incompatible with the common market, that it ‘distorts or threatens to distort competition’ and that it thus ‘affects trade between Member States’. The Commission found that, ‘by purchasing gas at tariff F, the Dutch nitrate fertilizer producers were not obtaining any advantage over their competitors in the other Member States’, because Gasunie supplied to the French, German and Belgian distribution companies quantities of gas intended for the nitrate fertilizer industry at ‘a price, referred to as a frontier price, that was more or less identical to tariff F’. ( 25 ) If an equivalent gas tariff were made available by Gasunie to Belgian and other ammonia producers (to the extent that a gas exporter can influence pricing decisions in other Member States), ( 26 ) it would not distort competition or threaten intra-Community trade. Such a tariff could thus escape prohibition under Article 92(1) of the Treaty, even if the price rebate were an aid, i. e. were conceded for reasons other than ordinary commercial ones.

20.

It is quite possible to envisage a preferential price being accorded only to the ammonia producers of one Member State and not to those of another because of the relative size and degree of weakness of the industry in the two countries. This would constitute normal price differentiation; such pricing would not constitute aid, even if it affected (as it must) the relative competitive positions of the undertakings concerned, so long as it were commercially justifiable. The Commission's commercial response ground would be satisfied. On the other hand, a commercially unjustifiable preferential price (i. e. an aid) might, however improbably, be extended to ammonia producers throughout the Community on equal terms without breaching Article 92(1) of the Treaty because there would be no threat to trade or competition within the common market.

21.

In theory, therefore, the Commission's two grounds should be independently capable of sustaining the decision. It has been established by the Court, in respect of the grant of an exemption under Article 85(3) of the Treaty, that, where the various conditions for a Commission decision are concurrent or alternative, the decision will be upheld if just one of these conditions is shown to have been satisfied. ( 27 ) Conversely, it follows logically that, where the conditions for a decision are consecutive or cumulative, as they are for a finding of the existence of a State aid prohibited under Article 92(1) of the Treaty, a decision declining to make such a finding must be sustained once it is clear that just one of those conditions has not been satisfied.

22.

However, there is a possible link, as a matter of evidence, between the Commission's two theoretically independent conclusions: the commercial response and frontier price grounds. This possible link is highlighted by the improbable conjecture in paragraph 20 above that a Member State might extend a commercially unjustified preferential price — an aid — to undertakings throughout the Community. As Member States are unlikely to grant Communitywide aid to undertakings, a finding that Gasunie provided gas at prices equivalent to tariff F for ammonia producers outside the Netherlands could suggest that the tariffin question was a normal response to market needs and commercial pressures rather than a politically motivated aid. It is neither necessary nor (inevitably) sufficient evidence of commercial justification, but it is clearly of circumstantial probative value in a field where precise quantification of commercial pressures is very difficult.

23.

The Commission took this stance in the oral hearing, although it probably overstated its case. The agent for the Commission stated that the commercial response and frontier price grounds were at the same time intimately connected and the one the corollary of the other; tariff F could not, therefore, be considered to fall under Article 92(1) of the Treaty in the case of a positive response in respect of the two issues. In response to a question from the bench, the Commission stated that the two issues were linked because, if Gasunie could not afford to lose the Dutch nitrate fertilizer market, it could equally ill afford to lose that of Belgium.

24.

This, in my view, is not entirely correct. The extension to other Member States of an equivalent preferential tariff serves to indicate that the price is probably commercial. However, the refusal of a frontier price, enabling the Belgian distribution company to offer a tariff equivalent to tariff F in Belgium, would not automatically indicate that the Dutch tariff was not a commercial response to market conditions, as market conditions in the two countries might be very different. The Commission was mistaken in suggesting an automatic link between the two. Its first conclusion may be, in evidentiary terms, a corollary of the second, but the reverse is hardly true. The commercial nature of the tariff in the Netherlands may be demonstrated (as a matter of probability) by its extension to other Member States, but a finding that the tariff was commercial does not lead inevitably to, nor depend upon, a finding of the equivalence of the frontier price of gas destined for ammonia producers in other Member States.

25.

The agent for the Commission conceded, in reply to a further question from the bench, that production features could vary from one country to another. However, he continued that, had the Commission found that gas was not being made available at a price equivalent to tariff F for the ammonia producers of other Member States, it would have had to enquire further into the reasons for this. He also pointed out that the decision on the raw material tariff, which has replaced tariff F, makes Commission approval expressly contingent on the equivalence of Dutch and export prices (irrespective, it seems, of any objective differences that may exist between the various markets). ( 28 )

26.

Thus, the Commission has made it clear that (even if it might in theory have been independently sustainable) its conclusion that tariff F was commercially justified was in reality dependent on its conclusion about the frontier price. In the event, it appears that the frontier price conclusion relates both to the possible distortion of competition and threat to trade, and to the prior question of the existence of an aid. It must therefore be fatal for both of the conclusions upon which the decision is founded, and so for the decision itself, if the Commission is found to have erred in respect of the frontier price. On the other hand, if the frontier price ground were sustained, the decision could survive a rejection of the commercial response ground by the Court for other reasons, related only to market conditions in the Netherlands. However, I have already remarked upon the improbability of such a finding: it postulates a Member State granting a State aid to producers in other Member States.

27.

It is also necessary to make some preliminary remarks about the burden of proof and the role of the Court in an annulment action such as this. The Court has no jurisdiction to find that a particular practice on the part of a State or a State-controlled authority or enterprise is a State aid prohibited under Article 92(1) of the Treaty. That function is exclusively reserved to the Commission and (in certain circumstances) the Council under Article 93(2) of the Treaty. ( 29 ) A judgment by the Court annulling a decision under Article 93(2) may leave the Commission (or the Council) with few substantive alternatives when it reconsiders its position, but that is another matter. The applicant, therefore, need not prove that tariff F constituted a State aid.

28.

The application of the State aids regime in Articles 92 and 93 of the Treaty entails the evaluation of complex economic situations, about which opinions may vary widely. The Commission is, therefore, entrusted with a wide power of appraisal, and the Court has made clear in its consistent case-law that it is not its role to substitute its assessment for that of the Commission. When examining the lawfulness of the exercise of the Commission's decision-making power in such fields, the Court restricts itself to examining whether a decision is procedurally flawed, contains a patent error of law or of fact, constitutes a misuse of power or exceeds the bounds of the Commission's significant freedom of evaluation. ( 30 )

29.

In order to establish that the Commission is guilty of a manifest error in the appraisal of the facts such as to justify the annulment of the impugned decision, the Belgian Government must satisfy the Court either that the decision was based on findings of fact which are objectively and evidently wrong, or drew from admitted facts secondary conclusions of fact which are objectively and evidently wrong. ( 31 ) A weighty burden of proof is imposed upon the applicant by virtue of the Court's very proper reluctance to substitute its evaluation for that of the Commission of factual issues on which differing views can legitimately be held. It must be demonstrated to the Court, to a reasonable degree of certainty, ( 32 ) that the Commission has committed errors of fact which undermine its ultimate conclusions.

30.

This does not mean that the Commission should simply wait in such cases to see if the applicant can adduce sufficient evidence to negate the presumption in favour of its factual assessment. The Court is entitled to the benefit of the best evidence that both of the parties can present to support their competing contentions. ( 33 ) In the instant case, for example, the Court posed a written question to the Belgian Government, in the light of a table introduced in evidence by the Commission, concerning the process of negotiation of the frontier price.

31.

The Belgian Government has introduced a number of related arguments in respect of the frontier price. It seeks to overturn the Commission's conclusion that tariff F did not favour Dutch ammonia producers compared with ammonia producers in the other Member States. ( 34 )

32.

The decision states at paragraph 8, fifth indent:

‘The French, Belgian and German distribution companies to which Gasunie supplied certain quantities of gas were charged, for the quantities intended for the nitrate fertilizer industry, a price referred to as a frontier price that was more or less identical to tariff F.’

The Belgian Government argues that the frontier price for Dutch gas exports to Belgium was, for considerable periods, higher than tariff F. This appears to be correct, but is not determinative of the issue. A misunderstanding has arisen from an infelicity of expression in the phrase just cited, on the part of the Commission. This should not, however, go to the validity of the decision.

33.

What is commonly referred to in the pleadings as the frontier price was a global price for all Dutch gas exports to Belgium, negotiated triennially: all of the gas passing through Gasunie's pipelines to the Belgian distribution company, Distrigaz, was sold at a single rate irrespective of its destination within Belgium. It is common cause that this price represented the culmination of a complex negotiation process. The frontier price was a volume-adjusted median dictated by the expected final selling price to various sectors in Belgium (domestic, commercial and industrial, broadly speaking) and the proportion of the total gas supply expected to be sold to each in the three-year period, along with an allowance for transport and other overheads as well as for Distrigaz' profit. The final selling price in each sector was largely dictated in turn by the price of the relevant competing petroleum-based fuel — gasoil in some cases, heavy oil in others — plus a premium for the greater advantages of gas. This price-fixing system is known as netback.

34.

As some customers (most notably domestic users) paid a much higher price for gas than others, the global or median frontier price would lie somewhere between the extremes of those charged in the different sectoral markets. It is not, therefore, surprising that it should at times have been higher than the final selling price to a lower-priced, high volume sector, such as industry. The Commission's reference to ‘a frontier price ... more or less identical to tariff F’ clearly means, upon examination, not the global frontier price, but that element in the calculation of that ultimate price which was based on predicted levels of gas sales and prices to the Belgian ammonia manufacturing industry. Whether the projected price for the Belgian ammonia industry, which was employed in devising the global frontier price, was, except for transport and other costs, more or less equivalent to tariff F is a separate, though central, issue which will be discussed below.

35.

The Belgian Government, the Commission and the Dutch Government are agreed on the basic criteria for the calculation of the frontier price, although the agents for both Belgium and the Commission indicated at the oral hearing that the mechanism described above presents a somewhat idealized picture of the negotiating process. In those negotiations, Gasunie would, not unnaturally, seek to justify, by reference to the netback criteria, a higher price, and Distrigaz a lower. The Belgian Government argues that the frontier price was a single price which could not, by a process of reverse reasoning, be separated out into its various constituent sectoral prices. Two reasons are given for this, namely the mode of calculation of the premium and the fact that, pursuant to the ‘take-or-pay’ principle, unclaimed quantities of gas had to be paid for at the frontier price, irrespective of the market segment for which they were initially envisaged.

36.

It may be observed, in passing, that the evidence given by Belgium on the calculation of the premium appears to be internally inconsistent, implying at one point that the premium was calculated per sector, and stating at another that it was the result of a commercial negotiation between the parties considering the market as a whole. ( 35 ) The ‘take-or-pay’ principle (by which Distrigaz effectively guaranteed the projected sales) might have caused inconvenience on occasion, if it had to be activated. None the less, and again in passing, the obligation to pay the frontier price for unused quantities of gas need not always have been less advantageous for Distrigaz than paying at a rate based on the final selling price in the underutilizing sector; this would depend on the level of that price relative to the frontier price. However, even if I accept that it is impossible retrospectively to deduce from the frontier price the elements initially employed in its calculation, this is hardly relevant so long as there is other evidence, whether from the negotiation process itself or elsewhere, that these elements included a final selling price for gas to the Belgian ammonia industry more or less equivalent to the Dutch tariff F.

37.

The Belgian Government argues that Belgian ammonia producers were threatened by the advantage accorded to their Dutch competitors by Gasunie, and that Distrigaz was compelled from 1986 to 1991 to sell gas to them, at a price equivalent to tariff F and at an annual loss of more than 100 million Belgian francs. ( 36 ) The Commission and the Dutch Government counter that these losses were more apparent than real, and were attributable only to Distrigaz' accounting practice. It has already been pointed out that the frontier price would quite normally have been greater than the final selling price of gas to certain sectors. Because Distrigaz' profit on gas sales was calculated, in all cases, by reference to the frontier price, it was inevitable that losses would be recorded on sales to these sectors taken in isolation. As long as the frontier price reflected the final selling price and sales volume in all sectors, these losses should have been negated by large profits in other sectors (e. g. the domestic) in which gas was sold at a price higher than the frontier price. We are thus drawn back to the question of the mode of formation of the frontier price. The Commission points out that Distrigaz made substantial and growing overall profits in the relevant period. The agent for Belgium responded at the oral hearing that Distrigaz' rate of profit (between 4% and 13% per annum) was greatly inferior to that of Gasunie (20%). Even if this is correct, however, too many imponderables (e. g. the higher price of Norwegian and Algerian gas imported by Distrigaz, the fact that Belgium, unlike the Netherlands, has no gas production capacity) intervene for this argument to lead to any conclusions about the relationship of the frontier price and tariff F.

38.

In the period between October 1984 and October 1986, Gasunie granted a rebate on a 20% portion, called the defensive portion (tranche défensive), of its total gas sales to Distrigaz. This rebate was used, on the recommendation of the Belgian price-fixing authority, the Comité de Contrôle, to underwrite the costs of according a tariff equivalent to tariff F to the Belgian ammonia producers. The Belgian Government initially maintained that there was no formal link between the defensive portion rebate, tariff F and the crisis in the ammonia industry; and that the defensive portion was conceded in response to a separate threat on the part of Distrigaz to avail of exceptionally cheap gas from the then Soviet Union.

39.

The Commission and the Dutch Government, on the other hand, stated that the defensive portion constituted approximately 7.5% of Belgian gas consumption, which corresponded to the quantities of gas utilized by the Belgian ammonia producers during the relevant period. The defensive portion was, therefore, evidence of efforts on the part of Gasunie to enable Distrigaz to charge a price equivalent to tariff F to ammonia producers in Belgium. If there was no formal link in the supply contract between the defensive portion and the price to be charged to ammonia producers, it was only because Gasunie was not empowered to dictate the marketing conditions applied by foreign distribution companies.

40.

The Belgian Government conceded implicitly in the latter part of the written pleadings and expressly at the oral hearing, that there was a connection between the defensive portion and the price of gas for ammonia producers, even if the threat of a Soviet purchase was used to strengthen Distrigaz' hand in the negotiations. However, Belgium now argues that it emerges implicitly from the Commission's position that it was only during the two-year period of application of the defensive portion that Gasunie sought to enable Distrigaz to match tariff F in the Belgian market without incurring losses; and that Distrigaz incurred such losses in the period after October 1986 when it continued to offer an equivalent price to the Belgian ammonia producers.

41.

This is denied by the Commission. It points to the variable character of the tariff F rebate, which was altered periodically in response to movements in the price of ammonia relative to that of gas for industrial users. The defensive portion was accorded by Gasunie when the tariff F rebate was at its maximum (5 cents/m3). It is implicit in the argument of the Commission that the frontier price took account of a final selling price for gas to Belgian ammonia producers which included a rebate, which rebate was lower than the maximum possible under tariff F in the Netherlands. When the tariff F rebate rose towards its maximum in the Netherlands, it was therefore necessary to supplement by some exceptional measure the rebate granted to Distrigaz through the frontier price. The defensive portion can be seen as such an exceptional measure.

42.

The Commission refers to a table outlining the level of the rebate granted to Dutch ammonia producers throughout the period of application of tariff F. ( 37 ) This shows that the maximum 5 cents/m3 rebate applied throughout the period of the defensive portion (with the exception of the last quarter), and did not apply afterwards, when, except for quite short periods, the rebate was either 2 or 2.5 cents/m3. It also shows that the maximum tariff F rebate applied throughout the period of about a year before that of the defensive portion, which is less helpful to the Commission's case. Given the circumstances in which tariff F was devised (in reaction to the Commission's disapproval of the preexisting pricing system), it is perhaps not surprising that there should have been a brief time-lag between its creation and its extension to Gasunie's foreign purchasers. ( 38 ) The argument of the Commission is otherwise persuasive, though contingent (yet again) on a conclusion in its favour on the separate issue of the mode of formation of the frontier price.

43.

The rudiments of the netback system of formation of the frontier price have already been described above, in a manner with which all of the parties concur. However, there is profound disagreement about its detailed application. The Belgian Government argues that purchasers of gas as a raw material rather than as a source of energy were not considered as a separate market sector. Thus, they were simply included in the general industrial market, for which the price of gas was fixed by reference to heavy fuel oil, and without regard to the price of competing raw materials (in the instant case, ammonia). Furthermore, the industrial segment of the frontier gas price was indexed to reflect movements in the price of heavy fuel oil, and not those of ammonia or other raw material prices. The Belgian submission that the frontier price was the volume-adjusted median of just two component prices, the domestic/commercial sector (calculated by reference to the price of gasoil) and the industrial, is founded on two documents supplied in its written pleadings. ( 39 ) However, both documents acknowledge that they give a simplified picture of the netback price formation mechanism. The agent for the Belgian Government added at the oral hearing that Gasunie had consistently refused to undertake to make special allowances for Belgian ammonia producers because it was only a partial supplier of the Belgian market, and did not feel that it should bear a cost not borne by Distrigaz' Norwegian and Algerian suppliers.

44.

The Commission argues, on the other hand, that the Belgian presentation of the formation of the frontier price on the basis of domestic/commercial and industrial prices is incomplete. The industrial sector embraced a number of distinct sub-sectors, among which was that of the ammonia producers. The Commission relies upon a table provided by the Ministry of Economic Affairs of the Netherlands to show that ‘fertilizers’ was a distinct sub-sector of the industrial sector. The market value of gas for the fertilizer sub-sector was determined (based for the largest part on the price of heavy fuel oil) separately from those for the paper, engineering, chemical and various other industrial sub-sectors. ( 40 ) The Commission submits that Gasunie applied a price more or less equivalent to tariff F to that portion of its gas sales to Belgium represented by the Belgian ammonia industry when the frontier price was being negotiated.

45.

The Court posed a written question to the Belgian Government, asking if the Commission's table could be employed as a basis for its decision in the instant case. This was denied by the Belgian Government, which repeated that no account was taken in the formation of the frontier price of distinct industrial sub-sectors. Furthermore, it stated, inter alia, that the table did not show that the gas price for the allegedly distinct fertilizer sub-sector was influenced by the price of ammonia as well as by that of heavy fuel oil; that the term ‘fertilizers’ was ambiguous, because manufacturers of certain fertilizers other than nitrate fertilizers could use gas as an energy source rather than as a raw material; that the table was only a theoretical model which did not reflect the actual negotiation process; and that even if correct, no provision was made for the indexation of gas prices for the fertilizer sub-sector to ammonia price movements after the negotiation of the frontier price. These are all, in themselves, valid observations.

46.

It must first be determined if the Commission table is an accurate (if admittedly theoretical) representation of the mode of formation of the frontier price. There is a direct conflict in this regard between the submissions of the Belgian Government, on the one hand, and the Commission, supported by the Dutch Government, on the other. Both sides have had ample opportunity to introduce evidence to support their competing submissions, supplemented by the Court's written question. The documentary evidence introduced by Belgium, which is expressed to be simplified in form, does not contradict directly the Commission's table, which purports to give a more detailed account of sectoral price formation in the process of formation of the frontier price. Notwithstanding the assertions of the Belgian Government, I must conclude that its case in this respect is not proven, and that the table gives an accurate account, as far as it goes, of the netback frontier price formation mechanism.

47.

However, I must also address the remaining Belgian arguments against reliance on the table to support the Commission's conclusion that Distrigaz was enabled to grant to Belgian ammonia producers a gas price more or less equivalent to the Dutch tariff F. It is immaterial that the table gives only a partial account of the price formation process, and cannot illustrate precisely how the various component prices were negotiated in practice, if it none the less outlines (which Belgium has been unable to disprove) the formal basis upon which the negotiations took place and to which they broadly adhered. I do not believe that the ambiguity of the term ‘fertilizers’ should impede us from reliance upon the table in the instant case, in the absence of evidence that manufacturers of non-nitrate fertilizers were included in this sub-sector, or were significant users of gas. ( 41 ) Nor do I believe that the failure of the table expressly to indicate the influence of the price of ammonia on that of gas destined for the fertilizer sub-sector is fatal. What is important is that the price of gas for the various sub-sectors is stated to be calculated (subject to the general effect of the price of heavy fuel oil) by reference to ‘market value per sub-sector’. It is acknowledged in the Belgian pleadings that a rebate similar to tariff F was accorded by Distrigaz to Belgian ammonia producers during much of the relevant period.

48.

There remains the difficult question of indexation. The Belgian Government appears to be correct in pointing out that no formal provision was made in the formation of the frontier price for the indexation to the shifting of ammonia prices relative to industrial gas prices of that component of the frontier price relating to the fertilizer sub-sector. ( 42 ) This is not surprising, as the price of heavy fuel oil (to which prices in the industrial sector were indexed) retained the most significant influence on gas prices in the industrial sector; the tariff F rebate, and any equivalent built into the frontier price for the ammonia producing sub-sector, was merely a palliative to temper the effect on manufacturers using gas as a raw material of a primarily energyoriented pricing system. The Commission does not address the question of indexation directly. Its position emerges by implication from its position in respect of the defensive portion. It argues that there was parallelism between the grant and revocation of the defensive portion, on the one hand, and the grant of higher and lower tariff F rebates in the Netherlands, on the other. As I pointed out above, it is implicit in the argument of the Commission that the frontier price took account of a final selling price for gas to Belgian ammonia producers which included a rebate, which rebate was lower than the maximum possible under tariff F in the Netherlands.

49.

I have already found the Commission's argument in respect of the defensive portion to be substantially correct, contingent on my conclusion on the wider issue of frontier price formation. In the absence of directly contradictory evidence from the Belgian Government, and in the light of my view that ammonia prices did influence the ultimate market value of gas for Belgian ammonia producers, I can now express my acceptance of the necessary implication, and condition, of that Commission argument: that the frontier price accommodated a rebate for those producers, which, circumstances strongly suggest, was lower than the maximum available under tariff F. That being the case, I conclude that the effect of indexation of the ammonia producing component of the frontier price to the movement of the price of ammonia relative to that of industrial gas was achieved, not by a formal indexation mechanism, but by an occasional exceptional measure, viz. the defensive portion.

50.

It is true that Distrigaz had to exert pressure on Gasunie to accord it the defensive portion. However, it would be unwise, in the context of commercial negotiations, to expect Gasunie to have volunteered further rebates to Distrigaz spontaneously, even in order to preserve the Belgian ammonia producing industry as an important outlet for its gas. This arises from the indirect manner by which Dutch gas reaches customers outside the Netherlands. The existence of an intermediary could have prompted Gasunie to seek to ensure that Distrigaz bore some of the cost (in lowered profits) of a rebate which both companies viewed as necessary. Distrigaz would naturally have sought to resist this tendency on the part of Gasunie. Distrigaz, as it claimed itself in its annual report for 1984, was largely successful in this regard, in winning the defensive portion. ( 43 ) Gasunie's commercial relationship with Distrigaz, and only indirectly with the Belgian ammonia producers, can similarly explain why minor defensive portions were not automatically extended to Distrigaz in response to the less significant fluctuations in the level of the tariff F rebate accorded by Gasunie between 1987 and 1991 to the Dutch ammonia industry, with which it had direct contractual links. This does not detract from the fact that Distrigaz was enabled by the rebate, included at all stages in the frontier price, to grant to the Belgian ammonia producers, during this period as well, a gas price roughly equivalent to that applicable to the ammonia industry in the Netherlands.

51.

The Belgian Government takes issue with the Commission's finding that tariff F was public and became a guide price on the Community's gas markets. Belgium argues that tariff F was not published unlike Gasunie's other tariffs, even though it was not confidential. The Commission relies, correctly in my view, upon the statement of the Court in its judgment in CdF Chimie, that ‘tariff F is a public tariff whose conditions of availability are public and perfectly open’; ( 44 ) the Belgian Government's argument should, accordingly, be rejected.

52.

Belgium introduced in the later stage of the written pleadings an argument that the price of gas for Dutch ammonia producers was further distorted by the repayment to them of the greater part of the environmental tax (milieuheffing) included in all of the Netherlands' gas tariffs. On the other hand, the frontier price included the environmental tax, which was thus imposed on all Dutch gas exports to Belgium, including those destined for the Belgian ammonia industry. Belgium also mentioned a supplement of 5 Belgian francs per gigajoule payable by the Belgian ammonia producers (without stating to whom it was payable). It is not clear why these matters were raised in this part of the pleadings.

53.

I believe that the Commission is correct to argue that in this instance, Belgium has introduced a new plea in law contrary to Article 42(2) of the Rules of Procedure of the Court. ( 45 ) That provision states:

‘No new plea in law may be introduced in the course of proceedings unless it is based on matters of law or of fact which come to light in the course of the procedure.’

The supplement mentioned by the Belgian Government, and the manner in which the environmental tax is imposed, may very well affect the price ultimately paid for gas by Dutch and Belgian ammonia producers, respectively. This is not sufficient to permit their introduction at the later stages (or at any stage) of pleadings relating to judicial review of the Commission's decision that tariff F did not constitute a prohibited State aid. Regarding the environmental tax, there was no indication in Belgium's initial pleadings that the combined effect of tariff F and another distinct rebate of a fiscal character was such as to undermine the Commission's conclusion that the former (as distinct from the latter) was not a prohibited State aid, or to show a factual error on the part of the Commission in its finding that Distrigaz was enabled to charge a price to Belgian ammonia producers more or less identical to tariff F. It is also clear that the milieuheffing rebate existed from 1988, and cannot be said to be a matter of fact which came to light in the course of the procedure. The same may be said of the supplement, about which virtually no information was provided by Belgium.

54.

If the Court were to examine the substance of the Belgian Government's argument about the environmental tax rebate, quite similar considerations should apply. The Commission submits that this rebate was a measure independent of tariff F, which could be considered in a separate procedure. Even if Gasunie's efforts to enable the Belgian ammonia industry to purchase gas at a rate similar to that applied to the industry in the Netherlands were undermined or countered by Dutch taxation policy, this would not reflect on tariff F or on Gasunie's export pricing, but on the fiscal measures in question. The circumstantial coincidence of tariff F and the environmental tax rebate — both affect the price paid for gas by ammonia producers — is not a legally relevant fact. They were at all times distinct measures. Therefore, the Commission was not in error for failing to consider the environmental tax rebate.

55.

The Belgian Government's final argument under this rubric is that, in setting its tariffs and thereby creating guide prices for the northwestern European gas market, Gasunie refused to take account of the difficulties occasioned to other gas distribution companies by these tariffs. This is essentially a reprise of the various pleas made in respect of the frontier price, and it is sufficient simply to refer to my conclusions in that regard in order to reject this argument. It is difficult to see how a judgment could be formed ori an objective basis that, in the give and take of export price negotiations, one party or the other achieved an unjustified advantage. The fact is that Distrigaz had the opportunity to make its case in the course of the negotiations and, as we have seen, had the capacity to extract significant concessions.

56.

This topic gives rise to a further criticism of the Commission's mode of expression of its reasoning in the impugned decision. The Commission states at paragraph 16 of the decision that, ‘Gasunie always achieved profits during the period in which tariff F was in force.’ The Belgian Government argues that this is irrelevant, as Gasunie's profit is fixed by contract at a constant 80 million Dutch guilders per annum. Gasunie receives its gas from NAM (Nederlandse Aardolie Maatschappij), a consortium owned jointly by Shell and Esso. ( 46 ) NAM exploits the Dutch gas fields on behalf of the Groningen Association (de Maatschap Groningen), in which NAM and the Dutch State have shares of 60% and 40% respectively. The Dutch State ultimately receives approximately 80% of the profits from gas sales. Gasunie pays NAM for the gas at a rate determined by its ultimate sale price on its various markets, with a deduction for Gasunie's transport and other costs and for Gasunie's agreed annual profit. The Belgian Government submits that Gasunie is only a cost centre {centre de coût) or a link in a chain, and that this arrangement cannot be described as a netback mechanism (as stated by the Commission in its pleadings) because the purchaser (Gasunie) bears no risk. As its costs and agreed profit are assured irrespective of the price charged to its customers, Gasunie's constant profit does not indicate that its prices are commercially justified.

57.

The Commission and the Dutch Government contest Belgium's submission that the relationship between NAM and Gasunie is not a netback arrangement. However, this semantic dispute does not as such concern us. More importantly, the Commission states that Gasunie's fixed annual profit of 80 million guilders represents simply the maximum that Gasunie can keep, any surplus being passed on to NAM. In fact, its effective profit during the period of tariff F was much higher, and it was to this effective profit that it referred in the decision. It will be observed that the effective profit is in fact (after Gasunie's 80 million guilders maximum is deducted) the Groningen Association's profit.

58.

Given the significance of the possible difference between contractual and effective profits, I must say that the Commission's statement in the decision about Gasunie's profit, quoted above, is unhelpfully opaque. However, I do not think it is fatally so (that is, so opaque as to amount to a failure to state a relevant reason for the decision). A person interested in the subject-matter of the decision is none the less alerted to this element in the Commission's reasoning. ( 47 )

59.

One may question as a matter of principle (although Belgium does not do so) the probative value of the finding that effective profits were made throughout the period of application of tariff F. This leaves unanswered the question whether any profits were foregone; ( 48 ) it also ignores the possibility that tariff F could have been a commercially justifiable device to safeguard long-term customers (who should eventually have given rise again to profitable sales) even if losses were made during that period. However, what is important is the degree of Gasunie's profit (or loss) relative to what an enterprise operating commercially under the conditions of the market in question could be expected to earn (or bear) while developing or seeking to preserve long-term prospects in that market.

60.

The Commission states, at paragraph 17 of the decision, that Gasunie's fixed and variable costs were well below tariff F, so that the company was able to increase its net revenue through sales at that price while also ensuring the maintenance of an important group of customers which it was in danger of losing. The Belgian Government points out that in its annual reports of 1990 and 1991, Gasunie stated its average purchase price for gas to have been a sum per cubic metre some 5 cents greater than tariff F.

61.

This argument is the obverse of Belgium's earlier argument in respect of Distrigaz' losses. It once again relies upon a global average cost of gas to measure the profit or loss generated by sales on a particular market at a particular price. In a netback system, as was pointed out above, losses in some sectors (measured against the average cost of gas) should be cancelled by profits in others. Any assessment of whether a real loss is made on any particular market should be made by reference to the cost of gas destined for that market, with an allowance also being made for management, infrastructural and other costs. The operation of the netback system ensured that Gasunie paid no more to NAM for gas destined for the Dutch ammonia producers than it received for the gas from those producers under tariff F, less a provision for its overheads and profit. Belgium's argument should, therefore, be rejected.

62.

The Commission explains Gasunie's need to maintain its customer base among ammonia producers in part by its loss of a substantial share of the French market in 1982. ( 49 ) The Belgian Government argues that this was a result of Gasunie's own policies, as the Netherlands had decided in the 1970s to reduce the volume of gas exported after the first oil crisis of 1973-74. The Commission and the Dutch Government counter that this decision was taken ten years before the introduction of tariff F; restricted only any increase in Dutch gas exports; and was unrelated to the loss of share on the French market, which was more likely to have been attributable to the trend towards diversification of sources of gas supplies. I agree with their submission that the Belgian Government has failed to show any relevant connection between the loss of French markets, and tariff F, on the one hand, and Gasunie's earlier export restraint, on the other. ( 50 )

63.

The Commission states, at paragraph 21 of the decision, that a nitrate fertilizer manufacturer has a choice either to produce ammonia (the chief raw material of such fertilizers) or to purchase it from other producers. ‘If the price of the gas which he uses in order to manufacture the ammonia he requires is too high, he will decide to purchase the ammonia, if possible, elsewhere at a lower cost than he would have to pay if he produced it himself (situation in Belgium, for example, in 1983).’ This is stated to have been the situation generally in the Community ammonia industry in the 1980s.

64.

The Belgian Government argues that Belgian nitrate fertilizer manufacturers annually import ammonia in order to cover a deficit between national production and consumption. However, the Commission has produced figures for Belgian ammonia imports from 1980-91, showing the proportion of Belgian ammonia needs supplied by imports to have risen from 38% in 1980 to 51% in 1983, and to over 70% in the early 1990s. ( 51 ) This demonstrates a considerable shift in the ammonia-sourcing practice of the Belgian nitrate fertilizer industry. The crisis in the Community ammonia industry to which tariff F was stated to be a response is documented by these figures, which show a massive decline in Belgian ammonia production even though Distrigaz applied for most of this period a tariff similar to the Dutch tariff F. It may be supposed that the decline would have been even more precipitate in the absence of such a concession.

65.

The Commission refers, at paragraph 22 of the decision, to the fact that, ‘from 1981 to 1991, the prices [for gas supplies to ammonia producers] found or estimated in the United States, Venezuela, Trinidad and Tobago and the Middle East have always been well below the Dutch tariffs’. The Belgian Government argues that these markets are not comparable with that in the Netherlands: while the low gas prices in the countries mentioned by the Commission represent the real value of gas in their respective markets, tariff F was lower than the value of gas on the European market. Evidence was presented that tariff F was consistently lower than the price of gas for industrial users in various Member State markets.

66.

The Commission, supported by the Dutch Government, submits that it did not seek to compare the various non-European markets with that in the Community, but to show that the ammonia producers of the former countries were in a position to pose a serious competitive threat to Dutch and other Community producers. Irrespective of the mode of gas price formation in the non-Community countries at issue, Belgium's argument does not undermine the Commission's conclusion that market conditions required a special price to be offered to Dutch ammonia producers — which price, however low, cannot be deemed to be artificial to the extent that it constituted a response to those market conditions.

67.

The Belgian Government also argues that ammonia producers have no available raw materials as an alternative to natural gas, and that, in this respect, they differ from enterprises which use gas mainly as a source of energy. Thus, it is claimed that they suffer greater technical and commercial difficulties in the event of interruption of supply and that gas must, therefore, have a higher intrinsic value for them than for other industrial sectors. Thus, price rebates were unjustified to a sector which is so exclusively dependent on natural gas. Belgium's argument about intrinsic value is really based on the ammonia producers' inability to abandon Gasunie in favour of a different provider of raw material. This argument is not persuasive. The manufacture of nitrate fertilizers in the Netherlands is not so dependent on Gasunie's supplies of natural gas as is the production of ammonia. Much of this case concerns the possibility of Dutch nitrate fertilizer manufacturers giving up their own ammonia production under pressure from cheaper imported ammonia. When such a decision is commercially feasible, or necessary, one cannot speak of gas being of greater intrinsic value to ammonia producers than for other industrial sectors; gas is not of value to the ammonia sector at all unless it can be employed in the production of ammonia competitive with other sources. Pricing (which allocates value in a market) is governed by effective demand rather than by a need-based idea of ‘intrinsic’ value. Effective demand is expressed through the ability and willingness to pay for purchases. It is irrelevant that the raw material needs of ammonia production can only be supplied by natural gas, while the energy needs of other sectors can be satisfied in a number of ways, if undertakings in other sectors can afford to purchase gas at the prevailing industrial tariff, while the competitiveness of ammonia producers is seriously undermined if they must do so. If the Commission's assessment that such was indeed the case is otherwise upheld, the Belgian argument just described must be rejected.

68.

The Commission states, at paragraph 24 of the decision, that compared with the maintenance of existing sales to Dutch ammonia producers, which requires no new marketing efforts, ‘[e] xport supplies, and in particular exports to Belgium, Germany and France, were less attractive in terms of price and in terms of the need to carry out new investment’. The Belgian Government argues that export prices during the relevant period were higher than tariff F, and that the existing infrastructure could have borne further exports without new investment. Furthermore, there was sufficient demand abroad for gas to justify a reorientation towards exports in line with any gradual decline in sales to Dutch ammonia producers. In the alternative, the Netherlands could have limited production, in order to prolong the life of its gas fields.

69.

No direct evidence was given that the decline in the custom of the Dutch ammonia producers would have been gradual (as Belgium asserts) rather than sudden (as the Commission believes), save the fact of the industry's significant investments in the Netherlands. The Commission points to the closure of numerous ammonia factories in the Community in the period 1992-93 when massive amounts of ammonia were imported from the former Soviet bloc. It is also possible to return to the evidence of the rapid decline of the Belgian ammonia industry in the 1980s, which decline would no doubt have been even more dramatic in the absence of a discounted tariff. The large size and small number of ammonia-producing operations in the Netherlands also suggests that any decline resulting in business failures or the abandonment by nitrate fertilizer manufacturers of that stage in their production would have had an immediately serious effect on Gasunie's sales volumes.

70.

The Commission and the Dutch Government recognize that the export price was, for reasons already discussed above, greater than tariff F. It does not follow that further large volumes of gas could have been exported at that price. In a netback system, the price Gasunie would have received would have depended on the sectoral market for which the gas was purchased, and on the price of gas on that market: that would have been the effective price received by Gasunie for the extra gas sold, less transport costs and distribution company profits, irrespective of how this might have been masked by an adjusted frontier price.

71.

They add that gas export contracts are concluded on a long-term basis in order to cater for predicted levels of demand, ( 52 ) so that it would have been very difficult at short notice substantially to augment supplies to another country. In the short to medium term, the export market was saturated, and export sales to markets in which long-term needs were already provided for could only have been increased on the basis of very low prices; the Dutch Government states that the most likely markets would have been the less profitable ones, such as that for electricity generation. The Commission pointed out at the oral hearing that Gasunie's gas sales to the Dutch ammonia producers were virtually equal in volume to all the gas sold in Belgium, which, according to the Dutch government, was already faced with an excess of Algerian gas. Increased exports would also have entailed technical problems relating to gas quality, pressure, and so on. As regards the possibility of restraining production, this would have resulted in a decrease in revenue, and therefore, in a less rapid payback on Gasunie's investments. The experts' report, commissioned by the Court in CdF Chimie, commented that while it is rational to discourage a too rapid depletion of domestic natural gas resources, the risk of losing present and potential future customers must also be considered.

72.

I am convinced by the submissions of the Commission and the Dutch Government that the Commission had ample grounds for concluding that Gasunie had little practical alternative to protecting its sales to Dutch ammonia producers by means of the tariff F rebate; Belgium's arguments should, therefore, be rejected.

73.

This is in reality the principal basis for the Commission's decision.

74.

The experts' report on tariff F, commissioned by the Court in CdF Chimie, stated that despite the economic and commercial advantages which tariff F may have secured for Gasunie, it ‘may as well at the same time have been politically decided in order to maintain special chemical productions within the Netherlands. ... [b] ut [that] from a commercial point of view that might also be in the economic interest of Gasunie’. ( 53 ) The experts also commented that, while there may have been an advantage for Gasunie in lowering certain sectoral prices in order to preserve market share and revenue, it was not in a position to assess ‘if the price level — although beneficial to Gasunie — could have been higher, and so wished by Gasunie — but not by the Government of the Netherlands’. ( 54 )

75.

The Belgian Government challenges generally the conclusion of the Commission that tariff F was commercially justified. It argues that tariff F was applied for political reasons, in order to confer an advantage on the Dutch ammonia industry. There are two ways of interpreting this argument, upon which the observations just quoted from the experts' report cast light.

76.

The first possible reading of the Belgian Government's argument is that political motivation or advantage fatally taints decisions taken within public sector enterprises, even when such decisions are also fully justifiable on commercial grounds. Such an argument is expressly made by Belgium later on in its pleadings. ( 55 ) If this argument were made, and upheld, the first quotation from the experts' report could be fatal for tariff F and, consequently, for the Commission's decision.

77.

The second possible argument is that while some tariff reductions might have been justified, and some advantage might have accrued thereby to Gasunie, tariff F was overgenerous and profits were foregone. In Van der Kooy, ( 56 ) the Court accepted the notion of competitive tariff reductions on the part of Gasunie (in that case for the Dutch horticultural industry), but warned cf the need to establish whether such competition is a real prospect, i. e., that, in effect, such reductions should be no more than is necessary, under the conditions of the market in question, taking account ‘not only of the different price levels, but also of the costs involved in conversion ...’.

78.

The first type of argument is not tenable. It is clear from the consistent case-law of the Court that the criteria for the identification of a State aid are objective. Once it is established that an investment, pricing decision, or other initiative is commercially justifiable, ( 57 ) the fact that the initiative also serves political interests is irrelevant. The Court stated in Deufil v Commission that Article 92 of the Treaty ‘does not ... distinguish between the measures of State intervention concerned by reference to their causűs or their aims but defines them in relation to their effects’. ( 58 ) Overtly political motives may prompt special scrutiny of the commercial arguments put forward for a particular aid-like practice, but there is no evidence in the instant case that the Commission did not investigate tariff F conscientiously and thoroughly.

79.

The second possible argument is consistent with the case-law of the Court, but has to be factually sustainable. Neither the experts commissioned by the Court, nor the Commission, found evidence that tariff F was anything other than a reasonable attempt on the part of Gasunie to derive maximum economic advantage from a difficult situation.

80.

The Belgian Government argues that the Commission should not have relied upon the second part of the experts' report, in which they moved on, unilaterally, from a ‘cost price’ analysis of the alleged savings achieved by virtue of the conditions of supply under tariff F, which was requested by the Court, to an ‘economic’ or ‘market price’ analysis of pricing decisions in response to market conditions. Belgium argues that the experts' findings were remarkably tentative and conjectural, being based on very little concrete evidence.

81.

The Commission counters that it was entitled to rely upon the market price analysis in the experts' report, which was referred to obliquely by the Court in CdF Chimie. ( 59 ) It continues that it also based its conclusions on evidence independently acquired by it in the course of its inquiries, relating to the difficulties in the ammonia industry and the connection between the cost of gas and that of ammonia production.

82.

Contrary to the submission of the Belgian Government, I do not believe that the mere invocation on the part of the Commission of the experts' report, some of the conclusions of which are rather speculative, should of itself be a ground for annulment of the impugned decision, as the Commission has clearly taken pains to substantiate the analysis canvassed in that report.

83.

Apan from the detailed arguments already discussed and rejected, the Belgian Government raises a factual argument, alleging a lack of synchronization between the movement of gas and ammonia prices and the fluctuation in the level of rebate granted under tariff F. ( 60 ) Two initial points should be made about this argument. First, it is clear that the Belgian Government does not challenge the principle that State-controlled enterprises can be commercially justified in responding to a threat to their sales by giving preferential prices to certain customers; it simply questions the proper observance of this principle in this case. The factual basis of any such argument must be examined thoroughly. Secondly, we are concerned in the present case with a system of rebates which applied over a number of years, and thus with the systematic observance of the principle of commercial justification. The Commission has to judge the effects of any system of alleged State aids. I believe that the exercise of its judgment should be overturned by the Court for manifest error of appraisal only where it is demonstrated that the measures in question, taken over the period of their application, are not systematically consistent with their stated justification. This is not to say that a short-term inconsistency will never be determinative; however, if any such lapse is minor in effect (in the context of the system as a whole), or can be plausibly explained in the circumstances of the case, the Court should not disturb the Commission's decision.

84.

The Commission presented in its written pleadings two tables, one charting the purchase price of imported ammonia in western Europe against the gas cost ( 61 ) of indigenous ammonia production between 1984 and 1991, the other setting out the levels of the Dutch industrial gas tariff (tariff E), the tariff F rebate, and the price per tonne of ammonia during the same period. ( 62 ) The agent for the Belgi n Government submitted at the oral hearing that these tables showed that tariff F was not a commercial response to market conditions. He argued that the maximum rebate was, on occasion, accorded to the Dutch ammonia producers at times when the price of competing imported ammonia was high, and a lesser rebate at other times when ammonia prices were low. ( 63 ) The agent for the Commission responded that the tariff F rebate had always varied, not in accordance with the absolute level of the price of ammonia, but by reference to the relative prices of industrial gas and imported ammonia. It must not be forgotten that the primary influence on the price of gas is that of its competing energy sources, namely oil products.

85.

Tariff F was indeed accorded to Dutch ammonia producers at its maximum level of 5 cents/m3 throughout 1984 and the first half of 1985, when the price of ammonia was at the highest level it reached in the 1983-91 period (at one point, some three times greater than its lowest price during this period). However, this was also the period when gas was most expensive, which appears to support the Commission's argument. Nevertheless, it is not sufficient to judge the matter simply by adopting an impressionistic approach to the figures; a more exact, if rather technical, analysis is demanded by the argument of the Belgian Government.

86.

By mapping the tariff F rebate figures on to the first table, charting relative gas and ammonia prices, I found that periods in which the gas cost of European ammonia production actually rose above the price of imported ammonia were those in which the maximum rebate applied, ( 64 ) or were those in which the rebate was increased from a low level to a higher one. ( 65 )

87.

There were also periods during which one is entitled to be somewhat more sceptical of the need for a rebate. ( 66 ) These periods correspond only in part to those mentioned by the Belgian Government, because some of the periods of high ammonia prices instanced by its agent witnessed almost simultaneous (and sometimes disproportionate) rises in gas prices, ( 67 ) the stability or gradual rise of gas prices as high ammonia prices began to fall, ( 68 ) or roughly equivalent falls in the high prices of both ammonia and gas, leaving the overall balance largely unchanged. ( 69 )

88.

However, the mere fact that the gas cost of European ammonia production was sometimes below the cost of imported ammonia does not automatically require the condemnation of rebates as uncommercial during such periods; while gas is the major cost in ammonia production, remaining processing expenses (between 10% and 25% of the total) ( 70 ) also had to be provided for.

89.

It is also illuminating to perform a simple mathematical operation on the basis of the figures supplied in the Commission's second table: to calculate the ratio of the price per tonne of ammonia to the price per cubic metre of gas at the tariff F rate. ( 71 ) It is impossible on the evidence before the Court to state what ratio represented breakeven point for European ammonia producers; this would have varied, of course, with the efficiency of plants in the utilization of their gas supplies in the production process. At the same time, it is clear that the higher the ratio, the better was the position of the European producer.

90.

Bearing in mind the hypothesis that gas pricing decisions followed ammonia prices (in so far as they were supposed, pursuant to the Commission's view, to respond to them), it can be observed that the level of rebate under tariff F tended towards decline when this ratio rose, unless such a rise was followed almost immediately by a decline in ammonia prices before Gasunie had responded. This is true, for example, of the cut in the rebate from 5 to 2.5 cents/m3 in the third quarter of 1986, when the price of gas was falling more quickly than that of ammonia, so that a more favourable ratio was obtained even on the lower rebate; conversely, increases in the rebate were required in response to difficult ratios in 1989-90, when gas prices rose while ammonia prices declined. This explains the rise in the rebate from 2 to 2.5 cents/m3 in the second quarter of 1989, even though the ratio of ammonia to gas prices had been the most favourable of the entire period of application of tariff F during the previous quarter.

91.

The one period when such responsiveness to the market is difficult to detect on the part of Gasunie is during 1984. Both 1983 and 1985 were very difficult years for European ammonia producers, judging by the Commission's price data, but the maximum 5 cents/m3 rebate applied throughout 1984 as well, when the ratio of ammonia to gas prices was quite high. It may be that even this relatively benign situation (relative, that is, to the crises of 1983 and 1985) still represented grave problems for the Dutch and other Community ammonia producers, so that a maximum rebate was justified even then. Belgium has given no evidence to the contrary, and this was the period in which Soviet bloc exporters made enormous strides. However, even if the level of rebate in 1984 were unjustifiable by the criteria just discussed (which possibility I raise only hypothetical^), two considerations must be noted: first, problems of acclimatization when applying a new pricing mechanism, which may excuse certain time-lags; ( 72 ) and secondly, the fact that the setting of the rebate by Gasunie became increasingly elastic over the rest of the decade in response to changes in market conditions. Thus, the tariff F price-fixing system can be said to have been, generally and systematically, commercially justified even if it suffered from some occasional problems of adjustment. This was the conclusion reached by the Commission in the exercise of its wide powers of evaluation of the operation over time of an exceptionally complex set of market forces. It cannot be said that the tariff F pricing system was manifestly unresponsive to the state of the markets for gas and ammonia. The Belgian Government's general argument to the contrary should therefore be rejected.

92.

The Belgian Government makes five distinct arguments under this rubric, for the annulment of the impugned decision.

93.

The Commission states, at the third indent of paragraph 8 of the decision, that ammonia producers were able to negotiate much more favourable terms with gas distribution companies in Italy, the United Kingdom and Ireland than were available in the Netherlands, Belgium and France. The Belgian Government submits that justifying tariff F by reference to such contractual regimes in other Member States is contrary to the position taken by the Court in Steinike and Weinlig v Germany, ( 73 ) that a Member State cannot excuse a violation of Article 92 of the Treaty by the circumstance that other Member States are also in breach of its provisions. The Belgian Government also states that the Commission failed to examine the specific conditions on the markets for gas or ammonia in Italy, the United Kingdom and Ireland.

94.

The Commission counters that it never suggested that the favourable contracts for the supply of gas concluded by the ammonia industry in these countries constituted State aid, nor that tariff F was justified for this reason. Indeed, the Commission pointed out in the text of the decision, that it considered these contracts to be commercially justifiable. As for the peculiar market conditions in those countries, it is not the reasons for lower gas prices, but rather, as with the third countries discussed above (the United States, Trinidad and Tobago, etc.), their effect on the relative competitive position of the Dutch ammonia producers which is relevant. Belgium's argument should, therefore, be rejected.

95.

The Commission quotes as follows from the experts' report commissioned by the Court in CdF Chimie, at paragraph 18 of the decision: ‘when natural gas is used as a feedstock in industrial processes, and when the price of the feedstock plays an essential role in determining the cost of the end product, not only prices of substituting feedstocks (or alternative production processes) play a role in determining the market price of the feedstock, but also the market price of the end product plays an essential role’. The Belgian Government argues that this is contrary to the position taken by the Court in United Brands v Commission, ( 74 ) that ‘the interplay of supply and demand should, owing to its nature, only be applied to each stage where it is really manifest’. Belgium submits that it is not permissible for the Commission to analyse Gasunie's tariff F by reference to price competition between Dutch, European and other undertakings on a different market, that for ammonia, which represents a different phase in marketing and production from that of gas employed in the manufacture of ammonia.

96.

Arguments were exchanged between the Belgian Government and the Commission on the applicability to a State aids decision of a dictum relating to Article 86 of the Treaty. I believe it is apparent that the reasoning of the passage quoted by Belgium is capable of extension to the field of State aids. If a State-controlled enterprise were essentially safe on its market from any adverse consequences of misfortunes suffered by its customers on their markets, special concessionary prices granted to its more vulnerable customers would be without commercial justification and would therefore be capable of constituting State aids prohibited under Article 92 of the Treaty.

97.

The circumstances in United Brands can be distinguished, however, from those in the instant case. The Court was not laying down any universal principle of law, other than the obvious one that, in assessing the effects of behaviour in markets, it is important to be clear that interplay between the supply and demand for different products is really present.

98.

In the instant case, the Commission's analysis turns on the degree of exposure of Gasunie to the risks borne by its purchasers (the ammonia producers) on a market ‘downstream’ from that for gas, viz. the market for ammonia. The Commission found there to have been, in effect, strong crosselasticity of demand on the part of nitrate fertilizer manufacturers (who tended also to be ammonia producers) between their own ammonia production and imported ammonia. Imported ammonia would be chosen if necessary in order to maintain their competitive position on the market for nitrate fertilizers, which would, in turn, gravely affect Gasunie's gas sales to ammonia producers. Therefore, ‘the interplay of supply and demand’ in respect of ammonia could ‘be really manifest’, to use the United Brands language, on the market for gas, and could legitimately influence the decisions of actors on that market. It was appropriate for the Commission to consider the connections between the two market stages in its State aids inquiry. Belgium's argument should, therefore, be rejected.

99.

The Belgian Government argues that the Commission failed in its reasoning to take account of the cost to nitrate fertilizer manufacturers of substituting purchases of imported ammonia for their own ammonia production. As was pointed out above, the Court considered in Van der Kooy that preferential gas tariffs (to supply the heating needs of the Dutch horticultural industry) could be objectively justified by economic reasons such as the need to resist competition on the same market from other sources of energy, the price of which was competitive. The Court continued: ‘In determining whether such competition is a real prospect account should be taken not only of the different price levels but also of the costs involved in conversion to a new source of energy, such as replacement and depreciation costs for heating equipment’. ( 75 )

100.

It is true that the Commission does not mention such conversion costs in the decision when it refers (e. g. paragraph 21) to the choice faced by nitrate fertilizer manufacturers between use of their own ammonia production and the purchase of imported ammonia. The Commission argues that it has shown sufficiently the existence of real competition between gas (when used in domestic ammonia production) and imported ammonia; and that substitution costs were negligible, as no conversion of nitrate fertilizer manufacturers' installations was entailed, but only the renunciation of one phase of production, subject to only marginal costs. This appears to be the case. The Belgian Government's argument should, therefore, be rejected.

101.

The Belgian Government argues that the Commission should have examined whether Gasunie renounced possible profits through the application of tariff F, irrespective of the profits it continued to make during the relevant period; and that the Commission failed to do so. Belgium also submits that any such commercial justification cannot relieve tariff F of its character as a politically motivated device to confer an advantage on the Dutch ammonia industry. However, no evidence or plea additional to the material discussed in great detail in section I(ix) above has been introduced; and I have already expressed my opinion on the correct approach to political motivation. ( 76 ) The Belgian Government's arguments were rejected at that point, and there is no need to examine them further.

102.

The Belgian Government submits that there is a contradiction between the fact that the conditions for the application of tariff F were objective, and the fact that it was, none the less, intended as a sectoral tariff to assist Gasunie's customers in the ammonia industry. Belgium argues that it was incumbent upon the Commission to explain this contradiction.

103.

The Commission responds that the grant of a special price to the ammonia industry was justified, while other industries, which were not in the same danger from competitors in third countries, did not need such an advantage, so that its grant to such sectors would have been illegitimate. This is consistent with the Commission's reasoning throughout, but it fails to address directly Belgium's point, which is that the (now) admittedly sectoral character of tariff F was not stated on its face, but was hidden behind purportedly neutral objective conditions relating to volume, interruptibility, and so on.

104.

Probably the best explanation of the conditions of application of tariff F, consistent with the reasoning of the Commission, is proffered in the experts' report:

‘If the tradition of the market prevents it, or because of publicly posted prices, it may not be possible to conduct a full price differentiation policy. In that case a block tariff system may be applied, as it is the case in the present Dutch pricing system. The block in the tariff system shall in that case be defined in such a way, that the gas company optimizes its revenue, coming as close as possible to the price differentiation situation. ... Posted prices can be combined with marginal cost related prices. In that case rebates or extra charges shall take into account the special conditions related to the delivery to certain customers [e. g. load factor, quantity, distance, seasonality]. ... However, these special conditions and rebates/extra charges can also serve as an indirect way to conduct price differentiation policy, although argued on a marginal cost basis.’ ( 77 )

105.

It is the norm in the Netherlands to sell gas by publicly posted tariffs. This situation can be contrasted with that, for example, in Germany, where full price differentiation is pursued through the negotiation of individualized contracts with major customers. Gasunie could seek to the best of its ability to replicate this position, within the constraints of the Netherlands market, by posting a tariff the objective conditions for which were designed with the ammonia industry in mind.

106.

I do not believe that the omission of such an explanation, or of any other which the Commission might have ventured, should lead to the annulment of the impugned decision for an error in the interpretation of Article 92 of the Treaty. It remains the case that the Commission indicated in its notice reopening the State aids inquiry into tariff F and seeking observations from interested parties that tariff F was essentially sectoral in character; ( 78 ) this resulted in turn from the findings of the Court in CdF Chimie. ( 79 ) Had the Commission taken the view that the tariff was not sectoral in nature (as it did in argument in CdF Chimie), ( 80 ) it would have had to justify the contradiction between the apparently neutral conditions for the application of tariff F and its actual effect in conferring an advantage almost exclusively on the ammonia sector. Having conceded that tariff F was sectoral in character, though justified by considerations specific to the sector in question, it was not incumbent on the Commission to address the matter further. As a matter of evidence, the suspicions of the investigator may be raised when appearance belies reality. Nevertheless, where the reality is explained and justified to the satisfaction of the investigator (and of the Court reviewing the investigator's decision), that initial suspicion, however natural, is allayed.

107.

The Belgian Government argues that the reasoning of the impugned decision does not conform to the requirements of Article 190 of the Treaty, because it is incomprehensible and insufficient. The Court has made clear in its consistent case-law that the Commission must support its decisions with a statement of the principal points of fact and of law which is clear and pertinent, even if succinct, and which should make apparent to interested parties and to the Court the reasoning underpinning the decision. ( 81 ) However, there are limits to the Commission's duty in this regard. As the Court pointed out in Remia v Commission:

‘[A] lthough under Article 190 of the EEC Treaty the Commission is required to state the factual matters justifying the adoption of a decision, together with the legal considerations which have led to its adopting it, the article does not require the Commission to discuss all the matters of fact and of law which may have been dealt with during the administrative proceedings. The statement of reasons on which a decision adversely affecting a person is based must allow the Court to exercise its power of review as to the legality of the decision and must provide the person concerned with the information necessary to enable him to decide whether or not the decision is well founded.’ ( 82 )

The statement about the interest of the Court and of affected persons in the Commission's reasoning imposes a duty on the Commission, but also places implicit limits on that duty: the Commission need not state either the obvious or the incidental or obscure in its decision, as neither is necessary to the review of the legality of the decision. Similarly, only sufficient detail need appear in the decision itself to alert the Court and individuals to the elements of the Commission's reasoning; these elements can be fleshed out in the exchange of evidence and argument in the course of annulment proceedings.

108.

The Belgian Government makes four arguments relating to defective motivation. The first is that the Commission does not set out in the text of the decision the incidents and conditions of application of tariff F. In the circumstances, this can hardly be fatal to the sufficiency of the Commission's reasoning. These details, including the finding that tariff F was essentially sectoral in character, were set out both in the judgment of the Court in CdF Chimie, and in the Commission notice reopening the State aids inquiry and inviting observations from interested parties. The Commission refers in the decision to both the judgment of the Court and the notice. Therefore, neither the Court nor interested parties can be said to have been deprived of any salient information about the tariff, and there is, accordingly, no defect in the reasoning of the Commission in this respect.

109.

Belgium's second submission is similar to the first: that the reasoning of the decision is flawed for its failure to point out that the tariff F rebate on tariff E became, after 1984, a variable rebate. This can be treated as a development of the first argument, rather than as a substantively new (and impermissible) one, as contended by the Commission. ( 83 ) However, it must fail for the same reason as the first: tariff F is described as variable both in the Opinion of Advocate General Mischo in CdF Chimie, ( 84 ) and in the notice reopening the Commission's inquiry under Article 93(2) of the Treaty.

110.

The Belgian Government argues, thirdly, that the Commission failed to explain how a tariff, which was available upon fulfilment of certain objective conditions, could be justified by reference to competition between gas and ammonia. This is, effectively, the same argument as that addressed above under the rubric of erroneous interpretation of Article 92 of the Treaty, under the heading ‘the specific character of tariff F’. It must be rejected in this context as well. The Commission was under no duty to explain this particular aspect of tariff F because it did not affect its ultimate decision.

111.

Belgium argues, finally, that the Commission failed, in its reasoning, to take account of conversion costs in the event of imported ammonia being substituted for their own ammonia production by nitrate fertilizer manufacturers. This is also a reprise of an argument already made in support of the contention of error of law, ( 85 ) and must again be rejected.

112.

The contentions of the Kingdom of Belgium that the decision challenged in the instant case is void for manifest error of appraisal of fact, for erroneous interpretation of Article 92 of the European Community Treaty and for defective reasoning should all be rejected. I conclude, therefore, that the Commission decision taken on 29 December 1992 terminating the procedure it had initiated pursuant to Article 93(2) of the Treaty regarding a preferential tariff system applied in the Netherlands to supplies of natural gas to Dutch nitrate fertilizer producers should not be annulled.

113.

The costs of these proceedings should be borne by the Kingdom of Belgium.

114.

The Kingdom of the Netherlands should bear its own costs.