Unilever Plc & Ors v Shanks

This appeal is from a decision of Mann J (3^rd December 2009, [2009] EWHC 3164 (Ch)) by which he reversed the decision of the hearing officer, Dr Elbro, acting for the Comptroller-General of Patents. Mr Daniel Alexander QC, assisted by Mr N Gardner of Herbert Smith, argued the case for the appellants whom I will collectively call “Unilever”. Mr Patrick Green argued the case for Professor Shanks.

Mann J describes the way in which the point in issue came before him. Fortunately it now does not matter, though I should observe that the course of argument before us has been rather different from that below.

Briefly the facts are these:

In 2006, after all but the US patent (expiry 2009) had expired, Professor Shanks made an application under s.40 of the Patents Act 1977 for what is colloquially called “inventor’s compensation.” Thus, when he made his application, the full history of exploitation of the patents was known. There remained no or no significant element of surmise about it.

The relevant part of s.40 and the other provisions in play in this case (ignoring subsequent amendments irrelevant here) read as follows:

40(1) Where it appears to the court or the comptroller on an application made by an employee within the prescribed period that the employee has made an invention belonging to the employer for which a patent has been granted, that the patent is (having regard among other things to the size and nature of the employer’s undertaking) of outstanding benefit to the employer and that by reason of those facts it is just that the employee should be awarded compensation to be paid by the employer, the court or the comptroller may award him such compensation of an amount determined under s.41 below.

…

41(1) An award of compensation to an employee under section 40(1) or (2) above in relation to a patent for an invention shall be such as will secure for the employee a fair share (having regard to all the circumstances) of the benefit which the employer has derived, or may reasonably be expected to derive, from the patent or from the assignment, assignation or grant to a person connected with the employer of the property or any right in the invention or the property in, or any right in or under, an application for that patent.

(2)

For the purposes of subsection (1) above the amount of any benefit derived or expected to be derived by an employer from the assignment, assignation or grant of—

(a)

the property in, or any right in or under, a patent for the invention or an application for such a patent; or

(b)

the property or any right in the invention

to a person connected with him shall be taken to be the amount which could reasonably be expected to be so derived by the employer if that person had not been connected with him.

(3)

Where the Crown or a Research Council in its capacity as employer assigns or grants the property in, or any right in or under, an invention, patent or application for a patent to a body having among its functions that of developing or exploiting inventions resulting from public research and does so for no consideration or only a nominal consideration, any benefit derived from the invention, patent or application by that body shall be treated for the purposes of the foregoing provisions of this section as so derived by the Crown or, as the case may be, Research Council. In this subsection "Research Council" means a body which is a Research Council for the purposes of the Science and Technology Act 1965.

…

44(7) In sections 40 and 41 above and this section " benefit" means benefit in money or money's worth.

(8)

Section 533 of the Income and Corporation Taxes Act 1970 (definition of connected persons) shall apply for determining for the purposes of section 41(2) above whether one person is connected with another as it applies for determining that question for the purposes of the Tax Acts.

The dispute before us is as to the meaning of “that person” in s.41(2). But before I can get to that it is necessary to consider the scheme as a whole.

Under s.40(1) the employee has two hurdles to get over: he must show that the patent is “of outstanding benefit to the employer” and if so that it “is just” that he should be awarded compensation. That compensation is to be paid by the employer.

Clearly this, key, subsection contemplates the inventor’s particular employer, not some notional employer. The fact that it refers to “the size and nature of the employer’s undertaking” makes this clear beyond any argument. It is the real actual benefit to the actual employer which is all that matters. If an invention had been immensely valuable and a patent for it could have been or could be (if still in force) exploited for a vast sum or have fetched vast royalties or a great sum on the open market, that is irrelevant. The inventor’s particular employer may not have exploited the invention well or at all. If so, the inventor cannot complain. The employer must be taken as it was, warts and all. The provision is not some kind of “best endeavours to exploit” requirement.

The whole scheme is built round a paradigm case, the case where the inventor worked for the same company as received all the benefit of the invention. That paradigm case is also reflected in s.41(3) where the Crown assigns the invention. Mr Green helpfully explained why s.41(3) was necessary. s. 522 of the Income and Corporation Taxes Act 1970 whose definition of “connected persons” is harnessed by s.44(8) to be used in s.41(2) does not apply to the Crown. So a separate provision was necessary. For present purposes what matters is that s.41(3) evidently is intended to see that a Crown inventor is treated in the same way as an inventor in the paradigm case. That makes it all the more unlikely that Parliament intended that where there has been an in-house assignment of a patent, the inventor should be treated wildly differently from an inventor in the paradigm case.

As to the amount of compensation, s.41(1) sets out the principles. Consider first the paradigm case. The inventor has surmounted the two hurdles of “outstanding benefit” and that it is “just” he should be paid compensation. S.41(1) says he is to have a “fair share” of the benefit. It is true that the subsection speaks also of the benefit which “may reasonably be expected to derive from the patent”. That is obviously there to cater for the case where the patents have yet to expire and further benefit may be expected to accrue. But it is the future benefit which may accrue to the particular employer which is to be considered. Not some notional general benefit in the market.

Thus in this case, if all the Unilever companies had been a single company, all the business had been just that of that company, Professor Shanks had been employed by that single company, and it was that company which had received the £23m, then there would have been no difficulty in assessing the benefit derived from the patent and invention. It would be the £23m.

In other cases the problem could be much more difficult – one would perhaps have to disentangle the benefit from the patent from the benefit obtained by such things as manufacturing expertise, other inventions or advertising. It would be not unlike the problem which arose in the now long-repealed provision for extension of the term of patent on the grounds of inadequate remuneration: the court had to have regard to “the nature and merits of the invention in relation to the public, to the profits made by the patentee as such, and to all the circumstances of the case” (s.23(5), Patents Act 1949).

The problem in this case is caused by the assignment from CRL to Unilever Plc. Rationally, that would make no difference. The question is whether the various assignments, and particularly the assignment from CRL to Unilever Plc, make a difference by virtue of the operation of s.41(2).

On the pleadings before the Comptroller both sides contended that they did. On Professor Shanks’ behalf it was contended that the benefit should be treated as a lot more than the £23m actually received by the Unilever Group, and on Unilever’s behalf it was contended that the benefit was near-negligible. There was a vast gulf between the parties’ positions, a gulf which narrowed in the argument before us.

Thus it was contended on behalf of Professor Shanks that his compensation should be based on a higher benefit than the actual benefit of £23m royalties. It was said that the effect of s.41(2) was two-fold. The first of these was that the threshold requirement of “outstanding benefit” was to be judged not by the actual benefit to the Unilever Group of companies but by the benefit which could have been obtained on the open market if the invention had been fully exploited by licensing earlier. This was called the “putative benefit.” The second effect was that this putative benefit was also the benefit of which Professor Shanks would be entitled to a “fair share”.

The second effect could lead to a bizarre result. Unilever (the Group) only in fact made £23m. But if the “benefit” was to be taken as US$1bn the “fair share” of that could be a large sum – even overtopping the actual benefit. Before us, Mr Green, recognising that that result would be outrageous, suggested it could be avoided or at least ameliorated by the requirement to have “regard to all the circumstances.” On this basis the “fair share” is determined by first determining the putative benefit and then the actual benefit to the group and cutting the former down. That seems a most unlikely intention of Parliament.

In oral argument Mr Green somewhat modified his submission: he suggested that the “putative benefit” applied essentially only at the threshold stage. Unilever were (and according to Mr Alexander still are) contending that although £23m royalties might be a lot for some companies, by Unilever standards it is not a lot and so the patent was not of outstanding benefit to Unilever. He pointed to the words in s.40(1) “having regard to the size and nature of the employer’s undertaking”, suggesting they meant that inventor/employees of big companies had to show a larger benefit to their employer than inventor/employees of smaller companies. I am far from convinced that Parliament meant that inventor/employees of large companies should get less or no compensation for a particular invention compared with what they would get if they had been employed by a small company. It may indeed be the other way round in that a large payment may be too much for a small company to able to afford and that was what Parliament had in mind. The point does not immediately arise – the Comptroller will have to consider it in due course if it is persisted in.

Mr Green’s “putative benefit” point was thus reduced in scope: principally to answer Unilever’s “we are a large company so you get nothing or less” point. Mr Green submitted that US$1b would be of outstanding benefit to any employer and so Professor Shanks got over the entrance hurdle. He accepted that once one came to determine the fair share, the actual benefit could be taken into account as essentially the determining benefit – though he did not go so far as to accept that it was only the actual benefit which counted.

I pause to point out the temporal element in Mr Green’s putative benefit approach. It involves a retrospective element: the tribunal asks what the benefit of the invention would have been, but knowing how successful it in fact was. It is put this way in the so-called “Supplementary Statement of Case”:

In view of the actual sales of products incorporating the invention, said royalties would have been at least hundreds of millions of $US and more likely than not, in excess of US$1 billion.

A temporal element also came into Unilever’s original position. Before the comptroller it contended that s.41(2) required the tribunal to postulate an actual arm’s length sale at the time of the actual assignment. The notional buyer is not some hypothetical company but the actual assignee. So the question is “what would Unilever Plc have paid CRL on an arm’s length basis for the benefit of the invention at the time of the assignment?” It contended that at that point there were a host of uncertainties – most particularly whether the invention had any commercial worth or whether any resulting patent would be valid. The subsequent actual success of the invention was to be disregarded because it would not have been predicted at the time of the notional sale. In the result Unilever Plc pleaded that it would not “have been prepared to pay more than a modest sum, which would not have been more than a few thousand pounds.”

Recognising that this position was deeply unattractive, Mr Alexander abandoned it before us. He accepted that the £23m royalties in fact obtained by the Unilever Group was the benefit obtained. But, he submitted, that was all. There was no room at any stage for any putative benefit: a benefit which would have been obtained by any arm’s length sale on the open market of the patents: s.41(2) did not call for the tribunal to imagine some sort of notional auction of the patents.

It is whether Mr Alexander is right about that to which I now turn. It depends on what s.41(2) means by the words “the amount which could reasonably be expected to be so derived by the employer if that person had not been connected with him.”

“The employer” in the last words goes back to “an employer” in the opening words. In the opening words that must mean the actual employer – the legal entity which was the real assignor. It is inescapable that the provision has in mind the real actual employer, not some notional employer – the real assignor. No one suggested otherwise.

What then about the assignee? Does the provision mean the real actual assignee – the actual “connected person?” That is the natural meaning of the language, “that person” and is what the Judge held:

[27]… On the words of the section as it stands, I do not think that it is capable of more than one literal meaning in this respect. It refers to a form of disposal, and identifies it in terms of the person to whom it is made - the actual person who is a connected person. It refers to a real disposal to a real person.

I agree with that. So the statute requires one to consider the actual assignment from the actual employer to the actual connected person: on the facts here from CRL to Unilever Plc.

What the statute then calls for is an assessment of what the assignor “could reasonably be expected to be so derived” if the assignee had not been connected with him. “So derived” takes one back to the opening words of the subsection: “any benefit derived or expected to be derived.” That is the same phrase as is used in s.41(1). In that context, the paradigm case, the assessment involves working out what benefit the employer has actually got in the past and any future benefit (if any) which that particular employer would get.

It seems to me that that is the same where there has been an assignment to a connected company. A temporal element comes in. The “benefit derived or expected to be derived” from the assignment is deemed to be the amount the employer could reasonably be expected to have derived if the assignee had not been connected, knowing what benefit was in fact obtained by the assignee. This is the benefit that must be considered for the purposes of s.40(1) and s.41(1). In this sense s.41 reaches through to s.40.

In the present case the benefit derived by the assignee, Unilever Plc, is £23m. Knowing what we now know, what would CRL reasonably have expected to derive from the assignment to Unilever Plc? Assuming (as Professor Shanks asserts) that Unilever Plc in effect got the £23m risk free and as “pure profit” the answer, as Mr Alexander conceded, the £23m.

Now it might be objected that s.41(2) in only referring “to the amount which could be reasonably be expected to be so derived” does not permit the tribunal to look back at what actually happened – the exercise is simply to ask what the assignee would have paid at the time of the assignment assuming it had no connection with the assignor. I reject such a construction. It leads in effect to asking what would have happened in a notional auction of the invention at the time of the assignment.

That in turn leads to such serious absurdities that the meaning cannot have been intended. Here they are:

The Judge only departed from the “natural meaning” of “that person” as meaning the actual assignee because he thought it could produce uncommercial results. But that was predicated on the basis that one is to exclude the known facts about exploitation. Once one brings them in the uncommercial results fall away. And in any event the Judge’s conclusion can lead to equally uncommercial results the other way – as I have outlined above.

In truth this is one of those provisions which is so ill-drafted (see the comments and citation from Hansard of Floyd J in Kelly v GE [2009] EWHC 181 (Pat)) that one has to be guided by its evident purpose (ascertainable from the paradigm case) to ascertain its meaning. My old head of Chambers, Thomas Blanco White QC, used to call this approach to construction of an ill-drafted provision “sewing the fly buttons on the statute.”

In reaching this conclusion I am reinforced by two other considerations:

In the result I hold that “that person” in s.41(2) means the actual assignee with its actual attributes. It follows that I would allow the appeal. The parties should be able to agree the consequential order. I hope they can also now agree on what amounts to a “fair share” for Professor Shanks.

I agree.

I also agree.