OPINION OF MR ADVOCATE-GENERAL ROEMER

DELIVERED ON 8 NOVEMBER 1967 (*1)

Mr President,

Members of the Court,

As in Cases 11/67 and 12/67, in the present Case 18/67, in which reference is made to the Court by the Belgian Conseil d'État, we are once again dealing with the regulations of the Council on social security for migrant workers, which we have already had to examine on several occasions. I shall start with the facts which are as follows:

The late husband of the plaintiff in the action pending before the national court worked in Italy, Germany and Belgium and in all three countries was affiliated to the social security system in force. It is clear that of these various periods of employment and of insurance the ones which matter most in respect of the action before the Conseil d'État are the periods completed in Belgium during 1927 and 1959 and for certain periods of time between 1 January 1940 and 31 December 1945, as well as those periods completed in Germany from 1 March 1941 to 4 September 1944. On the death of her husband on 17 July 1960, the plaintiff submitted a claim for a survivor's pension under the workers' scheme, which was rejected by ministerial decision on the ground that from 1926 to 1960 her husband had not been a wage-earner mainly in Belgium for a period of at least one year.

The Commission d'Appel Spéciale, to which this decision was referred, took a different view of the facts: it held that the plaintiff's husband had been in gainful employment in Belgium and Germany for a total period of eight years (96 months). It therefore granted the widow a pension but, as the period of 42 months covered by German insurance coincided in part with the period of 96 months covered by Belgian insurance, it based its calculations on 54/96ths of the full rate. The amount thus obtained was halved and, in accordance with Belgian legislation, this was the amount of the pension awarded by the Commission, on the ground that the plaintiff's husband had not been a wage-earner for at least half the period from 1926 to his death.

The Office National des Pensions pour Ouvriers was unwilling to accept this decision and referred the matter to the Commission Superieure des Pensions which ruled that the proportion of the survivor's pension for which Belgium was responsible should be determined on the basis of 2/34ths of the annual total of BF 18450. It arrived at this figure by following the direct method of calculation provided for by Article 29 (2) of Regulation No 4 and by taking into account only two insurance years (1927 and 1959) completed under Belgian legislation, and not the period from 1940 to 1945. The denominator of the fraction represents the number of years which elapsed between the entry into force of the first provisions relating to social insurance in Belgium and the year of the plaintiff's husband's death.

The plaintiff then asked the Conseil d'État to annul this decision, on the ground in particular that consideration of the matter should have started from the fact that the Belgian social insurance authorities had accepted her husband's contributions for a total period of eight years seven-and-a-half months.

In view of these facts and of the problems arising therefrom in relation to the interpretation of Community law, the Conseil d'État, after hearing the report of the Auditeur, Mr Dumont, decided on 11 May 1967 to suspend proceedings, in conformity with Article 177 of the EEC Treaty, and to ask the Court for a ruling on the following question :

‘When applying Regulation No 3, should the denominator of the fraction used in pro rata calculations be determined by reference to the same unit of time in every country in which the insured person has completed insurance periods? If the answer is in the affirmative, what should that unit be?’

In the course of the proceedings, only the Commission of the European Economic Community has submitted written and oral observations on this question. In view especially of the fact that the Conseil d'État has itself proceeded by way of allusions to make known its legal viewpoint on the facts outlined, we should first ask ourselves whether the interpretation advanced by the Commission is tenable.

Reply to the question put

It can be seen from the terms of the question, as from the Dumont report which has been sent to us, that, contrary to the Commission Superieure, the Conseil d'État seems inclined to take into account periods of employment and insurance periods completed not only in accordance with Belgian legislation but also in accordance with the legislation of a foreign country, more specifically that of Germany. It appears difficult, however, to reply to the question put, in particular because the summary of the facts does not enable us to ascertain exactly either which periods of employment completed in Belgium and Germany should serve as a starting point, or the importance of the presumption of Belgian law (cf. Royal Decree of 17 June 1955) that the years 1940 to 1945 are valid as assimilated insurance periods if it is proved that a certain period of employment was completed between 1 January 1938 and 10 May 1940. If we attempt immediately to form an opinion, without asking the Conseil d'État for more details, such an opinion — and I wish to emphasize this point — might very well be expressed in the terms put forward by the Commission during the oral procedure. Undoubtedly this would entail the risk of giving an incomplete reply which does not take precise account of the facts. It may be considered, however, that this risk is to be preferred to the delay in the proceedings which would be caused by a request for more details, all the more so as the Conseil d'État is free to put another question to us.

First of all, the Conseil d'État asks us to rule on the question of the denominator of the fraction to serve as the basis for the Belgian insurance authorities of the pro rata calculation of the pension and on the unit of time which should be used in the present case.

At first sight, the answer appears to present no difficulties. As the Commission emphasizes, for a logical and precise application of Regulation No 3, one should in principle use for this denominator the same unit of time as for the calculation of the pension for accounting purposes, that is to say, the benefit to which the insured person would be entitled if he had completed in Belgium all the insurance periods and assimilated periods aggregated under Article 27 of Regulation No 3. In Belgium pensions are calculated according to the number of calendar years (Law of 21 May 1955), without taking into account the years during which it is proved that the period of employment was less than 185 days, whereas a period of employment of 185 days or more is sufficient to qualify as a calendar year, in accordance with the Royal Decree of 17 June 1955. Therefore the denominator of the fraction used for pro rata calculations in Belgium should be expressed in calendar years. This means that there is no question of taking into account any actual insurance period under Belgian legislation totalling eight years seven-and-a-half months, as claimed by the plaintiff.

But we cannot rest content with such a reply, because the real problem arising from the summary of the facts obviously relates to the numerator of the fraction used for pro rata calculations and to the unit of time to be used in the present case, and because the answer to this problem may affect the denominator.

During the written procedure, the Commission put forward a number of hypotheses, as the summary of the facts left certain questions unanswered (calendar years involving periods of more than or less than 185 days completed entirely in Belgium; calendar years involving periods of employment in different Member States, in respect of which it is necessary to aggregate the different insurance periods in order to obtain a period of employment of 185 days; and other calendar years for which this is not necessary but which involve insurance periods actually completed abroad). Consequently, the Commission elaborated a detailed theory of the way in which insurance periods completed under Belgian legislation should be taken into account in order to arrive at the fraction to be used for pro rata calculations, of the need to deduct the insurance periods actually completed abroad during a calendar year from the insurance periods regarded by Belgian legislation as assimilated periods, and of the unit of time which is to be used. Although it is true, as the Office National des Pensions pour Ouvriers pointed out in a letter to the Court dated 13 October 1967, that these theories seem somewhat complicated, they appear nevertheless to take into account all the principles laid down by Regulation No 3 and to provide solutions adequate for the hypotheses put forward. However, when considering the Commission's written observations for the first time, I already had the feeling that the facts as outlined by the Conseil d'État hardly allowed us to adopt the general theories of the Commission in arriving at the preliminary ruling called for in this case, a ruling which certainly does not call for elaborate theories but merely for an interpretation in a specific case. My feeling on this point was reinforced when I found that the Commission's representative simplified its point of view considerably during the oral procedure. In fact, despite their lack of clarity, the facts which have been outlined to us show with sufficient precision that only a part of the elaborate theories first put forward by the Commission might be of interest to the Conseil d'État.

In substance, it is clear that, in so far as Community law is concerned, the only problems are as follows: The Conseil d'État is faced with the question of the entitlement to pension of the widow of a worker, who in all probability (according to the Dumont report which refers to recent Belgian case-law) will benefit from the legal presumption of the Royal Decree of 17 June 1955 to the effect that an insurance period for the years 1940 to 1945 exists if it is proved that a certain period of employment was completed before 10 May 1940. This insurance period represents, in part at least (in so far as there was no employment), an assimilated period within the meaning of Article 1 (r) of Regulation No 3. Furthermore, it is well established that during the same period the plaintiff's husband completed an actual insurance period of 42 months in Germany, in respect of which a claim may undoubtedly be made in Germany, from 1 March 1941 to 4 September 1944.

In order to avoid an accumulation of insurance benefits for one and the same period, which our case-law considers incompatible with the principles of Regulation No 3, the only possibility open is to deduct the insurance period actually completed in Germany from the assimilated insurance period under Belgian legislation and to take into account only the remainder of the assimilated period (possibly together with the insurance periods actually completed under Belgian legislation) in order to arrive at that proportion of the pension payable by Belgium. It is clear that in adopting such a method of calculation it is necessary to bear in mind that the unit of time used in Germany to calculate insurance periods is the month, and one must therefore convert into months, as provided for by Article 13 of Regulation No 4, the total insurance period (probably of eight years) calculated at first according to the Belgian rules in calendar years. Finally, in view of the fact that, after the necessary deduction of the insurance periods actually completed abroad, there still remains — unless I have made a mistake — a certain number of months belonging to the Belgian insurance period, which cannot be converted into calendar years, the denominator of the fraction used to calculate that proportion of the pension payable by Belgium should also be expressed in months for purely arithmetical reasons. In fact, the result thus obtained, provided that the figures quoted above correspond to the facts, is a fraction amounting to 54/96ths for the pro rata calculation, the figure on which the Commission d'Appel Speciale based its decision fixing the amount of the pension.

All in all, the answer to be given to the question put would therefore appear to be as follows:

In applying Regulation No 3, the denominator of the fraction to be used in pro rata calculations should be determined, as with the amount for accounting purposes, with regard to the unit of time provided for by the legislation of the Member State under which the insurance periods are taken into account.

The numerator of this fraction should also be determined with regard to that same unit of time for the periods during which the worker was subject solely to the legislation of that Member State.

When insurance periods considered as assimilated periods under the legislation of that Member State coincide with actual insurance periods under the legislation of another Member State, these actual insurance periods should be deducted. In this case, in conformity with Article 13 (4) of Regulation No 4, it is necessary to convert all units of time into the smallest units common to the Member States concerned, both for the denominator and the numerator of the fraction used for pro rata calculations.

* * *

(*1) Translated from the French version.