Oakley Inc v Animal Ltd & Ors

This case is part of a battle between rival makers of designer sunglasses. But it may have implications which go far beyond the worlds of fashion and designs.

A constitution tells us who has the power to make the laws that govern a country. This case raises a point about the constitution of the United Kingdom.

The UK is part of the European Union. The EU issues directives. They are proposed by the European Commission and approved by the European Parliament and Council. Article 249 of the Treaty Establishing the European Community explains that “A directive shall be binding, as to the result to be achieved, upon each Member State to which it is addressed, but shall leave to the national authorities the choice of form and methods”.

It follows that once a directive has come into force the Member States are bound to bring about the result that it requires to be achieved, and to do so by the date it specifies. This may require them to change their own internal laws. Should a Member State refuse or fail to do so it may be brought before the European Court of Justice, and fined. By remaining part of the European Union a Member State accepts that liability.

When we joined the European Community in 1972 it was appreciated that we would be required to implement directives. Article 189 of the Treaty of Rome said so, and its wording was identical to the current Article 249. So Parliament passed a very important constitutional measure. It made it possible for our law to be changed without need for any further Act of Parliament. It is section 2(2) of the European Communities Act 1972. It provides that a designated government minister or department shall have the power to make regulations. By section 2(4) those regulations not only have the force of law, they may even amend an Act of Parliament.

A question that arises in this case is: Can a government minister or department use these powers to make regulations which bring about a result that is not mandated by a European obligation? If there is such a power, how far does it go?

I have listened to arguments on behalf of the Secretary of State which satisfy me that the power that she claims must go very far indeed. Because regulations made pursuant to section 2(2) can amend even an Act of Parliament, as I have already observed, there is raised a question of some constitutional importance. (Footnote: ¹) Furthermore, and contrary to what was anticipated back in 1972, a very great use is now made of the powers conferred by section 2(2) by some government departments.

While this judgment was in preparation the European Court of Justice issued its decision in Case C-255/01 Markopoulos (7 October 2004). It raises yet another point of law – this time of European law – which may be of importance. It is about derogations. Therefore, I postponed the handing down of my judgment in order to receive arguments from the parties.

European directives frequently contain ‘derogations’. A derogation is an exception carved out of the broad scheme of the directive which Member States can avail themselves of if they want to. The case before me is about a derogation.

The question is: to what extent, if at all, may a Member State avail itself of a derogation after the final date by when the directive was supposed to have transposed the directive into its internal law as a matter of urgency?

The actual issue I have to decide could not be more mundane. Is Registered Design No. 2,059,677 invalid for prior publication? But the answer to that question may depend on which person or body had the power to make the relevant law about the validity of registered designs.

As I have said, the issue is part of a battle between rival makers of designer sunglasses. I shall call them “Oakley” and “Animal”. Oakley are suing Animal for infringing their registered design. They claim that Animal’s sunglasses are too close. Apart from other defences, Animal say: “Well, your registration is invalid anyway, and should be declared to be so, because your design was not new. You published the design yourselves, before taking proper steps to register it”.

This is the trial of a preliminary point of law. The parties have agreed certain facts which can be taken to be true for the purposes of this preliminary issue. They are very simple.

On 10 May 1996 Oakley sold in this country sunglasses made according to their disputed design. Of course, by doing this they disclosed the design to the public in this country. To use the jargon of registered designs law, they ‘published’ their design in the UK. One month later, on 10 June 1996, Oakley applied for design protection. (They did it in America; but nothing turns on this because Oakley in due time followed it up with a British application. So the British application takes the benefit of the American date, 10 June 1996).

Animal did not bring out their competing design – the one Oakley are complaining about – until much later. By that stage there had been important changes to our law of registered designs.

However, the fact remains that Oakley’s design was published in this country before Oakley applied for design registration. Therefore, say Animal, the Oakley design registration is invalid. But Oakley riposte: “According to which law?” For, as we shall see , there are two different laws that might have applied to the situation.

According to the version preferred by Oakley, the law allowed a 12-month grace period. During that period Oakley were entitled to sell their sunglasses without needing to initiate steps to register their design. They could, so to speak, test the market to see if it was worth their while to incur the expense of registration. Since 10 May 1996 was well within the grace period, it follows that the Registered Design is not invalid for prior publication.

According to the version preferred by Animal, the law allowed no such grace period. It follows that the Registered Design is invalid for prior publication.

Which version is right? The answer is said to depend on who had the power to make the law in the first place.

In this country there are at least two kinds of property in designs. I shall call them ‘Registered Design’ and ‘Unregistered Design Right’. In some ways they are similar, and it is possible to have both at the same time. Some of the main differences between them are set out in the following table.

Sometimes yet a third kind of protection is available for designs, which I shall call Industrial Copyright. In practice this is effective where the design is ‘for an artistic work or a typeface’ (Footnote: ²). An ‘artistic work’ could be a drawing, a photograph, a model for a building, and so forth (Footnote: ³). For example, the author of a sculpture (which counts as an artistic work) could license its reproduction in the form of garden gnomes (which is a form of industrial exploitation). Like Unregistered Design Right, copyrights cannot be registered: they arise automatically.

There is a fairly obvious relationship between these rights but sometimes it is quite complex. For example, if you infringe the Industrial Copyright in a design you do not infringe the Unregistered Design Right also (1988 Act, s.236). And in certain circumstances the Registered Design right will have expired when the copyright did (Registered Designs Act 1949, s.8(5)).

In order to obtain a British registered design you must fill in a form and send it off to the Designs Registry – a branch of the Patent Office – together with a depiction of your design. Officials at the Registry will then search their records to see if the design, or one similar to it, was already known. If they think your design is sufficiently new, and that your application complies with the other requirements of the law, they will cause it to be recorded on the Register. Your competitors are then entitled to search the Register (subject to certain exceptions which are not relevant here), and so they can take steps to avoid inadvertently infringing your right.

In practice many designs are registered which should not have been. (For example, the official search may have failed to spot that the design was not really new. It is easy for this to happen seeing that the files at the Designs Registry cannot possibly record all of the numerous designs that have been used in the past.) Therefore the law says that a person who disagrees with a registration – typically, a competitor – is entitled to apply to the Designs Registry to have it cancelled or, as we now say, declared invalid. Or he can wait until he is sued for infringing the registration and ask the court to do it. Of course, by waiting he risks losing vital evidence in the meantime or, I suppose, that the law might be changed. Either way, he must show that the design was not new, or otherwise failed to comply with the relevant requirements of the law.

A registered design is a monopoly. Other kinds of intellectual property include patents for inventions and registered trade marks, and they too are monopolies. Copyrights are yet another kind of intellectual property and although they are not, strictly speaking, monopolies, they are very like one. The same is true of Unregistered Design Right.

According to the constitution of the United Kingdom the government (i.e. the Executive) cannot grant or regulate monopolies, or rights in the nature of monopolies, unless and to the extent that it has been authorised to do so by Parliament. This was established long ago: The Case of Monopolies (1602) 11 Co Rep 84b; The Clothworkers of Ipswich Case (1615) Godbolt 252; The Statute of Monopolies 1623 (21 Jac 1 c 3) (Footnote: ⁴). Failure to respect that principle was one of the many causes of the English Civil War. Perhaps forgetfully, in the year 1964 counsel claimed for the Crown an inherent power to grant monopolies for broadcasting. The Court of Appeal roundly rejected that contention: British Broadcasting Corporation v. Johns (Inspector of Taxes) [1965] Ch 32. Diplock LJ said (at 79):

The framing of legislation which governs trading monopolies requires the making of difficult policy choices. There are competing constituencies and a delicate balance has to be struck between them. The inventor or designer wants to be rewarded for his skill, genius or efforts. If he sells his invention or design to an entrepreneur, that entrepreneur naturally wants the best possible monopoly he can get. It should be easy to enforce, and hard – or better still, impossible – to challenge. So that is one constituency. The public wants to be able to buy products manufactured to a high standard at the lowest possible prices and from a diversity of sources. That is another constituency. The manufacturer or trader who supplies the public wants to be able to compete in the market free from red tape and other people’s monopolies. That is yet another constituency. One and the same person may belong to more than one constituency. A designer wants the best possible protection for his designs. But he may create his designs using copyright-protected CAD software whose high price he laments.

The accommodation of all these interests is pre-eminently a matter for the legislature. Accordingly, patents, registered designs, copyrights and similar rights are regulated by Act of Parliament. For patents, we have the Patents Act 1977. For registered designs, we have the Registered Designs Act 1949. For copyrights or unregistered design rights, we have the Copyright, Designs and Patents Act 1988 (itself amending the Registered Designs Act 1949). For registered trade marks we have the Trade Marks Act 1994. All of those statutes are still in force.

However, in recent years the European Community has taken up the revision of intellectual property law. The result is a steady stream of directives issuing from the European Union. Member States, including the UK, are required to ensure that their laws comply with those directives. So they may have to change their internal laws. Who changes our internal laws is not a matter for the EU, but for our own constitution. Sometimes we do it by Act of Parliament. The European directive on registered trade marks was implemented by the Trade Marks Act 1994. But most usually it is done by regulations made by the Secretary of State for Trade and Industry under section 2(2) of the European Communities Act 1972, or purportedly so.

In general, Regulations made under the European Communities Act 1972 are laid before both Houses of Parliament and come into force unless annulled by one of them. But that is not at all the same thing as an Act of Parliament. One crucial difference is that a Bill in Parliament can be, and often is, amended in Committee. Anybody who has had much experience of “Pepper and Harting” a lengthy Act of Parliament is likely to have been impressed by the improvement to the quality of the legislation. In contrast, a set of regulations that is laid before both Houses of Parliament cannot be amended. It is all or nothing. (I wonder if they can receive much scrutiny at all, see paragraph 80 below.)

If you take a practitioner’s law book, particularly one on intellectual property and consult the various statutes printed at the end, you will often find that the Acts of Parliament are riddled with amendments made, not by Parliament, but by a Secretary of State. It is important to bear in mind that those texts are not necessarily the law. They are the Executive’s assertion of what the law may be. Parliament may make a law (primary legislation) and it may authorise the Executive to make subordinate legislation so as to change the law without need for a further Act of Parliament. A subordinate law – such as a regulation made by the Secretary of State – is presumed to be valid until the contrary is shown. Even so it is law if, but only if, the subordinate legislator had the power to make it and (what comes to the same thing) has properly exercised that power. If not, the regulation is said to be ultra vires: it is invalid. It is invalid because he had no inherent power to make law and Parliament did not authorise him to make that regulation.

A citizen who is affected by a regulation and can show that it is invalid can have it declared to be so by launching proceedings for judicial review. But he does not have to take proceedings for judicial review. In the general case he can wait until the point comes up in litigation to which he is a party and take it then: Boddington v. British Transport Police [1998] UKHL 13; [1999] 2 AC 143. But absent a court determination a regulation is to be treated as valid.

Often a regulation is not controversial: it plainly achieves the result the Directive intended us to achieve. But sometimes it is controversial. Then there may be concern that the maker of the regulations has failed to appreciate the true intent of the Directive, and has made a mistake. Or, that he appreciated it full well but has, for reasons best known to himself, purported to amend an Act of Parliament to achieve a result that was not required by the Directive. Either way, he has purported to achieve a result that was required neither by Parliament itself nor by the European Community.

Perhaps a single example may be mentioned here, to give something of the flavour. In 1995 a Secretary of State made a Regulation that claimed to extend the duration of copyright in ‘films’ as defined in the Copyright, Designs and Patents Act 1988. The new term was far longer than had been prescribed by Parliament or (I believe, though I am not deciding it) allowed by the European Community (Footnote: ⁵). I believe that whoever drafted that Regulation may have overlooked that ‘films’ are mere recordings, and are protected irrespective of any creative content, which may be nil (Footnote: ⁶). The authors of the regulation would have confused ‘films’ with ‘cinematographic or audiovisual works’, which do have to have a creative content and which the Directive had indeed required to bear a much longer term of copyright. This illustrates how easy it is for what may be a crucially important mistake to be made when there is no adequate machinery for adversarial scrutiny.

That Regulation – it may not even be the most striking example (Footnote: ⁷) – still purports to be in force, and in my experience continues to confuse those who come across it unawares. In this particular instance the result is, I am inclined to believe, contrary to the Directive, and so I think it is likely that the regulation is invalid. In other instances that could be given the result is less clear, because it is not actually contrary to the relevant European directive, but is not required by it either.

By the year 2000 the authors of the third edition of The Modern Law of Copyright and Designs (of whom I was one) thought they should devote a chapter to the topic of invalid regulations. It is chapter 2, and it was cited to me in this case. It was just a convenient way of citing a large body of case law and summarising some of the arguments because, of course, the book does not state the law, only someone’s opinion of what the law may be. The chapter identified 16 instances of intellectual property legislation which were “to put it neutrally, of doubtful validity” (Footnote: ⁸) and the list did not claim to be comprehensive.

What if it should be held that a regulation is ultra vires? Does this cause the whole of the subordinate legislative scheme to fail? By no means necessarily so. The court may be able to strike out just the offending part, or even to rewrite it so as to cure the defect (for there is no ‘blue pencil’ test in public law). However there is a limit to the court’s power and I shall return to this later.

As I have said, the law of registered designs is governed by the Registered Designs Act 1949. It was amended by Parliament, and very thoroughly, in 1988: the amended text is reproduced in Schedule 4 of the Copyright, Designs and Patents Act 1988. I shall call that text “the 1949 Act” (Footnote: ⁹).

Ten years later the European Union decided it was time to overhaul the law of registered designs because the laws were different in each Member State, causing trouble and confusion. It was necessary to harmonise them. They issued a directive for that purpose. It is Directive 98/71/EC of the European Parliament and Council of 13 October 1998 on the legal protection of designs. I shall call it “the Directive”.

So our law of registered designs had to be changed. This was done by the Secretary of State by regulations made by her pursuant to section 2(2) the European Communities Act 1972. Those are the Registered Designs Regulations 2001. I shall call them “the 2001 Regulations”. They came into force on 9 December 2001. It is the validity and effect of certain of those Regulations which I have to consider.

The Directive was approved in Luxembourg on 13 October 1998. It required Member States to transpose its provisions into their national law by 28 October 2001 at the latest. You would think it should have been time enough. Even more than at first sight appears, because representatives of our country must have been involved in the making of this European legislation long before October 1998. Anyway, the deadline was compulsory. Despite that, our 2001 Regulations did not come into force until 9 December 2001. The delay was presumably a breach of the Directive, and not for the first time in the intellectual property world.

The explanatory note attached to the 2001 Regulations claims that the 1949 Act “already provides a very similar protection to that required under the Directive”. That is a matter of opinion. I cannot really subscribe to it myself. Later in this judgment I shall show why I believe the Directive is a different animal. Be that as it may, the 2001 Regulations proceed to amend the 1949 Act almost on a line-by-line basis, employing the traditional techniques of the British parliamentary draftsman. So many, so detailed are the amendments that it is like a sculptor crafting a cat out of a dog.

Although I am not blaming any particular individual, I feel bound to remark that this is an extraordinarily untidy, confusing and risky way of going about things. I say so for several reasons.

The registered design which is before me in this case was applied for before 9 December 2001, i.e. before the 2001 Regulations came into force. I shall refer to registrations of that vintage as “existing registrations”, and there are a great many of them. Anyway, the right was applied for and granted under the 1949 Act as passed and amended by Parliament, and nobody else.

Animal contend that the registration is invalid because Oakley prior published the design themselves before taking any steps to register it. That will certainly be true if the law as to prior publication continues to be governed by the 1949 Act: see below. But Oakley contend that the law of prior publication is now governed by the scheme of the Directive, which contains a grace period of 12 months, ample to save them on the preliminary point of law I have to decide.

Oakley admit that according to the 2001 Regulations, if these are all valid, the benefit of the 12-month grace period is denied to proprietors of existing registrations. According to the Regulations the validity of existing registrations is governed by the 1949 Act, but the scope of any monopoly they enjoy is governed entirely by rules mandated by the Directive.

Oakley say this is bizarre anyway, because the result is a strange hybrid animal with uneven legs (see below). But, however that may be, Oakley go on to point out, quite correctly, that the Directive did not actually require the United Kingdom to do it. It is, say Oakley, an exception carved out of the scheme of the Directive on the say-so of the Secretary of State, and they claim that she did not have the power to make that decision.

Oakley admit that the Directive allowed Member States to retain their national laws about validity for existing registrations if they wanted to, by way of a national derogation. But they say that if the United Kingdom was to avail itself of this derogation at all the policy choice should have been made by Parliament, and not the Secretary of State. Furthermore, say Oakley, it should have been done in a timely manner, and not after the date by which the Directive should have been implemented, because this is a Directive which intends to achieve harmonisation of national laws within a stated time. In order to follow these arguments more closely I ought to review some of the leading features of the Directive and the 1949 Act, respectively.

Like any other law concerning registered designs, both the 1949 Act and the Directive contain certain requirements about validity and about infringement.

What does ‘validity’ mean? It means that invalid registrations cannot be enforced at all, and are liable to be declared invalid. I have already mentioned one requirement for validity: a design must be sufficiently ‘new’. But what counts as ‘new’ for the purposes of the 1949 Act is not the same as for the Directive. I shall examine that topic, and some of the other major differences with regard to validity, a little later in this judgment.

What does ‘infringement’ mean? For present purposes, it refers to the scope of the monopoly that is enjoyed by the proprietor of the registered design. Or to put it in layman’s terms, how far away you have to get in order not to infringe the registered design. Once again the rules about infringement as required by the Directive, when properly understood, are not the same as those laid down by the 1949 Act.

Validity and infringement may seem to be different topics, but that is not quite true. For example, it will be fairly obvious that a very innovative design ought to enjoy a wider monopoly than a design which is more commonplace (Footnote: ¹¹). It follows that the novelty of a design may determine whether the design should remain on the register at all, and it may determine the scope of the monopoly if the design does remain on the register. Likewise there may be other requirements of the law which go both to validity and to infringement.

Under the 1949 Act a registered design, in order to be valid, had to be ‘new’. Immaterial details or common trade variants did not count. That understood, it had to be new in comparison with anything previously published in the United Kingdom (Footnote: ¹²). Rather bizarrely, designs published abroad, but not here, did not count for that purpose. So I suppose you could go to Milan or New York, spot a likely design, fax a representation to your agent in London, and register that design here. The prior publication in Milan or New York did not prevent your registered design from being ‘new’. Not in point of law, anyway.

On the other hand (and subject to limited exceptions I shall not go into now) any kind of prior publication of your design in the UK would prevent it from being new. Even if you published it yourself (Footnote: ¹³). Thus you could not try the water by putting your product on the market to see how it got on – to see if it was worth the expense of registration. If you did, you would have invalidated your own registration. This was always a problem in some industries: there were torrents of new designs, but it was hard to identify the few that were going to be popular enough to justify the expense of registration (Footnote: ¹⁴). Furthermore, sometimes a single act of publication, however obscure, was fatal. Even the sale of a length of curtain material to a housewife in Polperro would be enough to do it (Footnote: ¹⁵).

The rules according to the Directive are different. In principle, a prior publication in any country of the world might do. It is true that a publication “which could not reasonably have become known in the normal course of business to the circles specialised in the sector concerned, operating within the Community” does not count: Article 6.1. But this may be an exception directed to lack of notoriety, and not necessarily to place of publication. I suppose it might be held to apply to certain transactions within the UK, as in my Polperro example.

Furthermore, the Directive contains an exception in favour of those who publish their own designs. Article 6.2 says:

Under the 1949 Act a registered design, in order to be valid, had to appeal to the eye of the customer. Section 1(1) said:

Furthermore, eye-appeal (or lack of it) might affect the scope of the monopoly. Suppose you registered a design for an article, part of which did have some eye-appeal but part of which did not. Then your competitor came out with an article which was identical to yours in all respects, except that the part that had eye-appeal was completely different. You would expect the law to say: no infringement (Footnote: ¹⁶).

In contrast, Recital (14) of the Directive specifically says that although features of a design which are dictated solely by a technical function should not be protected, “it is understood that this does not entail that a design must have an aesthetic quality”. So you can now register a design which has no eye-appeal at all; and this design may be valid and infringed. What is more, an existing design that barely scraped through the validity requirements of the 1949 Act (because of its marginal eye-appeal) can, just by staying on the register, acquire a very different significance under the new law. It is as if a team entered a soccer competition but then the referee imposed rugby rules.

According to the 1949 Act a design could not validly be registered for a part of an article which was not made and sold separately (Footnote: ¹⁷). You can see the broad justice of that principle if confined to parts which ‘must match’. If the off-side front panel of your car has taken a bash you will need a replacement; and one that does not match the appearance of the opposite side will not do. Thus the 1949 Act contains a ‘must match’ exception (Footnote: ¹⁸) in order to prevent car manufacturers and others from obtaining a monopoly in such “compulsory” replacement items. However, the 1949 Act was held to go much further. It was held that you could not validly register a design for a component part which nobody would want to buy for its own sake (Footnote: ¹⁹), even though there was no need for visual matching, and even though the component part had appeal in itself. (I suppose one example would be an ornamented door for a domestic fire).

In contrast, the Directive requires such component parts to be protected if they are visible in normal use. Article 1 says:

The upshot is that the two philosophies are not the same. For instance, upon transposing the Directive into UK law an existing registered design right could turn into a different animal. Oakley say it would be an odd animal if its validity were to be governed under one philosophy and its infringement by another but that, at any rate, any decision to create such an animal should have been a matter for Parliament, and not the Secretary of State.

You cannot properly understand a directive unless you know its purpose, and this is set out in the Recitals at the beginning. An important purpose of the EU is to have a market without internal frontiers, much as they do in the United States. The purpose of our Directive was to harmonise the law of registered designs throughout the EU to that end.

This is plain from its Recitals. Recitals (1) to (3) tell us that it is necessary to approximate the laws of the Member States on the legal protection of designs because the existing differences can affect the proper working of the internal market and distort competition. Recitals (5) and (6) tell us that it is unnecessary to undertake a full-scale approximation of those laws for it is enough to limit it to the aspects which are causing most of the harm. A distinction is made between substantive and procedural matters: the latter can be left to the Member States themselves. Recitals (7) and (8) – which I shall return to – tell us that this Directive is not attempting to harmonise the law of unregistered designs (what I have called ‘Unregistered Design Right’) or what I have called Industrial Copyright.

Recitals (9) and (10) are important. They say:

Recital (19) tells us that “the rapid adoption of this Directive has become a matter of urgency for a number of industrial sectors”, and goes on to explain that this is why one topic has been deferred for the time being – I need not go on to explain why.

It will be obvious that the Directive had to deal with what I have called existing registrations: roughly speaking, designs already on the registers of one or more Member States. In theory the Directive might have done nothing about these and just confined itself to future registrations. But then it would have failed to achieve its main object, that is, to harmonise registered designs law throughout the EU within a reasonable time – nay, urgently, see Recital (19). A recent registration might have remained in force for another 25 years and so it would have taken that long for harmonisation to work through the system.

Instead, and subject to limited exceptions, the Directive applies to all registered designs, new or already existing. This is apparent from Article 2.1. It says:

Furthermore Article 12, having prescribed what rights shall be enjoyed by the holder of a design registration, continues thus:

The purpose of this provision is to preserve legitimate expectations. It is there to protect lawful competitors – traders whose products did not infringe under the old national law and, but for that provision, might have been changed into infringers overnight. They are granted a personal defence. Apart from those parties, the law of infringement as prescribed by the Directive applies to registered designs of all vintages. Animal do not claim that Article 12.2 applies to them. They introduced these sunglasses after the new law had come into force.

Article 11 of the Directive lists the grounds upon which a design may be refused registration or, if already registered, declared invalid. They include: failure to be a ‘design’ as defined; lack of novelty; lack of individual character; wrong proprietor registered; and so on. Then Article 11.8 says:

So the United Kingdom, regarded as a State, had the option of availing itself of this derogation if it wanted to. A major point in this case is that Oakley say that, if that option was to be exercised at all, it should have been done by Act of Parliament, in the light of the relevant policy implications, and not by a mere regulation made by the Executive: in this case, Regulation 12. So Oakley contend that Regulation 12 is contrary to the constitution of the United Kingdom, and is ultra vires. Irrespective of that, say Oakley, if the option was to be exercised at all, it had to be done by a certain date, namely, 28 October 2001, the deadline by which the Directive was required to be transposed into national law as a matter of urgency. Since the 2001 Regulations were not made until after that date the purported exercise of the derogation is contrary to the law of the European Union – again with the result that Regulation 12 is ultra vires.

Regulation 12 deals with existing registrations of Oakley’s vintage. In pertinent part it reads as follows:

In other words, if this Regulation is valid the United Kingdom has availed itself of the derogation mentioned in Article 11.8: the validity of existing registrations continues to be governed by the old law – what I have labelled “the 1949 Act”. I must now consider whether the Regulation is valid. This depends on whether the Secretary of State had the power to make it. And that depends on the true construction of section 2(2) of the European Communities Act 1972.

This is an extraordinarily important constitutional provision. But it is written in English that is quite inaccessible to the man on the Clapham omnibus. What am I to do? I must do my best to explain my understanding of it. Now, the general idea is not so hard to understand after all if you get hold of one simple concept. It is as follows.

Basically, the European Union makes two kinds of laws. The first kind is supposed to apply automatically – to bind citizens without further ado. It is said to be ‘directly applicable’. The second kind is typified by a directive, like the Directive in our case. A directive binds Member States, not citizens. Thus it is not automatically part of a State’s internal law. Unless it should happen that the State’s internal laws already comply, that State will be required to enact specific legislation in order to put it into force in its territory: to ‘transpose’ the directive into national law, as they say.

Section 2(1) of the European Communities Act 1972 means to deal with the first kind of European law: the sort that is directly applicable. It takes 93 words to tell us that directly applicable laws shall come into force in this country without further ado (Footnote: ²⁰). It describes such a law as an “enforceable Community right”. Obviously we needed something along those lines if we were to join the European Community, or else the directly applicable Community laws would not have been so in this country.

Section 2(2), on the other hand, means to deal with matters which may require specific legislation if they are to be part of our internal law. It provides machinery for that purpose. Put simply, it empowers the Executive to make orders or regulations to achieve it – subordinate legislation. Of course, the 2001 Regulations about registered designs are an illustration in point.

With that introduction, let us visit the language of section 2(2), which I hope will now be easier to follow. I shall emphasise the most immediately relevant parts in bold type, just to get started, although in the end I must consider the lot.

So: a minister or department is nominated; and he, she or it can make regulations. Section 2(4) provides that those regulations shall have the force of an Act of Parliament. To be precise, it says:

In my judgment section 2(4) must be regarded as subordinate to section 2(2). ‘Includes’ does not broaden out section 2(2) so that the Executive can make a regulation which has the force of an Act of Parliament even though that regulation is not within the purposes of section 2(2). On the contrary the Executive must first bring itself within section 2(2) and only then can it point to section 2(4). Compare Minister of Health v. The King, Ex parte Yaffé [1931] AC 494. If that were not so a designated Minister or government department would have a general and unfettered power equal to that of Parliament itself, subject only to Schedule 2 which, as I shall show, is not all that much of a limitation. They could even abolish the Parliament Acts.

Even so, it is clear that the combined effect of section 2(2) and (4) is to enable the Executive, in appropriate circumstances, to make legislation with all the force of an Act of Parliament, and even to amend an existing or future Act of Parliament. Thus section 2(2) combined with section 2(4) is an instance of what is sometimes known as a “King Henry VIII clause” (per Laws LJ in Thoburn v. Sunderland City Council [2002] EWHC 195 (Admin); [2003] QB 151 at paragraphs 13 and 38). It is a power granted by Parliament to the Executive to make subordinate legislation which itself counts as if it were primary legislation. The name is a reference to that monarch’s supposed avidity for absolute powers.

Now, whatever may be the precise constitutional position concerning King Henry VIII clauses, it will be obvious that powers of that sort have to be watched rather carefully. Therefore when Parliament does delegate to the Executive the power to amend primary legislation the courts scrutinise that power with care and consider that it should not receive anything but a narrow construction. Any doubt is to be resolved against the Executive. Of course if there is no doubt the provision must be applied as it stands. See R v. Secretary of State for the Environment, Ex parte Spath Holme Ltd [2001] 2 AC 349 at 382, 400, H.L.; R. v. Secretary of State for Social Security, Ex parte Britnell [1991] 1 WLR 198, 204, H.L.; see also Thoburn v. Sunderland City Council [2002] EWHC 195 (Admin); [2003] QB 151 at paragraph 73. In Spath Holme at page 382 Lord Bingham of Cornhill cited with approval this passage from the judgment of the Lord Donaldson M.R. in McKiernon v. Secretary of State for Social Security (1989, unreported):

Now, is that a paradox? For Parliament is supposed to be supreme. (I am here ignoring the secondary point that Parliament has chosen to subordinate itself to European law, because that is an artefact of our membership of the European Union and Parliament could change that tomorrow.) Therefore, when Parliament has enacted a King Henry VIII clause, by what right do the courts apply a restrictive interpretation? How is it that any doubts must be resolved against the Executive, seeing that Parliament has said that the Executive can make law as if it were Parliament itself?

I believe it all depends on what it means to say that Parliament is supreme and on what Parliament is actually taken to have claimed. At bottom, all effective constitutions depend on legitimacy. In the United Kingdom the pre-eminent source of legitimacy is Parliament – what is called the Queen in Parliament – and that has been taken to be so for 350 years at least. Because of that fact it would be extraordinarily difficult to overthrow our constitution by force, yet it is one of immense flexibility and on the whole it has served to protect fundamental rights. In the last resort even a written constitution like the Constitution of the United States of America has constitutional force only because everybody takes it to have that force. One should not confuse the institution with the document. Some South American republics have copied the document almost word for word but their institutions have not achieved the same legitimacy.

But to say that Parliament is supreme does not mean that Parliament is taken to have, and to assert, all the powers of an absolute despot. Despite the oft-repeated phrase that Parliament can do anything, that is subject to certain reservations and, in my respectful opinion, a certain amount of harm is done by pretending otherwise. Any exception calls into question the rigidity of the supposed rule, and here are two:

(1) Judicial Review. Respected commentators have pointed out that in order to prevent inferior tribunals from exceeding their powers our courts have not infrequently defied the language of Acts of Parliament; and that they have been doing it, in one way or another, for hundreds of years. See the 7th edition of Wade & Forsyth, ‘Administrative Law’, pp.731-8, whose authors added:

(2) Parliament Cannot Abdicate. In Thoburn v. Sunderland City Council [2002] EWHC 195 (Admin); [2003] QB 151 the Court rejected the proposition that Parliament, in enacting section 2 of the European Communities Act 1972, had set permanent limits on its own sovereignty. See paragraphs 53-64. This was so even though the ECJ had interpreted the Treaty to mean that member states, by joining the Community, have permanently limited their sovereignty. It is enough to quote one sentence from the ECJ’s judgment in Case 6/64 Costa v. ENEL [1964] ECR 585:

I do not press the above matters. For present purposes it is enough to proceed on the basis that King Henry VIII clauses attract a restrictive interpretation, any real doubt being resolved against the Executive.

Schedule 2 contains certain backstops: things the government is not allowed to do by regulation, and which would require a further Act of Parliament. Thus the government cannot put up taxes, or pass retrospective legislation, or (in general) delegate these powers to anybody else, or create a new criminal offence landing you with more than two years in prison.

That is reassuring, for it would be rather surprising if the Executive could do those things. But there is nothing in Schedule 2 itself to stop a government from doing other things which would be just as surprising, or more so. I shall choose fanciful examples, and then tell you why I have done so. Such as: making regulations about dissolving your marriage; or about depriving you of your citizenship; or about taking away your right to vote; or about confiscating your property; or about conscripting you into the army; or about controlling your freedom of speech. Suppose the European Community required none of those things, and then ask yourself: are those fanciful examples? Of course they are. That is why I chose them. For, as I shall show, if it is the case that a government is not allowed to make such regulations, the limit must lie in section 2(2), and nowhere else.

Mr David Pannick QC, who appeared for the Secretary of State at my invitation, and for whose arguments I am greatly indebted, contended that the matter was as clear as could be. Here is a Directive that tells us we can have a derogation (Article 11.8): we can keep our old law on validity if we want to. Well, the Secretary of State has chosen to exercise that option. She had ample power to do it under section 2(2) of the 1972 Act. There are two ways she could have done it.

The first way (he says) is under section 2(2)(a). This allows the Secretary of State to make regulations ‘for the purpose of implementing any Community obligation’. We were indeed under a Community obligation: we had to implement the Directive. Yes, agrees Mr Pushpinder Saini, who argued this point for Oakley. But (continues Mr Saini) we did not have to implement the derogation. That was not a Community obligation. To the contrary: the “default” position was that we would implement the whole lot, with no derogation at all. That would not have been a breach of our Community obligation. Indeed if we had done that the Community would have been better pleased, because our law would be more perfectly harmonised.

To which Mr Pannick replies: we had to do the one or the other. Either we had to take the Directive lock, stock and barrel, and implement that; or we had to take the Directive with the exception permitted by Article 11.8. What we could not do was to do nothing. Our Community obligation required us to do something. It follows that Regulation 12(2) and (3) was made ‘for the purpose of implementing a Community obligation’.

It is an elegant argument, but I am not persuaded by it. I believe it proves too much. It implies that whenever a Community obligation affords a derogation the decision – whether to adopt it or not – can always be made by the Executive alone, merely by making a regulation, irrespective of the public importance of the matter, which might bear comparison to keeping the pound sterling. I am reluctant to believe this absent clear and compelling statutory language, and in my judgment such is not to be found in section 2(2)(a) of the European Communities Act 1972. As I have said, the courts presume that Parliament does not lightly take the exceptional course of delegating to the Executive the power to amend primary legislation and any such power is to be scrutinised carefully and should receive a strict and narrow construction. Any doubt should be resolved against the Executive. After all, if Parliament wanted to confer the wider power it could have used clear and unambiguous language in order to do so, and must be taken to have known it.

It was undoubtedly the general purpose of the 2001 Regulations to implement a Community obligation. But we are concerned with the purpose of Regulation 12. It is the latter alone whose validity is challenged by Oakley. If you go through the 2001 Regulations one by one and ask, each time, “Was the Community obliging us to do this?” the answer, when you get to Regulation 12, is “Certainly not”. In my judgment, Regulation 12 was not made for the purpose of fulfilling a Community obligation. If anything, it was the opposite.

It is, at any rate, certainly not an abuse of language to say that the purpose of Regulation 12 was, not to implement a Community obligation, but to implement a derogation. In my judgment the Secretary of State’s argument is founded on an ambiguity or lack of compulsive force in the language of section 2(2)(a) and depends for its success on resolving that ambiguity or lack of compulsion in favour of the Executive.

Membership of the Community imposes obligations but may confer certain rights too. Although it was not stressed in argument, section 2(2)(a) says that regulations may also be made for the purpose of ‘enabling any rights enjoyed … by the United Kingdom under or by virtue of the Treaties to be exercised’. But is a derogation like Article 11.8 a ‘right’ enjoyed by the UK under the Treaty? I believe not, for if the UK had not signed the Treaties it would have enjoyed the ‘right’ anyway. We could have had any registered designs law that we liked. The benefit did not arise from joining the club. Derogations were well known in 1972 and clearer language would have been employed if it had been intended to refer to them.

I have not overlooked that ‘Community obligation’ is defined – in Schedule 1 to the European Communities Act 1972 – as “any obligation created or arising by or under the Treaties, whether an enforceable Community obligation or not”. I do not believe it helps me in what I have to decide. I doubt that it could cover a mere gentleman’s agreement at diplomatic level which the ECJ would not recognise. The obvious case that is covered by “or not” is the implementation of a directive before the final deadline. We do not have to wait until the last day. If we do it a month before, we are nevertheless implementing a Community obligation, though the Community could not have enforced it at that stage. But even if ‘Community obligation’ includes a moral obligation, not legally binding, it makes no difference in this case. We were not under any obligation, moral or otherwise, to avail ourselves of the Article 11.8 derogation.

Mr Pannick QC says that if the case is not covered by section 2(2)(a) it is certainly covered by section 2(2)(b). Regulation 12 was made ‘for the purpose of dealing with matters arising out of or related to’ a Community obligation. What was the purpose of Regulation 12? To avail ourselves of a derogation to an obligation imposed by the Directive. Surely that was a matter ‘related to’ the obligation. Or, if not, it was a matter ‘arising out of’ it.

This argument, although plausible at first sight, caused me some concern. I wanted to know how far it went. What kind of relationship or connection is contemplated by s.2(2)(b)? Can the Executive make regulations – even changing an Act of Parliament – so long as they are in any way ‘related to’ a Community obligation, even though not required by it?

During the course of argument I put some instances to Mr Pannick. I asked Mr Pannick whether he thought the Secretary of State, while she was implementing our Community obligation with respect to Registered Designs, could have overhauled our law with respect to Unregistered Design Right and Industrial Copyright. There would be no Community obligation to do that, quite the contrary (see Recitals (7) and (8) of the Directive), but would it be enough that her purpose was to deal with matters ‘related to’ the obligation it did impose? For to an intellectual property lawyer those are indeed matters which relate to Registered Designs, see paragraphs 17 to 19 of this judgment. Mr Pannick was cautious, but I understood him to claim that she did indeed have that power. I also asked him what would happen if a directive told the UK to extend the term of copyright to a certain minimum level, not forbidding us to go further if we wanted to. Could the Secretary of State go further? (Footnote: ²¹) Mr Pannick’s reply was adroit. He said he did not concede that the government lacked that power.

Mr Pannick QC was prepared to admit that there must be some outward limit to the Executive’s power. He said that the making of a regulation dealing with parental discipline, for instance, obviously would not be for a purpose ‘related to’ or ‘arising out of’ our Directive. But he added that I did not have to worry about where to draw the line. All I had to decide was that Regulation 12 was related to the Directive. What closer relationship could be imagined? I should not speculate about other cases. Like Cardinal Newman, I was not to ask to see the distant scene – one step enough for me.

But, as it seems to me, the government’s argument contains a silent premise. It is that, with the limited exception of Schedule 2, the only outward limit on the powers conferred by section 2(2)(b) is set by the words ‘matters arising out of or related to’ a Community obligation and that any discernible kind of relationship will do. I believe that premise ought to be examined, not assumed.

For what the Secretary of State’s claim seems to me necessarily to entail is a right to amend any Act of Parliament (without needing to persuade Parliament to pass a further Act specially made for the purpose) so long as her purpose is to deal with a matter ‘related to’ an obligation owed by the UK to the European Union, and her regulation is not actually forbidden by Schedule 2. And if the Secretary of State for Trade and Industry has that power, so must any other designated government minister or department. Since nowadays there are few branches of UK law which do not, in a broad sense, somehow ‘relate’ to EU obligations, the principle underlying the claim would seem to go very far indeed. It is not an answer to say that the actual claim in this case is “only a little one” (which, by the way, I do not think it is). For that would be to solve the problem in this case by reference to no known judicial standard and to allow a creeping erosion of the constitution of the United Kingdom. Each step in that process might be described as “only a little one”.

I reproduce the following extract from the home page of the European Union’s website (Footnote: ²²). What it outlines is known as the ‘acquiscommunautaire’.

The expression ‘acquis communautaire’ is conveniently defined in a Glossary which is also available on the EU’s website (Footnote: ²³), as follows:

The list of topics I have reproduced gives one an adequate idea of the range of Community obligations, but perhaps not of their depth or detail. I take a couple of examples at random. Council Directive 94/80/EC concerns the right to vote in local elections and to stand for office. Council Regulation 2201/2003 (Footnote: ²⁴) concerns what courts have the power to dissolve your marriage or rule on the custody of your children (and, therefore, whose law to apply); and there is another Council Regulation concerning maintenance for children and spouses. I could multiply other examples at length. The general point, however, is simple. Nowadays there are few aspects of the law of this country which are not in some sense ‘related to’ our Community obligations. But to say that, therefore, the Executive has the power to change our law at will – including our Acts of Parliament – even when this is not required by any Community obligation would seem remarkable. For most practical purposes it would be a right to rule by decree.

For example, could a government department have said: “Since we have to change part of the law concerning elections to local authorities – in order to comply with the European Directive – why not overhaul the rest of our local electoral law while we are at it? Why bother to get an Act of Parliament? Surely the whole ‘relates to’ the part.” Or, upon the coming into force of the Council Regulation about divorce courts, could the appropriate government department say: “Now, this is an instance of ‘the operation from time to time’ of section 2(1) of the European Communities Act. So why not make regulations which overhaul the rest of our law of divorce, child custody and maintenance? We can truly say they are ‘for the purpose of dealing with matters related to’ the Council Regulation”.

If the answer is, “Yes, you have the right to do it”, then, for all that I can see, there is almost no limit to the power of the Executive in this country, save for what is called ‘Wednesbury unreasonableness’. But if the answer is, “Obviously not; you need to get an Act of Parliament”, then I want to know why not. I want to know the legal principle. It is not enough just to assert, “Oh, they would not do that”.

Surprisingly, there is not a great deal of case law on the true scope of this extremely important provision, section 2(2)(b) of the European Communities Act 1972, and there has been some judicial controversy.

The point was first mentioned in Hayward v. Cammell Laird Shipbuilders Ltd [1988] AC 894, 903, H.L. where a woman complained that she was not being paid as much as male colleagues who were doing work of equal value. An Act of Parliament had made certain provisions in that regard. Later, that Act had been amended for the purpose of complying with Community law and this had been done by regulations made under section 2(2) of the European Communities Act 1972. Although the House of Lords did not have to decide the following point, Lord Mackay of Clashfern LC (with whom Lord Bridge, Lord Brandon and Lord Griffiths agreed), said:

The main English authority is R v. Secretary of State for Trade and Industry ex parte Unison [1996] ICR 1003 but I should say at once that it has not been followed in Scotland. The European Communities Act 1972 is a United Kingdom statute and it cannot have one meaning in England and another in Scotland. So somebody must be wrong.

I shall examine the Unison case first. Its circumstances are not very easy to understand, for me anyway, because it concerns detailed provisions of employment law. But by scrutinising not only the law report but also the legislation itself (of which there were several items) and two decisions of the ECJ which our court took into account, I discern as follows.

Directive 75/129/EEC required workers’ representatives to be consulted in cases of redundancy. (This was not just a formality: the right to be consulted was actually worth something.) However this directive was confined to collective redundancies – where 10 or more employees lost their jobs. It did not require anything to be done if just a single employee was made redundant, or just a few.

In summary, then, so far as single or few redundancies were concerned the 1980 Act took away the employee’s right to representation (but not his abstract right to be consulted); the 1992 Act consolidated that; and then the 1995 regulations purported to take away even his right to be consulted in person. The question before the Divisional Court was whether the Secretary of State had the power to take away that last-mentioned right just by making regulations under the European Communities Act 1972, or whether it would have required an Act of Parliament in that behalf.

The Divisional Court held that the regulations were valid. I do not presume to comment on the actual result regarded as a point of employment law. Furthermore, although the Divisional Court was a branch of the High Court and so the decision is not binding on me in strict law (see R v. Greater Manchester Coroner, Ex parte Tal [1985] QB 67 at 76-81) I would certainly be slow to depart from it, or indeed question its underlying reasoning if that case had stood alone. In fact, it has been called in question: see below. However what exercises me in the present case is the principle that is said to govern the interpretation of section 2(2)(b) of the European Communities Act 1972.

At page 1013 of the report Otton LJ (with whom Newman J agreed) said:

Pausing there, I am not quite sure that I follow that last sentence, because a member state would also have had the right to introduce a measure in favour of workers even if it had nothing to do with the Directive – more paid holidays, say. Continuing:

Again I do not quite follow the significance of the fact that the Commission did not complain that the earlier measures did not “relate to” the Directive. For I would have thought that what was before the court now was the validity of this new regulation, which purported to remove the right to consultation in cases of single or few redundancies, which was not the subject of Community law at all. Nor do I see that there was any occasion for the Commission to consider the meaning of “related to” in the context of section 2(2)(b) of the European Communities Act 1972, which was scarcely its business. I venture to make those observations only because they seem to colour the last paragraph of the judgment which I have quoted above.

Be that as it may, it appears that the court in Ex parte Unison proceeded on the basis that anything is “related to” a Community obligation so long as it is not distinct, separate or divorced from it. What does seem clear, however, is that the court was not referred any of the authorities applicable to the interpretation of so-called King Henry VIII clauses. So far as I can tell none of these appear in the list of authorities cited by counsel and no argument appears to have been addressed to that topic at all. Nor did the court enquire into the purpose of section 2(2)(b). The court just interpreted it acontextually.

Now, if any regulation made by a designated Minister or government department is to have all the force of an Act of Parliament, subject only to the limited exceptions provided in Schedule 2, provided only that it is not distinct, separate or divorced from a Community obligation, the powers that those persons and bodies enjoy must not only be very wide, they must be utterly extraordinary. Yet these implications do not seem to have been considered by the court in Unison at all. This is somewhat perturbing. I wonder why it has been thought necessary, every time there has been an amendment to the Treaty establishing the European Community, and every time there have been new accessions to that Treaty, to deal with the matter by Act of Parliament. Why not just pass a regulation? Surely those matters are not “distinct, separate or divorced from” our existing Community obligations. Or, to come to the future, one day this country may have to decide whether to ratify the European Constitution. Could this be done by a mere regulation? Surely our opportunity to ratify the European Constitution is not distinct, separate or divorced from our existing Community obligations either.

Ex parte Unison has attracted extraordinarily little interest outside specialist employment law circles. It is not reported in the main law reports. It is not commented upon in any article in Halsbury’s Laws of England on constitutional law, European law or (except in a different context) administrative law. It is not mentioned in Bennion on Statutory Interpretation. However, it has been referred to in other case law, which I shall now mention.

Some six months after the judgment in Ex parte Unison was delivered the point fell to be considered in Scotland: Addison v. Denholm Ship Management (U.K.) Ltd [1997] ICR 770. In that case an issue before the Employment Appeal Tribunal was whether certain employment-protection regulations made by a Minister under section 2(2) of the European Communities Act 1972 covered the crew of a floating hotel located in the North Sea oilfield, which was held to be a ship (see at 783F). In essence, it was held that according to European law the U.K. did not have to confer employment protection on this crew; but neither were we were prevented from doing so. It was contended by the employers that our regulations did not intend to confer employment protection upon the crew but that if, on their proper construction, our regulations did purport to do so, they would be ultra vires since this was not a Community obligation. If it was to be done at all it should have been done by Act of Parliament, and not by mere secondary legislation. The judgment of the EAT was delivered by Lord Johnston. At page 784 he said:

It should be stressed that Lord Johnston was not querying the vires of the regulations which were before him on the basis that they might be contrary to the Directive. He was doing so on the basis that Community law permitted them to be made: it was up to each individual member state. Although the passage I have quoted was not essential to the Tribunal’s decision it is plain that Lord Johnston was critical of the notion that the Executive could make a regulation under section 2(2) in order to achieve a result that was not mandated by a Directive.

In Betts v. Brintel Helicopters Ltd [1997] EWCA Civ 1340 the Court of Appeal noted the divergence of opinion between Unison and Addison but said that in the circumstances of that particular case it was not necessary to resolve the conflict.

In R v. Secretary of State for Trade and Industry, Ex parte Orange Personal Communications Ltd, Administrative Court, 25 October 2000 it was held that certain regulations made under section 2(2) of the European Communities Act 1972 which, if valid, took away valuable rights of Orange which they had enjoyed under the Telecommunications Act 1984, were ultra vires; but only upon the narrow ground that the regulations had failed explicitly to state that rights enjoyed under primary legislation were being taken away. The Unison case was distinguished on that ground, but otherwise approved. However the court does not seem to have been referred to Lord Johnston’s criticism of Unison.

In Thoburn v. Sunderland City Council [2002] EWHC 195 (Admin); [2003] QB 151 it was held that regulations which were made under section 2(2) of the European Communities Act 1972 in compliance with a European directive were effective to amend the Weights and Measures Act 1985, thus abolishing the right to use imperial units of measurement. The argument that the regulations were ultra vires because they purported to amend primary legislation passed after the date of the European Communities Act 1972 was rejected. Also rejected was the argument that King Henry VIII clauses are in principle only to be used to effect minor changes. In that case the United Kingdom had no option but to prohibit the use of imperial units of measurement. The Executive was not attempting to achieve a result which was not mandated by a Community obligation. Hence the present point on section 2(2)(b) did not arise; and, although the Unison case was cited to the court, the judgment of the court did not need to consider it and did not do so.

In Perth & Kinross Council v. Donaldson, EAT, Scotland, 30 October 2003, Lord Johnston returned to the point. A European Directive (“ARD”) intended to protect the rights of employees in the event of a transfer of a business to another employer. The Directive was transposed into U.K. law by certain regulations (“TUPE”) made under section 2(2) of the European Communities Act 1972, or purportedly so. An employment tribunal held that TUPE applied to the facts and circumstances of that particular case, where the employer became irretrievably insolvent. On an appeal to the EAT it fell to be considered whether, assuming that TUPE did indeed cover the case, it was to that extent ultra vires, seeing that the Directive did not actually require the U.K. to afford that measure of employment protection in the case of an irretrievable insolvency. It was held that the relevant provision was ultra vires and that, if it was desired to cover a business in that situation, recourse would have to be had to primary legislation. Lord Johnston said:

For the reasons I have given in many places in this judgment, I respectfully prefer the approach of Lord Johnston in the Scottish cases to that of Otton LJ in the Unison case. But that does not relieve me of my task of interpreting section 2(2) of the European Communities Act 1972 for myself and to that I now turn.

I start by enquiring: what was the object which Parliament may have had in mind when it enacted section 2(2) of the European Communities Act in 1972? Plainly that object must have concerned the implementation of obligations or rights under European law which were not automatically effective in the United Kingdom without more, those being the province of section 2(1). Section 2(2) is telling us that in certain circumstances the Executive has power to make secondary legislation. By section 2(4) it shall have the force of an Act of Parliament: hence it may amend or repeal primary legislation. Now, it would anyway be possible to deal with a Community obligation or right by enacting primary legislation. So what is the reason this alternative route – secondary legislation – is provided? Surely it must be because it would be inconvenient to have to get an Act of Parliament every time we needed to implement a Community obligation.

Practical convenience, then, must be an object of section 2(2). But that cannot supply the whole answer, because it would always be a great convenience for an Executive to be able to legislate at its pleasure at any time on any subject without getting an Act of Parliament. Yet obviously no Parliament would ever concede that measure of power to an Executive – under normal peacetime conditions, anyway. What, then, is special about Community obligations that might, on any rational view, make all the difference? In my judgment the answer is to be found by examining the context in which section 2(2) was enacted.

Article 189 of the Treaty of Rome (now repeated in Article 249) said that directives were binding as to the result to be achieved, but left to the national authorities the choice of forms and methods. In any given case the achieving of the result would be a Community obligation; and, since it would be a foregone conclusion anyway, it would be pointless, and a great waste of Parliamentary time, to be obliged to do it by the making of primary legislation. But the adoption of this or that form and method would not be a Community obligation (always provided it achieved the required result and did so effectively, of course). For instance, if a European directive were to require us to ban smoking in public places, it would not prescribe the penalty, and might not lay down that there should be inspectors, vicarious liability for publicans, and so forth.

It is therefore a perfectly rational hypothesis to say that the purpose of section 2(2) was to enable the Executive not only to achieve the required result but to choose the methods appropriate to that end, so long as those methods did not breach Schedule 2. (The “forms” are, of course, prescribed by Parliament and are to consist of an Order in Council or, more commonly, regulations to be made by a designated minister or government department.)

On that hypothesis the existence of section 2(2)(b) may be accounted for without making the extravagant assumption that Parliament has chosen to confer upon the Executive a free-standing power to legislate upon any subject at will, with all the force of an Act of Parliament, so long as it somehow ‘relates to’ a Community obligation, even though the result which is achieved by that legislation is not actually required by the Community at all. I have already illustrated how that would produce results best described as startling and absurd.

So far in this judgment I have slightly simplified the task that must have confronted the framers of section 2(2), because I have referred to European directives as if those were the only Community instruments that require a measure of transposition into national law. In fact, directives are not the whole story, because there may be other Community acts which could have binding force yet which might have to be dealt with by Member States. Thus Article 249 of the Treaty says that the European Parliament together with the Council, or the Council or the Commission alone, may make ‘regulations’. It says

Let me quote the relevant language again. According to section 2(2):-

In my judgment, if this were not part of a King Henry VIII clause, section 2(2)(a) by itself might have implied a power on the part of the Executive to adopt methods for achieving a result required by Community law, even though those particular methods were not themselves required by Community law. That is the so-called rule in A-G v. Great Eastern Railway Co (1880) 5 App Cas 473 which, according to Lord Blackburn, is that “those things that are incident to, and may reasonably and properly be done under the main purpose [of an enactment], though they may not be literally within it, would not be prohibited [i.e. be ultra vires]”. See 44(1) Halsbury’s Laws of England §1335 and Bennion, ‘Statutory Interpretation’, pages 429-430. In modern statutes it is commonplace for such incidental executive powers to be set forth expressly; and, since section 2(2)(a) is indeed part of a King Henry VIII clause and hence falls to be construed strictly, it is not unreasonable to infer that section 2(2)(b) may have been inserted out of an abundance of caution.

At any rate, I would hold that section 2(2)(b) does not enable the Executive to make secondary legislation, with the potential to amend or repeal even an Act of Parliament, and to achieve a result not required by a Community obligation, just because its purpose is in some way related to or arises out of that obligation.

Although I do not have to decide it in this case, I should explain that neither do I believe that section 2(2)(b) enables the making of secondary legislation which is not required in order for us to enjoy a right under or by virtue of the Treaties.

It should be noted in passing that, in fact, section 2(2)(b) has two limbs. The second limb is “or the coming into force, or the operation from time to time, of subsection (1) above”. The words “arising out of or related to” apply to the first limb, not the second.

For those who may prefer the wider interpretation of section 2(2) the crucial word in that entire subsection is the word “or” that separates subparagraphs (a) and (b), which I suppose they take to be a strong disjunction. I presume that their argument might be as follows. If, to put it with some simplification, subparagraph (a) is about things the United Kingdom must do (say, a Community obligation) or needs to have, then, since (b) is preceded by the word “or”, it must be about things we do not have to do, or do not need to have. Thus there is conferred on the Executive a wide discretion to make secondary legislation even where the EU is not saying we must achieve a certain result or where we need nothing. On that view the word “or” must surely be the most significant particle in the history of British constitutional legislation. But I do not think so. I believe the fallacy consists in treating the English word “or” as if it were necessarily equivalent to a strong logical operator, such as the Boolean “Exclusive OR”. In fact, as the Concise Oxford English Dictionary confirms, in natural English the word has a range of meaning.

In short, if section 2(2) is given a narrow but purposive construction, as it must be, it does not enable the attainment of a substantive result that is required neither by the Community nor by Parliament itself.

The effect of Regulation 12 as worded would be to retain the old (1949 Act) law of validity in respect of existing registrations. Subject to precisely what is meant by “existing registrations”, 11.8 of the Directive permitted this, but it did not require it. Thus it was not a Community obligation. Nor, of course, was it required by Parliament. Nor was the derogation a ‘right to be enjoyed by the United Kingdom’ for the reason I have already explained in paragraph 95 above.

The decision (whether or not to retain the old law of validity in respect of existing registrations the scope of whose monopoly was anyway going to be different) was a significant policy choice and, in my judgment, had to be implemented by primary legislation. Thus I hold that the Secretary of State did not have the power to make that decision – one way or the other – and to implement it by secondary legislation. Such a legislative act would be ultra vires.

If part of a secondary legislative scheme is ultra vires the courts have the power to remove it, or in appropriate circumstances even to re-write it, provided that offending part is ‘substantially severable’. This means that the proposed alteration will not affect the substantial purpose and effect of the impugned legislation as a whole. Otherwise the court has no power to strike out or modify just the offending part, and must treat the subordinate legislation as a whole as being invalid.

In DPP v. Hutchinson [1990] 2 AC 783, 804 Lord Bridge (with whom Lord Griffiths, Lord Oliver and Lord Goff agreed) said as follows:

But after reviewing a number of authorities Lord Bridge continued at page 811:

Here are two contrasting illustrations. In Daymond v. South West Water Authority [1976] AC 609, H.L., subordinate legislation sought to make local householders pay a charge to meet the cost of sewerage. The provision was ultra vires as worded because, although it did not refer to them in terms, it purported to charge even those people whose houses were not connected to the sewers. But it was held that the courts could save the legislation as a whole by notionally amending it – by writing the unconnected householders out of the story. (See the explanation of that case in the next-mentioned authority.) On the other hand, in DPP v. Hutchinson [1990] 2 AC 783, H.L., byelaws which made it an offence for members of the public to enter Greenham Common were too wide because numerous local people had the legal right to go onto the Common anyway. But this time it was held impossible to cure the defect by cutting out the excess and so the byelaws as a whole were held invalid. The reason for the difference is that in the first case the purpose of the legislation was to raise revenue to pay for the sewerage, which purpose could substantially be achieved by exempting the minority who were not getting the services. Even though that required some notional re-writing of the secondary legislation. But in the second case the purpose of the legislation was the security of a nuclear airbase and this could not be substantially achieved by cutting out the excess, seeing that numerous local people could still have wandered over Greenham Common at will without producing identification. A fresh legislative scheme, substantially different in its effect, would have been required.

I venture to suggest that the test of severability, even non-textual severability, will more easily be satisfied where it can be seen that the purpose of the secondary legislator was to transpose a European directive. Even if he has exceeded his powers in one respect, such as by mistaking the true scope of the directive, or by legislating contrary to the directive, or by attempting to produce a substantive result required neither by the Community nor by Parliament, it will normally be clear that the substantial purpose of the rest of the instrument was to implement a Community obligation. That being compulsory anyway, it would be strange if the court were to feel obliged to cut down the whole. It could instead delete the offending part or even rewrite it so as to make it conform to the directive.

Mr James Mellor and Mr Pushpinder Saini for Oakley submitted that if the Secretary of State did not have power to make regulation 12 I should strike regulation 12(2) and (3) out of the text of the Registered Designs Regulations 2001, without more. The result would be that the new law of validity would apply to existing registrations and so Oakley would take the benefit of the 12 month grace period afforded by Article 6.2 of the Directive or, as transposed into British law, section 1B(6)(c) of the 1949 Act as substituted by regulation 2 of the 2001 Regulations.

The upshot would be that Oakley’s prior publication of their own design on 10 May 1996, and thereafter, would not have served to invalidate their registered design. Therefore the preliminary point of law should be decided in favour of Oakley. The decision would not harm third parties, if such there were, who had already started to use the design before the 2001 Regulations came into force, see regulation 14(3) as authorised by Article 12.2 of the Directive.

Miss Fiona Clark, who appeared for Animal, did not agree that the preliminary point of law should be decided in Oakley’s favour even if, contrary to her main contention, regulation 12 should be held to be ultra vires. Article 11.8 of the Directive afforded a derogation. The United Kingdom could avail itself of that derogation, or not, as it pleased. That was a policy decision. If the Secretary of State did not have the power to make that policy decision and herself implement it by secondary legislation, neither would she have had the power to implement the whole of the Directive without the derogation. For that would have been to make the opposite policy choice and to implement it by secondary legislation. That too should have been a matter for Parliament. What was sauce for the goose was sauce for the gander.

In reply Mr Pushpinder Saini said that it would have been a valid act of secondary legislation for the Secretary of State to transpose the Directive into our national law without the derogation. The derogation was an exceptional feature carved out of the scheme of the Directive – what he called the “default” position. Therefore all I had to do was to strike out the words of regulation 12(2) and (3) and apply the rest of the 2001 Regulations.

I believe Miss Fiona Clark is right. Whether the United Kingdom was to avail itself of the derogation afforded by Article 11.8 of the Directive, or not, was clearly a significant policy choice. The power to make that choice and implement in our national law may have lain with Parliament alone (as I have held), or it may have been available to the Secretary of State. But what is absolutely clear is that it did not lie with the courts. Were I to accede to Oakley’s submissions the end result would be that we would have a law of registered designs which, as regards the validity of existing registrations, was chosen neither by Parliament or (if that were relevant) the Secretary of State, nor required by the European Community.

That said, I do not hold that the 2001 Regulations as a whole must be treated as ultra vires. Insofar as these Regulations intend to bring in the new law of infringement in respect of registered designs of all vintages, and insofar as they intend to bring in the new law of validity in respect of new registrations, that intent is correct and indeed its due implementation in our law was a Community obligation. Compare paragraph 141 of this judgment. The remaining question is what law – the new or the old – applies to the validity of existing registrations.

The 2001 Regulations purport to amend the text of the 1949 Act. That text was passed and amended by Parliament, and by Parliament alone. Thus Parliament enacted a particular legislative scheme which was to govern at least the validity of existing registrations. I have held that the Secretary of State did not have the power to make regulation 12(2) and (3). In my judgment the somewhat paradoxical consequence is that the old law of validity as defined by the 1949 Act will continue to apply to existing registrations, because that is what Parliament provided, just as it would have continued to apply if, instead of purporting to adopt the derogation, the Secretary of State had purported to eschew it. Therefore, if that were all I would decide the preliminary point in favour of Animal.

However, what I have just written is subject to the Markopoulos point, which concerns directives which are sought to be implemented out of time. To that point I now turn.

This last issue concerns whether the United Kingdom was entitled to avail itself of the derogation afforded by Article 11.8 of the Directive after the time had expired when the Directive was required to have been transposed into our national law and, if not, with what consequences.

Oakley contend that, after 28 October 2001, any attempt to enact the derogation into United Kingdom law would be contrary to Community law, and ineffective. Therefore, by waiting until after that date the U.K. lost that option. Consequently, the new law of validity applies to all existing registrations. So Oakley are entitled to rely on the grace period of 12 months, and the preliminary point of law ought to be decided in their favour.

I believe that the purpose of the Directive was to achieve a measure of harmonisation of the law of registered designs and to do so as a matter of urgency: see paragraphs 60-64 of this judgment. In particular Recital (10) informs us that it was considered to be essential, in order to facilitate the free movement of goods, to ensure in principle that registered design rights confer equivalent protection in all Member States. Recital (19) informs us that the rapid adoption of the Directive had become a matter of urgency. Article 19 required that Member States should bring into force the laws, regulations or administrative provisions necessary to comply with the Directive “not later than 28 October 2001”. The 2001 Regulations were made on 8 December 2001 and purported to come into force on the next day. They cannot by any process of judicial construction be deemed to have come into effect any earlier than 8 December because Schedule 2 paragraph 1(1)(b) so prohibits.

Oakley do not contend that every time a Member State chooses to transpose a European directive out of time it thereby loses the option to make use of a derogation contained in that directive. There are directives and directives. But they say their point is good with respect to our Directive, because its purpose was to achieve harmonisation of the law as a matter of urgency and by a stated time. The Directive sought to strike a balance between its ultimate purpose – to harmonise registered designs law urgently – and a measure of deference to those Member States who wanted to keep their old law on validity for existing registrations. The cut-off point was set at 28 October 2001. It is true that in our case the 2001 Regulations were only a few weeks late, but Oakley say that cannot affect the principle. If the Regulations had been many months, or even years, late – and delay of that order has been known to happen – the point would be the same, although more evident. There is nowhere you can draw the line except at the date expressly laid down in the Directive itself, namely in Article 19.

Animal point out that Article 11.8 of the Directive does not itself lay down a definite date by which the derogation had to be implemented in national law. It says that “the grounds for … invalidation in force in that State prior to the date on which the provisions necessary to comply with this Directive enter into force” may be caused to apply to existing registrations. “Enter into force” means enter into force in national law. That would cover a date later than 28 October 2001.

For my part I do not think that much can be derived from the language of Article 11.8 on this point. First, because it may presuppose that Member States will anyway have complied with their obligations by the date prescribed in Article 19. Secondly, because the use of words I have italicised (instead of an express reference to Article 19) may be explained on the basis that a Member State might choose to transpose the Directive into its national law earlier than is required by Article 19. If anything, the words “necessary to comply with this Directive” suggest a terminal date not later than is prescribed by Article 19.

In Case C-255/01 Markopoulos, 7 October 2004, the ECJ had to consider the interpretation of Article 15 of a directive (84/253/EEC) concerning who could be allowed to audit companies’ annual accounts. The general purpose of this directive was to harmonise the national laws of the member states on that point and to require that none but approved persons could carry out the work. It was the responsibility of each member state to approve suitable persons. The “default” position, to adopt Mr Saini’s phrase in this case, was stated in Articles 2 and 4 which, put shortly, provided that nobody could be approved unless he had certain passed certain university examinations.

However in Greece there were not a few professionals who had many solid years experience of auditing company accounts under then-existing national law but had not passed these particular examinations. What was to happen to them? Could they be approved too, or should they be forced to sit the examinations like fresh-faced youngsters?

Recital (6) recognised that “Member States should … be authorised to adopt transitional provisions for the benefit of professional persons”. And Article 15 said:

By Article 30:-

In its judgment the ECJ said [references to the “Eighth Directive” are to the auditing directive I have been discussing]:

The substance of that last paragraph was repeated in the ECJ’s formal ruling.

Animal and the Secretary of State say that was a different kind of case, because Article 15 expressly stated that that particular derogation – to allow member states to approve someone as an auditor though he had not passed the examination – applied ‘until one year after the application of the provisions referred to in Article 30(2)’, and Article 30(2) said that member states could, as it were, disapply the provisions until 1 January 1990. Therefore all the ECJ decided was that it was incompetent for a member state to exceed the ultimate time limit – 1 January 1991 – by a late transposition of the directive. That ruling, they said, had no meaningful application to the present case.

I am not persuaded by that argument. First, the fact that the time clause in that particular directive (Article 30) was broken down into two parts – transpose the directive into your national law before 1 January 1988, but then disapply your new national law until 1 January 1990 if you wish – does not seem to me to make any real difference. I can see why it would be so: the Commission would want to see that the member states had “done their homework” by January 1988, compare the wording of Article 30(1), even though any State was entitled to delay the actual implementation until 1990. But it still remained the case that the purpose of Article 15, read together with Article 30(2), was to achieve a balance between, on the one hand, the overall purpose of the directive (harmonisation of the auditing rules) and, on the other hand, the legitimate protection of seasoned professional auditors out on the ground. That balance would be upset if, because of a late transposition, professionals were let into the fold after the ultimate date of 1 January 1991. But it would equally be upset if the same result were to be achieved by other machinery, for example, a national law that disapplied the “exam” requirement after that date.

In that same case the Advocate-General said:

The provision before me in this case is a transitional provision. It allows member states to keep their old law on the validity of registered designs in respect of existing registrations. It is true that that is a long transitional provision, in the sense that it might take as much as 25 years for harmonisation to work through the system, seeing that a registered design may be on the Register for such a period. But, in my opinion, that fact should not be allowed to obscure the proposition that the Directive was required to be transposed into our law no later than 28 October 2001. The Directive does not tell me that a still later cut-off point can be achieved under Article 11.8 by a belated transposition.

In summary, then:

Now, in point of Community law the Registered Design in our case is not itself a victim of this breach of the Directive, because Oakley applied for it well before 28 October 2001. However, that is not quite the point. The point is whether, assuming what I have said in the preceding paragraph is correct, the Secretary of State had power to make Regulation 12(2) and (3) on 8 December 2001. (For present purposes I shall ignore that the Regulations actually came into force the next day.)

By making this Regulation more than a month after the prescribed deadline the Secretary of State was inevitably spreading her net too wide. She was catching a class of registered designs which Community law forbade her to catch. Namely, all those designs whose owners applied for registration after 28 October 2001 and before 9 December 2001. According to Regulation 12 the validity of all those designs falls to be considered under our old law. But according to what was mandated by the Directive it ought to be considered under the new law.

The last point is whether Regulation 12 may be saved all the same by notionally removing the excess. If I could amend Regulation 12 so that, in place of the words “before the coming into effect of these Regulations”, there were inserted the words “before 28 October 2001”, the revised wording would comply with the Directive. There would remain no class of registered designs whose validity wrongly continued to be governed by the new law. No doubt revisions along similar lines would be required to Rules 10, 11, 13 and 14, although those are not before me now. Although it would not satisfy the “blue pencil” test, nevertheless I could make that notional revision in accordance with the principle in DPP v. Hutchinson provided that the substance of what remained (i.e. in the 2001 Regulations as a whole) would be essentially unchanged in its in its legislative purpose, operation and effect.

In my judgment, United Kingdom law allows me to make that notional revision and requires me to do so unless forbidden by Community law. If Regulation 12 were to speak from the appropriate date, 28 October 2001, the 2001 Regulations regarded as a substantive whole would achieve their evident legislative purpose, operation and effect. And although Schedule 2 to the European Communities Act 1972, paragraph 1(1)(b), forbids the making of any provision taking effect before the date when the instrument was made, in my judgment the excision I have in mind would not breach that rule. The surgery is deemed to have been effective from the beginning, that is, since 8 December 2001.

In conclusion, then, Regulation 12 as currently worded would be ultra vires because the Secretary of State has no power under section 2(2) of the European Communities Act to make regulations that breach Community obligations. Parliament itself could do it, but then it would have to be an enactment expressly defying the 1972 Act or repealing it. However, because of the doctrine of substantial severability I consider that Regulation 12 takes effect subject to the revision I have adumbrated and, thus revised, is valid so far as United Kingdom internal law is concerned.

In holding so, I have not quite met Oakley’s point, which is that according to EU law a member state could not in the first place purport to adopt the derogation afforded by Article 11.8 of the Directive after the final date prescribed by Article 19. Oakley have deployed the Markopoulos judgment, not just to say that Regulation 12 as worded goes too far, but that after 28 October 2001 it was too late to make it at all. Oakley’s point is that under the Directive the very existence of the option to exercise the derogation vanished on 28 October 2001.

Thus Oakley’s point presumes that Community law may have an interest in forbidding the late implementation of the Directive which goes beyond that which may be cured by subsequent judicial intervention, even though, technically, that intervention is deemed to have had effect from the beginning. Or, as described by the Advocate-General in Markopoulos in paragraphs 61-63 of his Opinion, an object is to

If that is arguable, a point of Community law would arise which might be expressed along the following lines: “With regard to Directive 98/71/EC of the European Parliament and Council of 13 October 1998 on the legal protection of designs, could a national measure adopted by a Member State, after the date mentioned in Article 19, to the extent that the said measure purported to exercise the derogation mentioned in Article 11.8, nevertheless be partly valid according to Community law? Or did the opportunity which was available to a Member State to adopt the derogation conferred by Article 11.8 by a national measure permanently and wholly cease to exist after the date mentioned in Article 19?”

That question, though barely hinted at in Markopoulos, and perhaps weakly supported in the present state of Community law, is, in my judgment, not acte claire. It concerns legal certainty and transparency, and not just the theoretical legal effect of Regulation 12 discovered after the event. If, therefore, Oakley wish to continue to avail themselves of this point, despite the rest of this judgment, I would propose to refer to the ECJ under Article 234 of the Treaty a question along the lines I have set out above. Otherwise I shall give judgment on the preliminary point of law in favour of Animal.

After the draft of my judgment had been sent to counsel in the usual way for identification of typographical errors but before delivery of the corrected version in open court, Mr Mellor on behalf of Oakley quite properly sought to contend that my judgment contained an error of substance which I ought to correct before the order was drawn up. That error, if corrected, would lead to judgment on the preliminary point of law in favour of his clients. For the purpose of making his submission Mr Mellor sent in a short supplementary skeleton argument, copied to the other parties.

Mr Mellor’s point concerns paragraphs 152-154 above. In essence, he submits that Parliament never mandated that the old law of validity should apply to registered designs whose infringement would be governed by the new law. This was a policy choice and one cannot assume that Parliament would have favoured it. As Mr Mellor puts it, “Parliament has in fact been silent on this issue… It is not the task of the Court to fill any perceived gap caused by the fact that regulations 12(2) and (3) are ultra vires by making assumptions as to how Parliament would have addressed the policy issues – that would be impermissible judicial legislation.” Hence he submits that the proper course is notionally to delete Regulation 12(2) and (3) and to apply “what remains on the statute book” after the deletion.

I do not permit further argument on this point because I am satisfied that Mr Mellor’s point, if fully developed, would not assist his clients. His reasoning if otherwise correct would lead, not to judgment in favour of Oakley, but to a finding that the 2001 Regulations were ultra vires in their entirety. They would be ultra vires in their entirety because Regulation 12(2) and (3) would not be capable of substantial severance, see paragraph 145 above. The result would be that section 1B(6)(c) and (d) of the 1949 Act, as substituted by the 2001 Regulations, would not exist. Therefore there would be no 12-month grace period for anyone, including Oakley.

The reasons why I do not consider that the 2001 Regulations are ultra vires in their entirety have been explained in this judgment, of which paragraph 147 should be noted.

In the alternative Mr Mellor asks for permission to appeal. I shall hear the parties if so desired, but it is fair to say that in a case of this sort permission to appeal would very readily be granted.

In conclusion I wish to say that I am greatly indebted to all counsel for the excellence of their arguments. They were outstanding.