61989C0152

++++

My Lords,

1. In these cases, Cases C-152 and C-153/89, the Commission has brought proceedings under Article 169 of the EEC Treaty for declarations that Luxembourg and Belgium respectively have failed to fulfil their obligations under Articles 95 and 96 of the Treaty as a result of certain features of the system by which excise duties are charged on beer in those countries. I will briefly describe the impugned system.

2. In Belgium and Luxembourg, pursuant to arrangements under the Belgo-Luxembourg Economic Union, excise duty on beer is levied under legislation which has been enacted in Belgium and has been incorporated by Luxembourg so as to apply also in Luxembourg territory. The duty is calculated on the basis not of the final product but of an intermediate product, namely the hot wort. In the course of the manufacturing processes by which the wort is transformed into beer a certain amount of liquid will inevitably be lost. That much seems to be common ground: 100 litres of hot wort, of a given density, will produce less than 100 litres of beer of the same density. The amount that is lost may be described as the wastage factor. The precise percentage is in dispute. Figures as low as 2% have been advanced by the Commission, while the defendant Member States maintain that the amount may be in excess of 10%.

3. In such a system the amount of duty borne by each litre of beer will depend on the efficiency of the brewery. A brewery that obtains 98 litres of beer from 100 litres of hot wort will pay less duty per litre of beer than a brewery that obtains only 90 litres of beer from 100 litres of hot wort.

4. In the context of a purely national market such a system, which appears to have been adopted in one form or another in many countries and whose origins are of considerable antiquity, may be perfectly satisfactory. Indeed, it may even be said to possess the merit of encouraging the efficient use of resources. But in the context of a common market, in which it is axiomatic that trade should not be distorted by peculiarities of the respective fiscal systems, such a method of calculating excise duty raises severe difficulties.

5. Some of the beer produced in Belgium and Luxembourg is exported to other Member States. When that happens, the exporter is entitled to reclaim the excise duty paid in respect of the beer in question. The amount of duty actually paid in respect of each litre of beer exported will of course depend on the wastage factor. Owing to the difficulty of evaluating that factor, the legislation in force in Belgium and Luxembourg has recourse to a flat-rate method of calculation, under which it is assumed that it takes 100 litres of hot wort to make 90 litres of beer of the same density. In other words, allowance is made for a wastage factor of 10%. If 90 litres of beer are exported, the amount of duty reimbursed is the amount that would have been charged on 100 litres of hot wort. That means of course that in the case of a particularly efficient brewery, with a wastage factor of less than 10%, the amount of duty reimbursed may exceed the amount that was charged. The Commission maintains that that is contrary to Article 96 of the Treaty, which provides that:

"Where products are exported to the territory of any Member State, any repayment of internal taxation shall not exceed the internal taxation imposed on them whether directly or indirectly."

6. Similar problems arise when beer is imported into Belgium and Luxembourg from other Member States. In order to tax imported beer in the same way as domestic beer it would be necessary to determine the quantity of hot wort from which the imported beer had been produced. In view of the obvious difficulty of doing that, the relevant national legislation again has recourse to a flat-rate method of calculation. But in this case, instead of 10%, a different wastage factor is taken into account. It is assumed that 100 litres of imported beer was produced from 105 litres of hot wort and excise duty is charged on the basis of that assumption. Thus, the volume of the end-product is increased by 5%. This gives a wastage factor of 5/105 or 4.7619%. A lower rate is of course more favourable to the imported product. But the result is that, if there are breweries in Belgium and Luxembourg with a wastage factor of less than 4.7619%, the amount of duty charged on each litre of beer brewed by them will be less than the amount charged on each litre of imported beer. The Commission maintains that that is contrary to the first paragraph of Article 95 of the Treaty, which provides that:

"No Member State shall impose, directly or indirectly, on the products of other Member States any internal taxation of any kind in excess of that imposed directly or indirectly on similar domestic products."

7. The cases thus raise two distinct questions concerning the compatibility of the Belgian and Luxembourg legislation with Articles 95 and 96 respectively. In my view, those two questions need to be examined separately, because the issues raised by them are not identical, even though they are obviously related. To treat them together, as has at times been done in these proceedings, creates confusion and makes it more difficult to identify the relevant issues and the appropriate criteria for the application of Articles 95 and 96. I shall therefore deal with these two aspects of the cases separately. Before doing so, I must deal with an issue of admissibility.

Admissibility

8. Luxembourg submits that the action is inadmissible on account of certain discrepancies between the reasoned opinion and the application. The reasoned opinion accused Luxembourg of infringing Article 96 by basing the repayment of excise duty on a wastage factor that exceeded the average wastage factor existing in the brewing industry of Luxembourg and of infringing Article 95 by taxing imported beer on the basis of a wastage factor that exceeded the average wastage factor existing in Luxembourg and in the countries that export to Luxembourg. The application seeks a declaration that Luxembourg has infringed Articles 96 and 95 by reimbursing excise duty on exported beer and charging duty on imported beer on the basis of a wastage factor that exceeds the average wastage factor in Luxembourg and exceeds in any event the wastage factor of certain breweries in Luxembourg.

9. The application thus differs from the reasoned opinion in two respects. First, it abandons the suggestion that the wastage factor in the exporting country might be relevant under Article 95. Secondly, it qualifies the idea that the flat-rate system need only take account of average wastage factors and refers in the alternative to the wastage factors of "certain" breweries in Luxembourg (presumably the most efficient ones). It is to be observed that the first point arises only in relation to Article 95; the second point arises in relation to both Articles 95 and 96.

10. Luxembourg is right to draw the Court' s attention to the above discrepancies and it cannot be denied that the Commission has caused considerable confusion by its inconsistency as regards the relevant criteria. The question whether that renders the application inadmissible should in my view be approached in the following way.

11. In proceedings under Article 169 of the Treaty, the Commission' s reasoned opinion serves to define the issues before the Court, and the Commission cannot raise, in proceedings before the Court, any alleged infringements which go beyond those recorded in the opinion. However, I do not consider that Luxembourg' s objection to the admissibility of the action can be sustained as regards the first discrepancy mentioned above, namely the reference in the operative part of the reasoned opinion, but not in the application, to the wastage factor in the country of origin for beer imported into Luxembourg. I observe that that reference appears only in the reasoned opinion addressed to Luxembourg, and not in the reasoned opinion addressed to Belgium. The explanation for that curious inconsistency appears to be that in the preliminary exchanges with the Commission Luxembourg, but not Belgium, invoked the wastage factor in the country of origin as the appropriate criterion. I shall consider the appropriateness of that criterion when examining the substance of the alleged infringement. But, as regards admissibility, it is sufficient to note that, although the operative part of the reasoned opinion refers to the wastage factor in the country of origin, the substance of the opinion does not relate to that factor. Indeed the Commission expressly states that the wastage factor on beer exported to Luxembourg is not to be taken into account. The reference to it in the operative part was obviously a mistake, which should not have induced any misunderstanding on this point.

12. The position is less clear-cut in relation to the second discrepancy mentioned above, namely the reference in the application to the wastage factor of certain Luxembourg breweries, which was not contained in the reasoned opinion. Here the Commission appears to have introduced a stricter criterion in the application than was contained in the reasoned opinion, which referred only to the average wastage factor in Luxembourg. Despite the Commission' s contention to the contrary, I think it is impossible to read the reasoned opinion as referring to anything other than average rates.

13. If the stricter criterion were retained, it might be found that the tax offended against that criterion, even though it was based on a wastage factor which did not exceed the average wastage factor in Luxembourg. It could be said that, in that event, an infringement would be established which was not that alleged in the opinion, contrary to the requirements of the Article 169 procedure which obliges the Commission to identify in its reasoned opinion the precise infringement to be established by the Court.

14. It is to be observed that this second discrepancy, unlike the first, arises also in the proceedings against Belgium, so that if the objection raised by Luxembourg were well-founded, the Court would have to consider whether it should examine the same objection of its own motion in the Belgian case, although it has not been raised by Belgium.

15. Although the matter is not free from doubt, I am of the opinion that the objection should not be allowed. I say so for two reasons.

16. First, the substance of the infringement alleged is that Luxembourg makes allowance for too high a wastage factor when reimbursing duty on exported beer and when charging duty on imported beer. There is room for argument, as is shown by the submissions of the parties on the substance of the case, about the appropriate criterion to be used in assessing the wastage factor, but that does not affect the essential issue, which is whether the wastage rate used is too high.

17. That point can be demonstrated by the fact that, if the infringement is established, it will be sufficient for the declaration to be made by the Court to take the form that Luxembourg has infringed Article 95 and 96 respectively by reference to the way in which the amount of the duty and the amount of repayment of the duty are actually assessed, without specifying how those amounts ought to be assessed. The reasoning leading to that declaration will of course make it clear what the appropriate criterion is: it might be the average wastage rate in the domestic industry, or the wastage rate in certain domestic breweries, or even some other criterion: that issue must be addressed on the substance of the case. The issue is not in my opinion foreclosed by the terms of the view recorded in the Commission' s reasoned opinion, when the opinion does record the essential allegation that the system of taxation infringes Articles 95 and 96 in that the wastage rate used is too high.

18. Secondly, I do not think that the procedural rights of Luxembourg have been prejudiced. Throughout the protracted pre-litigation procedure - the first letter from the Commission to the Luxembourg authorities was sent on 9 February 1982 - the parties were able to address all the issues exhaustively. The reference to the average rate arose from the fact that that was the basis on which Luxembourg defended the rate which it used - as also did Belgium. Although the debate focused on the average rate, there was debate also on what the lowest rate might be: thus, in their responses, dated 23 January 1984, to the Commission' s letter formally inviting them to submit their observations, both Governments denied that the wastage rate could descend to as little as 2%. In the proceedings before the Court, it was open to the Luxembourg Government to respond to both issues raised by the Commission, namely the average rate and the lowest rate, and to adduce evidence on both. In fact, however, although the Government has advanced certain arguments of a general nature, it has adduced no evidence of any kind, either in relation to the average rate or in relation to any other rate. Belgium, it is true, has done so, in relation to the average rate, but I shall consider on the substance of the case whether Belgium' s procedural rights have been infringed, even though Belgium has raised no objection to the admissibility of the action.

19. In any event I consider for the reasons set out above that Luxembourg' s objection to the admissibility of the action must be rejected. So I turn to the substance, and first to Article 95.

The compatibility of the Belgian and Luxembourg tax system with Article 95

20. The disparities between national systems for charging excise duties on beer and other alcoholic beverages have long been a matter of concern to the Commission. As early as 1972 it proposed a Council Directive on the harmonization of excise duty on beer (Journal Officiel 1972 C 43, p. 37). Under that directive the excise duty would have to be calculated on the basis of the final product. The fifth recital in the preamble states that "la neutralité de la concurrence, tant sur le plan national que sur le plan communautaire, peut être le mieux assurée par un système d' accise basé sur le produit fini". Similar considerations are mentioned in the Commission' s most recent proposal on the subject (Official Journal 1990 C 322, p. 11). According to an internal memorandum attached to the applications in the present cases, the Commission sees the acceptance of its proposals harmonizing excise duty on alcoholic beverages as "the first priority in the move towards the elimination of fiscal frontiers in the excise field". However, throughout the present proceedings the Commission has stated that it does not challenge the principle of charging duty on the basis of an intermediate product; it simply maintains that the wastage factors allowed for by Belgium and Luxembourg are too high.

21. It is true that Article 95 leaves Member States free to choose the system of taxation that they consider most suitable. However, that freedom must be subject to certain limits. In particular, the system of taxation used in each Member State must be transparent, at least to the extent that it must be possible to determine objectively whether the tax burden falling on imported products exceeds that falling on similar domestic products. Moreover, the system must be capable of even-handed application to domestic and imported products. There must be some doubt whether those criteria are fulfilled by the system in use in Belgium and Luxembourg.

22. Even if it is not objectionable in principle, the element giving rise to difficulties in the impugned system, from the point of view of Article 95, is that the basis of assessment used for imported products is different from that used for similar domestic products. Domestic beer is taxed on the basis of the quantity of hot wort used, without regard to the amount lost in transforming the hot wort into beer. The more beer that can be obtained from a given quantity of hot wort, the lower the duty is on each litre of beer. As a result, the efficient domestic producer enjoys a fiscal reward. The efficient producer in another Member State does not enjoy any such reward, because he - unlike his counterpart in Belgium and Luxembourg - is not taxed on the basis of the amount of hot wort used; instead, he is taxed on the basis of the quantity of the final product. Admittedly, that quantity is adjusted so as to take into account the notional amount of hot wort that will have gone into the production of the beer. But the adjustment is made on a flat-rate basis that takes no account of the actual efficiency of the foreign brewer.

23. The question then is whether the above situation can be reconciled with the Court' s case-law on Article 95. In Case 45/75 REWE v Hauptzollamt Landau [1976] ECR 181 the Court held that:

"... the first paragraph of Article 95 is infringed where the taxation on the imported product and that on the similar domestic product are calculated in a different manner on the basis of different criteria which lead, if only in certain cases, to higher taxation being imposed on the imported product.

This finding cannot be refuted by the claim that although the imported product is taxed at a flat rate whilst the domestic product is taxed according to a sliding scale this is because the investigations which would be necessary in the former case could not be carried out.

Even though it might indeed be impossible to introduce the same sliding scale for the increase or reduction of taxation on both domestic and imported products, it is nevertheless possible to impose a single flat rate or fixed charge on both products in order to observe the prohibition on discrimination laid down in Article 95."

24. In Case 127/75 Bobie v Hauptzollamt Aachen-Nord [1976] ECR 1079 the Court held that:

"The levying by a Member State of a tax on a product imported from another Member State in accordance with a method of calculation or rules which differ from those used for the taxation of the similar domestic product, for example a flat-rate amount in one case and a graduated amount in another would be incompatible with the first paragraph of Article 95 of the EEC Treaty if the latter product were subject, even if only in certain cases, by reason of graduated taxation, to a charge to tax lower than that on the imported product."

25. Both those cases resembled the present cases inasmuch as domestic products were taxed according to a graduated system, while imported products were taxed according to a flat-rate system. The Bobie case concerned excise duties on beer. Under the legislation then in force in Germany a graduated system was applied to domestic products whereby the rate of duty increased in proportion to the annual output of the brewery. After observing that Article 95 does not restrict a Member State' s freedom to establish the system of taxation which it considers the most suitable, the Court went on to state that:

"... it is the system of taxation chosen by each Member State in relation to a specific domestic product which constitutes the point of reference for the purposes of determining whether the tax applied to the similar product of another Member State complies with the requirements of the first paragraph of Article 95 or not.

If therefore a Member State has elected to apply to home-produced beer a graduated tax calculated on the basis of the quantity which each brewery produces in one year, the first paragraph of Article 95 is only fully complied with if the foreign beer is also taxed at a rate, the same or lower, applied to the quantities of beer produced by each brewery during the period of one year."

26. The following principles are established by the cases cited. First, the Member States are, as I have already observed, in principle free to choose the system of taxation that they consider most suitable. Secondly, the system applied to domestic goods constitutes the point of reference for determining whether imported products are taxed more heavily than similar domestic products. Thirdly, Article 95 is infringed even though it is only in isolated cases that the imported product is taxed more heavily than the domestic product; it is no defence to say that the heavier taxation charged on imported products in some cases is offset by the lower burden falling on them in other cases. Fourthly, Community law looks with suspicion on national legislation that taxes domestic products according to a graduated system and imported products according to a flat-rate system; if for reasons of policy a Member State decides that domestic production shall be subject to a graduated system of taxation (i.e. a system under which the amount of tax per unit varies in accordance with some factor peculiar to the producer concerned, such as his total output or wastage factor), then there are only two ways in which it can avoid falling foul of Article 95: either it must apply the same graduated system to imported products; or the rate applied to imports must be the lowest that is charged under the graduated system applicable to domestic products. Otherwise, the Member State concerned will be unable to avoid taxing imported products more heavily than domestic products in some cases.

27. The question that arises next is how those principles are to be applied to the present cases. According to the Commission, the relevant criterion is the wastage factor attained by breweries in Belgium and Luxembourg. If in those countries there exist breweries with a wastage factor of less than 4.7619%, then the amount of duty charged on each litre of beer produced by those breweries will be less than the amount charged on each litre of imported beer, for which a wastage factor of 4.7619% is assumed. The Commission maintains that there are breweries in Belgium and Luxembourg with a wastage factor of less than 4.7619% and that Article 95 is therefore infringed.

28. Belgium agrees that the relevant criterion is the wastage factor achieved by brewers in the importing State, though it does not admit that the figure of 4.7619% is excessive, having regard to the particular characteristics of the Belgian brewing industry. Luxembourg, on the other hand, contends that the relevant criterion is the wastage factor obtaining in the exporting country. It suggests that the legislation of Belgium and Luxembourg proceeds on the same assumption, inasmuch as it allows for a wastage factor of 4.7619% when taxing imported beer as against the 10% that is allowed for when repaying internal tax on exported beer. According to Luxembourg, the difference between the two rates is intended to take into account the greater efficiency of some brewers established in other Member States. As I have already pointed out, although the Commission now maintains that the wastage factor in the exporting State is irrelevant, it created the opposite impression in the operative part of the reasoned opinion delivered to Luxembourg (though not in the reasoned opinion delivered to Belgium).

29. Luxembourg' s argument is not without logic. Certainly, one way in which the defendant States could, at least theoretically, comply with Article 95 would be to extend to imported beer the graduated system applied to domestic production. If it were technically possible to calculate the amount of wort from which each consignment of imported beer had been produced, domestic and imported beer could be taxed in exactly the same way. The fiscal advantage enjoyed by efficient domestic brewers and the fiscal penalty suffered by inefficient domestic brewers could thus be extended to foreign brewers. The fact that each litre of beer produced by a foreign brewer with a wastage factor of 7% was taxed more heavily than each litre of beer produced by a domestic brewer with a wastage factor of 3% would not be contrary to Article 95. Further support for Luxembourg' s argument may be derived from the Bobie judgment (cited above), where the Court held that the "point of reference" for determining whether the discrimination prohibited by Article 95 exists is "the system of taxation chosen by each Member State in relation to a specific domestic product".

30. However attractive the solution described above might be in theory, there is clearly no possibility of it being applied in practice, because there is no reliable means of establishing the wastage factor to be taken into account in respect of each consignment of beer that is imported into Belgium and Luxembourg. Moreover - and this is plainly the decisive consideration - the method of taxation actually used does not attempt to establish the actual wastage factor for imported beer; instead, it takes a notional figure, which is applied to all imports.

31. In those circumstances the approach proposed by the Commission (and accepted by Belgium, subject to the dispute about the correct percentages) is clearly correct. That approach is to look at the imported product, namely beer, to ascertain the amount of tax charged on that product and to examine whether a smaller amount of tax is charged on the domestic product.

32. In comparing, in such circumstances, the amounts of tax imposed on the imported product and on the domestic product respectively, it is necessary to take the lowest rate borne by any part of the domestic production: in the present case, by beer with the lowest wastage rate. Neither Belgium nor Luxembourg accepts that approach; they continue to rely on the average rate. But the approach which I adopt follows, in my view, from the Court' s case-law as set out above. I mention, in passing, that average rates were expressly authorized by Article 97 of the Treaty in relation to the turnover taxes which were widely used in the Member States before the entry into force of the Community provisions on value added tax. However, Article 97, in authorizing the use of average rates for products or groups of products, both in the case of internal taxation on imported products and in the case of repayments on exported products, expressly requires that there must be no infringement of the principles laid down by Article 95 and 96.

33. The essential question under Article 95 is therefore whether the amount of tax charged on the imported product exceeds the amount of tax charged on any part of the domestic production. How then is this to be established?

34. In my view, it is not necessary for the Commission to prove that the tax system in force in Belgium and Luxembourg has actually resulted in some domestic beer being taxed at a lower rate than beer imported from other Member States. It is sufficient for the Commission to establish that the system is liable to have that result. If the Commission succeeds in establishing that, then it becomes incumbent on the defendant Member States to show that in no case does the system actually have that result. That transfer of the burden of proof is necessary, in my view, because of the lack of transparency of the system of taxation used in Belgium and Luxembourg. Where a Member State uses a system of taxation that makes it impossible to compare precisely the fiscal burden falling respectively on domestic and imported products, the onus is on that Member State to show that the system cannot result in a breach of Article 95. Support for that view may be found in the judgment in Case 45/64 Commission v Italy [1965] ECR 857, in which the Court held that, where a Member State introduces a flat-rate system for determining the amount of internal taxation repayable upon exportation to another Member State, it is for the Member State to show that the system always remains within the mandatory limits of Article 96. Although that case was concerned with Article 96, the same rule must also apply in the context of Article 95.

35. It is objected on behalf of Belgium and Luxembourg that they cannot be required to prove a negative proposition - a proof described as a "preuve diabolique". I do not consider that objection well-founded. There are different ways in which they could provide the necessary evidence. One way would be to establish the wastage rates for each of their breweries; it might result from those figures that in no case was imported beer taxed more heavily than the domestic product. Another way would be to establish, on the basis of technical reports, the minimum wastage rate which might be attained by the most efficient brewery; that might show that no brewery could have a lower wastage rate than that assumed in taxing imported beer, and hence that no domestic brewery is likely to be advantaged. Neither of these methods really involves proving a negative proposition, since although the final result may be framed in negative terms, the evidence would lead to affirmative findings. Nor does either method impose an intolerable burden on the defendant Governments.

36. I turn then to the evidence. This consists of four experts' reports - two submitted by the Commission (drawn up by Dr C. E. Dalgliesh and by Professor L. Narziss) and two by Belgium (drawn up by the Centre Technique et Scientifique de la Brasserie, de la Malterie et des Industries Connexes - C.B.M. and by Dr Wittmann of the Versuchs- und Lehranstalt fuer Brauerei in Berlin). In addition, supplementary reports were produced by Professor Narziss and Dr Wittmann. All the reports address the issue of a normal wastage factor, and I therefore consider the evidence first in that context. Perhaps only one thing is established clearly by the reports - namely, that it is extremely difficult to say categorically what can be regarded as a normal wastage factor. There are too many variables and there is insufficient clear information in the public domain. Indeed, the difficulty of establishing any average rate was repeatedly stressed by Belgium at the hearing. Moreover, there are considerable difficulties in determining the quantity of hot wort used even in relation to a specific quantity of beer. As for ascertaining the average wastage factor, the evidence shows that there are very wide variations depending on such factors as the type of beer in question, the age and condition of the plant, operational and management efficiency, the size of the brewery and the range of beers produced.

37. In the circumstances, the question may even arise whether the system of taxation in force in Belgium and Luxembourg is contrary to Article 95 simply because it does not possess sufficient transparency to permit a comparison to be made between the respective tax burdens falling on domestic and imported products. By choosing to tax domestic products on the basis of the hot wort and imported products on the basis of the end-product, Belgium and Luxembourg have made it so difficult to compare the incidence of taxation that even a plethora of experts' reports have not been able to resolve the problem. That alone might be regarded as sufficient to constitute a breach of Article 95. However, I think it is preferable to try to reach a view on the evidence before the Court, and to do so, for the reasons I have given, on the basis of the lowest rate likely to be attained in Belgium and Luxembourg.

38. I think it is clear from the experts' reports that the figure of 4.7619%, though perhaps not unreasonable as an average figure if the notion of an average can sensibly be used, does not represent the limit of technical achievement in the brewing industry and can be bettered by certain breweries producing certain types of beer. Thus Dr Dalgliesh states that "If administrative convenience demands a single value representative of good manufacturing practice in a reasonably well-equipped modern brewery, then 5% would be generous, and 4% would not be too low". Professor Narziss, in a more elaborate report that gives separate figures for each stage of the brewing process, concludes that an average brewery might attain a figure of 5% for ordinary beer. He states that the values given by him "could still be reduced somewhat, although this would require a great deal of technical work". It is clear from both those reports that the figure of 4.7619% will be bettered by the most efficient breweries. It must be assumed, unless evidence is adduced to the contrary, that at least some of the beer produced in Belgium and Luxembourg is brewed in such breweries.

39. The two experts' reports submitted by the Belgian Government do not, in my view, prove the contrary. The report drawn up by the Centre Technique et Scientifique de la Brasserie, de la Malterie et des Industries Connexes focuses on the Belgian brewing industry and emphasizes the special features that tend to lead to a higher wastage factor. It concludes, not surprisingly, that a minimum figure of 10% would be "more than reasonable". However, the objectivity of that report must be doubtful in view of the obvious interest of the authors in demonstrating that the 10% export refund accorded to Belgian brewers is justifiable. Moreover, the report is concerned primarily with the general situation of the Belgian brewing industry: it does not address the question what the minimum levels of wastage are that can be attained by a particularly efficient brewery either in Belgium or elsewhere. The last point applies equally to Dr Wittmann' s report, which concludes that a wastage factor of 10.25% can be regarded as appropriate for the Benelux brewing industry. That report is based on the author' s personal investigations and on information published in a German periodical in 1977. Dr Wittmann' s supplementary report examines the performance of four Belgian breweries that account for 70% of Belgian production. It might therefore be relevant to the establishment of an average figure, although there is no precise indication of how the sample was chosen; nor indeed is it clear that the basic data were established independently. I do not see how any of the reports submitted by Belgium can be said to have established that none of the domestic beer is produced with a wastage factor of less than 4.7619%.

40. I conclude, on the balance of probabilities, that the amount of tax charged on imported beer is likely to be greater than the amount of tax charged on some part of the domestic production, and therefore that the system of taxation must be held to infringe Article 95.

41. The survey of the evidence also leads to a further conclusion, which is that the defendant Governments' case has not in fact been prejudiced by the emphasis earlier placed by the Commission on average wastage rates. Since the average figure appears to be close to the flat-rate figure, and since there are acknowledged to be wide differences for different breweries and for different types of beer, it is abundantly clear that the Governments would not have been able to establish that the amount of duty imposed on imported beer is no higher than the amount imposed on any part of domestic production. The same holds good, mutatis mutandis, for Article 96.

The compatibility of the Belgian and Luxembourg tax system with Article 96

42. Article 96 provides that, where products are exported to another Member State, any repayment of internal taxation must not exceed the internal taxation imposed on them, whether directly or indirectly. A comparison must therefore be made between the internal taxation charged on beer produced in Belgium and Luxembourg and the amount that is reimbursed upon exportation. Whenever the latter amount exceeds the former, there is a breach of Article 96. That is so even though it is only in a limited number of cases that the repayment exceeds the taxation imposed. That is clear from the judgment in Case 45/64 (already cited), in which the Court held that the Member State that introduces a flat-rate system must "show that the system always remains within the mandatory limits of Article 96" (emphasis added). The point is confirmed by the REWE and Bobie cases cited above.

43. Moreover, in my opinion, it is sufficient once again for the Commission to show that the system is liable to have the result that the repayment will in some cases exceed the taxation imposed; it is not necessary for the Commission to show that the repayment actually exceeds the taxation imposed in a specific case. Although the burden of proof lies on the Commission, all that it need prove is that the system in force is liable to result in an excessive repayment of tax. Once the Commission has discharged that burden, for example by showing that it is technically possible to attain a wastage factor of less than 10%, it is then for the Member State to show that no brewer in its territory actually achieves such efficiency and that as a result the repayment of tax never in fact exceeds the amount imposed. Again the justification for taking a strict view lies in the lack of transparency of the tax system used in Belgium and Luxembourg; the same considerations apply as under Article 95. (In one respect the issues under Article 96 might differ from those under Article 95: the wastage rates are, as I have mentioned, different for different types of beer, and not all types of beer are exported. So the wastage rates might be higher on exported beers. But as the defendant Member States have not proved the point, it can in my view be disregarded.)

44. It is clear from what I have said in relation to Article 95 that the figure of 10% greatly exceeds the minimum wastage factor that is capable of being attained by a particularly efficient brewery. Neither of the defendant Member States has succeeded in proving that no brewer in its territory attains a wastage factor of less than 10%.

45. On the contrary, it is clear from the experts' reports that, even if 10% is not unreasonable as an average figure for the brewing industry of Belgium and Luxembourg as a whole, there must be breweries in those countries that attain a lower figure in respect of at least some of their production.

46. Thus the figure of 10.25% given by Dr Wittmann, which appears in the report submitted by Belgium and which focuses specifically on the brewing industry in the Benelux countries, is clearly an average figure indicative of what might be attained in a typical Benelux brewery. For each stage of the brewing process Dr Wittmann gives minimum and maximum figures and then a representative figure falling somewhere between the two extremes. The representative figures total 10.25%. If, however, the minimum figures are added up, they produce a total figure of 6.5%. I have already suggested that the figures in this report may be too high, but even if they are not it is clear, even from the evidence adduced by Belgium, that a brewery in Belgium or Luxembourg that maximizes efficiency will achieve a wastage factor of less than 10%. The reports drawn up by Dr Dalgliesh and Professor Narziss confirm even more clearly that an efficient brewery can attain a wastage factor well below 10%.

47. In view of the above finding it seems more than likely that some of the beer exported from Belgium and Luxembourg will qualify for a refund of taxation in excess of the amount imposed. Consequently, on the view I take, it is clear that Belgium and Luxembourg have failed to comply with their obligations under Article 96.

Conclusion

48. In conclusion I am of the opinion that the Court should:

(1) Declare that, by assessing excise duty on domestically produced beer on the basis of the amount of hot wort and by assessing duty on beer imported from other Member States on the basis of the volume of the end-product, increased by a flat-rate figure of 5% to arrive at the notional amount of hot wort used to produce the imported beer, Belgium and Luxembourg have failed to fulfil their obligations under Article 95 of the Treaty;

(2) Declare that, by calculating the amount of excise duty to be reimbursed when domestically produced beer is exported to other Member States on the basis of an assumption that 10% is lost in converting the hot wort into beer, Belgium and Luxembourg have failed to fulfil their obligations under Article 96 of the Treaty;

(3) Order the defendants to pay the costs.

(*) Original language: English.