OPINION OF MR ADVOCATE GENERAL WARNER
DELIVERED ON 27 NOVEMBER 1980
My Lords,
This case comes before the Court by way of a reference for a preliminary ruling by the Gerechtshof of Amsterdam. It arises from the prosecution of a Dutch cheese manufacturer, the Koninklijke Kaasfabriek Eyssen BV (which I shall call “Eyssen”), for offences against Dutch legislation relating to the addition of preservatives to processed cheese; and it raises questions as to the compatibility of that legislation with Articles 30 to 36 of the EEC Treaty.
The Dutch legislation about additives to foodstuffs is based on what has been described to us as the “positive list” system, i.e. a system under which certain additives, which are reckoned to be harmless, are expressly authorized, and all other additives are, by and large, forbidden. That system prevails in a number of Member States and has been adopted as the basis of the Council's “harmonization” directives on the subject — see the Directive of 23 October 1962 on colouring matters, the Directive of 5 November 1963 on preservatives, the Directive of 13 July 1970 on anti-oxidants and the Directive of 18 June 1974 on emulsifiers, stabilizers, thickeners and gelling agents.
The offences with which Eyssen was charged were that it had in stock, with a view to their sale for human consumption, quantities of cheese spread and of what is described as “rindless cheese” to which an antibiotic called “nisin” had been added as a preservative, contrary to Article 8 (h) of the “Smeltkaasbesluit (Warenwet)”, which is a Dutch statutory instrument relating to processed cheese.
A good deal of information about nisin has been placed before the Court. I do not propose to go into it all. I found particularly helpful what we were told at the hearing by the Head of the Foodstuffs Division of the Commission's Directorate-General III (Internal Market and Industrial Affairs), Mr Kinch. This complemented information contained in the Gerechtshof's file and information given to us by Eyssen, by the Dutch Government and by the German Government.
It appears that nisin occurs naturally in cheese. It can also be added to foodstuffs, particularly to processed cheese and to canned foods, as a preservative. We were told by the German Government that its chemical structure was not yet completely known. Be that as it may, nisin has been the subject of a Report made by the Joint FAO/WHO Expert Committee on Food Additives (WHO Tech. Rep. Ser. 1969, No 430). From experiments conducted on animals, that Committee deduced that a daily intake of 3300000 international units of nisin per kg of their body weight had no adverse effect. To determine the acceptable daily intake for human beings, the “a.d.i.”, they divided that figure by a “safety factor” of 100, which took into account, among other things, the different ages of people who might be consuming the product. The Committee thus obtained an a.d.i. of 33000 international units, equivalent to 0.78 mg., per kg. of body weight. The problem then is to translate that a.d.i. into acceptable levels of nisin in particular kinds of food. That problem was described by Mr Kinch as one of “enormous difficulty”. Its solution depends, to a large extent, upon establishing the composition of people's diet, which, Mr Kinch said, was “a very long and complicated job”. His Division had started on it but was far from having completed it. The job is made the more complex by the fact that tastes vary from country to country. That was illustrated by some statistics quoted by the Dutch Government, which showed for instance that the consumption of cheese per head of the population in the Netherlands is more than twice what it is in the United Kingdom. In answer to a question of mine Mr Kinch confirmed that there is no authoritative international guide as to what amount of nisin in cheese is safe. The furthest he could go in that respect was to refer to a recommendation, which was mentioned to us also by other parties, emanating from the joint FAO/WHO. “Codex Alimentarius” Commission to the effect that 12 1/2 mg. of nisin per kg. of cheese (including the nisin naturally present) should be regarded as safe. Mr Kinch told us, however, that the “Codex” procedure concerning that recommendation was not yet complete, from which I infer that there is no means of knowing whether or not the recommendation will eventually be accepted.
So far as Community law is concerned, there is the Council Directive of 5 November 1963 to which I have already referred on “the approximation of the laws of the Member States concerning the preservatives authorized for use in foodstuffs intended for human consumption” — Directive 64/54/EEC. All it says, however, about nisin is: “This Directive shall not affect the provisions of national laws concerning ... nisin” — see Article 6 (b).
The laws of the Member States about the addition of nisin to processed cheese fall into three groups. In France and in the United Kingdom its use is permitted without limit, subject, in the case of the United Kingdom, to an obligation to specify on the label that the cheese “contains permitted preservative”. In Belgium, Denmark, Ireland and Italy, the use of nisin is permitted up to specified limits, which vary. In Germany, Luxembourg and the Netherlands, the addition of nisin is forbidden except in products destined for export and except, in the case of the Netherlands, where specific authorization has been obtained from the competent minister on grounds of technical necessity.
It appears that an expert consultative body in the Netherlands (the “Adviescommissie”) has recommended a relaxation of Dutch law so as to permit the addition of nisin up to a specified limit and that that body has drafted new regulations to give effect to its recommendations. Those recommendations and drafts have, however, no legal effect.
Eyssen had not, when the proceedings against it were instituted, in December 1977, applied for any authorization from the Minister, though is has done so since. Nor is there any suggestion that the cheese that Eyssen had in stock was for export, though is appears from what is said by the Gerechtshof in its order for reference that Eyssen in fact manufactures both for the Dutch market and for export. It also appears from the order for reference that the quantity of nisin in Eyssen's products was within the 12 1/2 mg. limit recommended by the “Codex Alimentarius” Commission, which the Gerechtshof refers to as “the recommendation of September 1976 by the FAO/WHO committee of government experts”.
The proceedings came at first instance before the Arrondissementsrechtbank of Alkmaar. That court, by a judgment dated 14 November 1978, found the facts alleged against Eyssen proved, but acquitted it on the grounds that the amount of nisin that it had added to its products was desirable in order to prevent their deterioration and presented no danger to health; that Article 6 (b) of Directive 64/54/EEC left Member States free to authorize the addition of nisin; and that the Dutch legislation under which Eyssen had been charged was incompatible with Articles 30 and 34 of the Treaty.
The prosecution having appealed to the Gerechtshof, that court has referred to this Court a question whereby, after reciting the salient facts, and in particular that “nisin must be considered to be not absolutely but relatively (slightly) harmful to human health”, it asks, in short, whether, despite what Article 36 says about prohibitions justified on grounds of the protection of the health and life of humans, the provisions of the Treaty as to the free movement of goods should be interpreted as meaning that a provision such as that in Article 8 (h) of the Smeltkaasbesluit, forbidding added nisin in processed cheese except where an exemption is granted, is unlawful, and whether it makes any difference if such exemption is granted for processed cheese that is clearly intended for export.
The reference there to the exemption for processed cheese clearly intended for export reflects, I think, an argument that was put forward on Eyssen's behalf in the Dutch courts, to the effect that the practical result of the Dutch legislation was that a Dutch cheese manufacturer must set up two production lines, one for cheese destined for the Dutch market and one for cheese destined for export to countries where the addition of nisin is permitted, and that the requirement that he should do that amounted to a measure having equivalent effect to a quantitative restriction on exports of a kind forbidden by Article 34 of the Treaty. In this Court, however, Eyssen has relied more on Article 30 of the Treaty as interpreted by the Court in Case 8/74 Procureur du Roi v Dassonville [1974] 1 ECR 837, Case 104/75 De Peijper [1976] 1 ECR 613, Case 120/78 Rewe v Bundesmonopolverwaltung für Branntwein (the “Cassis de Dijon” case) [1979] ECR 649 and Case 788/79 Gilli (26 June 1980, not yet reported). Eyssen's argument, shortly stated, is that the Dutch prohibition on the addition of nisin is a measure having equivalent effect to a quantitative restriction on imports from Member States where such addition is permitted.
Before I consider that argument I must deal with two minor points raised by the Commission.
First the Commission referred to Article 22 (1) of Council Regulation No 804/68, the basic Regulation on the common organization of the market in milk and milk products. That provision forbade, in the internal trade of the Community, “any quantitative restriction or measure having equivalent effect”. The reference to it was, however, in my opinion, inapposite, because since the end of the transitional period, when Articles 30 to 36 of the Treaty took direct effect, provisions such as Article 22 (1) have become spent — see for instance Case 251/78 Denkavit Futtermittel v Minister fiir Ernährung, Landwirtschaft und Forsten [1979] ECR 3369 (paragraph 3 of the judgment).
Secondly, the Commission submitted that Eyssen could not avail itself of Article 30 because it was not an importer. The Commission referred in that connexion to Case 68/79 Just v Danish Ministry of Fiscal Affairs [1980] ECR 501, where it was held that Article 95 of the Treaty protected only products imported from other Member States and could not be invoked by a domestic producer. That authority is not, however, in my opinion in point. It is well established that Article 95 forbids the imposition by a Member State of higher taxation on products imported from other Member States than it imposes on domestic products, but does not forbid the converse — see Case 86/78 Peureux v Directeur des Services Fiscaux [1979] ECR 897 (paragraph 32 of the judgment). That that is so is indeed a conclusion inescapably to be drawn from the very wording of Article 95. Article 95, in other words, is one of the rare provisions of the Treaty that permits “reverse discrimination”. Article 30 is different, and much wider in scope. In the familiar words of the judgment of the Court in the Dassonville case, it forbids “all trading rules enacted by Member States which are capable of hindering, directly or indirectly, actually or potentially, intra-Community trade”. Thus Article 30 does not merely forbid discrimination against goods from other Member States. It can apply even where there is no such discrimination. That is illustrated by the Gilli case, where the prohibition in question, that of the marketing in Italy of apple vinegar, applied whether the vinegar was made in Italy or in another Member State. The point was particularly emphasized by Mr Advocate General Capotorti, who said:
“The character of a national rule falling within the scope of Article 30 lies in the fact of its being a complete barrier to the importation from other Member States of a given product: that fact is enough for it to be regarded as offending against the principle of the free movement of goods even if the national rule prevents the marketing of the product in question irrespective of the country in which it has been manufactured”.
I do not overlook that, in the Gilli case, the accused were importers. But the decision of the Court did not turn on that. It turned on the circumstance that the Italian rule, looked at objectively and independently of the facts of the particular case, constituted an obstacle to intra-Community trade. Similarly, in my opinion, the fact that Eyssen is a Dutch manufacturer and not a Dutch importer cannot affect the answer to be given by Your Lordships to the Gerechtshof's question in this case.
So I turn to that question, which in the strict analysis embodies three questions :
(i)
Is a provision such as that in Article 8 (h) of the Smeltkaasbesluit prima facie forbidden by Article 30 of the Treaty?
(ii)
Is such a provision prima facie forbidden by Article 34 of the Treaty?
(iii)
If such a provision is prima facie forbidden by those articles or either of them, is it saved by Article 36 of the Treaty?
I need not detain Your Lordships for long on the first two questions.
Eyssen and the Commission both submitted that Article 8 (h) was prima facie invalidated by Article 30. No-one argued the contrary. It seems to me clear that, on that point, Eyssen and the Commission were right.
Article 8 (h) has the consequence that processed cheese containing added nisin made in a Member State where such addition is permitted may not be sold in the Netherlands. Manifestly that restricts intra-Community trade in such cheese.
That being so it is not in my opinion necessary to come to a concluded view on Article 34. The Dutch Government and the Commission both submitted that Article 34 was inapplicable since exports from the Netherlands were exempted from the prohibition on the addition of nisin and since there was no legal requirement in any Member State that nisin should be added to processed cheese. I am not sure that that is necessarily the end of the matter. One can imagine a situation in which, for instance, a Dutch manufacturer, wishing to export processed cheese to the United Kingdom, found that, without the addition of nisin, his product would be uncompetitive there because it would deteriorate faster than the local product. There would then be an economic necessity for the Dutch manufacturer to add nisin to processed cheese destined for export to the United Kingdom. Eyssen's “two production lines” argument might then become valid. Whether that is so or not, however, depends on questions of fact which are for the Dutch courts.
So I come to the real problem in this case, which is as to the effect of Article 36.
Eyssen and the Commission contended that that provision could not save Article 8 (h) from invalidity; the Dutch and German Governments contended that it did. The arguments of Eyssen and of the Commission in support of their contention coincided on some points, but diverged on others. The arguments of the Dutch and German Governments to the contrary were much the same, except that those of the German Government were, in some respects, more far-reaching than those of the Dutch Government. Your Lordships have read and heard those arguments, and I do not think that any useful purpose would be served by my rehearsing them.
There is a familiar line of authority in this Court on which the arguments of the Commission were for the most part based, and on which those of Eyssen were largely based. It is the line of cases in which the Court has drawn conclusions from the general proposition, which is not challenged in the present case, that the purpose of Article 36 of the Treaty is not to reserve certain matters to the exclusive jurisdiction of the Member States, but merely to allow national legislation to derogate from the principle of the free movement of goods to the extent to which such derogation is and remains justified on the grounds mentioned in that Article. Those conclusions, so far as here relevant, have been:
(i)
that derogations from the principle of the free movement of goods enacted by the legislation of a Member State for the protection of the health and life of humans are compatible with the Treaty only if “necessary” for that purpose;
(ii)
that the burden of showing that such derogations satisfy that test lies on the Member State concerned; and
(iii)
that where the validity of such derogations is challenged before a court or tribunal of a Member State, it is for that court or tribunal to apply the test and decide the question.
In no way do I propose to suggest to Your Lordships that the Court has been wrong in drawing those conclusions. It is, however, noteworthy that they were drawn in cases each one of which was concerned with administrative procedures or requirements prescribed by the legislation of a Member State to guard against an undoubted danger to public health. Thus, for instance, the De Peijper case (already cited) was about documentation prescribed by Dutch legislation to guard against the marketing in the Netherlands of defective pharmaceutical products, Case 35/76 Simmenthal v Italian Minister for Finance [1976] 2 ECR 1871 was about veterinary and public health inspections designed to protect the Italian public against imports of diseased meat, Case 153/78 Commission v Germany [1979] ECR 2555 was about a requirement of German law as to the origin of the ingredients of meat products, of which the (alleged) purpose was to safeguard consumers from unwholesome meat, and Case 251/78 the Denkavit Futtermittel case (already cited) was about a complex German system of certificates, inspections and licences intended to avoid the presence of pathogens in animal feeding stuffs.
This case is different. No-one alleges that nisin is inherently harmful, as are defective pharmaceutical products or pathogens in meat, meat products or animal feeding stuffs. Nisin occurs naturally in cheese. What may be harmful is the consumption of an excess of it. No-one can say for sure how much of it in cheese is too much. Nor does the Dutch legislation here in question prescribe a procedure to guard against the presence of nisin in cheese. It simply forbids the sale in the Netherlands of cheese containing added nisin, unless an exemption has been granted on technical grounds. The purpose of that legislation is plainly to limit the consumption of nisin by people in the Netherlands.
The question is whether such legislation is “justified” within the meaning of that term in Article 36. The test of necessity laid down by the Court in the cases to which I have referred is in my opinion here inappropriate. Nor, in my opinion is it right to deduce from those cases either that, as was suggested by both Eyssen and the Commission, though Counsel for the Commission resiled from the submission at the end of the hearing, the onus is on the Dutch Government to show that nisin is harmful, or that, as was suggested by the Commission, but disputed by Eyssen, the question must be left to the Dutch courts to decide.
In a case where the question is whether an administrative procedure designed to guard against an undoubted danger to public health is “justified” within the meaning of Article 36, it makes sense to hold, as this Court has held, that it is for the competent national court to decide whether, in all the circumstances, that procedure is necessary for the purpose; and to hold also that the burden of showing that the procedure is necessary lies on those who assert it. But in a case like this, concerning a substance that is potentially harmful if taken in excessive quantities, where there is no generally recognized standard as to the amount of it that is acceptable in particular foodstuffs, and where the question is whether national legislation designed to limit its consumption is “justified”, that approach is in my opinion inapt. It is inapt because no-one can then say what is “necessary”, let alone discharge the burden of proving that it is necessary, and inapt also because to leave the decision of such cases to national courts must almost inevitably result in the very lack of uniformity in the interpretation of the Treaty that Article 177 is intended to avoid. The Gerechtshof was therefore, in my opinion, right in thinking that, in such a case, it is for this Court to rule. It is, to my mind, of some relevance in that connexion that the Court has, in at least two cases, taken upon itself to state that a particular product could not be regarded as harmful to health: the “Cassis de Dijon” case and the Gilli case.
Should Your Lordships, then, hold that legislation such as the Dutch legislation here in question cannot be justified on grounds of the protection of the health of humans? I think not. There is, so far as I can discern, no reason for saying that that legislation constitutes “a means of arbitrary discrimination or a disguised restriction on trade between Member States”. Nor is there, in my opinion, any reason for giving effect to Eyssen's argument that, since the unrestricted use of nisin is allowed in France and in the United Kingdom, and since the French and British authorities may be assumed to be as responsible as the Dutch, the restrictions imposed by the Dutch legislation cannot be justified. It is, at the very least, possible that the Dutch authorities are wiser than the British and the French. In any case, this is a matter over which, plainly, pending the enactment of a “harmonizing” directive, or at all events the adoption of a generally accepted international standard, Member States retain a measure of discretion; and I do not think that the Netherlands can be held to have exceeded the bounds of that discretion in this instance. Much less do I think that Your Lordships should give effect to the Commission's argument that, if this Dutch legislation is held to be justified, the incentive for Member States to agree to further harmonization legislation will be reduced. Your Lordships' duty is to decide cases that come before this Court in accordance with what Your Lorships conceive the law to be, and not with an eye on such political considerations.
Your Lordships should, however, in my opinion, decide this case on the narrow ground that, in the light of the evidence placed before the Court about nisin, and in view of the neutral stance as to nisin adopted by the Council in Article 6 (b) of Directive 64/54/EEC, the Dutch legislation relating to that substance is compatible with Community law as it stands at present. I say that because arguments were submitted to us on behalf of the German Government of which the purpose was, so it seemed to me, to secure Your Lordships' acceptance of a much wider proposition, namely that, having regard to the general pollution of the environment that is now rife, and to the threat to public health thereby caused, a Member State should be held to be entitled, under Article 36 of the Treaty, to forbid the use in products sold on its territory of any substance that has not been proved to be harmless either alone or in combination with other substances. That may well be right, but it is not, in my opinion, necessary to go that far in the present case. In any event, I doubt if a ruling on such a general proposition could properly be given by a Chamber of the Court.
In the result I am of the opinion that, in answer to the question referred to the Court by the Gerechtshof of Amsterdam, Your Lordships should rule that, at present, a provision in the legislation of a Member State forbidding, subject to exceptions, the addition of nisin to processed cheese is compatible with the provisions of the Treaty relating to the free movement of goods.
Full & Egal Universal Law Academy