JUDGMENT OF 21. 2. 1984 — CASE 202/82 COMMISSION v FRANCE

In Case 202/82

THE COURT composed of: J. Mertens de Wilmars, President, T. Koopmans, K. Bahlmann and Y. Galmot (Presidents of Chambers), P. Pescatore, Lord Mackenzie Stuart, A. O'Keeffe, G. Bosco, O. Due, U. Everling and C. Kakouris, Judges, Advocate General: G. F. Mancini Registrar: H. A. Rühi, Principal Administrator

gives the following

JUDGMENT

Facts and Issues

I — Facts and procedure

In France and Italy the manufacture and marketing of pasta products containing common wheat are prohibited. The authorities of each country have laid down methods of analysis allowing them to detect the presence of common wheat in such products. The methods are different, but both rely on the same phenomenon, namely the presence of a protein which exists only in common wheat. It follows that every sample in which that protein can be found necessarily contains common wheat.

The Commission, considering that the difference between the methods of analysis and between the tolerances accepted by the two Member States where a small quantity of common wheat was found had the result that pasta products, legally produced and marketed in Italy, could not be legally imported into France, initiated the procedure laid down by Article 169 of the EEC Treaty. It made an application to the Court, which was lodged at the Court Registry on 9 August 1982. The written procedure followed its normal course. The Italian Republic was given leave to intervene in support of the conclusions of the French Republic.

Upon hearing the Report of the Judge-Rapporteur and the views of the Advocate General, the Court decided to open the oral procedure without any preparatory inquiry.

II — Conclusions of the parties

The Commission claims that the Court should:

1. Declare that the French Republic, by applying to imported pasta products manufactured exclusively from durum wheatmeal and legally marketed in another Member State a method for determining their common-wheat content and tolerances regarding the presence of common wheat which are liable to constitute obstacles to imports, has failed to fulfil its obligations under Article 30 of the EEC Treaty;

2. Order the French Republic to pay the costs. The French Republic contends that the Court should: 1. Declare that the Commission's application is unfounded; 2. Order the applicant to pay the costs. The Italian Republic contends that the Court should dismiss the application and order the applicant to pay the costs.

III — Submission and arguments of the parties

1. The method of analysis to be applied to imported pasta products

The Commission states that, according to the information at its disposal, the application of the legislation which is at present in force in France has caused the authorities responsible for the prevention of fraud to bring criminal proceedings against certain importers. In several cases, there have been criminal convictions for the importation into France of pasta products of Italian origin containing common wheat, even though compliance with the Italian provisions was not only required by the contract but also constantly monitored by both Italian and French laboratories. The effect of this situation is to create an unjustified barrier to intra-Community trade.

The Commission considers that where the products in question have been legally manufactured and marketed in another Member State, the method of analysis to be applied in order to verify that the legal requirements have been complied with is that of the product's country of origin, in this case Italy; since the substantive rules governing the composition of a product are those of the country in which the product was legally manufactured, it is hard to see why the same principle should not also apply to the procedural rules governing the verification of compliance with the substantive rules.

The Commission contests the French Government's assertion, made in the course of the procedure prior to the application to the Court, that the two methods will always give identical results. It refers to a study (Ring Test) carried out by the Lord Rank Research Laboratory in 1978 on behalf of the Union of Organizations of Manufacturers of Pasta Products in the EEC (UNAFPA). According to the Commission, that study shows a difference in the results obtained by the two methods when applied to the same samples and, what is more, that the same method applied to the same samples by different laboratories frequently gives different results.

The French Government considers that it cannot be concluded from that study that the application of the same method to the same samples by different laboratories will frequently produce different results. In any event, if the French laboratories were imprecise in their analysis, they would be equally so if they were using the Italian method.

The experts who have studied the two methods consider them to be equivalent. The only notable technical difference between the French and Italian methods is that the French method is less sensitive to small percentages of common wheat. The French method cannot detect the presence of common wheat at levels of less that 4 to 5%, whilst the Italian method can detect it at levels of 1 to 2% and above.

It therefore cannot be concluded that the use of the French method entails more stringent monitoring of the presence of common wheat in pasta products such as might create barriers to the importation of Italian pasta products into France.

In the view of the Commission, the question whether a product has been legally manufactured and marketed in a Member State must necessarily be decided by reference to the rules of the country of origin, including the rules governing, for example, the carrying out of analyses. As regards the reliability of the two methods of analysis, the minimum requirement is that the different methods of analysis should be equivalent in every respect and always give identical results. That is not in fact the case as regards the French and Italian methods for determining the common-wheat content of pasta products. The Commission analyses in detail the results of the two “Ring Tests” carried out on behalf of UNAFPA in 1978 and 1980 and refers to the criticisms of the two methods made by the German and Netherlands Associations of Manufacturers of Pasta Products, maintaining in particular that the methods may not be used either for pasta dried at high temperature or for pre-cooked pasta.

According to the results of the analyses presented by the Commission, the two methods have a margin of error so considerable as to render them extremely unreliable as a means of determining the exact composition of pasta products. Consequently, the method used by the French authorities is liable to constitute a barrier to imports.

2. The tolerances

The Commission maintains that it is administrative and judicial practice to accept a certain common-wheat content (called “tolerance”). This tolerance is 3 to 5% in France and 7% in Italy. These different levels of tolerance constitute another source of obstacles to the free movement of goods. Where imported products have been legally manufactured and marketed in another Member State, the French authorities should apply to them the tolerances accepted in the country of origin.

The French Government considers that it is dangerous to extend the principles laid down by the Court in the “Cassis de Dijon” case (Case 120/78, REWE ν Bundesmonopolverwaltung für Branntwein, [1979] ECR 649) to behaviour which is both illegal and hard to prove, since this is likely to give rise to difficult problems of legal certainty and to force Member States to allow products to be marketed which are not in conformity with the legislation of their country of origin.

In the alternative, the French Government contests the facts alleged by the Commission. There certainly is a degree of tolerance in France, but it is 5%. More precisely, the official laboratories do not consider that pasta products are not in conformity with the law unless the common-wheat content is at least equal to 8%, plus or minus 3%, the purpose of the safety margin of 3% being to allow for the errors which are always possible in the method of analysis or in the laboratory's techniques (“technical tolerance”). This practice is more liberal than that recommended by the trade associations in five Member States (including Italy), which consider a common-wheat content higher than 4%, plus or minus 2%, to be unacceptable.

The French Government denies that the level of tolerance in Italy has been fixed at 7% by the Italian courts. There was in the past a 7% administrative tolerance, based on the percentage of common wheat allowed in durum wheat offered to the intervention agencies. That percentage is at present 4%. The level of tolerance has thus been reduced in line with it to 4%, that is to say, to a slightly less favourable level than in France.

Furthermore, for budgetary reasons, there is no quality control of pasta products exported to France; this contradicts the Commission's claim that the conformity of Italian pasta products with the Italian legislation is “constantly monitored” by Italian laboratories.

The French Government concludes, that the Commission's supposition that the different levels of tolerance constitute a source of obstacles to the free movement of pasta products is not borne out by the facts.

In its reply, the Commission states that the existence of tolerances cannot be regarded as illegal, in view of the fact that their purpose is to take account of the natural presence of common wheat in durum wheat, the uncertainties (reliability margins) which those who have developed the methods of analysis recognize as being inherent in them and the margins of error in the performance of the analyses.

If, however, the accepted tolerances are not the same in two countries, this may have the result of creating barriers to trade, because a product which is considered to have been legally produced in one Member State — having regard to the accepted tolerance in that Member State — might not be so considered in another Member State where a less favourable tolerance is applied.

The Commission does not accept the French Government's allegation that the accepted tolerance is 5% or 8%, plus or minus 3%. In the first place, the French Government meant to say 5%, plus or minus 3%, so that the accepted maximum is 8%. In the second place, the Commission cites two judicial decisions which refer to a 3% tolerance accepted by the French administration and which are therefore fully at variance with the assertions of the French Government in the present case.

As regard Italy, the Commission does not accept the distinction made in the defence between administrative tolerances and those laid down by the courts. It is wrong to claim that the latter do not exist in Italy.

In any event, the situation described shows that there is sufficient legal uncertainty with regard to the accepted tolerance to give rise to barriers to intra-Community trade.

The Commission maintains that, when it stated that the conformity of Italian pasta products with the law was constantly monitored, it was referring mainly to voluntary checks carried out at the request of manufacturers and importers.

The French Government, in its rejoinder, states that the laboratories which participate in the prevention of fraud consider samples to be:

“in conformity with the law” if their common-wheat content is less than 5%, plus or minus 3%;

“to be examined further” if their common-wheat content is at least equal to 5%, plus or minus 3%, but less than 8%, plus or minus 3%;

“not in conformity with the law” if their common-wheat content is equal to or above 8%, plus or minus 3%.

When a sample is “to be examined further”, additional samples are taken to see if there is systematic fraud.

The “administrative” tolerance (the first percentage indicated) is intended to ensure that a trader whose products accidentally contain common wheat is not prosecuted. The presence of common wheat in such cases is not “natural”, but is due solely to the handling of the products.

The “technical” tolerance (plus or minus 3%) is intended to take account of the inaccuracies which can take place during the visual implementation of the method of analysis. It is in no way intended to allow for a lack of precision in the method itself, which cannot indicate the presence of common wheat in pasta products containing no common wheat and which becomes relatively inaccurate only if the common-wheat content exceeds 15%.

The judgments cited by the Commission which mention a tolerance of 3% refer to events which took place before 1978, that is to say, before the present tolerances were put into effect. In view of the results obtained, the trader in question would have been prosecuted even under the present tolerances.

The French Government considers it curious that the Commission should refeito a lack of legal certainty supposedly created by the Italian tolerances. The sole purpose of the tolerances is to take account of mistakes which may be made by traders. Traders may not deliberately include common wheat in the pasta products which they manufacture and, from that point of view, their decisions are not affected by any lack of legal certainty.

The French Government also refers, as regards the administrative tolerance of 4°/o, to the observations submitted by the Italian Government in support of its intervention. An annex to those observations refers to errors of plus or minus 1 or 2%; consequently, the limit of the tolerance in Italy can, according to the French Government, fall below 4%, whilst in France it never falls below 5%.

3. The reliability of the French method

(The arguments put forward on this subject in the application, the defence and the reply have been summarized on pages 5, 6 and 7.)

In its rejoinder, the French Government emphasizes that the two methods of analysis have been developed in order to verify that the prohibition of the inclusion of common wheat in pasta products is observed. The essential purpose of the French method is to detect the presence of common wheat (qualitative analysis) and not to determine the quantity thereof once its presence has been detected (quantitative analysis). Whilst the qualitative reliability of the method is absolute, its quantitative reliability can vary according to the percentage of common wheat present in the samples being analysed. The method is very precise where the common-wheat content is no more than 15%, but is only “semi-quantitative” beyond that.

The French Government criticizes the Commission's analysis of the results of the “Ring Test” carried out by the Lord Rank Research Centre and contests the relevance of the criticisms made of the French and Italian methods by the Netherlands and German associations. Since the legislation of certain European countries allows common wheat to be used in the manufacture of pasta products, it is understandable that the associations in question expect a European method to have a very high degree of quantitative accuracy.

As regards the implementation of the French method, it should be noted that the common-wheat content of a sample is evaluated visually. This is the reason for certain minimal inaccuracies and why a margin of plus or minus 3% is applied to the results obtained.

The French Government also points out that an accused has the right to ask for a second expert opinion.

In conclusion, the French Government considers that all precautions have been taken to prevent a trader from being convicted in France for marketing pasta products which do not contain common wheat. The absence of barriers to imports is confirmed by the fact that the Italian Government recognizes the reliability of the controls.

4. The need to use the Italian method in France

The French Government does not share the Commission's doubts about the reliability of the Italian method, but raises the question why the Commission wishes to see this method used in France if it is of such doubtful validity. It wonders whether the Commission's purpose is not to induce the Court to decide the theoretical problem of whether the “Cassis de Dijon” decision, cited above, applies to procedural rules governing the verification of compliance with substantive rules, by making use of an alleged Franco-Italian problem.

If the theoretical problem were to arise, the French Government would oppose such an extension of the “Cassis de Dijon” decision to methods of analysis. It cites the judgment of 17 December 1981 in Case 272/80 (Biologische Producten, [1981] ECR 3277). In the alternative, the French Government considers that, if the “Cassis de Dijon” decision is applicable, any barriers which exist are justified in the interests of fair trading. The difference between the price of durum wheat and common wheat, which is about 50%, makes fraud extremely lucrative even where the quantities involved are relatively small. There is no other way of determining whether or not pasta products imported from a foreign country contain common wheat.

IV — Observations of the intervener

1. The method of analysis to he applied to imported pasta products

The Italian Government considers that it cannot be concluded from the Court's case-law that, in verifying compliance with the substantive rules in force in the exporting country, the importing country must employ the exporting country's methods of analysis.

The Court did not consider that question in its decision in the “Cassis de Dijon case. On the contrary, in its judgments of 30 September 1975 (Joined Cases 89/74, 18 and 19/75, Arnaud, [1975] ECR 1023; Joined Cases 10 to 14/75, Lahaille [1975] ECR 1053), the Court by implication accepted the principle that, in the absence of a common method of analysis, the importing country could legitimately apply its own method of analysis.

Furthermore, the solution proposed by the Commission would in reality entail impractical and absurd consequences. The experts responsible for carrying out the verifications in a Member State would be obliged to learn the methods of analysis of all the other Member States and the State would be obliged to acquire all the additional scientific instruments and to provide the necessary training.

The Italian Government points out that the methods of analysis are only one of a number of means used to establish the existence of fraud, and that a criminal court can make use of other means of investigation to form its own opinion.

It should be emphasized that it is not technically possible to obtain perfect results with such methods of analysis. The reliability of a method must be considered in the light of the result which is sought to be achieved. The French and Italian methods are more than adequate to ensure that an analysis of pasta products manufactured exclusively from durum wheatmeal does not indicate that it contains common wheat. This is confirmed by the results of the UNAFPA studies carried out in 1978 and 1980. A minimal difference between the results does not undermine the reliability of the two methods, if account is taken of the errors which analysts without special training will have made in carrying out the tests or in reading the results thereof.

As regard the criticisms made by the Commission in its reply, the Italian Government refers to a report, annexed to its observations, prepared by the Institute of Agricultural Industries, attached to the Faculty of Agriculture of the University of Milan.

Furthermore, the Italian Government denies that pasta products cooked at high temperature and pre-cooked pasta products play any role in trade between the two countries, which consists of ordinary dried pasta products.

It adds that exports to France represent 50% of Italian exports of pasta products to the EEC and that French production does not cover the needs of its own domestic market.

2. The tolerances

Italian law does not allow for any tolerance, even where common wheat is deliberately added.

The situation is different where the presence of common wheat is due to a natural impurity in the durum wheat meal. In such a case, the authorities take account of the lack of any deliberate intent on the part of the producer and a minimal presence of common wheat is not punished. It is only in that sense that there is any tolerance.

Pasta containing common wheat cannot be legally produced or marketed in Italy. Italian law makes no provision for tolerances and, like the French Government, the Italian Government considers dangerous the Commission's view that Article 30 of the EEC Treaty should extend to facts considered contrary to legislation but not punishable by the courts.

The Italian Government refers to the Community rules which allowed impurities in the durum wheat supplied to the intervention bodies. Those rules were the basis for accepting a tolerance equal to a 7%, now 4%, common-wheat content.

V — Oral procedure

At the sitting on 23 November 1983 oral argument was presented for the Commission by R. Wägenbaur, acting as Agent, for the French Republic by B. Botte, acting as Agent, for the Italian Government by I. Braguglia, acting as Agent, and Professor P. Resmini, in the capacity of expert.

The Advocate General delivered his opinion at the sitting on 14 December 1983.

Decision

1. By application lodged at the Court Registry on 9 August 1982, the Commission of the European Communities brought an action under Article 169 of the EEC Treaty for a declaration that the French Republic, by applying to imported pasta products manufactured exclusively from durum wheatmeal a method for determining their common-wheat content and tolerances regarding the presence of common wheat which were liable to constitute obstacles to imports, had failed to fulfil its obligations under Article 30 of the EEC Treaty.

2. The Commission's action concerns barriers which France is said to have raised to the importation of pasta products legally manufactured and marketed in Italy. The legislation of both those countries requires that pasta products should be manufactured solely from durum wheat and prohibits the manufacture or marketing of pasta products containing common wheat. The Commission's action is not directed against the prohibition itself, which, it is agreed, does not create a barrier to the importation of pasta products legally manufactured and marketed in Italy, but rather against the rules adopted in France to ensure compliance with the prohibition.

3. The French and Italian authorities have each adopted an official method of analysis which allows them to detect the presence of common wheat in pasta products. The French method, called the “Montpellier” method, and the Italian method, called the “Resmini” method, differ in detail but are both based on the identification of a protein which exists only in common wheat. Every sample in which the presence of that protein can be detected necessarily contains common wheat.

4. The Commission has claimed that the disparity between the two methods of inspection is such as to create a barrier to the importation of pasta products legally manufactured and marketed in Italy. In the view of the Commission, goods produced in another Member State must be presumed to have been legally manufactured and marketed, and it is for the importing Member State which wishes to prohibit the marketing of those goods to prove the contrary on the basis of the substantive rules and rules governing inspections in force in the Member State in which the goods were produced. The Commission has also expressed doubts regarding the reliability of the French method and its correct application by the specialized laboratories in France.

5. The Commission's main argument must be rejected. It is apparent that the importation into France of Italian pasta products manufactured exclusively from durum wheat is not liable to be hindered in any way, since the results of any analysis carried out by the French authorities responsible for the prevention of fraud can only confirm the absence of common wheat. Furthermore, although Article 30 obliges Member States, in certain circumstances, to take active steps to ensure the free movement of goods legally produced and marketed in other Member States, in particular by accepting certificates issued by the competent authorities in other Member States, that obligation does not go so far as to require them to carry out inspections according to the methods laid down by the legislation of the other Member States. Such a requirement would entail an obligation to acquire the appropriate additional equipment and to employ specialized staff. That solution, costly in economic terms, is of no value in ensuring the free movement of goods, particularly when this can be effectively guaranteed by other means.

6. As regards the doubts expressed by the Commission regarding the reliability of the French method and its correct application, it is clear from the explanations provided by the French Government and the Italian Government, which has intervened in support of the conclusions of the French Government, that the objective of the two methods is to detect the presence of common wheat in pasta products and not, primarily, to determine the exact proportion of common wheat in a given sample. In that respect the two methods appear to be equally reliable, and therefore the small differences between them are of no consequence in the context of these proceedings. Moreover, it would appear that the solution proposed by the Commission would lead to results less certain than those produced by the present situation, inasmuch as the staff responsible for carrying out the inspections would lack experience in the use of the method in force in the other Member States.

7. Finally, the Commission has claimed that the French authorities are stricter than the Italian authorities as regards the tolerances which they accept. Thus, according to the Commission, Italy accepts in practice a common-wheat content of 7%, below which level a trader will not be prosecuted or, at least, will not be convicted, whilst in France the tolerance is only 4%.

8. It must be held that the Commission's complaint has not been proved. The explanations provided by the French and Italian Governments show that the situation is in fact the reverse, inasmuch as the tolerances applied in Italy are, at present, stricter than those in force in France.

9. Furthermore, in so far as it has been neither proved nor contended that a prohibition on the importation of pasta products containing common wheat is contrary to Community law, it cannot be considered that judicial or administrative practices designed to combat breaches of that prohibition come within the ambit of Community law, at least in the absence of any discriminatory treatment of imports.

10. It follows from the foregoing considerations, taken as a whole, that the Commission's application must be dismissed.

Costs

11. Article 69 (2) of the Rules of Procedure provides that the unsuccessful party is to be ordered to pay the costs. Since the Commission has failed in its submissions, it must be ordered to pay the costs, including those incurred by the Italian Government.

On those grounds, THE COURT hereby:

1 Dismisses the application;

2 Orders the Commission to pay the costs, including those incurred by the Italian Government.