My Lords,

This case comes before the Court by way of a reference for a preliminary ruling by a Social Security Commissioner, sitting in London. “Social Security Commissioner” is, so we are told, the new name of what was until recently called a National Insurance Commissioner. I shall refer to him simply as “the Commissioner”.

The appellant in the proceedings before the Commissioner is Mr Maurice Galinsky, a United Kingdom national living in the Netherlands. The respondent is the Insurance Officer. The dispute between them is about a claim by Mr Galinsky for an increase in his British retirement pension in respect of his dependent children.

The facts of the case are these.

Mr Galinsky was born in England on 29 June 1905. He married in 1953 and has four children born in 1955, 1956, 1958 and 1961 respectively. Until 1964 he worked in Great Britain as a self-employed person and was compulsorily insured as such under the British national insurance scheme. In December 1964 he moved with his family to the Netherlands, where he took up employment and became compulsorily insured under the Dutch social security legislation. It is not in doubt that Mr Galinsky has rightly been regarded as a “worker” as defined by Article 1 (a) of Council Regulation (EEC) No 1408/71 in the Netherlands.

When Mr Galinsky went to the Netherlands it ceased to be obligatory for him to pay contributions under the British scheme, but he was permitted voluntarily to continue such contributions as a “non-employed” person. This would improve his “contribution record” in Great Britain in relation to retirement pension. Having made enquiries about his rights in that respect, Mr Galinsky decided to pay “non-employed” person's contributions until he attained pensionable age at 65. He thereby rendered himself entitled to a British pension at the full normal rate upon retirement.

Shortly before he attained the age of 65, Mr Galinksy was informed by the competent British institution, the Department of Health and Social Security, that he might be entitled to an increase in pension in respect of his children. For the reason that I will explain in a moment, that information was correct as the law then stood, having regard to a bilateral Convention on social security between the Netherlands and the United Kingdom signed at the Hague on 11 August 1954, the provisions of which were incorporated into British law by the National Insurance and Industrial Injuries (Netherlands) Order 1955 (SI 1955 No 874).

Mr Galinsky attained the age of 65 on 29 June 1970. He did not however then become entitled to a British retirement pension because, under the relevant British legislation, it was and is a condition of entitlement to a retirement pension for a man under the age of 70 that he should actually have retired from regular employment. Mr Galinsky in fact continued in regular employment in the Netherlands after attaining the age of 65 and indeed beyond the age of 70.

No such condition applies, it appears, to the receipt of a Dutch old-age pension. Mr Galinsky was awarded such a pension under the Dutch “General Old Age Act” (the Algemene Ouderdomswet or AOW), but not of the full amount. A reduction of 36% was applied, the reason being, so I understand, that, although Mr Galinsky had voluntarily paid back contributions for the period since the AOW came into force, he could not, owing to his British nationality, be credited with notional contributions for any period before that.

On 1 April 1973, following the accession of the United Kingdom to the Community, Regulation No 1408/71 became applicable to United Kingdom nationals. On 16 April 1975 Mr Galinsky applied for the consequent revision of the amount of his Dutch pension. He was found to satisfy the conditions of paragraph H.2. of Annex V to that Regulation, and his pension was revised to the full amount, with one year's retroactive effect.

On 29 June 1975 Mr Galinsky attained the age of 70, whereupon his continued employment ceased to disqualify him from receiving a British retirement pension. He was thenceforth paid such a pension at the full rate. He was entitled to it under the relevant British legislation alone, by virtue of the contributions that he had paid as a self-employed person and later as a “non-employed” person.

Mr Galinksy's Dutch pension and his British pension are both payable at the increased rates appropriate to a married man. No question arises as to that. He has at all material times received in the Netherlands allowances in respect of his children. It is not in doubt that those allowances are “benefits” of a kind to which Article 77 of Regulation No 1408/71 applies.

At the same time as he claimed payment of his British retirement pension, Mr Galinsky claimed an increase in that pension in respect of his three younger children — the eldest being by then too old to qualify. Mr Galinsky was not entitled to such an increase under the relevant British legislation alone, because his children were not in Great Britain. There is no doubt, however, that, before 1 April 1973, he would have been entitled to the increase by virtue of the bilateral Convention that I have mentioned, Article 35 of which provides:

“In all cases where, under the legislation of one contracting party, any cash benefit would have been paid in respect of a dependant if the dependant had been in the territory of that party, such benefit shall be paid if the dependant is in the territory of the other party.”

Mr Galinsky's claim was rejected on the grounds that by virtue of Article 6 of Regulation No 1408/71 the provisions of that regulation had replaced those of the Convention and that, by virtue of Article 77 (2) (b) (i) of the regulation, he was entitled to benefits in respect of his children only in the Netherlands, that being the Member State in which he resided.

Before the Commissioner Mr Galinksy contested that conclusion mainly on the ground that, since he was compulsorily insured in the United Kingdom only as a self-employed person, he was not in relation to United Kingdom benefits a “worker” within the meaning of the regulation, so that its provisions did not apply.

It appeared to the Commissioner that there were two ways in which that contention might be countered. First it could be argued that since Mr Galinsky was a “worker” in the Netherlands he must be regarded as a “worker” throughout the Community. Secondly it might be argued that, since under the British scheme employed persons and self-employed persons are both compulsorily insured against the same contingencies (save that only employed persons are insured against unemployment and occupational accidents and diseases), Mr Galinsky, as a person compulsorily insured under that scheme, was for that reason a “worker” within the meaning of Article 1 (a) (i) of Regulation No 1408/71. It also appeared to the Commissioner, however, that, even if Mr Galinsky were a “worker” within the meaning of the regulation, he might not, in relation to his British pension, be a “pensioner” within the meaning of Article 77. Lastly it appeared to the Commissioner, having regard to a view that I had expressed in Case Giuliani v Landesversicherungsanstalt Schwaben, that the question arose whether Article 6 of Regulation No 1408/71 was valid in so far as it operated to deprive Mr Galinsky of a benefit to which he was entitled under the bilateral Convention as incorporated into British law by the National Insurance and Industrial Injuries (Netherlands) Order 1955.

Such are the circumstances in which the Commissioner has referred to this Court the following questions:

“I — Whether in relation to social security benefits provided under the legislation of a Member State a person who has been compulsorily insured as a self-employed person (but not as an employed person) in that Member State should be regarded as a worker within the meaning of Article 1 (a) of Council Regulation (EEC) No 1408/71 in that Member State by reason either:

(a) that he is a worker under the legislation of some other Member State; or

(b) (if the answer to (a) is negative) that he has been so insured as self-employed under a social security scheme for one or more of the contingencies covered by a social security scheme for employed, self-employed and non-employed persons.

II — (If the answer to either Question I (a) or 1 (b) is affirmative) whether a person who has been insured as a self-employed person (but not as an employed person) in a Member State should in relation to a pension payable under the legislation of that Member State be regarded as a pensioner for purposes of Article 77 of the said regulation by reason that he is or has been a worker for purposes of that regulation and/or by reason of the fact that his pension (though unaffected in the competent State by the provisions of that regulation) has been subject to revision in another Member State under that regulation.

III — (If the answers to Question I (a) or (b) and to Question II are affirmative) whether Article 6 of the said regulation is valid in so far as it operates to deprive a person of any right to benefit derived from a provision of national law of a Member State giving effect to a reciprocal convention with one or more other Member States.”

Although Question III, as the Commissioner has formulated it, arises only if the answers to Questions I and II are affirmative, I propose to deal with it first. It goes to the extent of the powers of the Council under Article 51 of the Treaty and is of more far-reaching importance than either Question I or Question II. The Commissioner indeed indicates, in his order for reference, that, had it not been for the importance of that question, he probably would not have referred this case to this Court at all.

In the Giuliani case, in discussing a question that had been referred to the Court by the Sozialgericht of Augsburg, but which had been only superficially argued before us, I expressed the opinion that the rationale of the well-known rulings of the Court to the effect that Article 51 of the Treaty does not empower the Council to legislate to the detriment of migrant workers was equally applicable whether the advantage of which the legislation of the Council purported to deprive such a worker was derived from national law alone or from national law incorporating an international convention. The Court did not, in the events, find it necessary to deal with that question in its judgment.

At that time the only authorities in this Court on the question were Case 34/69 the Duffy case [1969] ECR 597 and Case 82/72 the Walder case [1973] 1 ECR 599. The Duffy case was an early case in the development of the Court's case-law to the effect that Article 51 of the Treaty does not empower the Council to take away rights that a migrant worker derives from national law alone. In that case the Court expressly assimilated to such rights those that a migrant worker derived from “special conventions between Member States” — see paragraph 9 of the judgment. The Walder case was one where only the interpretation, not the validity, of the relevant Council legislation was in issue. The Court naturally held that that legislation must be interpreted as meaning what it said. It has been suggested to us that other judgments of the Court have a bearing on the point, in particular those in Case 28/68 the Torrekens case [1969] ECR 125 and in Case 187/73 the Callemeyn case [1974] 1 ECR 553. It does not seem to me however that they do.

We have had in this case the advantage of very full arguments on the point, developed by the Insurance Officer, the United Kingdom Government, the Council and the Commission. They were unanimous in saying that the conclusion I expressed in the Giuliani case was wrong. Unfortunately no argument was presented to us on behalf of Mr Galinsky, who might have contended for the contrary view. This was despite the fact that, as the Commissioner records in his order for reference, the Department of Health and Social Security was prepared to bear the expense of Mr Galinsky's representation before this Court.

At all events I have been convinced by those arguments that I was indeed wrong in the Giuliani case. Undoubtedly Article 51 of the Treaty does not empower the Council to take away from a migrant worker rights that he has under the law of a Member State alone. No-one disputes that. But we are here concerned with a different question : whether Article 51 empowered the Council to substitute, for the bilateral and multilateral conventions about social security that existed between Member States before the Treaty, a single system for the coordination of the social security schemes of Member States, even though that might result in particular workers finding themselves, in particular respects, in a less favourable position. Two main reasons have led me to the conclusion that that question must be answered in the affirmative.

The first is that most of the provisions of those bilateral and multilateral conventions operate by reference to the nationality of the person concerned in any particular case. That is true indeed of the bilateral Convention between the Netherlands and the United Kingdom that is here in question. Only Dutch nationals and citizens of the United Kingdom and Colonies can benefit from it. Thus the effect of keeping such provisions alive is inevitably, in many cases, to maintain discrimination of a kind forbidden by Article 7 of the Treaty. The authors of the Treaty cannot have intended such a result.

Secondly the number, the complexity and the variety of those bilateral and multilateral conventions are such that it would make administrative nonsense to require the social security institutions of Member States to consider, in the case of every migrant worker, not only his rights under national law and under Community law, but also his possible rights under such conventions. I have been persuaded that the authors of the Treaty cannot have intended such a result either. In saying this I do not overlook that, by virtue of Article 7 of Regulation No 1408/71, some international conventions and some provisions of other international conventions are excepted from Article 6, or that by virtue of Article 8 Member States may, as need arises, still conclude conventions with each other “based on the principles and in the spirit of” the regulation. I do not, however, think that that detracts from the force of the practical considerations that the authors of the Treaty must be taken to have had in mind when they framed Article 51.

It was, however, common ground between the United Kingdom Government, the Council and the Commission — and this is important — that Article 6 of Regulation No 1408/71 could apply only in a case where the provisions of that regulation effectively replaced the pre-existing conventions.

I propose to deal next with the Commissioner's Question I (b), which, Your Lordships remember, is directed to determining whether, having regard to the comprehensive nature of the British national insurance scheme, covering as it does the employed, the self-employed and others, Mr Galinsky is to be regarded as a “worker” within the meaning of Regulation No 1408/71 in Great Britain, independently of his status in the Netherlands.

The United Kingdom Government, in its written observations, pressed the Court to answer that question and to answer it affirmatively, because, it said, such an answer would greatly facilitate the administration of social security in the Community in relation to persons who have at one time been insured under such a comprehensive scheme. On 27 November 1980, however, after those observations had been lodged, the Council announced that it had agreed upon a regulation extending to self-employed workers moving within the Community the system of coordination of national social security schemes already applicable to employed workers. The regulation is to enter into force once the necessary adaptations to Regulation No 574/72 (the “implementing Regulation”) have been agreed upon. Counsel for the United Kingdom Government told us at the hearing that, as a result, the present question had lost some of its importance.

The question is essentially one of interpretation of Article 1(a) of Regulation No 1408/71, which defines the term “worker” for the purposes of the regulation. We were referred to a number of authorities on the interpretation of the phrase “travailleurs salariés ou assimilés” in the old Regulation No 3. Those authorities are not, however, in my opinion in point because Regulation No 1408/71 uses different language. Article 1 (a) reads as follows:

“‘Worker’ means :

(i)subject to the restrictions set out in Annex V, any person who is insured, compulsorily or on an optional continued basis, for one or more of the contingencies covered by the branches of a social security scheme for employed persons;

(ii)any person who is compulsorily insured for one or more of the contingencies covered by the branches of social security dealt with in this regulation, under a social security scheme for all residents or for the whole working population if such person:

—can be identified as an employed person by virtue of the manner in which such scheme is administered or financed, or

—failing such criteria, is insured for some other contingency specified in Annex V under a scheme for employed persons, either compulsorily or on an optional continued basis;

(iii)any person who is voluntarily insured for one or more of the contingencies covered by the branches dealt with in this Regulation, under a social security scheme of a Member State for employed persons or for all residents or for certain categories of residents if such person has previously been compulsorily insured for the same contingency under a scheme for employed persons of the same Member State.”

The only provision of Annex V that appears to be in point is paragraph I (1) which provides, so far as material, that all persons who are “employed earners” within the meaning of the legislation of Great Britain shall be regarded as “workers” within the meaning of Article 1 (a) (ii). That provision was held in Case 17/76 the Brack case [1976] 2 ECR 1429 not to restrict but to clarify the scope of Article 1 (a) (ii) in relation to the British legislation.

The Commissioner's suggestion was, Your Lordships remember, that Mr Galinsky might be regarded in Great Britain as a “worker” by virtue of Article 1 (a) (i). That would involve treating the British scheme as a “scheme for employed persons” within the meaning of that phrase in Article 1 (a). It seems to me impossible to do that if only because the phrase is there used in contrast to such phrases as “a social security scheme for all residents or for the whole working population” (in Article 1 (a) (ii)) and “a social security scheme of a Member State for employed persons or for all residents or for certain categories of residents” (in Article 1 (a) (iii)). Article 1 (a) must in my opinion be interpreted as a whole and it appears to me manifest that it is Article 1 (a) (ii) that fits a person insured under the British scheme. That accords with the approach of the Court in the Brack case (albeit that the Court there left the question open) and also with its approach in Case 84/77 the Recq case [1978] ECR 7 and in Case 143/79 the Walsh case [1980] ECR 1639. Indeed, if one treated a comprehensive scheme such as the British as a “scheme For employed persons”, Article 1 (a) (ii) would become largely otiose and its provisions restricting the definition of “worker”, in the case of a person insured under such a scheme, to one who “can be identified as an employed person by virtue of the manner in which such scheme is administered or financed, etc.” would be rendered futile.

I conclude that Mr Galinsky is or has been a “worker” within the meaning of Article 1 (a) only by virtue of his having been insured in the Netherlands.

On that footing I turn, lastly, to the Commissioner's Questions I (a) and II, which I propose to deal with together and shortly.

The fact that Mr Galinsky is or has been a “worker” in the Netherlands is clearly sufficient to make him a person to whom Regulation No 1408/71 applies — see Article 2 (1) of the Regulation. It does not, however, in my opinion, mean that he was a “worker” in Great Britain. Nor does it mean that each and every provision of the regulation can be applied to him in the same way as if he had been a “worker” in both Member States. That depends, in the case of each such provision, on its content.

It was common ground between those who made submissions to us on this point (the Insurance Officer, the United Kingdom Government and the Commission) that Article 77 of the regulation could only be applied to Mr Galinsky in Great Britain if he was a “worker” there. The Commissioner, so it appears from his order for reference, shared that view. I share it too, not least because any other view could lead to a pensioner falling between two stools in a case where the proviso to Article 79 (1) of the regulation had to be applied in one of the Member States concerned. Nor does it appear to me that the circumstance that the amount of Mr Galinsky's Dutch pension was revised pursuant to the regulation can have any bearing on the question. I therefore think that Article 77 is inapplicable to Mr Galinsky in Great Britain. If that is so its provisions do not, as regards his rights there, replace those of the bilateral Convention, with the consequence that Article 6 of the regulation is also inapplicable.

In the result, I am of the opinion that, in answer to the questions referred to the Court by the Commissioner, Your Lordships should rule that a person who has been compulsorily insured as a self-employed person (but not as an employed person) in a Member State is not to be regarded in that Member State as a “worker” within the meaning of Article 1 (a) of Council Regulation (EEC) No 1408/71 by reason only of the fact that he is a “worker” under the legislation of another Member State, or by reason of the fact that he has been insured in the first-mentioned Member State under a social security scheme covering employed, self-employed and other persons.

The Commissioner has so formulated his questions that, if that is the correct answer to Question I, no answer to Questions II and III are called for. Your Lordships know, however, what my answers to those questions would be.