DELIVERED ON 8 OCTOBER 1981 (1)

Mr President,

Members of the Court

1. This reference for a preliminary ruling concerns the interpretation of various aspects of Note 5 (B) to Chapter 85 of the Common Customs Tariff relating to electronic micro-circuits (micro-assemblies and integrated circuits). The purpose of that note is to facilitate the application of tariff subheading 85.21 D which covers electronic micro-circuits.

The facts of the case may be summarized as follows.

Between 1971 and 1973 and subsequently in 1977 Analog Devices, a Munich undertaking, imported into the Federal Republic of Germany various modules (electronic circuits) intended for automatic data-processing machines and other equipment. After checking the tariff classifications determined on the occasion of each importation, the German customs authorities decided to classify the goods imported in 1971 under subheading 85.21 C of the Common Costoms Tariff in force at the time, entitled “Transistors and similar mounted semi-conductor devices”, and to classify those imported subsequently under subheading 85.21 D of the Common Customs Tariff in force as from 1 January 1972.

The importing undertaking considered that classification to be inappropriate and submitted an objection to the decision of the customs authority, which was however dismissed. The company therefore instituted proceedings before the Finanzgericht München [Finance Court, Munich]. By order of 10 April 1980, that court stayed the proceedings and referred to the Court of Justice for a preliminary ruling a series of questions all of which concern the interpretation of Note 5 (B) to Chaper 85 of the Common Customs Tariff.

2. It is necessary to consider in the first place the tariff provisions most directly relevant for the purposes of this case.

Heading 85.21 of the Common Customs Tariff, as amended by Regulation (EEC) No 1/72 of the Council of 20 December 1971 groups together in paragraph D “diodes, transistors and similar semiconductor devices” as well as, and this is the central issue, “electronic micro-circuits”. Goods classified under that subheading attract a fairly high rate of duty. Note 5 to Chapter 85 (entitled “Electrical machinery and equipment; parts thereof”) specifies in paragraph (B) which articles are included in the concept of “electronic micro-circuits”. There are three categories: subparagraph (a) covers micro-assemblies consisting of “discrete ... miniaturized components which are combined and interconnected”; subparagraph (b) covers monolithic integrated circuits in which the circuit elements are created in the mass and on the surface of a semiconductor material and “are inseparably associated”; and subparagraph (c) cover hybrid integrated circuits, that is to say, those in which elements obtained by various technologies are combined, to all intents and purposes indivisibly, on a single insulating substrate and which may also include miniaturized discrete components.

3. The first question submitted by the Finanzgericht München concerns the concept of “miniaturized discrete component” which is used in two passages in Note 5 (B), namely in subparagraph (a) and in the last sentence of subparagraph (c). The order making the reference to the Court provides a clearer explanation of what the German court wishes to ascertain concerning the concept of discrete component: whether it covers exclusively components which are mounted as such separately (and provided with electrical connections), whether reference should be made to other criteria or whether, in addition to components consisting of a single electric circuit element, the same concept may also include circuits made up of several elements, such as monolithic integrated circuits or ceramic plates with a printed resistor network.

In technical terms, the expression “discrete component” (in French “composant discret”, in Italian “componente discreto” and in German “Einzelbauelement”) is used to denote an entity which forms part of a more complex circuit but is an individual physical entity. It is therefore possible to state, in approximate terms, that the discrete components of a given module (electronic circuit) are basic units employed in the manufacture of the module itself.

However, as regards the interpretation of the expression “discrete components” in connection with the above-mentioned Note 5 (B), the position of the Commission diverges markedly from that of the plaintiff in the main action. Counsel for the plaintiff argues that discrete components are independent units forming part of the electronic circuit (that is to say, diodes, transistors, resistors, condensers and the like) characterized not only by their original physical individuality but also by the fact that they each fulfil a single function. Instead, the Commission takes the view that a discrete component is to be understood as meaning any element of the circuit which constitutes an indivisible physical unit. Each of these interpretations results in a different answer to the question raised by the German court: if the Commission's view is accepted, a monolithic integrated circuit may be an element in an electronic circuit inasmuch as, though itself composed of different elements, it constitutes a unit from a physical point of view, regardless of whether or not it has a single function. If, on the other hand, we adopt the argument adduced by the plaintiff in the main action, that possibility is ruled out on the ground that a monolithic integrated circuit does not always fulfil a single function but may have several functions, particularly in view of the fact that it may be used in conjunction with different items of equipment.

I should like to state straight away that the Commission's interpretation strikes me as the more convincing of the two primarily because it is in conformity with the general criteria on which the Common Customs Tariff is based. It is well known that in order to be correctly applied, the tariff headings must identify on the basis of clear and objective factors the goods which come within their respective spheres. That enables the customs authorities to carry out the necessary checks swiftly and at the same time guarantees importers a reasonable degree of certainty in legal relationships. The Court has recently reaffirmed that concept by stating that “the principle of legal certainty requires that rules imposing charges on the taxpayer must be clear and precise so that he may know without ambiguity what are his rights and may take steps accordingly” (judgment of 9 July 1981 in Case 169/80, Administration des Douanes ν Gondrand and Garancini, [1981] ECR paragraph 17 of the decision). If the classification of electronic circuits were to be influenced by the varying functions for which they were intended, the factors to be taken into account by the customs authorities in determining the appropriate heading would include one which is difficult to assess, and that would also complicate the carrying out of the necessary administrative checks.

That argument accords with the last paragraph of Note 5 which states that: “For the classification of the articles defined in this note, heading No 85.21 shall take precedence over any other heading in the tariff which might cover them by reference to, in particular, their function”. That implies that, at least in relation to the articles covered by heading 85.21, their function is not relevant for the purposes of thier classification.

Additional evidence in support of the interpretation which I favour may be derived from the notes to the Nomenclature of the Customs Cooperation Council established by the Brussels Convention. In that connection I would point out, to begin with, that those notes may be used for the interpretation of the Common Customs Tariff, but within certain limits. The Court has had occasion to state that: “Where no Community explanatory notes have yet been issued in respect of the tariff headings to the Common Customs Tariff, the observance of these explanatory notes and opinions” (issued by the Customs Cooperation Council in Brussels) “is a useful means of ensuring that the common external tariff is uniformly interpreted and applied at all the frontiers of the common market” (judgment of 8 December 1970 in Case 14/70 Bakels [1970] ECR 1001, paragraph 10 of the decision). Consequently, where, as in the present case, explanatory notes have been issued by the Community, it is inevitable that the notes drawn up by the Customs Cooperation Council will not be as significant as they would be in cases where no Community explanatory notes whatsoever exist and, in any event, notes originating in the Community must take precedence. Having said that, I would point out that, according to the second paragraph of the note to Section XVI, Chapter 85, paragraph F, “Electronic micro-circuits fall into two groups, according to manufacturing process: micro-assemblies and integrated circuits”. The expression “according to manufacturing process”, which constitutes a clear reference to the manner in which they are constructed, implies that the classification of micro-circuits into groups has been carried out by reference to physical criteria. Therefore, in my opinion, there can be no doubt that recourse to the notes of the Brussels Convention is justified for the purpose of interpreting the Common Customs Tariff since the division of electronic micro-circuits into micro-assemblies and integrated circuits, as set out in Note 5 (B) to the Common Customs Tariff, is based without variation on the identical division set forth in the classification of the Customs Cooperation Council.

There is therefore more than one argument militating in favour of the Commission's view that circuits (whether monolithic or hybrid) which are “integrated” and thus constitute inseparable physical units come within the concept of “discrete components” and may therefore form part of a micro-assembly as defined in subparagraph (a) of Note 5 (B).

Against that, the plaintiff in the main action relies above all on an argument deduced from the context of Note 5 (B) to the Common Customs Tariff. Subparagraph (c) of that note specifies that hybrid integrated circuits may also include elements obtained by semiconductor technology, for example, monolithic integrated circuits. It follows, according to the plaintiff, that in the absence of a similar express provision, monolithic integrated circuits cannot be discrete components of the micro-assemblies covered by subparagraph (a) of the same note. In my opinion, however, such a peripheral point in the text cannot have any significance for the purposes of interpretation. One need only consider that, in defining micro-assemblies, subparagraph (a) merely states that they consist of discrete, active or both active and passive, components without identifying such components, unlike subparagraph (c) which does so, albeit in brackets and by way of illustration, in the case of the elements of hybrid integrated circuits. Therefore, in my opinion there is nothing in the text to preclude a monolithic integrated circuit incorporated in a micro-assembly from being regarded as a discrete component.

Next, the plaintiff relies on the notes to the Brussels Nomenclature. Section XVI, Chapter 85, paragraph (F), (I) (micro-assemblies) lists the discrete components (diodes, transistors, capacitors, inductances and resistors) in paragraph (1) but makes no mention of integrated circuits, whether monolithic or hybrid. The plaintiff seeks to deduce from that omission that integrated circuits cannot be included in the list of discrete components. That argument seems to me to carry little weight for the purpose of clarifying the concept of discrete component. It is necessary to bear in mind that the list contained in the aforesaid paragraph does not refer to discrete components in general but has been incorporated in the context of paragraph (1), which covers only a particular category of modules (cordwood or fagot modules).

The plaintiff undertaking also refers to another aspect of the notes to the Brussels Nomenclature. It observes that the fourth subparagraph of Section XVI, Chapter 85 (F) (II) (2), contains an exception to the effect that “assemblies formed by adding to an electronic micro-circuit other devices or other micro-circuits of the same or different type”, leaving aside hybrid integrated circuits consisting of combinations which are to all intents and purposes indivisible, are excluded from the heading “Electronic micro-circuits”. In the plaintiff's opinion, the inclusion of an integrated circuit as a component of a micro-assembly would result in an “assembly” of the type envisaged in the aforesaid passage. That assembly could not therefore be classified as a micro-assembly consisting of discrete components. If the contents of the relevant note to the Brussels Nomenclature are applied to the interpretation of the concept of electronic micro-circuit contained in heading 85.21 D of the Common Customs Tariff, the resultant conclusion should be that a micro-assembly which includes an integrated circuit amongst its components cannot be classified as an electronic micro-circuit under the above-mentioned heading.

That argument is also open to criticism. I would point out that the exception referred to in the note to the Brussels Nomenclature applies only to assemblies which constitute a complete machine or appliance or machine parts. The fifth paragraph of Part F makes clear that assemblies which cannot be defined as electronic micro-circuits are to be classified as machines or machine parts; in other words, such assemblies, unlike electronic micro-circuits, are identified essentially by reference to their function. We have seen, however, that the last paragraph of Note 5 to the Common Customs Tariff gives precedence to tariff heading 85.21 over any other heading capable of being applied to the same article if the function for which it is intended is taken into account. In my opinion, that rule on the order of priority must, in view of its general nature, also apply within the framework of the Common Customs Tariff to those combinations which would not be classifiable as micro-circuits if the notes to the Brussels Nomenclature referred to above (Section XVI, Chapter 85 (F) (II) (2), fourth subparagraph) were to be employed. To put it another way, the rule on the order of priority, established by the Community note, does not permit account to be taken of the approach discernible in the notes to the Brussels Nomenclature on “assemblies” since that approach lays emphasis on the function for which the articles in question are intended.

4. I have already stated that the first question submitted by the court hearing the main action concerns the concept of miniaturized discrete components. It therefore remains to be seen how the requirement relating to miniaturization, expressly laid down by subparagraphs (a) and (c) of Note 5 (B), is to be interpreted. In that connection, the court wishes to ascertain in particular whether miniaturization implies merely a reduction in size of the individual components or the arrangement of the components themselves in a certain density, whether reduction in size and density may be determined in absolute terms, and, finally, whether the degree of miniaturization which is customary in the electronics industry is to be taken into account.

I would observe first of all that Note 5 (B) associates the requirement of miniaturization with discrete components. Furthermore, it must be assumed that when the same note refers to electronic micro-circuits in the title of paragraph (B), and to micro-assemblies in subparagraph (a) thereof, it is by implication referring to the same requirement but in relation to the circuit in its entirety.

With regard to the meaning of miniaturization, the plaintiff in the main action and the Commission agree that it is a relative concept, linked to the level of technical development attained at the time of importation. That strikes me as a reasonable and sound approach. It would in fact be inconsistent to define miniaturization in absolute terms, in view of the speed of technical progress particularly in the electronics sector. However, there are differences of opinion between the parties concerning the selection of a criterion as a point of reference to establish whether or not an individual part is miniaturized. Counsel for the plaintiff firm has argued that account should be taken of the size of elements of the same kind which were generally produced at the time of importation. According to that criterion, elements which are smaller in size than the average dimensions of those produced in the sector may be regarded as miniaturized. The Commission, on the other hand, argues that account should be taken only of the fact that, having regard to the technical possibilities in existence at the time of manufacture, the elements have been made in accordance with the prevailing tendency to construct elements of ever smaller dimensions.

I believe the latter argument is to be preferred. It must be borne in mind, and it was also made clear in the course of the informal meeting involving the participation of experts, that the production of elements of ever smaller dimensions reflects a widespread tendency in the field of electronic circuits. Consequently, if the average size of the elements produced at the time of importation were taken as a criterion, it would probably be very difficult to identify electronic circuits of below-average dimensions and thus the application of the tariff heading in question would constitute an exception. In my opinion, therefore, the Commission's interpretation, which has the merit of taking into account the relative value of the concept of miniaturization but at the same time allows reasonable scope for the application of that tariff heading covering electronic micro-circuits, is more consistent and more in conformity with the function of the Common Customs Tariff.

The significance of the arrangement of the various elements in a certain density lies in the fact that it is of course a requirement which relates to the entire circuit. However, it must be recognized in addition that the density factor is to be taken into consideration alongside that of a reduction in size of the individual elements. My general observations on the concept of miniaturization also apply in this matter. Thus the density of the elements, as well as their size, has a relative value and as such must be determined having regard to the level of technical development attained at the time of importation.

5. The second question concerns certain expressions contained in subparagraph (a) of Note 5 (B) two of which refer to micro-assemblies (“moulded module types”, “similar types”) and the other to discrete elements (“which are combined”). Essentially, the processes concerned involve the inter-connection of different parts to form an electronic micro-circuit. The German court wishes to ascertain in the first place what moulded-module technology is to be understood as meaning: whether it is only the process whereby the individual components are moulded in a block and then connected up electrically or whether it includes processes whereby the components are mounted on a circuit board fitted with a printed circuit and the completed circuit is then put into a housing. The order making the reference to the Court also raises the problem of the significance of the components' dimensions, their arrangement on the carrier board, the manner in which the circuit is secured and how the elements are interconnected.

In the opinion of the plaintiff undertaking, the expression “moulded modules” refers solely to modules manufactured with the special so-called “simi-block” process. That process involves casting the elements of the module in resin contained in a mould which is subsequently removed, abrading the faces of the block until the connecting wires are free and, after that, connecting up the components electrically with each other by coating with copper and etching the conductor elements. On the other hand, the concept of “moulded module” does not encompass modules consisting of circuit boards provided with printed circuits even if after being assembled they have been cast in a housing containing resin. The Commission, however, maintains that the modules resulting from both of those processes come within the concept of moulded module. That opinion strikes me as the sounder of the two.

The notes to the Brussels Nomenclature (Section XVI, Chapter 85, paragraph (F), (I) (2) explain the meaning of moulded module and I believe that, in the absence of other information in the notes to the Common Customs Tariff, recourse may be had to notes to the Brussels Nomenclature as a valid instrument of intrepretation. They define “moulded modules” as those in which “the components are encased in a block (cube, parallelepiped, hemisphere, etc.) generally of artificial resin”. That definition therefore accords both with the “simi-block” manufacturing process, which is characterized by the fact that the resin cast is used not only as a protective casing but also as a base for the individual parts, and with the process whereby a micro-circuit already mounted on its own base is cast in a housing containing resin.

The element common to both processes is the indivisibility of the module after being cast. That characteristic may also be observed in the other two types of module referred to in subparagraph 5 (B) (a), namely the “fagot” and the “micro-module”which are also defined in the notes to the Brussels Nomenclature. I would add that the interpretation proposed by the Commission is more in conformity with the requirement of legal certainty which I have already emphasized, because it enables the customs authorities to identify the characteristics of the goods, on the basis of easily discernible objective factors, with a view to their classification for tariff purposes.

As regards the significance of the expression “similar types”, I am of the opinion that the essential characteristic for establishing whether or not a micro-assembly may be considered as being of a similar type, compared with modules constructed by specific processes, lies in the indivisibility of the parts in relation to the whole. We have already seen that this feature is common to the three micro-circuits expressly mentioned in the note. Therefore I believe it stands to reason that the same feature is also of fundamental importance when it comes to identifying “similar types”. It is of course necessary to understand what is meant by the concept of indivisibility. I do not believe it is reasonable to consider indivisibility merely from a technical point of view in the strict sense of the word. Some micro-assemblies may, in theory, be repaired by removing and replacing damaged components. However, the operations involved are very costly, require specialized staff and are therefore economically disadvantageous in the sense that it will be quicker and cheaper to replace the entire module. In my opinion, therefore, recourse should rather be had to the factor of economic expediency to establish whether a module is indivisible and comes within the tariff heading in question. According to that criterion, even modules encased for protective purposes in a plastic housing may constitute micro-assemblies of a “similar type” for tariff purposes. For this to be so, it is sufficient if the separation of the individual components is economically inexpedient.

Finally, as regards the expression “which are combined”, used in the subparagraph in question with reference to the discrete components making up a micro-assembly, I would observe that there does not appear to be any justification for considering the adjective “combined” (réunis) as having a meaning of its own and ignoring the other accompanying adjective “connected” (connectés). Clearly, the purpose of those adjectives is to express the idea of a technical process whereby individual elements are assembled into one unit. That in consequence confirms the criterion of inseparability which I stated was, in my opinion, significant in relation to micro-assemblies.

6. The third question concerns hybrid integrated circuits, which are covered by subparagraph (c) of Note 5 (B), and in particular the following sentence: “in which ... elements, some obtained by ... semi-conductor technology ... are combined, to all intents and purposes indivisibly”. In that connection, the first query raised by the court hearing the main action is whether elements manufactured by semi-conductor technology must in the main be non-discrete elements or whether hybrid circuits may, in addition to passive elements obtained by thin- or thick-film technology, also be made up exclusively of discrete (individual) components.

I would point out that hybrid integrated circuits are characterized by the fact that the passive components are secured to the base by means of vacuum metallization, by means of thin- or thick-film technology. That particular process requires the utilization of glass, ceramic or some other substance displaying the same characteristics as a base for the individual elements. With regard to active components (diodes, transistors, monolithic integrated circuits), Note 5 (B) merely establishes in general terms that they are obtained by semi-conductor technology. The same note proceeds to explain in the last sentence of subparagraph (c) that the circuits in question “may also include miniaturized discrete components”. Since that sentence permits the utilization of discrete components without laying down any restrictions, I am of the opinion that such components may be employed as semi-conductors and that the circuits in question may also be made up exclusively of discrete elements (in addition, of course, to the passive elements secured to glass or ceramic).

Another query raised by the German court concerns the inseparable connection between the elements of the circuit: is an electrical connection required or is it sufficient to encase the circuit in a special housing? Further, must the connection be so solid that it cannot be severed without destruction of the components or is it sufficient that the separation of the individual components leads to the destruction of the housing? Finally, does the requirement of inseparable combination also apply to the individual components of a hybrid circuit? The highly complex and technical nature of these problems impairs comprehension of their legal significance. I would observe, however, as far as passive components are concerned, that there is a logical argument which suggests the answer to the question. In view of the fact that, in hybrid integrated circuits, the passive components are assembled in such a way as to form a single unit with the glass or ceramic base, it is impossible to separate the components from the base without damaging them. Consequently, as regards the passive components, indivisibility consists in the physical impossibility of separating them from the unit. However, as far as the other components (diodes, transistors, monolithic integrated circuits) are concerned, they are merely fused with conductor boards with the result that, at least in principle, it is possible to disconnect them without damaging them or the base. It is true that those components too, obtained by semiconductor technology, are frequently protected by a film of resin, with the result that the detachment of one component, involving the destruction of that protective coating, affects the unity of the module as a physical entity.

Finally, I am of the opinion that a satisfactory comprehensive solution must be based on the same concept of indivisibility which I have adopted in connection with subparagraph (a) of Note 5 (B). By the application of that concept, it is possible to state that the individual elements of a hybrid integrated circuit are indivisible either when the individual parts cannot be separated without damage to the physical unity of the whole or when it is economically inexpedient to separate them.

7. In conclusion, I propose that the Court should answer the questions formulated by the Finanzgericht München by order of 10 April 1980 by ruling that Note 5 (B) to Chapter 85 of the Common Customs Tariff must be interpreted in the following manner:

1. The expression “discrete components” used in subparagraphs (a) and (c) denotes any component of a micro-assembly which consists of an indivisible physical unit. The integrated micro-circuits, whether monolithic or hybrid, covered by subparagraphs (b) and (c) may constitute discrete components of a micro-assembly in so far as they are indivisible or virtually indivisible units.

2. Compliance with the requirement of miniaturization laid down by sub-. paragraphs (a) and (c) in respect of discrete components must be determined by the application of the criterion of the level of technical development attained at the time of importation. Consequently, components which are manufactured in accordance with the prevailing tendency to construct elements of smaller dimensions and are arranged in a density corresponding to the technical level attained at the time of importation are to be regarded as miniaturized.

3. The expression “moulded modules” in subparagraph (a) is intended to denote micro-assemblies obtained by the “simi-block” process, whereby a block of resin is used as a base for, and as a housing to protect, the various components, as well as by the process which consists in casting in a housing containing resin or some other protective substance a micro-circuit already assembled on its own base. The expression “similar types” in subparagraph (a) denotes a micro-circuit assembled by processes other than those relating to moulded modules, micro-modules and fagot modules but characterized, as in the case of those three types of micro-assemblies, by the fact that it constitutes a single physical entity which is indivisible on grounds of technical impossibility or economic inexpediency. The expression “which are combined” reflects the idea of unity between discrete components, achieved in such a way as to make them inseparable in the sense described above.

4. The hybrid integrated circuits referred to in subparagraph (c) may be assembled by utilizing as semi-conductors miniaturized discrete components either in addition to, or to the exclusion of, other elements. The clause “which ... are combined, to all intents and purposes indivisibly, on a single insulating substrate” in subparagraph (c) means that the individual elements are indivisible either when the individual parts cannot be separated without damage to the physical unity of the whole or when it is economically inexpedient to separate them.

(1) Translated from the Italian.