SmithKline Beecham plc and others v Apotex Europe Ltd
[3.MB]KWA WAKUBWA TU ,KUANGALIA VIDEO YAO WAKIBANDUANA CHUMBANI BONYEZA HAPA CHINI 18++
KUTIZAMA PICHA ZAKE NA VIDEO ZAKE ALIZO PIGA AKIWA UCHI NI RAHISI BONYEZA HAPA CHINI
[2005]
EWHC 1655 (Ch)
ADMINISTRATION OF JUSTICE; Other: CIVIL PROCEDURE
CHANCERY DIVISION
LEWISON
J
15,
18, 19, 26 JULY 2005
Injunction – Interlocutory – Undertaking as to
damages – Cross-undertaking as to damages – Enforcement of undertaking – By
whom and for whose benefit undertaking enforceable – Third parties – Amendment
of cross-undertaking – Slip rule – Estoppel by convention – Restitution –
Whether cross-undertaking in favour of third parties who were not defendants to
the proceedings could be implied – Whether undertaking could be amended under
the ‘slip rule’ in favour of third parties – Whether undertaking having
prospective or retrospective effect – Whether claim in restitution existing –
Whether claimants estopped by convention from denying third party entitlement
under undertaking – Whether draft pleading liable to be struck out or having no
reasonable prospect of success – CPR 3.4 – CPR Pt 24 – CPR PD 25A, para 5.1.
The claimants, members of the GSK group of companies,
manufactured a pharmaceutical product, Seroxat, widely used for the treatment
of psychiatric disorders. GSK was the
proprietor of a patent which claimed a process for producing paroxetine
hydrochloride anhydrate, the active ingredient in Seroxat. The defendants were a generic pharmaceuticals
company, distributor and parallel importer.
Following revocation proceedings brought by the defendants, the
claimants commenced infringement proceedings against the defendants alleging
infringement of its patented process and applied for an interim injunction
restraining the defendants from disposing of, or offering to dispose of, any
pharmaceutical preparation containing paroxetine hydrochloride anhydrate in the
United Kingdom. In fact the paroxetine
was manufactured and supplied by two other companies (the Canadian companies),
neither of whom was a claimant in the revocation proceedings nor formal
defendant in the infringement proceedings.
In the course of the proceedings the judge granted an injunction,
subject to the claimants’ cross-undertaking as to damages. At trial, the judge dismissed the action and
revoked the patent. However, he
continued the injunction until the effective hearing of a further application,
on the basis of a further cross-undertaking.
Both cross-undertakings referred only to the loss of the defendants or
‘the Apotex Parties’, which did not include the Canadian companies. The claimants indicated that they would not
seek to maintain the interim injunction during the pendency of an appeal but,
one day before inevitable discharge of the injunction, the Canadian companies
applied to be joined as defendants, with the presumed objective of giving those
parties the benefit of the cross-undertaking with retrospective effect. On appeal, the Court of Appeal reversed the
judge’s revocation but held that the patent had not been infringed. The defendants sought to enforce the
cross-undertaking in favour of the Canadian companies, submitting that the
cross-undertakings could be enforced by them or for their benefit on the
grounds, inter alia, that (i) the cross-undertakings be
53
amended under the ‘slip’ rule so as to conform with
the practice direction to CPR Pt 25 on interim injunctions; (ii) the Canadian
companies were entitled to recover the losses they sustained by virtue of a
restitutionary cause of action; (iii) the claimant was estopped by convention
from denying that the Canadian companies were entitled to claim their own
losses under the cross-undertakings; (iv) the first defendant was entitled to
claim the losses of the Canadian companies as ‘third party’ losses recoverable
by them. CPR PD 25A, para 5.1a provided: ‘Any order for an injunction, unless
the court orders otherwise, must contain: (1) an undertaking by the applicant
to the court to pay any damages which the respondent(s) (or any other party
served with or notified of the order) sustain which the court considers the
applicant should pay…’
________________________________________
a Paragraph 5, so far as material, is set out at
[24], below
________________________________________
Held – (1) As at 2002, it could not have been
said that it was implicit in any application for an interim injunction, other
than a freezing or search order, that the applicant would have been taken to
have offered a cross-undertaking in favour of third parties who were not
defendants to the proceedings. It might
be that, following the March 2005 revision to the standard from of interim
injunction order, the practice might change, but it could not realistically be
said to have changed when the interim injunctions in the present case were
granted. Moreover, this was not a case
in which no undertaking was offered; it was a case where a limited form of
undertaking (in what used to be the conventional form) was offered and accepted
both by the defendants and the court. In
such a case an express offer had to displace any implied offer. In addition, the more limited form of
undertaking was what the judge had in fact ordered on both occasions; and
consequently the judge had ‘otherwise order[ed]’, so as to displace the
presumption in the practice direction (see [57]–[59], below); Imutran Ltd v
Uncaged Campaigns Ltd [2001] 2 All ER 385 and Miller Brewing Co v Mersey
Docks & Harbour Co, Miller Brewing Co v Rhui Enterprises Ltd [2004] IP
& T 542 considered.
(2) The slip rule allowed the court to correct an
‘accidental’ error or omission. In the
instant case the form of the cross-undertaking could not be said to be an
accidental error. It was a
cross-undertaking deliberately given in the form in which it was intended to be
given. It was embodied in an order
settled by junior counsel for each party, and approved by the judge. While forgetting to ask for an order which
one intended to ask for could (just) be described as an accidental error, the
same could not be said of a case where it did not occur to anyone that a
particular form of relief might be available.
Further, if the practice direction had been drawn to the attention of
the judges who granted the interim injunctions, the court could not say with
confidence what cross-undertaking they would have required, or the claimants
would have been willing to give. At the
time of the injunctions it would not have been implicit in any application that
the form of undertaking in the practice direction was being offered and, even
if it was, any implied offer had to be displaced by an express offer accepted
both by the defendants and the court. If
the court were to amend the form of cross-undertaking it would be retrospectively
imposing on the claimants an undertaking that they had not given and could not
have been made to give. Accordingly, the
application under the slip rule would be dismissed and the parts
54
of the points of claim that relied upon the slip rule
would be struck out (see [63], [65], [66], below).
(3) On the assumption that the benefit of a
cross-undertaking given in favour of the ‘defendants’ would extend to those who
were added as defendants during the period when the injunction was in force, it
would have only prospective, rather than retrospective effect. The injunction itself could not have been
enforced directly against added defendants until they became defendants; and
their previous acts would not retrospectively become contempt of court. In addition, in the circumstances of the
instant case, the application for joinder was made after judgment at first
instance and after GSK had indicated that it would not seek to maintain its
injunction pending appeal. Accordingly,
if the Canadian companies were joined as defendants to the infringement action,
they would only be entitled to the benefit of the cross-undertaking (at best)
for two days. The joinder application
would also be dismissed (see [49], [67], below); Ketteman v Hansel
Properties Ltd [1988] 1 All ER 38 and Berkeley Administration Inc v
McClelland [1996] IL Pr 772 considered.
(4) The power to strike out had to be exercised
sparingly in a case which involved developing areas of law: it was preferable
that novel points of law were decided on real rather than assumed facts. On the other hand, it was no part of the
function of a first instance judge to apply the law as it might one day
become. His job was to apply the law as
it was (or arguably was). In the instant
case, a restitutionary claim by the Canadian companies to recover a benefit
enjoyed by the claimants which they would not have enjoyed if the injunction
had not been granted was an attempt to outflank the well-established principle
that, in the absence of a cross-undertaking, damages were not recoverable for
loss suffered as a result of the grant of an interim injunction. It followed that, applying the law as it was
and not as it one day might become, the claim in restitution was legally
unsustainable and would accordingly be struck out pursuant to CPR 3.4 (see [21],
[75], [76], below).
(5) The legal basis of the plea for estoppel by
convention was one that was sustainable.
If the court were to go down the route of examining in detail the
particular documents on which the defendants relied, it would be conducting a mini-trial
which current practice under CPR Pt 24 forbade.
Accordingly, the court would not strike out those parts of the points of
claim which relied on convention estoppel, nor give summary judgment in favour
of the claimants on that issue (see [72], [73], below).
(6) The claim that the defendants wished to advance
on behalf of the Canadian companies came nowhere near the Linden
principle relating to third party loss.
The principle was predicated on a breach of contract. No cause of action arose at all unless B had
broken his promise; if B had not broken his promise, discussion of what the
cause of action might encompass if he had done was pointless. In the instant case, the claimants promised
to abide by any order of the court requiring them to compensate the defendants
(or the ‘Apotex parties’ as defined) for any loss that they had suffered. They did not promise to pay the defendants
for loss that other people had suffered.
No amount of development in the law relating to the scope of damages
that could be awarded for breach of contract could result in a party to a
contract recovering damages for something which was not a breach of contract at
all. Accordingly, this argument was
legally unsustainable and
55
would be struck out (see [81]–[84], below); Linden
Gardens Trust Ltd v Lenesta Sludge Disposals Ltd [1993] 3 All ER 417 and
Panatown Ltd v Alfred McAlpine Construction Ltd [2000] 4 All ER 97
distinguished.
Notes
For undertaking as to damages, see 24 Halsbury’s Laws
(4th edn reissue) para 982.
Cases referred to in judgment
A-G v Albany Hotel Co
[1896] 2 Ch 696, CA.
Allied Irish Bank v
Ashford Hotels Ltd, Ashford Hotels Ltd v Higgins [1997] 3 All ER 309, CA.
Berkeley
Administration Inc v McClelland [1996] IL Pr 772, CA.
Colledge v Crossley (1975)
Times, 17 March, CA.
Commodity Ocean
Transport Corp v Basford Unicorn Industries Ltd, The Mito [1987] 2 Lloyd’s
Rep 197.
Fritz v Hobson (1880)
14 Ch D 542, [1874–80] All ER Rep 75.
Galaxia Maritime SA v
Mineralimportexport, The Eleftherios [1982] 1 All ER 796, [1982] 1 WLR 539,
CA.
Hillgate House Ltd v
Expert Clothing Service & Sales Ltd [1987] 1 EGLR 65.
Hoffmann-La Roche
(F) v Secretary of State for Trade and Industry [1974] 2 All ER
1128, [1975] AC 295, [1974] 3 WLR 104, HL.
Imutran Ltd v Uncaged
Campaigns Ltd [2001] 2 All ER 385.
Inchcape, Re,
Craigmyle v Inchcape [1942] 2 All ER 157, [1942] Ch 394.
Ketteman v Hansel
Properties Ltd [1988] 1 All ER 38, [1987] AC 189, [1987] 2 WLR 312, HL.
Linden Gardens Trust
Ltd v Lenesta Sludge Disposals Ltd [1993] 3 All ER 417, [1994] 1 AC 85,
[1993] 3 WLR 408, HL.
Lokumal (K)
& Sons (London) Ltd v Lotte Shipping Co Pte Ltd [1985] 2
Lloyd’s Rep 28, CA.
Miller Brewing Co v
Mersey Docks & Harbour Co, Miller Brewing Co v Rhui Enterprises Ltd
[2003] EWHC 1606 (Ch), [2004] IP & T 542.
National Australia
Bank Ltd v Bond Brewing Holdings Ltd [1991] 1 VR 386, Vic SC.
Norwegian American
Cruises A/S (formerly Norwegian American Lines A/S) v Paul Mundy
Ltd, The Vistafjord [1988] 2 Lloyd’s Rep 343, CA.
Oberrheinische
Metallwerke GmbH v Cocks [1906] WN 127.
Panatown Ltd v Alfred
McAlpine Construction Ltd [2000] 4 All ER 97, [2001] 1 AC 518, HL.
R v Medicines Control
Agency, ex p Smith & Nephew Pharmaceuticals Ltd (Primecrown Ltd
intervening) [1999] RPC 705.
The Bank v A Ltd (23
June 2000, unreported).
The Indian Endurance (No
2), Republic of India v India Steamship Co Ltd [1997] 4 All ER 380,
[1998] AC 878, [1997] 3 WLR 818, HL.
Tucker v New
Brunswick Trading Co of London (1890) 44 Ch D 249, CA.
W v H (Family
Division: without notice orders) [2001] 1 All ER 300.
Z Ltd v A [1982]
1 All ER 556, [1982] QB 558, [1982] 2 WLR 288, CA.
Applications
In proceedings for infringement of their process patent
relating to paroxetine hydrochloride anhydrate, the claimants, (1) SmithKline
Beecham plc, (2) GlaxoSmithKline UK Ltd and (3) Glaxo Group Ltd, applied for an
interim
56
injunction restraining the defendants, (1) Apotex Europe
Ltd, (2) Neolab Ltd and (3) Waymade Healthcare plc, from disposing of, or
offering to dispose of, any pharmaceutical preparation containing paroxetine
hydrochloride anhydrate in the United Kingdom.
The judge granted an injunction, subject to the claimants’
cross-undertaking as to damages ([2002] EWHC 2556 (Pat), [2002] All ER (D) 436 (Nov))
and, following trial ([2003] EWHC 2939 (Ch), [2003] All ER (D) 411 (Dec)), and
continued the injunction until the effective hearing of a further application,
subject again to a cross-undertaking as to damages. On appeal, the Court of Appeal found the
patent not to be infringed ([2004] EWCA Civ 1568, [2004] All ER (D) 431 (Nov))
and the defendants sought to enforce the cross-undertaking for the benefit of
two other companies, not parties to the proceedings, namely Apotex Inc (a
company incorporated under the laws of Canada) and Apotex Pharmachem Inc
(formerly named Brantford Chemicals Inc) (a company incorporated under the laws
of Canada) (the Canadian companies).
Three applications were made to the court: (a) an application to amend a
‘cross-undertaking in damages’ embodied in two orders for interim injunctions,
under the ‘slip rule’; (b) an application to join additional parties to the
action in which the interim injunctions were granted, with the objective of
giving those additional parties the benefit of the cross-undertaking, with retrospective
effect; and (c) an inquiry whether draft pleading would be liable to be struck
out as disclosing no reasonable cause of action; or plead a claim that has no
reasonable prospect of success. By order
of Pumfrey J on 26 May 2005 the third application was to be determined
according to the criteria under CPR 3.4 and CPR Pt 24. The facts are set out in the judgment.
Andrew Waugh QC, Marcus
Smith, Justin Turner and Geoffrey Pritchard (instructed by
Simmons & Simmons) for the claimants.
Antony Watson QC,
Alain Choo Choy and Thomas Mitcheson (instructed by Taylor
Wessing) for the defendants.
Cur adv
vult
26 July 2005.
The following judgment was delivered.
LEWISON J.
57
INTRODUCTION
[1] I have to decide three applications. The first is an application to amend a
‘cross-undertaking in damages’ embodied in two orders for interim
injunctions. The application is made
under the slip rule. The second is to
join additional parties to the action in which the interim injunctions were
granted, with the objective of giving those additional parties the benefit of
the cross-undertaking, with retrospective effect. The third is an unusual application. I have to decide whether certain parts of a
draft pleading, if served in the form in which they currently are, would be
liable to be struck out as disclosing no reasonable cause of action; or plead a
claim that has no reasonable prospect of success.
[2] The third application comes about because
of an order made by Pumfrey J on 26 May 2005 on the claimants’ application for
the trial of a preliminary issue. The
order provides:
‘The Claimants’
application for the hearing of the preliminary issues be refused and instead it
be directed that the question of whether the case as pleaded in the draft
amended Points of Claim raises triable issues in relation to the two Canadian
companies (Apotex Inc. and Apotex Pharmachem Inc.)
58
is to be determined
according to the criteria under CPR Parts 3.4 and 24 at the hearing currently
listed for 3 days floating from 14th July 2005 all other directions in the
enquiry to await determination of these issues.’
[3] In para [31] of the judgment that led to
the making of that order ([2005] EWHC 1122 (Pat), [2005] All ER (D) 430 (May))
Pumfrey J said that he was satisfied that the purely legal argument that would
have to be advanced ‘on the basis that all the facts pleaded are true’ would
take three days.
BACKGROUND
[4] The two claimants in the action are
members of the GSK group of companies.
They manufacture a pharmaceutical product under the brand name
Seroxat. It is widely used for the
treatment of psychiatric disorders such as depression; and is one of GSK’s
best-selling products. The active
ingredient in Seroxat is paroxetine hydrochloride anhydrate. GSK is the proprietor of a patent which
claims a process for producing paroxetine hydrochloride anhydrate substantially
free of bound organic solvent. GSK had
(as one might expect) been protective of their patent. In the two years preceding the current
litigation GSK had sought and obtained interim injunctions against two generics
companies (Generics (UK) Ltd and Alpharma Ltd) restraining alleged
infringements of the patent.
[5] The immediate background to the current
proceedings was described by Mr Reilly of GSK in his witness statement as
follows:
‘7.1 The First Defendant
(“Apotex”) is an affiliate of Apotex Inc., a multinational generic
pharmaceutical company with headquarters in Canada. The Second Defendant (“Neolab”) is a UK-based
distributor of generic pharmaceutical products.
The Third Defendant (“Waymade”) is a parallel importer of branded
pharmaceutical products and a distributor of generic pharmaceutical products.
7.2 In mid-November 2001
it came to GSK’s attention that Neolab had announced an intention to sell a
generic paroxetine product in the UK.
GSK instructed Simmons & Simmons to write to Neolab and, after some
correspondence between solicitors, on 28 November 2001 Neolab gave an
undertaking through its solicitors not to launch any product containing
paroxetine hydrochloride without giving GSK three weeks’ advance notice. In late August 2002 it came to GSK’s
attention that a marketing authorisation for paroxetine 20mg tablets had been
granted to Apotex on 30 July 2002, with Neolab named as a distributor of the
product. For the reason given in
paragraph 8.9 below, this means that Apotex must have applied for its UK
authorisation by January 2002. Copies of
Apotex’s patient information leaflet and Summary of Product Characteristics are
now produced to me marked MR-3.
7.3 On 29 August 2002
Simmons & Simmons wrote to Apotex and the solicitors who had earlier
represented Neolab, requesting confirmation that the undertakings previously
given by Neolab would be adhered to and requesting samples of the product which
Apotex intended to market through Neolab.
On 06 September solicitors on behalf of Neolab and Apotex confirmed that
the undertaking previously given by Neolab would be adhered to by both their
clients. On 09 October 2002 the
solicitors gave notice on behalf of Apotex and Neolab, as well as on behalf of
Waymade, of
59
their joint intention to
launch a paroxetine hydrochloride product in three weeks’ time i.e. on 30
October 2002. GSK were also notified of
the commencement of proceedings by the Defendants for the revocation of the
anhydrate patent and EP (UK) 0223 403 (which covers paroxetine hydrochloride
hemihydrate).’
[6] The revocation proceedings were in fact
begun on 9 October 2002. In response,
GSK began its own proceedings alleging infringement of its patented process. GSK applied in its own proceedings for
interim injunctions restraining the defendants from disposing of, or offering
to dispose of, any pharmaceutical preparation containing paroxetine
hydrochloride anhydrate in the United Kingdom.
The three defendants to those proceedings were Apotex Europe Ltd; Neolab
Ltd and Waymade Healthcare plc. They
were also the three claimants in the revocation proceedings.
[7] In fact the paroxetine was manufactured in
Canada by Brantford Chemicals Inc (which is now called Apotex Pharmachem Inc
and is part of the Apotex group); and some of the finished product was to be
supplied by Apotex Inc (another part of the group) to Neolab. The terms of the supply agreement were such
that the price that Neolab were to pay Apotex Inc consisted of a fixed price
and a percentage share of profit.
[8] Neither Brantford Chemicals Inc nor Apotex
Inc (which I shall call the Canadian companies) were claimants in the
revocation proceedings or formal defendants in the infringement proceedings. Nor did the interim injunction prohibit the
manufacture of paroxetine in Canada; or its import into the United Kingdom (so
long as it was not disposed of or offered for disposal).
PROCEDURAL HISTORY
[9] The first hearing of GSK’s application was
disposed of on 29 October 2002 before Jacob J on the basis of undertakings and
cross-undertakings. The defendants
undertook that they would not until the determination of GSK’s application or
further order whether acting by themselves or through their employees,
directors or agents or otherwise howsoever dispose or offer to dispose of, any
pharmaceutical preparation containing paroxetine hydrochloride in the United
Kingdom. In return GSK gave a
cross-undertaking in the following terms:
‘AND UPON the Claimants
(and each of them) undertaking by their Counsel to comply with any order this
court may make if the court later finds that the undertaking recorded in the
following paragraph has caused loss to the Defendants and decides that the
Defendants should be compensated for that loss …’
[10] On 28 November 2002 ([2002] EWHC 2556
(Pat), [2002] All ER (D) 436 (Nov)) Jacob J granted an injunction in
substantially the same terms as the undertaking until trial or further
order. This time GSK’s cross-undertaking
was in the following terms:
‘AND UPON the Claimants
agreeing to comply with any order this court may make if the court later finds
that this order for an injunction has caused loss to the Defendants and decides
that the Defendants should be compensated for that loss …’
60
[11] The trial of the action resulted in a
judgment of Pumfrey J given on 5 December 2003 ([2003] EWHC 2939 (Ch), [2003]
All ER (D) 411 (Dec)). He dismissed the
action and revoked the patent. On the
handing down of the judgment Pumfrey J continued the injunction until the
effective hearing of a further application on the following cross-undertaking:
‘AND UPON the SmithKline
Beecham Parties undertaking to comply with any order this court may make if the
court later finds that the order for an injunction in paragraph 1 of this Order
has caused loss to the Apotex Parties and decides that the Apotex Parties
should be compensated for that loss …’
[12] The ‘Apotex Parties’ were defined in the
order as Apotex Europe Ltd; Neolab Ltd and Waymade Healthcare plc.
[13] The contemplated further application was
an application for permission to appeal.
That was due to be heard on 18 December 2003. On 16 December 2003 GSK wrote to say that it
would not seek to maintain the interim injunction in force during the pendency
of any appeal. Thus it became inevitable
that the injunction would be discharged two days later. On 17 December 2003 the Canadian companies
applied to be joined as defendants to the action. So far as I can see the only purpose of the
application was to obtain the benefit of the cross-undertaking. In a witness statement made in support of the
joinder application Mr Cohen (the defendants’ solicitor) said:
‘4. I have reviewed the
Witness Statement of [Mr Freeland] in relation to GSK’s licensing arrangements
and their application to add a further Glaxo company to this action. We have also reviewed this issue to ensure
that all the appropriate parties are in the action.
5. I consider that it
would be appropriate for BCI and Apotex Inc (AI) to be joined as parties to
this action in addition to Apotex Europe.
BCI is a company in the Apotex group.
6. The bulk active
Paroxetine is manufactured by BCI in Ontario Canada. This was the site of the lengthy inspection
carried out by GSK. BCI send the active
to Apotex Inc for tableting and packaging.
7. …
8. Both BCI and AI have
abided by the terms of the interim injunction granted and “voluntarily”
provided disclosure in the case. In
addition the process inspection which was ordered in the action took place over
several weeks at BCI’s facility. Further
if, which my clients deny, Apotex’s products have been an infringement of the
patent in suit, both BCI and AI would have been joint tortfeasors.’
[14] On a subsequent appeal to the Court of
Appeal that court reversed Pumfrey J’s revocation of the patent; but held that
it had not been infringed (see [2004] EWCA Civ 1568, [2004] All ER (D) 431
(Nov)).
[15] The cross-undertakings given by GSK on
the grant of the various interim injunctions have now become enforceable. In so far as the court has a discretion
whether or not to permit enforcement, that discretion has already been
exercised by Pumfrey J in the defendants’ favour. An inquiry into damages will, therefore, take
place.
61
THE ISSUE
[16] In their draft points of claim the
defendants seek to enforce the cross-undertaking in favour of the Canadian
companies. The points of claim allege
that the cross-undertakings can be enforced by them or for their benefit on
four separate grounds: (i) the cross-undertakings in the orders of 28 November
2002 and 5 December 2003 should be amended under the ‘slip’ rule so as to
conform with the Practice Direction to CPR Pt 25 on interim injunctions; (ii)
GSK is estopped by convention from denying that the Canadian companies are
entitled to claim their own losses under the cross-undertakings; (iii) the
Canadian companies are entitled to recover the losses they have sustained by
virtue of a restitutionary cause of action; (iv) Apotex Europe is entitled to
claim the losses of the Canadian companies as ‘third party’ losses recoverable
by them.
[17] GSK alleges that even on the assumption
that the facts alleged in the points of claim are true, none of these claims is
legally sustainable.
[18] The Canadian companies also apply to be
joined as additional defendants to the infringement action.
THE TEST
[19] It is common ground that, as regards the
first two applications (amendment under the slip rule and joinder of additional
parties) I must decide the applications, rather than merely decide whether they
are arguable.
[20] Pumfrey J’s order requires me to apply
the criteria under CPR 3.4 and Pt 24 to the remaining applications. Under r 3.4 the court may strike out a
statement of case if it discloses no reasonable grounds for bringing the
claim. Under Pt 24 the court may give
summary judgment against a claimant on a claim or issue if it considers that
the claimant ‘has no real prospect of succeeding on the claim or issue’.
[21] The power to strike out should be
sparingly exercised in a case which involves developing areas of law. It is better for novel points of law to be
decided on real, rather than assumed, facts.
On the other hand, it is no part of the function of a first instance
judge to apply the law as it might one day become. His job is to apply the law as it is (or
arguably is).
[22] In order to defeat an application for
summary judgment, the claim must be shown to have a real prospect of success;
that is to say one that is more than merely arguable. But an application for summary judgment must
not be allowed to become a mini-trial.
The purpose of Pt 24 is to dispose of cases that are not fit for trial
at all.
THE SLIP RULE
The rule
[23] CPR 40.12(1) says: ‘The court may at any
time correct an accidental slip or omission in a judgment or order.’
The argument
[24] The defendants’ argument runs as follows:
(i) the practice direction (PD 25A, para 5.1) dealing with the grant of interim
injunctions says:
‘Any order for an
injunction, unless the court orders otherwise, must contain: (1) an undertaking
by the applicant to the court to pay any damages
62
which the respondent(s)
(or any other party served with or notified of the order) sustain which the
court considers the applicant should pay …’
(ii) The cross-undertakings given by GSK do not, on their
face, extend beyond the defendants to the action; (iii) this was because the
judge’s attention was not drawn to the contents of the practice direction on
either of the two occasions on which interim injunctions were granted, and the
practice direction had been overlooked by solicitors and counsel for both
sides; (iv) in the absence of an express order of the court (‘unless the court
orders otherwise’) the cross-undertakings should have extended to anyone
(whether or not technically a party to the action) who was served with or
notified of the order; (v) no such express order having been made, because of
the accidental omission to draw the judges’ attention to the practice
direction, the cross-undertaking should be amended under the slip rule.
The development of cross-undertakings
[25] It seems that what is conveniently called
the cross-undertaking in damages was invented by Knight-Bruce LJ when he was
Vice-Chancellor, and was originally inserted only in ex parte orders for injunctions. From there the practice developed of
requiring a plaintiff to give a cross-undertaking in damages as the price of
any interim injunction. There are two
related reasons why the court requires the giving of a cross-undertaking in
damages as the price of an interim injunction.
The first is that at the time when it grants the injunction the court
has not finally determined the rights of the parties. The modern practice is to decide whether
there is a serious issue to be tried. If
there is; and if damages would not be an adequate remedy, the court proceeds to
determine the balance of convenience. Of
course, when the facts are finally investigated, the claim may fail; with the
consequence that the defendant may have been temporarily stopped from doing
something that he was entitled to do all along.
That is why the grant of an interim injunction has sometimes been called
a ‘damage limitation exercise’. The
second, and related, reason is that if an injunction is discharged at trial,
the court has no power to award damages to the successful defendant for any
loss he may have suffered while it was in force.
The existence and extent of a cross-undertaking
[26] In some cases the court will be able to
say that a cross-undertaking has been implicitly given. In The Bank v A Ltd (23 June 2000,
unreported) the bank was suspicious about the propriety of a number of large
transfers into the account of A Ltd in its books. The bank was concerned that the transfers
represented the proceeds of fraud and that, if such activity were permitted to
continue, it might be held liable to the beneficiaries as a constructive
trustee of the funds. On the other hand,
the bank felt unable to raise the issue with A Ltd for fear that it might
thereby commit the offence of tipping off under the relevant legislation. Accordingly, the bank applied without notice
to A Ltd and in private for directions from the court as to what it should
do. It did not apply for an
injunction. The judge suggested that an
order freezing A Ltd’s accounts would be the best way forward and the bank took
up the offer. The freezing order, as
originally granted and subsequently varied, did not contain an express
cross-undertaking in damages. When the
case subsequently came back to court Laddie J said that a cross-undertaking was
demanded as of course. He approved
63
the following passage in the then current edition of
Mr Gee QC’s book Mareva Injunctions and Anton Piller Relief (4th edn,
1998) p 155:
‘Whenever the court
grants an Anton Piller order or an interlocutory injunction, unless the
contrary is expressly said, the plaintiff will be taken to have given the usual
undertaking in damages by implication.
In the Queen’s Bench Division it is usually the responsibility of the
plaintiff to provide a draft order to the court. That order ought to contain the usual
undertaking in damages. If, through
inadvertence, it is not included, the plaintiff could not resist an application
by the defendant to rectify the order either under RSC Ord 20 r 11 or under the
inherent jurisdiction of the court.’
[27] Laddie J then referred to Colledge v
Crossley (1975) Times, 17 March in which Lord Denning MR was reported as
saying—
‘that it was unfortunate
that the undertaking in damages was not put into the original order. It was an automatic undertaking which was
invariably inserted when an interim injunction was granted. If afterwards at the trial there was shown to
be no right to the injunction the plaintiff would have to pay damages as the
price of getting the interim injunction.
Undoubtedly the failure to include the undertaking came within the slip
rule. The undertaking should be
inserted.’
[28] I would infer from this quotation that
the case concerned a cross-undertaking that was in fact given, but which for
one reason or another was not recorded in the order. However, it is not possible to be sure from
the brief report in The Times; and it may well be a case in which no
cross-undertaking was expressly offered and accepted. However, I do not think that this matters.
[29] Laddie J concluded as follows:
‘When an ex parte
injunction or order is made other than against the Crown, it is automatically
subject to a cross-undertaking in damages.
So that the parties are made fully aware of their respective rights and
liabilities, that cross-undertaking should be set out in terms in the order,
but if it is not, it is to be implied none the less. It is difficult to imagine any circumstances
in which it would be appropriate for a court to refuse to impose a
cross-undertaking when it is making an order in the absence and without the
knowledge of the target of the injunction, not least because the imposition of
such a cross-undertaking does not mean that the court will inevitably enforce
it against the claimant if he fails at the trial. The cross-undertaking only requires the
claimant to compensate the defendant if, in its discretion, the court decides
that such compensation is appropriate.
In the unlikely event that a court, after argument from the claimant,
decides not to impose a cross-undertaking, that should be stated
expressly. The failure of the court to
expressly deal with this issue in the November and January orders was not
because of an unwillingness to impose a cross-undertaking but because the
necessity of such an undertaking was so obvious that it did not need to be
stated … Any party who seeks interlocutory relief must be taken to know and
appreciate that a cross-undertaking in damages will be the price it is
64
expected to pay and the
advocate appearing on his behalf will be taken to have authority to give the
undertaking. Even in a case like this,
where the November order was offered by the court rather than asked for by the
party, acceptance of the offer subjects the party to exactly the same
conditions and obligations as if he had asked for it in the first place. He has to pay the same price. The need to protect the interests of the
absent party are the same wherever the idea for the injunction came from. If the offer of more than was asked for
places the party’s advisers in difficulties they should either ask for a short
adjournment to seek instructions or decline the offer. If neither of these courses are adopted and
the offered injunction is accepted, the automatic obligation to submit to a
cross-undertaking must be explained to the client as soon as possible. If he is not prepared to pay that price, he
must return to court at the earliest opportunity—particularly before the order
is served or executed—to ask for it to be modified either by expressly
excluding the cross-undertaking or removing the injunction.’
[30] Although Laddie J was dealing with an
application for an injunction made without notice, I do not consider that his
observations were intended to be confined to such a case, since the rationale
for the requirement of a cross-undertaking applies equally to the grant of an
interim injunction on notice. However, I
do not read his observations as suggesting that the court can impose a
cross-undertaking on a party unwilling to give it; merely that it is implicit
in an application for an injunction (or, on the unusual facts of that case, the
taking of the benefit of an injunction) that a cross-undertaking will be
given. As Laddie J himself recognised in
the last of the quoted sentences, if the claimant is not willing to pay the
price he must ask for the cross-undertaking to be removed or for the injunction
to be discharged. To hold otherwise
would be contrary to authority which does not appear to have been cited to
Laddie J. I will develop this point
later.
[31] The principle that the offer of a
cross-undertaking is taken for granted has a long history. As long ago as 1896 North J said in A-G v
Albany Hotel Co [1896] 2 Ch 696 at 699–700:
‘Upon drawing up an
order for an interlocutory injunction the registrar invariably inserts such an
undertaking on the part of the plaintiff, even though, as frequently happens,
it has not been mentioned in court, but has been taken for granted. Of course such an undertaking must be
voluntary: the Court cannot compel a person to give an undertaking; and, if the
plaintiff declines to give it, either in court or before the registrar, the
order will not be made, or, if pronounced, will not be drawn up. If in the exercise of his discretion a judge
should think fit to dispense with such an undertaking he could of course do so,
and there are cases in which judges have done so; but this would only be under
special circumstances. In the absence of
any express direction of the judge to the contrary, the undertaking will always
be inserted in the order.’
[32] In consequence, the court would enforce
an implied cross-undertaking for damages even if it had not been included in
the order, unless the contrary had been agreed and expressed at the time: see
Oberrheinische Metallwerke GmbH v
65
Cocks [1906] WN 127 (see W v H (Family
Division: without notice orders) [2001] 1 All ER 300 at 319 where Munby J
sets out the history of the practice).
[33] These are cases in which no
cross-undertaking was expressly given or recorded in the order. What if a limited cross-undertaking is given
and accepted by the court?
[34] In Tucker v New Brunswick Trading Co
of London (1890) 44 Ch D 249, an action was brought against three
defendants, Matthews, Lamplough and the New Brunswick Trading Co of London, to
restrain the company from confirming in general meeting certain agreements
between the company and Matthews and Lamplough.
An interim injunction was granted against Matthews and the company. Matthews asked for the usual undertaking in
damages, to which the plaintiff’s counsel replied that it would of course be
given. Lamplough had not been served and
did not appear. When the order was drawn
up, the undertaking it contained was confined to damages sustained by the
company, and it was passed and entered in this form. Matthews and Lamplough appealed, asking that
the undertaking might be extended to damages sustained by them
respectively. The Court of Appeal held
that since Matthews had applied for an undertaking, which had in fact been given,
the order was wrong in not extending the undertaking to damages sustained by
him and could be corrected. Indeed that
correction could have been made by the first instance judge under the slip
rule; and did not require an appeal.
However, the same did not go for Lamplough, who had not applied for an
undertaking and not received one. Cotton
LJ said (at 252):
‘As regards Lamplough,
I am of opinion that his appeal fails; for we cannot impose on the Plaintiff
any undertaking which he has not given.
If a defendant applies for an undertaking, the plaintiff may decline to
take any order. The Court only makes the
undertaking a condition of granting an injunction; if the plaintiff refuses to
give it the Court can refuse the injunction, but it cannot compel the plaintiff
to give an undertaking.
As a general rule, I
think that when an injunction is granted the undertaking as to damages ought
not to be confined to the persons restrained.
In Pemberton on Decrees (4th edn 435), it is said: “The
undertaking applies to all the Defendants, although one or more only may be
restrained.” Mr. Pemberton does
not refer to any authority for this; but I consider it to be a correct
statement of the practice.’
[35] Lindley LJ said (at 253):
‘The cases of the two
Appellants are distinguishable. Matthews
asked for an undertaking and got it. An
undertaking is the price of an injunction, and if a man gets an injunction he
must pay the price. Lamplough did
not ask for an undertaking, and for anything we can tell, if he had done so the
Plaintiff would have declined to take an injunction. I think, therefore, that the undertaking can
only be extended to Matthews.’
[36] Lopes LJ agreed.
[37] It is noteworthy that although Cotton LJ
agreed with the statement in Mr Pemberton’s book that a cross-undertaking ought
to extend to all defendants,
66
whether or not restrained, the cross-undertaking
given in that case did not in fact extend to Lamplough. In the light of that fact, the Court of
Appeal decided that the cross-undertaking could not be extended to him. Thus the Court of Appeal did not elide the
two questions: what should the cross-undertaking contain and what did the
cross-undertaking contain. Even though
the cross-undertaking should have extended to Lamplough, it was not retrospectively
extended. It seems very unlikely that
the practice of the court had changed radically between 1890 (Tucker’s
case) and 1896 (the Albany Hotel case); and no one suggested that it
did. In the light of Tucker, it seems to
me that if a limited cross-undertaking is offered and accepted by the court,
there is in general no room for implying some further offer of an undertaking
beyond that which is expressly offered and accepted.
Imposition of a cross-undertaking
[38] The cross-undertaking has been described
as the ‘price’ of an injunction. If the
applicant is unwilling to pay the price, he does not get the injunction. But as Tucker’s case demonstrates the
court has no power to compel the giving of a cross-undertaking. Its only choice, if no cross-undertaking is
given, is to withhold the injunction.
[39] This is borne out by a number of
cases. In F Hoffmann-La Roche v
Secretary of State for Trade and Industry [1974] 2 All ER 1128 at 1134,
[1975] AC 295 at 341 Lord Reid said that a claimant ‘cannot be compelled to
give an undertaking but if he will not give it he will not get the
injunction’. In the same case Lord
Diplock said ([1974] 2 All ER 1128 at 1150, [1975] AC 295 at 361): ‘The court
has no power to compel an applicant for an interim injunction to furnish an
undertaking as to damages.’
[40] In Commodity Ocean Transport Corp v
Basford Unicorn Industries Ltd, The Mito [1987] 2 Lloyd’s Rep 197 at
199–200 Hirst J said:
‘When such security is
originally sought it is sought as a condition for the grant of the injunction,
in other words the plaintiff is told: “if you want this injunction you have got
to pay the price by fortifying the undertaking as to damages”. The plaintiff can then either agree or disqualify
himself from obtaining the injunction … [Counsel for the defendants] says that
the plaintiff has already paid a price here when the cross-undertaking was
given, which is perfectly correct so far as it goes; but the plaintiffs did not
ever agree nor were they ever asked to pay the extra price, that is the
fortification of the undertaking by security.
If they had been asked to do so it may very well be that they would …
“have declined to take an injunction”.
Of course, [counsel for the defendants] accepts, as he must, that the
Court has no power to impose an undertaking on the plaintiffs; and here I think
that if I were to make this order I would in essence, ex post facto, be
imposing an additional term to the undertaking, without any knowledge one way
or the other as to what the situation would have been if it had been sought by
the defendants in the first place. That
is something which I think it is wrong in principle to do.’
[41] Since a cross-undertaking cannot be
imposed, it follows that a fortiori it cannot be imposed retrospectively.
67
Nature of the cross-undertaking
[42] It is important to recall at the outset
that a cross-undertaking is not given to a party to the proceedings; it is
given to the court. As Lord Diplock
explained in the Hoffmann-La Roche case ([1974] 2 All ER 1128 at 1150,
[1975] AC 295 at 361):
‘The undertaking is not
given to the defendant but to the court itself.
Non-performance of it is contempt of court, not breach of contract, and
attracts the remedies available for contempts; but the court exacts the undertaking
for the defendant’s benefit. It retains
a discretion not to enforce the undertaking if it considers that the conduct of
the defendant in relation to the obtaining or continuing of the injunction or
the enforcement of the undertaking makes it inequitable to do so; but if the
undertaking is enforced the measure of the damages payable under it is not
discretionary. It is assessed on an
enquiry into damages at which principles to be applied are fixed and clear. The assessment is made on the same basis as
damages for breach of contract would be assessed if the undertaking had been a
contract between the plaintiff and the defendant, that the plaintiff would
not prevent the defendant from doing that which he was restrained from
doing by the terms of the injunction.’
[43] It follows from this that the proper
interpretation of a cross-undertaking is not a question of divining the mutual
understanding of the parties to the litigation, for the terms of the
cross-undertaking are a matter for the court.
Equally, because the cross-undertaking is given to the court, it may be
enforced by one who is not a party to the action, if the cross-undertaking is
given for his benefit. I will return to
the question of the protection of third parties. It also follows from the nature of the
cross-undertaking that a claim to recover on the cross-undertaking is not a
claim for damages at all (even though the cross-undertaking is conveniently
referred to as a cross-undertaking ‘in damages’). The modern form of cross-undertaking rightly
describes it as ‘compensation for loss’.
[44] There is another piece of judicial
shorthand which also requires some explanation.
Judges have said from time to time that a cross-undertaking is needed to
guard against injustice that might result if at trial it is decided that the
interim injunction was granted ‘wrongly’ or ‘in error’, or
‘improvidently’. These and similar
adverbs do not, in my judgment, imply that the court’s decision to grant an
interim injunction was wrong or unjust when it was made. If it was, then it could have been the
subject of a successful appeal. Rather,
what is meant is that with the benefit of hindsight and after investigation of
all the facts, the court at trial may decide that the claimant (in whose favour
the injunction was granted) is not entitled to the relief that he claimed. Had it known all the facts when it made the
order for the interim injunction it would have refused the injunction; but
equally it would not have exacted the cross-undertaking. The cross-undertaking is an integral part of
the whole package. The package as a
whole cannot fairly be described as erroneous or wrong. The points of claim describe the injunctions
as the ‘Wrongful Injunctions’ but that, in my judgment, is a misconception.
The benefit of the cross-undertaking
[45] Tucker’s case shows that a
cross-undertaking might not extend to one of several defendants to the
proceedings. However, this is not the
modern practice.
68
For many years, the conventional cross-undertaking
has been given to ‘the Defendants’. Even
in Tucker’s case itself, Cotton LJ said that the undertaking ought to
extend to all defendants, even though not all of them were restrained.
[46] In Berkeley Administration Inc v
McClelland [1996] IL Pr 772 Scott V-C referred to a note in the then
current Supreme Court Practice and said (at 792):
‘[67] These statements
of principle justify, in my judgment, the proposition that (subject to any
direction to the contrary a court may in a particular case give):
(i) advantage can be taken
of a cross-undertaking in damages by every defendant who was party to the
action when the undertaking was granted;
(ii) advantage cannot be
taken of the cross-undertaking by persons who are not parties to the action,
or, at least, do not become parties until after the order has been discharged.
[68] That leaves
outstanding the position regarding defendants who are joined as parties during
the currency of the order. That state of
affairs does not apply in the present case.
I do not regard the correct answer as being clear from the decided
cases, although I would, for my part, wish to extend the benefit of a
cross-undertaking in damages to all defendants who became parties while the
undertaking was in force.’
[47] However, he continued (at 793):
‘[71] … There is, in my
judgment, no legally acceptable basis on which the benefit of an undertaking,
to which a member of a group of companies is entitled, may be claimed on behalf
of the group as a whole.’
[48] Roch and Potter LJJ agreed.
[49] Scott V-C did not expressly consider the
question whether the benefit of a cross-undertaking would operate in favour of
an added defendant prospectively from the time at which he was added as a
defendant, or retrospectively from the time at which the cross-undertaking was
given to the court. Mr Watson accepted
that if a non-party applied to be joined as a defendant, the claimant could
elect to have the injunction discharged; in which case the cross-undertaking
would not apply to the would-be defendant.
But, he said, if the claimant chose to maintain his injunction, the
cross-undertaking would have retrospective effect. However, he could find no authority to
support that submission; and it seems to me to be contrary to the statements of
principle that I have already quoted. In
addition, in my judgment Mr Watson’s submission is inconsistent with the
repudiation of the ‘relation back’ theory of joinder of parties by the House of
Lords in Ketteman v Hansel Properties Ltd [1988] 1 All ER 38, [1987] AC
189. In my judgment, on the assumption
that the benefit of a cross-undertaking given in favour of the ‘defendants’
would extend to those who are added as defendants during the period when the
injunction is in force, it has only prospective, rather than retrospective
effect. The injunction itself could not
have been enforced directly against added defendants until they became
defendants; and their previous acts would not retrospectively become contempt
of court. If, as I think, the cross-undertaking
is aptly described as the price of the injunction, one would expect the two to
correlate temporally. In addition, in
the facts, of this case, the
69
application for joinder was made after judgment at
first instance and after GSK had indicated that it would not seek to maintain its
injunction pending appeal. It is,
therefore, on all fours with the situation in which Mr Watson accepted that the
cross-undertaking would not have retrospective effect. I hold, therefore, that if the Canadian
companies are joined as defendants to the infringement action, they will only
be entitled to the benefit of the cross-undertaking (at best) for two days.
Protection of third parties
[50] It is common ground that the origins of
the requiring of undertakings to be given to the court for the benefit of third
parties can be traced to the development of freezing orders. Within a few years after the jurisdiction to
make such orders (then known as Mareva injunctions) was discovered (or
invented) it had become standard practice for the court to require cross-undertakings
to be given by applicants in favour of banks and others who would be put to
expense by the execution of the order.
[51] The jurisdiction of the court to require
undertakings to be given for the benefit of third parties is not confined to freezing
orders. In Allied Irish Bank v
Ashford Hotels Ltd, Ashford Hotels Ltd v Higgins [1997] 3 All ER 309 the
question arose whether the court had power to require a cross-undertaking in
favour of third parties as a condition of appointing a receiver. Phillips LJ said (at 317):
‘The Mareva injunction
is a comparatively recent addition to the armoury of the court. Having discovered the existence of, or some
would say invented, this weapon, the court went on to invent the ancillary
weapon of the cross-undertaking in damages for the benefit of third parties
(see Z Ltd v A [1982] 1 All ER 556, [1982] QB 558). In that case the cross-undertaking approved
by the court was one designed to protect third parties from the consequences of
compliance with the injunction but the scope of the protection of the
undertaking has since been expanded to embrace third parties adversely affected
by the injunction.
For myself, I cannot
accept that the jurisdiction of the court to require such an undertaking only
exists where a Mareva injunction is ordered.
Once the cross-undertaking for the benefit of third parties became a
recognised feature of the court’s jurisdiction in that context, it necessarily
followed that the court could make use of it when granting other discretionary
relief, at least where that relief was empowered under the same statutory
provision.’
[52] However, the court decided on the facts
of the case that, as a matter of discretion, such a cross-undertaking should
not be required. It cannot be said, in
my judgment, that on the basis of this case it had become standard practice (at
least in 1997) for such cross-undertakings to be required in cases that were
not cases of freezing orders.
The introduction of the CPR
[53] At the heart of the defendants’ case is
that all this was radically changed by the introduction of the practice
direction supplementing CPR Pt 25. The
CPR came into force in April 1999, and the practice direction was issued at the
same time. The practice direction does
not explain why what had hitherto been the standard form of cross-undertaking
was to be expanded so as to include ‘other
70
parties’. Nor
is it clear from the terms of the practice direction whether ‘other parties’
means other parties to the litigation, or other persons who are not parties to
the litigation at all. It may be that
the Rules Committee would consider clarifying the scope of the required
cross-undertaking. However, I am content
to assume (without deciding) that the practice direction requires that a
non-party to the litigation is entitled to the benefit of the cross-undertaking
unless the judge otherwise orders.
[54] In Miller Brewing Co v Mersey Docks
& Harbour Co, Miller Brewing Co v Rhui Enterprises Ltd [2003] EWHC 1606
(Ch), [2004] IP & T 542 an injunction had been granted in 2002 preventing
the movement of beer which was aboard a ship.
A company which was liable to pay demurrage while the injunction was in
force applied for compensation under the cross-undertaking. The cross-undertaking had not been given for
the benefit of third parties; but the company argued that, by analogy with the
standard practice in freezing order cases, it should have been. Neuberger J dealt with the point as follows:
‘[44] … None the less,
while this court should not be over-indulgent to a person seeking an
interlocutory injunction, it seems to me that it would be a strong thing to
require him to sign not merely a blank cheque in favour of the defendant, if it
turned out that he should not have been granted the injunction, but a series of
blank cheques in favour of third parties of whose very existence and interest
he may be unaware and for whose losses he may find himself liable even though
he is entitled to his injunction.
[45] As Mr Howe says, if
a third party is detrimentally and unfairly affected by an interlocutory
injunction, his interest can either be put before the court by the defendant,
as happened in [Galaxia Maritime SA v Mineralimportexport, The Eleftherios [1982]
1 All ER 796, [1982] 1 WLR 539], or he can himself make an application to the
court to vary or discharge the interlocutory injunction or to extend to him the
benefit of the cross undertaking.
[46] If it could be said
that the court was overburdened with such applications by third parties or that
there were many cases of injustice to [a] third party because of the absence of
such a wide cross undertaking, then there might be more in this point. However, I have not had my attention drawn to
any case which has led to a procedural problem, nor any reports, whether in the
law reports or legal commentaries, of a perceived injustice in this connection.
[47] Accordingly, I am
not persuaded that it would be appropriate to impose the sort of wide cross
undertaking indemnities on a claimant seeking an interlocutory injunction. That is not to say that the court should
never consider imposing a wider cross undertaking in damages than that usually
extracted from an applicant for an interlocutory injunction. It is fair to say that the argument in this
case has ensured that I will henceforth think a little more carefully about the
terms of any cross undertaking in damages to be extracted from a claimant in
return for the grant of an interlocutory injunction.
[48] Quite apart from
this, it seems to me that Miss Heal’s contention suffers from a further
problem, namely whether the effect of the wider cross undertaking which she
seeks can properly be imposed. The fact
that it may have been open to Bahr Behrend to apply for an extension of the
cross
71
undertaking, or even if
the court could and should have imposed a wider cross undertaking on 11 or 25
January 2002, it does not mean that the court can impose such a cross
undertaking retrospectively.’
[55] It does not seem, from the terms of
Neuberger J’s judgment, that his attention was drawn to the terms of the
practice direction. Mr Watson QC argues
that this is a fatal flaw in the judge’s decision; and that I should not follow
it. He buttresses his submission by
referring to Mr Gee QC’s book on Commercial Injunctions (5th edn, 2004)
pp 298–299 (para 11.015) where the learned author says that the Miller
case was wrongly decided. He also
referred me to the decision of Sir Andrew Morritt V-C in Imutran Ltd v
Uncaged Campaigns Ltd [2001] 2 All ER 385 in which Sir Andrew Morritt V-C
said that the cross-undertaking in the form contained in para 5.1 of the
practice direction should be incorporated into the standard form of interim
injunction. This has now been done in
the latest revision to the form in March 2005.
However, I think, with respect, that Sir Andrew Morritt V-C was in error
in saying that the continued use of previous forms was merely the ‘suggestion’
of the editors of Civil Procedure (White Book). In fact it seems that use of the previous
forms was prescribed by CPR Pt 4 (at least as regards the ‘N’ forms used in the
county court, which do not contain the wider form of cross-undertaking). It is also the case that although Sir Andrew
Morritt V-C said that the standard form should contain the wider form of
cross-undertaking, he did not say that the use of that form was already
standard practice; and he expressly rejected the suggestion that the order of
Hart J (which did not contain it) was ‘defective’.
[56] It seems to me therefore, that at least
as at 2002: (i) the court had jurisdiction even in non-freezing order cases to
require cross-undertakings to be given for the benefit of third parties; (ii)
such cross-undertakings, although routinely required in freezing order cases,
were not in fact routinely required in other cases, even though the practice
direction appeared to require it unless the judge otherwise ordered; (iii) if
no such cross-undertaking was given it should not be imposed, let alone imposed
retrospectively; (iv) a third party adversely affected by an injunction could
apply to the court for the discharge of the injunction, unless the
cross-undertaking were extended for his benefit.
[57] It follows from my conclusions thus far
that, at least as at 2002, it could not be said that it was implicit in any
application for an interim injunction (other than a freezing or search order)
that the applicant must be taken to have offered a cross-undertaking in favour
of third parties who were not defendants to the proceedings. I might also add that in the present case Pumfrey
J himself described the cross-undertaking that was in fact given as the
‘standard’ form. The Chancery Guide (see
Civil Court Practice (2005) vol 1, p 2115 (para 5.22) (Green Book))
describes the implied form of cross-undertaking as an ‘undertaking in damages
by the party applying for the injunction in favour of the other’ (my
emphasis). Snell’s Equity (31st
edn, 2005) p 413 (para 16-25) also describes the undertaking as one ‘to pay
damages for any loss to the defendant if at the trial it appears that
the injunction was wrongly granted’ (my emphasis). It seems to me, therefore, that the position
in 2002 remained substantially unchanged during 2003. I rather suspect that the existence of the
wider form of cross-undertaking embodied in the practice direction has been
overlooked by the profession. It may be
that, following the March 2005 revision to the standard form of order the
practice will change; but I
72
do not consider that it can realistically be said to
have changed when the interim injunctions in the present case were
granted. That, no doubt, is why the
wider form of cross-undertaking in the practice direction never occurred to
either party.
[58] Moreover, this is not a case in which no
undertaking was offered. It is a case
where a limited form of undertaking (in what indisputably used to be the
conventional form) was offered and accepted both by the defendants and the
court. In such a case Mr Waugh submits,
that an express offer must displace any implied offer: expressum facit cessare
tacitum. I agree.
[59] In addition, it seems to me that the more
limited form of undertaking was what the judge in fact ordered on both
occasions; and consequently the judge did ‘otherwise order’, so as to displace
the presumption in the practice direction.
The scope of the slip rule
[60] In his book on Commercial Injunctions (5th
edn, 2004) p 298 (para 11.015) Mr Gee QC is critical of Tucker’s
case. He says:
‘That case [ie Tucker’s
case] depended upon the practice then followed in providing undertakings. The scope of an undertaking depends upon what
would reasonably have been understood by the court which heard the
application. The then prevailing
practice provided the context for understanding the words used by counsel. If the same case happened today, the
assumption would be that, unless counsel clearly excluded the undertaking set
out in para. 5.1 of the practice direction, this was impliedly being offered by
the claimant … The order could therefore be amended under the slip rule.’
[61] In view of Cotton LJ’s statement of ‘the
correct practice’ as described by Mr Pemberton in his book, I am not convinced
that this explanation of Tucker’s case is correct. I am also not convinced by Mr Gee’s assertion
that a court hearing an application today (or at least in 2002 or 2003) would
inevitably understand a cross-undertaking as being offered in the form of para
5.1 of the practice direction.
[62] Mr Watson QC fastened on the judgment of
Lindley LJ in Tucker’s case and his statement that the plaintiff might
have declined to take the injunction.
Here, he said, it was obvious that GSK would have given the extended
cross-undertaking, rather than decline the injunction. However, it does not seem to me that the
court should embark on a hypothetical inquiry of this kind. The point is that the hypothesis did not
arise. In the present case, Mr Waugh QC
also says with force that although it is possible that GSK might have agreed to
extend the cross-undertaking to the Canadian companies, they would not have
agreed to the cross-undertaking in the form of the practice direction; at least
not without very careful consideration of who might be covered by it.
[63] The slip rule allows the court to correct
an ‘accidental’ error or omission. Was
the form of the cross-undertaking an accidental error? At first blush the answer must be ‘No’. It was a cross-undertaking deliberately given
in the form in which it was intended to be given. It was embodied in an order settled by junior
counsel for each party; and approved by the judge.
[64] Mr Watson relies on the decision of
Morton J in Re Inchcape, Craigmyle v Inchcape [1942] 2 All ER 157,
[1942] Ch 394. In brief, what had
happened was that
73
Morton J had decided the question of the domicile of
Lord Inchcape at the date of his death.
At the conclusion of the hearing counsel had asked for a direction that
the costs of all parties should be paid out of the estate. However, costs had been incurred before the
issue of proceedings; and these were not covered by the order. An application to amend the judge’s order was
made under the slip rule; and the slip was said to be counsel’s accidental
omission to ask for the costs in question to be paid out of the estate. Morton J referred to Fritz v Hobson (1880)
14 Ch D 542, [1874–80] All ER Rep 75, a decision of Fry J, and continued
([1942] 2 All ER 157 at 159–160, [1942] Ch 394 at 397–398):
‘In the course of the
argument before Fry J, counsel submitted, at p 560:
“Ord. 41A does not apply. This is not an accidental slip or
omission. Those words mean only an
accidental slip or omission to embody in the order something which the court in
fact ordered to be done; they do not apply to an accidental omission of counsel
or solicitor to ask for, or of the court to provide for, something which ought
to have been provided for.”
That argument precisely
expresses the doubt which I felt when the case first came before me and before
the authority had been cited; because the error, which it is now sought to set
right, if it can be properly described as an error, did not arise from an
omission to embody in the order something which I, in fact, ordered to be done,
but arises from an accidental omission of counsel to ask that a particular
thing might be done. However, Fry J,
after dealing with the other grounds upon which he had power to correct the
order, said this, at p 561:
“There is another ground on which, in my opinion, I
have jurisdiction to make the order asked for, namely, under Ord. 41A. In my view the error in the present case has
arisen from the accidental omission of counsel to call my attention to the
adjourned motion when I pronounced my judgment, an omission very natural at a
time when counsel’s attention was directed to matters of greater
importance. In substance the motion was
before me at the trial, for my attention was called to the affidavits made upon
it. I recollect quite enough of what
took place, and I am confirmed in my recollection by the notes which I made at
the time, to know that all the affidavits upon the motion were in substance
before me at the trial, and that the various witnesses who made those
affidavits were examined in the course of the proceedings. On this ground, therefore, I think I have
jurisdiction to make the order.”
It might be said that
there is a distinction between Fritz v Hobson and the present case, in
that in Fritz v Hobson Fry J was dealing with the costs of a motion
which, as he says, was in substance before him, whereas I am asked now to deal
with costs incurred before the issue of the summons; but I think that in
substance the reasoning in Fritz v Hobson applies to the present
case. I think that Fry J had sufficient
recollection of the whole matter in Fritz v Hobson to feel sure that he
would have made the order if he had been asked to do so. So in the present case, I have a sufficiently
clear recollection of the evidence which was produced before me as a result of
those researches to feel sure that I would have made the order, if I had been
asked to do so.’
74
[65] There are, however, two reasons why, in
my judgment, the principle thus stated does not apply to the present case. First, while forgetting to ask for an order
which one intended to ask for can (just) be described as an accidental error, I
do not consider that the same can be said of a case where it did not occur to
anyone that a particular form of relief might be available. Second, if the practice direction had been
drawn to the attention of the judges who granted the interim injunctions, I do
not feel able to say with confidence what cross-undertaking they would have
required; or GSK would have been willing to give. As I have said, I do not consider that at the
time these injunctions were granted, it would have been implicit in any
application for an interim injunction that the form of undertaking in the
practice direction was being offered; and even if it was, any implied offer
must, in my judgment, have been displaced by the express offer that was
accepted both by the defendants and by the court. If I were now to amend the form of the
cross-undertaking, I would be retrospectively imposing on GSK an undertaking
that they did not give; and could not have been made to give.
[66] I therefore dismiss the application under
the slip rule. It follows that those
parts of the points of claim that rely on the slip rule must, likewise, be
struck out.
THE JOINDER APPLICATION
[67] Mr Watson said that in the event that I
concluded that a cross-undertaking in favour of an added defendant would not
have retrospective effect, he would not press the application to join the
Canadian companies. I have so
concluded. Accordingly, I dismiss the
joinder application.
ESTOPPEL BY CONVENTION
[68] The species of estoppel on which Apotex
relies is estoppel by convention. I can
take the principle (which is sufficient for present purposes) from the speech
of Lord Steyn in The Indian Endurance (No 2), Republic of
India v India Steamship Co Ltd [1997] 4 All ER 380 at 391, [1998] AC 878 at
913:
‘It is settled that an
estoppel by convention may arise where parties to a transaction act on an
assumed state of facts or law, the assumption being either shared by them both
or made by one and acquiesced in by the other.
The effect of an estoppel by convention is to preclude a party from
denying the assumed facts or law if it would be unjust to allow him to go back
on the assumption (see K Lokumal & Sons (London) Ltd v
Lotte Shipping Co Pte Ltd, The August Leonhardt [1985] 2 Lloyd’s Rep 28; Norwegian
American Cruises A/S (formerly Norwegian American Airlines A/S) v
Paul Mundy Ltd, The Vistafjord [1988] 2 Lloyd’s Rep 343 and Treitel
Outline of the Law of Contract (9th edn, 1995) pp 112–113). It is not
enough that each of the two parties acts on an assumption not communicated to
the other. But it was rightly accepted by counsel for both parties that a
concluded agreement is not a requirement for an estoppel by convention.’
[69] Apotex relies on a number of documents
(including correspondence and orders of the court) to support one of two
alternative estoppels: (i) that there was a common communicated assumption
that, despite the fact that the Canadian companies were not defendants to the
infringement action, they were
75
nevertheless entitled to the benefit of the
cross-undertaking; or (ii) that there was a common communicated assumption that
the Canadian companies were to be treated as if they were in fact defendants to
the infringement action with the consequence that they fell within the
description ‘Defendants’ in the cross-undertaking.
[70] Mr Waugh submitted that, on examination,
none of the documents relied on supported the existence of the alleged
assumptions; still less that the assumption had been communicated. All were explicable on other grounds. He relied strongly on Mr Cohen’s statement
that the Canadian companies had given ‘voluntary’ disclosure. This, he said, demonstrated that the Canadian
companies did not regard themselves as bound by any order of the court to give
disclosure, otherwise the disclosure could not have been described as
‘voluntary’.
[71] I have considerable sympathy with this
submission; and Mr Waugh’s submissions on the detailed contents of the
documents on which Apotex relied were powerful.
[72] In the end, however, I have come to the
conclusion that the legal basis of the plea is one that is sustainable. If I were to go down the route of examining
in detail the particular documents on which Apotex relies, I would be
conducting the mini-trial which current practice under CPR Pt 24 forbids. Nor, as it seems to me should I be tempted by
Mr Waugh’s submission that since the common assumption must be communicated,
disclosure and oral evidence can produce nothing more of any value. I say this, not because he is necessarily
wrong; but because to travel down that road would be to go behind Pumfrey J’s
refusal to order the trial of a preliminary issue. That decision has not been appealed, and
binds the parties.
[73] Accordingly, I am not prepared to strike
out those parts of the points of claim which rely on the alleged estoppel; or
to give summary judgment in favour of GSK on that issue. That being so, I think that it would be
inappropriate for me to express any views on the details of the submissions,
since to do so might embarrass the trial judge.
RESTITUTION
[74] It is common ground that, in the absence
of a cross-undertaking, a person against whom an interim injunction is granted
has no claim to recover damages for loss suffered during the currency of the
injunction if that injunction is discharged at trial. But Mr Watson submits that this does not
preclude the Canadian companies from advancing a restitutionary claim to
recover a benefit enjoyed by GSK which it would not have enjoyed if the
injunction had not been granted. He
accepts that such a claim is novel; and relies on no authority in support of
it; other than general statements to the effect that the categories of
restitution are not closed.
[75] Mr Waugh submitted that this claim was no
more than an attempt to outflank the well-established principle that damages
are not recoverable for loss suffered as a result of the grant of an interim
injunction. That principle is the very
foundation for the practice of requiring a cross-undertaking in damages. If Mr Watson’s submission is correct, there
is not (and never has been) any need for such a cross-undertaking. In addition, Mr Waugh submitted that the
so-called claim for restitution is simply a claim for damages under another
name.
76
[76] I accept Mr Waugh’s submission. Suppose that during the currency of the
injunction GSK had made profits of, say, £2m in selling Seroxat to a multiple
pharmacist. Part of the profit would
have been due to the effect of the injunction in enabling GSK to keep up the
price of its branded product. Suppose
that, absent the injunction, Apotex would have sold paroxetine to the same
pharmacist, but at reduced prices such that it would have earned profits of
£1m. Apotex does not claim £2m (which is
GSK’s benefit). It only claims £1m which
is its own claimed lost profit. In other
words, the claim is not measured by GSK’s gain (which is the hallmark of a
restitutionary claim); but by Apotex’s own loss (which is the hallmark of a
claim for damages). Apotex’s own loss,
moreover, is not the loss of some identifiable piece of property, but simply
the loss of an opportunity to compete in the market. This is quite different to the power of an
appeal court, on reversing a judgment of a lower court, to order the repayment
of money paid under a judgment that has been set aside. The payment of such moneys, or the return of
property transferred under an order that is later set aside, could be enforced
through a writ of restitution; but that is not the same as a restitutionary
claim, as that concept has developed in the modern law.
[77] It is also inherent in Mr Watson’s submission
that the court’s order was itself unjust.
For the reasons I have explained, I do not accept this premise. It is well established that a person who acts
on an order of the court (even if it is later set aside) commits no wrong while
the order is in force: see Hillgate House Ltd v Expert Clothing Service
& Sales Ltd [1987] 1 EGLR 65.
The purpose of requiring the cross-undertaking in damages was to prevent
injustice. In addition the injunction as
granted did not bind the Canadian companies, so the direct causal link is
missing.
[78] I was referred to the very learned
judgments of the Supreme Court of Victoria in National Australia Bank Ltd v
Bond Brewing Holdings Ltd [1991] 1 VR 386.
The case concerned the appointment of a receiver. The court had made an order for the
appointment of a receiver without requiring a cross-undertaking in
damages. The order was subsequently set
aside, and the question was whether compensation could be awarded in the
absence of a cross-undertaking. The
court plainly considered that if it had any power to award compensation it
should do so (see in particular [1991] 1 VR 386 at 582 per Murphy J). Nevertheless, after an exhaustive review of
authorities from three continents, the court unanimously concluded that the
court had no such power; and that a person against whom an injunction is
granted but later discharged is ‘without remedy’ in the absence of a
cross-undertaking. Had there been a
remedy available through the means of a restitutionary claim, there is no doubt
in my mind that the court would have found it.
[79] Mr Watson urged on me that I should not
strike out a claim based on a developing field of jurisprudence. However, I must apply the law as it is; and
not as it one day might become. In my
judgment this claim is legally unsustainable; and should therefore be struck
out.
THIRD PARTY LOSS
[80] The claim under this head depends, among
other things, on there being no direct route of recovery for the Canadian
companies. Mr Watson argues that in this
situation Apotex Europe Ltd is entitled to recover, under the
cross-undertaking, losses suffered by its affiliated Canadian companies. He runs
77
into the difficulties that: (i) in Berkeley
Administration Inc v McClelland [1996] IL Pr 772 Scott V-C said that there
is no legally acceptable basis on which the benefit of an undertaking, to which
a member of a group of companies is entitled, may be claimed on behalf of the
group as a whole; and (ii) that an argument to precisely this effect was
rejected by Jacob J in R v Medicines Control Agency, ex p Smith & Nephew
Pharmaceuticals Ltd (Primecrown Ltd intervening) [1999] RPC 705.
[81] Nevertheless, he says that the claim
falls within the principle established by the House of Lords in Linden
Gardens Trust Ltd v Lenesta Sludge Disposals Ltd [1993] 3 All ER 417,
[1994] 1 AC 85 and Panatown Ltd v Alfred McAlpine Construction Ltd [2000]
4 All ER 97, [2001] 1 AC 518. Both these
cases concerned the fact-situation where A entered into a building contract
with B. B failed to do what he promised
to do, with the result that there was a breach of contract. But, by the time of the breach, A no longer
owned the land on which the building stood; because he had transferred it to C. So although there was a breach of the
contract between A and B, the resulting loss was suffered by C. The question for the House was whether, in
those circumstances, A was entitled to sue B for damages for breach of contract
whose quantum was to be measured by C’s loss.
The House held that, under some circumstances, A could maintain this
claim.
[82] However, the claim that Apotex wishes to
advance on behalf of the Canadian companies comes nowhere near this
principle. The principle is predicated
on a breach of contract. No cause of
action arises at all unless B has broken his promise. If B has not broken his promise, discussion
of what the cause of action might encompass if he had done is pointless. So the first question must always be: what
has B promised to do?
[83] In the present case, GSK promised to abide
by any order of the court requiring them to compensate the defendants (or the
‘Apotex parties’ as defined) for any loss that they had suffered. They did not promise to pay the defendants
for loss that other people had suffered.
This is the reason why Jacob J rejected a similar argument in the
Primecrown case; and in my respectful opinion, he was right to do so. As Jacob J put it (at 715):
‘Whether the recoverable
damage is that which is foreseeable by the plaintiff or that which is directly
caused by the injunction is not in point.
None of the differing views expressed in the cases go so far as to say
that the injunctee can claim for damage not suffered by him. Nor do the very words of the undertaking
(which is the foundation of the jurisdiction) suggest that he can recover more
than that which he has suffered, whether that damage is foreseeable by the
injunctor or not. Thus while I have
sympathy with Mr Howe’s [counsel for Primecrown] “flexible approach” I do not
think it can go so far as to require the “wrongful injunctor” to pay for damage
not suffered by the injunctee at all.
I think this
consideration also disposes of Mr Howe’s Linden Gardens point. In that case the House of Lords held that
damages for breach of a contract between a developer and a builder should
include the damage suffered by the purchaser from the developer. The parties could be treated as having
entered into the contract on the basis that the developer would be entitled to
enforce its contractual rights on behalf of [the] purchaser who suffered the
actual damage. The case depended on the
parties having full knowledge that the developer was going to pass the property
on to the purchaser, so the
78
builder knew exactly who
would be suffering if his work was inadequate.
Mr Howe suggested that in this case there is a parallel in that S&N
expected to have to pay for trading losses.
So they did, but they did not undertake to pay for trading losses, they
only undertook to pay for Primecrown’s losses.
The analogy with Linden Gardens breaks down.’
[84] Mr Watson did not argue that the
cross-undertaking could be construed as though the reference to the defendants
included a reference to the Canadian companies; and in my judgment he was right
not to do so. Again Mr Watson urged on
me that I should not strike out a claim based on a developing field of
jurisprudence. But in my judgment no
amount of development in the law relating to the scope of damages that can be
awarded for breach of contract can result in a party to a contract recovering
damages for something which is not a breach of contract at all. I conclude therefore that this argument is
also legally unsustainable and should be struck out.
THE PROFIT SHARE
[85] This part of the dispute relates to
Neolab. As I have said, the arrangement
under which Neolab was to supply paroxetine involved its purchase from Apotex
Inc. The purchase price was in part a
fixed price; and in part a percentage of profits. The cross-undertaking requires GSK to
compensate Neolab for the loss that it has sustained by reason of the making of
the order for the injunction. Is the
loss its loss of profit or is it some other loss?
[86] Mr Waugh’s argument is a simple one. Neolab is entitled to be compensated for its
loss of profit. Its loss of profit will
be measured by its assumed turnover (on the footing that the injunction had not
been granted) minus its cost of sales.
One of the costs of sale will be the price for the goods that it sold
that it had to pay Apotex Inc. It does
not matter whether that price is a fixed price or a price based on a percentage
of profits. What matters is that
Neolab’s loss is what it would have kept in its own corporate pocket if the
injunction had not been granted. Any
money that it would have had to have passed on to others (whether they were
suppliers of product or suppliers of services necessary to get the product to
market) is not loss that it has suffered itself.
[87] Mr Watson’s response was that that part
of the agreement that required Neolab to share profit with Apotex Inc was in
the nature of a partnership. Had the
cross-undertaking been expressed in favour of a partnership, it would have been
irrelevant what were the profit shares of each partner. The purpose of the injunction was to keep
Apotex’s product off the market; and it was entirely foreseeable that if the
injunction were discharged, it would cause loss to Apotex Inc. The loss that Neolab suffered was in effect
its loss of turnover; and that is what would be recoverable on the
inquiry. Moreover, even if I were
attracted by Mr Waugh’s submission, the case was not so clear as to justify
striking out or summary judgment.
[88] In my judgment one must be careful not to
confuse questions of foreseeability with the scope of the promise. GSK’s promise was to compensate Neolab for
its own loss; not to pay compensation for someone else’s loss. So foreseeability can, in my judgment, be put
to one side. Is the cost price that
Neolab had to pay its supplier part of its own loss? In my judgment No. I cannot
79
see that it matters whether the price is expressed as
a fixed price or a price calculated by means of a formula, even if that formula
is a percentage of profit. Neolab is not
part of the Apotex group; and the arrangement between it and Apotex Inc is one
at arms’ length. The case is not one of
partnership in the true sense; and I do not consider that the principles of
partnership can be applied by analogy.
[89] In my judgment there is no real prospect
of Neolab succeeding on this issue; and GSK is entitled to summary judgment.
RESULT
[90] Accordingly: (i) I dismiss the
application under the slip rule, and strike out those parts of the points of
claim that depend on it; (ii) I dismiss the joinder application; (iii) I strike
out (or give judgment in favour of GSK on) those parts of the points of claim
that depend on the argument relating to third party loss; (iv) I strike out (or
give judgment in favour of GSK on) those parts of the points of claim founded
on restitution; (v) I strike out (or give judgment in favour of GSK on) those
parts of the points of claim that seek to claim the amounts that Neolab would
have had to pass on to Apotex Inc; (vi) I decline to strike out (or give
judgment in favour of GSK on) those parts of the points of claim that rely on
estoppel.
[91] I will hear counsel on the form of order
I should make to give effect to this ruling; and on what (if any) directions I
should give for the future conduct of the inquiry.
Order accordingly.
Victoria
Ellis Barrister.
80
[2006] 2 All ER 81
Pirelli Cable Holding NV and others v Inland Revenue
Commissioners
[2006]
UKHL 4
EUROPEAN COMMUNITY; Taxation: QUANTUM: TAXATION;
Corporation Tax
HOUSE OF LORDS
LORD
NICHOLLS OF BIRKENHEAD, LORD HOPE OF CRAIGHEAD, LORD SCOTT OF FOSCOTE, LORD
WALKER OF GESTINGTHORPE AND LORD BROWN OF EATON-UNDER-HEYWOOD
23,
24 NOVEMBER 2005, 8 FEBRUARY 2006
Damages – Measure of damages – Income tax –
Corporation tax – Advance corporation tax – European Court of Justice holding
denial of exemption from paying advance corporation tax to United Kingdom
companies with parent companies established in other member states constituting
discriminatory treatment bringing right of compensation or restitution –
Whether receipt by parent companies of payments from United Kingdom Revenue
pursuant to double taxation agreements extinguishing or reducing compensation
or restitution – Whether advance corporation tax a withholding tax – Income and
Corporation Taxes Act 1988, ss 231, 247, 788(3) – Double Taxation Relief (Taxes
on Income) (Netherlands) Order 1980, SI 1980/1961, art 10(3)(c) – Double
Taxation Relief (Taxes on Income) (Italy) Order 1990, SI 1990/2590, art
10(3)(c) – Council Directive (EC) 90/435.
The P parent companies were two companies resident in
the Netherlands and a company resident in Italy. Together they owned all of the shares in P
plc, a company resident in the United Kingdom which held all the shares in P
(General) plc which was also resident in the United Kingdom (the P subsidiary
companies). Between May 1995 and March
1999 P plc paid dividends to the parent companies. It was required to pay advance corporation
tax (ACT) (which was deducted from the dividends paid) which it was able to set
off against its subsequent liability to mainstream corporation tax. In certain circumstances s 231(1)a of the Income and Corporation Taxes Act 1988
provided that the recipient of the dividend would be entitled to a tax credit
in an amount corresponding to the amount of ACT. Section 231 did not apply where the dividend
had been paid pursuant to a group income election under s 247b of the 1988 Act which permitted the payment of
dividends between United Kingdom group companies without the requirement to pay
ACT. Section 231 also applied only where
the paying and recipient companies were resident in the United Kingdom. The United Kingdom had double taxation
agreements with the Netherlands (the Double Taxation Relief (Taxes on Income)
(Netherlands) Order 1980) and Italy (the Double Taxation Relief (Taxes on
Income) (Italy) Order 1990) which were effective under United Kingdom domestic
law pursuant to s 788c of the 1988 Act. Section 788(3)(d) enabled double taxation
agreements to contain provision for conferring on non-United Kingdom residents
the right to a tax credit under s 231.
Article 10(3)(c)d of the double
taxation agreements conferred on a non-United Kingdom resident company
receiving
________________________________________
aSection 231, so far as
material, is set out at [66], below
bSection 247, so far as
material, is set out at [66], below
cSection 788, so far as
material, is set out at [51], below
dArticle 10, so far as
material, is set out at [52], below
________________________________________
81
dividends from a United Kingdom subsidiary an
entitlement to a tax credit equal to one-half of the tax credit to which a
United Kingdom resident individual would have been entitled to receive had he
received ‘those dividends’. Accordingly,
the P parent companies claimed and received tax credits from the Revenue. In 2001 the Court of Justice of the European
Communities held that the denial of the right of election under s 247(1) to
holding companies established in member states other than the United Kingdom
and their United Kingdom subsidiaries was contrary to the right of freedom of
establishment under art 52 of the EC Treaty (now art 43 EC). The court further held that companies which
had been disadvantaged as a result of having to pay ACT were entitled under
Community law to compensation or restitution and that an effective legal remedy
to obtain reimbursement or repatriation of the financial loss suffered should
be afforded to resident companies and their non-resident parent companies. A number of actions were launched in the
United Kingdom courts by non-United Kingdom resident companies and their United
Kingdom subsidiaries, including the P companies, claiming relief by way of
damages or restitution for ACT which they had paid in the past and which they
contended they would not have paid if they had been able to join with their
parent companies in making s 247 elections.
The various actions were brought together within a group litigation
order. The Revenue submitted that, in
calculating any compensation or restitution that might be payable it was
necessary to have regard to any payments made to the P parent companies
pursuant to the double taxation agreements since, if United Kingdom law had
permitted the dividends paid by the subsidiary companies to be paid pursuant to
a group income election and therefore without the payment of ACT, the P parent
companies would not have been entitled to the tax credits. In the High Court the judge rejected the
Revenue’s submissions holding that even if the P companies had jointly made a
group income election with the result that no ACT would have been payable in
respect of dividends paid to the P parent companies, the latter would have
nevertheless been entitled to tax credits pursuant to art 10(3)(c) of the
double taxation agreements, and that in any event compensation payable to the P
subsidiary companies should not be reduced by reference to the value of the tax
credits received by the P parent companies since their separate economic
identity could not be ignored. The judge
therefore found it unnecessary to determine the further question whether ACT
constituted a withholding tax within the meaning of Council Directive (EC)
90/435 (the Parent/Subsidiary Directive) and so unlawful in its application to
United Kingdom subsidiaries of parent companies of other member states. The Court of Appeal upheld the judge’s
decision. The Revenue appealed to the
House of Lords.
Held – (1) On the true construction of art
10(3)(c) of the double taxation agreements the P parent companies would not
have been entitled to tax credits in respect of dividends paid to them by the
United Kingdom holding company as group income under a s 247(1) group income
election. In order to apply the
hypothesis prescribed by art 10(3)(c), which was that entitlement of to a s 231
tax credit of a P parent company that had received dividends that had been paid
under a group income election depended upon the entitlement to a s 231 tax
credit of a United Kingdom-resident individual had he received ‘those
dividends’, either the individual had to be taken to have received dividends different
from those received by the P parent company, that is the individual had to be
supposed to have received dividends not paid under a group income election and
in respect of which ACT had been paid or the individual had to be attributed
with the ability
82
to receive dividends paid under a s 247 group income
election. The second solution did less
violence than the first to the language of art 10(3)(c). The dividends assumed to have been received
by the P parent companies and in respect of which it was to be assumed that
they had claimed tax credits were dividends assumed to have been paid under a
group income election jointly made by the subsidiaries and the parents. The reference in art 10(3)(c) to ‘those dividends’
that the hypothetical individual was to be taken to have received should be
dividends of the same assumed character and therefore the individual had to be
attributed with the ability to receive dividends of that character. On that basis, the individual who had
received dividends of that character would not have been entitled to a s 231
tax credit and, consequently, nor would the P parent companies in the instant
case. The alternative construction of
art 10(3)(c) involved overlooking the
statutory definition of ‘tax credit’ as a ‘tax credit under section 231’ and
attributing to the individual, and thence to the companies, the right to a tax
credit that an application of s 231 would reject (see [14]–[16], [30], [31],
[38], [39], [67]–[72], [103], [105], [107], [108], below).
(2) The art 52 infringements of which the P companies
complained were infringements of the P parent companies’ rights to freedom of
establishment in the United Kingdom. The
primary measure of the loss thus caused was the financial detriment to the
United Kingdom subsidiary caused by its liability to pay ACT and its inability
to make group income elections but the benefit of the tax credits that the P
parent companies would not have received but for the art 52 infringements had
to be brought into account. Logically, the
compensation should be compensation to the group for the detriment suffered by
the group. Compensation that
concentrated on the financial detriment of the subsidiary and ignored the
financial benefit of the parent overlooked the nature of the art 52 infringement. The identity of the recipient or recipients
of the compensation payable for the loss suffered by the group could be left to
be decided by the group, but could not affect the quantum of the compensation (see
[19]–[22], [25], [26], [30], [41], [43], [44], [77]–[82], [107], [108], below);
Metallgesellschaft Ltd v IRC, Hoechst AG v IRC Joined cases C-397/98 and
C-410/98 [2001] All ER (EC) 496 applied.
(3) ACT was not a withholding tax under the
Parent/Subsidiary directive. Accordingly, the proceedings would be remitted to
the Chancery Division to decide the unresolved factual question whether, had
group income election been available to the P group, the group would have
elected to have the United Kingdom subsidiaries pay the dividends in question free
of ACT or would have chosen that the United Kingdom subsidiaries should pay the
dividends outside group income elections, thus enabling the overseas parents to
receive convention tax credits (see [27]–[30], [46]–[48], [89]–[91], [107],
[108], below); Océ van der Grinten NV v IRC Case C-58/01 [2003] STC 1248
applied.
Notes
Advance corporation tax was abolished by s 31 of the
Finance Act 1998 with effect from 6 April 1999.
Section 247 of the Income and Corporation Taxes Act
1988 was repealed by the Finance Act 2001, ss 85(5), (6), 110, Sch 33, Pt
2(10), with effect in relation to payments made after 11 May 2001.
83
For the Double Taxation Relief (Taxes on Income)
(Netherlands) Order 1980, SI 1980/1961 and the Double Taxation Relief (Taxes on
Income) (Italy) Order 1990, SI 1990/2590, see 19 Halsbury’s Statutory
Instruments (2004 issue) 115.
For the Income and Corporation Taxes Act 1988, ss
231, 788, see 44(1) Halsbury’s Statutes (4th edn) (2005 reissue) 253,
1142.
Cases referred to in opinions
Adams v Cape Industries
plc [1991] 1 All ER 929, [1990] Ch 433, [1990] 2 WLR 657, CA.
Athinaïki Zithopiia
AE v Greece Case C-294/99 [2002] STC 559, [2001] ECR I-6797, ECJ.
Bachmann v Belgium
Case C-204/90 [1994] STC 855, [1992] ECR I-249, ECJ.
EC Commission v
Belgium Case C-300/90, [1992] ECR I-305, [1993] 1 CMLR 785 ECJ.
Food Distributors Ltd
(DHN) v Tower Hamlets London BC [1976] 3 All ER 462, [1976] 1
WLR 852, CA.
Metallgesellschaft
Ltd v IRC, Hoechst AG v IRC Joined cases C-397/98 and C-410/98 [2001] All
ER (EC) 496, [2001] Ch 620, [2001] 2 WLR 1497, [2001] ECR I-1727, ECJ.
Ministério Público,
Fazenda Pública v Epson Europe BV Case C-375/98 [2002] STC 739, [2000] ECR
I-4243, ECJ.
NEC Semi-Conductors
Ltd v IRC [2003] EWHC 2813 (Ch), [2004] STC 489; affd [2006] EWCA
Civ 25, [2006] STC 606.
Océ van der Grinten
NV v IRC Case C-58/01 [2003] STC 1248, [2003] ECR I-9809, ECJ.
Ord v Belhaven Pubs
Ltd [1998] 2 BCLC 447, CA.
R v IRC, ex p
Commerzbank AG [1991] STC 271, DC.
Salomon v A Salomon
& Co Ltd [1897] AC 22, [1895–9] All ER Rep 33, HL.
Salomon v Comrs of
Customs and Excise [1966] 3 All ER 871, [1967] 2 QB 116, [1966] 3 WLR 1223,
CA.
Staatssecretaris van
Financiën v Verkooijen Case C-35/98 [2002] STC 654, [2000] ECR I-4071, ECJ.
Cases referred to in list of authorities
Air Canada v British
Columbia (1989) 59 DLR (4th) 161, Can SC.
Albacruz (Cargo
Owners) v Albazero (Owners), The Albazero [1976] 3 All
ER 129, [1977] AC 774, [1976] 3 WLR 419, HL.
Amministrazione delle
Finanze dello Stato v San Giorgio SpA Case 199/82 [1983] ECR 3595, [1985] 2
CMLR 658, ECJ.
Asscher v
Staatssecretaris van Financiën Case C-107/94 [1996] All ER (EC) 757, [1996]
ECR I-3089, ECJ.
Baars v Inspecteur
der Belastingdienst Particulieren/Ondernemingen Gorinchem Case C-251/98
[2000] ECR I-2787, [2002] 1 CMLR 1437, ECJ.
Bachmann v Belgium
Case C-204/90 [1994] STC 855, [1992] ECR I-249, ECJ.
Bank of Tokyo Ltd v
Karoon [1986] 3 All ER 468, [1987] AC 45n, [1986] 3 WLR 414n, CA.
Barclays Mercantile
Business Finance Ltd v Mawson (Inspector of Taxes) [2004] UKHL 51,
[2005] 1 All ER 97, [2005] 1 AC 684, [2004] 3 WLR 1383.
Bosal Holding BV v
Staatssecretaris van Financiën Case C-168/01 [2003] All ER (EC) 959, [2003]
ECR I-9409, ECJ.
84
Brasserie du Pêcheur
SA v Germany, R v Secretary of State for Transport, ex p Factortame Ltd
Joined cases C-46/93 and C-48/93 [1996] All ER (EC) 301, [1996] QB 404, [19963]
2 WLR 506, [1996] ECR I-1029, ECJ.
British Steel plc v
Customs and Excise Comrs [1997] 2 All ER 366, CA.
Cie de Saint-Gobain,
Zweigniederlassung Deutschland v Finanzamt Aachen-Innenstadt Case C-307/97
[2000] STC 854, [1999] ECR I-6161, ECJ.
Deutsche Morgan
Grenfell Group plc v IRC [2005] EWCA Civ 78, [2005] 3 All ER 1025, [2006] 2
WLR 103.
Dextra Bank &
Trust Co Ltd v Bank of Jamaica [2001] UKPC 50, [2002] 1 All ER (Comm) 193.
EC Commission v
Belgium Case C-300/90 [1992] ECR I-305, [1993] 1 CMLR 785, ECJ.
EC Commission v
France Case 270/83 [1986] ECR 273, [1987] 1 CMLR 401, ECJ.
Eurowings
Luftverkehrs AG v Finanzamt Dortmund-Unna Case C-294/97 [2001] [1999] ECR
I-7447, 3 CMLR 1669, ECJ.
Imperial Chemical
Industries plc v Colmer (Inspector of Taxes) [2000] 1 All ER 129,
[1999] 1 WLR 2035, HL.
IRC v Océ Van Der
Grinten NV [2000] STC 951.
Istituto
Chemioterapico Italiano SpA and Commercial Solvents Corp v EC Commission Joined
cases 6/73 and 7/73 [1974] ECR 223, [1974] 1 CMLR 309, ECJ.
Kleinwort Benson Ltd
v Birmingham City Council [1996] 4 All ER 733, [1997] QB 380, [1996] 3 WLR
1139, CA.
Lipkin Gorman (a
firm) v Karpnale Ltd [1992] 4 All ER 512, [1991] 2 AC 548, [1991] 3 WLR
10, HL.
Manninen (Proceedings
brought by) Case C-319/02 [2005] All ER (EC) 465, [2005] Ch 236, [2005] 2
WLR 670, ECJ.
Marks & Spencer
plc v Halsey (Inspector of Taxes) Case C-446/03 [2006] STC 237, ECJ.
Niru Battery
Manufacturing Co v Milestone Trading Ltd [2003] EWCA Civ 1446, [2004] 1 All
ER (Comm) 193, [2004] QB 985, [2004] 2 WLR 1415.
Ostime (Inspector
of Taxes) v Australian Mutual Provident Society [1959] 3 All ER 245,
[1960] AC 459, [1959] 3 WLR 410, HL.
Provimi Ltd v Aventis
Animal Nutrition SA [2003] EWHC 1211 (Comm), [2003] All ER (D) 59 (Jun).
Sempra Metals Ltd
(formerly Metallgesellschaft Ltd) v IRC [2005] EWCA Civ 389,
[2005] STC 687, [2005] 3 WLR 521.
Union Texas Petroleum
Corp v Critchley (Inspector of Taxes) [1988] STC 691; affd
[1990] STC 305, CA.
Viho Europe BV v
European Commission Case C-73/95P [1997] All ER (EC) 163, [1997] ICR 130,
[1996] ECR I-5457, ECJ.
Woolfson v
Strathclyde Regional Council (1978) 38 P & CR 521, HL.
Woolwich Equitable
Building Society v IRC [1992] 3 All ER 737, [1993] AC 70, [1992] 3 WLR 366,
HL.
Appeal
The Commissioners of Inland Revenue appealed with
permission of the House of Lords Appeal Committee given on 6 May 2004 from the
decision of the Court of Appeal (Peter Gibson, Laws LJJ and Sir Martin Nourse)
on 17 December 2003 ([2003] EWCA Civ 1849, [2004] STC 130) dismissing their
appeal from the decision of Park J on 22 January 2003 ([2003] EWHC 32 (Ch),
[2003] STC 250)
85
which determined, as a preliminary issue, that any
compensation or restitution to which Pirelli Cable Holding NV, Pirelli Tyre
Holding NV, Pirelli SpA, Pirelli General plc and Pirelli UK plc might have been
entitled as a result of making payments of advance corporation tax between May
1995 and March 1999 pursuant to a provision of United Kingdom law which, in
2001, had been declared to be contrary to Community law by the Court of Justice
of the European Communities, was not extinguished or reduced by virtue of the
fact that Pirelli Cable Holding NV, Pirelli Tyre Holding NV, and Pirelli SpA
had received payments from the United Kingdom Revenue pursuant to art 10 of the
Double Taxation Relief (Taxes on Income) (Italy) Order 1990, SI 1990/2590 and
art 10 of the Double Taxation Relief (Taxes on Income) (Netherlands) Order
1980, SI 1980/1961. The facts are set
out in the opinion of Lord Scott of Foscote.
Ian Glick QC, David
Ewart and Kelyn Bacon (instructed by the Solicitor for Revenue
and Customs) for the Crown.
Graham Aaronson QC,
David Cavender and Paul Farmer (instructed by Dorsey &
Whitney) for the respondents.
Their Lordships took time for consideration.
8 February 2006.
The following opinions were delivered.
LORD NICHOLLS OF BIRKENHEAD.
[1] My Lords, the primary issue on this appeal
concerns the application of two double taxation conventions in circumstances
not envisaged when they were made. The
problem arises from the unforeseen impact of Community law on the partial
‘imputation’ system of corporation tax operated within the United Kingdom
between 1973 and 1999. This system was
introduced by the Finance Act 1972 as a means of avoiding the perceived double
taxation of distributed profits, once in the hands of the company and again in
the hands of its shareholders. The
central principle of the new taxation scheme was that part of the corporation
tax paid by a company was ‘imputed’ to its shareholders by giving them an
appropriate tax credit. The means
adopted for this purpose was that a company was required to pay a tax on its
dividends known as advance corporation tax, or ACT in short, and its
shareholders were given a corresponding tax credit. Nowhere did the legislation state that
liability to pay ACT was a precondition of entitlement to a tax credit. But this unspoken linkage lay at the heart of
the scheme, and the legislation was drawn in a form which achieved this
result. This linkage is central to the
first issue raised by this appeal.
[2] In broad outline the legislation provided
as follows. Where a company resident in
the United Kingdom paid a dividend to its shareholders it became liable to pay
ACT in respect of the dividend. ACT was
set against the company’s liability to ‘mainstream’ corporation tax. A recipient of the dividend, if resident in
the United Kingdom, became entitled to a tax credit. The amount of the tax credit corresponded to
the current rate of ACT. In the case of
an individual a tax credit was utilised primarily as a credit against his
income tax. Any excess was paid to the
individual. Where the recipient of the
dividend was a company the amount of the dividend plus the amount of the tax
credit constituted franked investment income.
This could be used to frank dividends paid by the company
86
so the company would not be liable to ACT on its
dividends (see ss 14, 231 and 238–241 of the Income and Corporation Taxes Act
1988).
[3] This basic scheme was subject to an
important exception. Special provision
was made for dividends paid by one company to another company in the same
group. These companies could jointly
make a group income election under s 247 of the 1988 Act. A group income election had a twofold effect,
reflecting the linkage between payment of ACT and entitlement to a tax
credit. Dividends paid by a subsidiary
to its parent while the election was in force, conveniently described as
‘election dividends’, did not trigger liability to pay. Nor did their receipt trigger entitlement to
a tax credit. So election dividends did
not constitute franked investment income of the parent. They constituted, in the language of the
legislation, ‘group income’ of the receiving company: hence the description
‘group income election’.
[4] On the face of the legislation the right
to make a group income election was expressly confined to cases where both
companies were resident in the United Kingdom.
In March 2001 the Court of Justice of the European Communities ruled
that affording groups of companies the right to make a group income election
where they were resident in the United Kingdom but denying them that right
where the parent companies were not resident in the United Kingdom was contrary
to art 52 of the EC Treaty (now art 43 EC) providing for freedom of
establishment (see Metallgesellschaft Ltd v IRC, Hoechst AG v IRC Joined
cases C-397/98 and C-410/98 [2001] All ER (EC) 496, [2001] Ch 620). The court held that the claimant companies in
these cases were entitled to compensation for loss of the use of the money paid
as ACT between the date of payment and the date when the ACT was used by being
set off against the subsidiary company’s mainstream corporation tax liabilities.
[5] This ruling had a widespread effect in
practice. Shortly stated, the effect of
this decision was that wherever group companies had been denied the opportunity
to make a group income election, the United Kingdom government became liable to
compensate them for the loss they suffered if the denial was an infringement of
art 43 EC as interpreted by the court in Metallgesellschaft Ltd v IRC,
Hoechst AG v IRC.
[6] Not surprisingly, many groups of companies
sought to take advantage of this ruling.
The present appeal concerns the assessment of compensation payable to
one such group, the Pirelli group, chosen as a test case in one category of
claims within a group litigation order.
For the purposes of this appeal the detailed facts, summarised in the
speech of my noble and learned friend Lord Scott of Foscote, are not
material. The essential facts are that
dividends were paid by a subsidiary company resident in the United Kingdom to
parent companies resident in Italy or the Netherlands. Group income election was not available,
because the parent companies were not resident in the United Kingdom. So ACT was paid by the subsidiary company in
respect of the dividends. For the same
reason the parent companies were not entitled to tax credits under s 231 of the
1988 Act. But under double taxation
conventions made by this country with Italy and the Netherlands the parent
companies became entitled to tax credits of a reduced amount calculated as set
out in those conventions. I shall call
these ‘convention tax credits’.
THE FIRST QUESTION
[7] In the present case the first question
arising on the assessment of compensation is this. If the United Kingdom legislation had
permitted parent
87
companies resident in other member states of the
European Community to make a group election, and if an election had been made
in respect of the dividends in question, would the parent companies have been
entitled to payment of the convention tax credits they in fact received under
the double taxation conventions? If they
would have been so entitled then no deduction should be made in respect of
these tax credits when calculating the compensation. If, however, the parent companies would not
have been so entitled, then in principle—and subject to the other issues raised
on this appeal—due allowance should be made in respect of these tax credits
when calculating the compensation. Due
allowance should be made because, in this event, the convention tax credits
received by the parent companies comprise financial benefits they would not
have received had the group been able to make a group income election and had
the group duly done so.
[8] Whether the parent companies would have
been entitled to convention tax credits in this hypothetical circumstance
depends upon the proper interpretation of the relevant double taxation
conventions. The two conventions are the
Netherlands Convention of 7 November 1980 (see the Double Taxation Relief
(Taxes on Income) (Netherlands) Order 1980, SI 1980/1961) and the Italian
Convention of 21 October 1988 (see the Double Taxation Relief (Taxes on Income)
(Italy) Order 1990, SI 1990/2590). Both
conventions have effect in United Kingdom law pursuant to Orders in Council
made under s 788 of the 1988 Act.
[9] In all respects material on the present
issue the two conventions are in similar form.
For convenience I will refer only to the Netherlands convention. The relevant provision in the Netherlands
convention is art 10(3). This was
applicable as long as an individual resident in the United Kingdom was entitled
to a tax credit in respect of dividends paid by a company resident in the
United Kingdom. Article 10(3)(c)
entitled a company resident in the Netherlands which received dividends from a
company resident in the United Kingdom to a tax credit calculated and payable
as follows:
‘… a tax credit equal to
one half of the tax credit to which an individual resident in the United
Kingdom would have been entitled had he received those dividends, and to the
payment of any excess of that tax credit over its liability to tax in the
United Kingdom.’
‘Tax credit’ in this subparagraph has the same meaning as
in the 1988 Act (see art 3(2) of the convention). Entitlement to a convention tax credit was
subject to tax at a rate not exceeding 5% on the total amount of the dividend
and the tax credit (see art 10(3)(a)(ii)).
Nothing turns on this additional provision.
[10] The first point to note is that, as would
be expected, art 10(3)(c) accords with the scheme of the underlying legislation
whereby entitlement to a tax credit marched hand-in-hand with liability to pay
ACT. Article 10(3)(c) did not, in terms,
exclude election dividends from its scope.
But this was not necessary. On
the face of the legislation a parent company resident in the Netherlands could
not make a group income election. By
definition, an election dividend was payable only to a company resident in the
United Kingdom. So, on the face of the
legislation, there could be no question of a Netherlands resident company ever
being in a position to claim a convention tax credit on a dividend whose
payment did not attract ACT.
88
[11] This analysis of art 10(3)(c) accords
with the intended purpose of the Netherlands convention as stated in s
788. The object of s 788 is to enable
effect to be given to arrangements made with the governments of other countries
‘with a view to affording relief from double taxation’. Section 788(3) provides that, notwithstanding
anything in any enactment, arrangements such as the Netherlands convention
shall have effect in relation to income tax and corporation tax:
‘in so far as they
provide … (d) for conferring on persons not resident in the United Kingdom the
right to a tax credit under section 231 in respect of qualifying distributions
made to them by companies which are so resident.’
In s 231(1) entitlement to a tax credit was expressed to
be subject to s 247. The effect of s 247
was to exclude an election dividend from entitlement to a tax credit under s
231 and from attracting liability to ACT under s 14(1) as a qualifying
distribution.
[12] So much is clear. The question before the House, however,
requires one to inhabit a new and different world, a world where an election
dividend may be paid to a non-resident company.
How is art 10 to be read and understood in this new world? If art 10 is to be read literally I would
agree with Park J (see [2003] EWHC 32 (Ch), [2003] STC 250) and the Court of
Appeal (Peter Gibson, Laws LJJ and Sir Martin Nourse) (see [2003] EWCA Civ
1849, [2004] STC 130) that a Netherlands resident parent would be entitled to a
convention tax credit on an election dividend.
That is the effect of the literal interpretation of art 10. But is this the proper interpretation?
[13] Article 10, like all documents, must be
interpreted purposively. So, in
answering these questions it is important to have in mind how they come to
arise at all. They arise from the
explosive effect of Community law on a prime feature of the legislation
regarding groups of companies. On its
face s 247 precluded non-resident parent companies from making a group income
election. The Netherlands convention was
drafted and agreed on this basis. But
this residence limitation, in its application to companies resident in member
states of the European Community, was negated by the Metallgesellschaft Ltd
v IRC, Hoechst AG v IRC Joined cases C-397/98 and C-410/98 [2001] All ER
(EC) 496, [2001] Ch 620 decision.
Consequent upon this decision s 247 fell to be applied in a circumstance
for which it was not designed. This had
a knock-on effect on the application of art 10 of the Netherlands
convention. The further effect was that
art 10 fell to be applied in a circumstance for which it too was not designed,
namely, that a Netherlands resident company could receive a dividend which had
not attracted payment of ACT.
[14] So where does this leave art
10(3)(c)? In my view no one could
suppose that in the new world ushered in by the Metallgesellschaft Ltd v
IRC, Hoechst AG v IRC decision a group with a Netherlands resident parent could
elect to avoid payment of ACT by making a group income election pursuant to the
Metallgesellschaft Ltd v IRC, Hoechst AG v IRC ruling and, at the same
time, remain entitled to a tax credit under the Netherlands double taxation
convention. That would be to entitle a
company resident overseas to a convention tax credit where a company resident
in this country was not entitled to a tax credit. That would put a Netherlands resident parent
in a better position than a United Kingdom resident parent. That would fly in the face of the stated
purpose of art 10(3)(c), namely, to entitle a Netherlands resident parent to
part (‘one half’) of the tax
89
credit to which a United Kingdom resident would have
been entitled had he received the dividend.
[15] An interpretation of art 10 having this
effect would comprise such a gross and obvious departure from the evident
purpose of art 10(3)(c), and from a fundamental feature of the tax credit
scheme on which art 10(3)(c) is superimposed, that in my view art 10(3)(c)
cannot be so read. The Netherlands
convention assumes that the dividend whose receipt attracts a convention tax
credit will also have attracted liability to ACT. That is an assumption implicit in art
10(3)(c). When interpreting art 10(3)(c)
in the post- Metallgesellschaft Ltd v IRC, Hoechst AG v IRC world effect
should be given to this implicit assumption.
Article 10(3)(c) is to be read as not applying to election dividends.
[16] Accordingly, on the first issue raised on
this appeal I would declare that if the United Kingdom-resident subsidiary of a
parent resident in another member state of the European Union had paid a
dividend to its parent while a group income election was in force the parent
would not have been entitled to a tax credit in respect of the dividend under
double taxation conventions in the form of the Netherlands and Italian
conventions.
THE SECOND QUESTION
[17] The Pirelli group have another string to
their bow. Mr Aaronson QC submitted
that, even so, when assessing the compensation payable to a United
Kingdom-resident subsidiary for the wrongful denial of an opportunity to
exclude payment of ACT by making a group income election, no account should be
taken of the convention tax credits received by the overseas parent companies. Even though (on this footing) the parent
companies were not entitled to these tax credits and the consequential
payments, these should be left out of account because a subsidiary and its
parent are separate legal entities. The
harm suffered in consequence of the breach of Community law was suffered by the
United Kingdom subsidiary, whereas the countervailing advantages were enjoyed
by the parent companies.
[18] I have to say I find this argument of
Pirelli no more attractive than their submission on the first question. Here again, Pirelli are seeking to have the
best of both worlds. The outcome for
which they contend is so artificial that it cannot be right.
[19] A group income election is a group
election. A group income election cannot
be made by a subsidiary alone. It is an
election made jointly by the subsidiary paying the dividend and the parent
receiving the dividend. By making such
an election both companies seek the fiscal consequences of making the
election. One consequence is that by
making the election the subsidiary will obtain the advantage of not paying ACT
in respect of the relevant dividend.
Another consequence is that the subsidiary will obtain this advantage at
the cost of depriving the parent of a tax credit in respect of the
dividend. These two fiscal consequences
are inextricably linked. You cannot have
one without the other. That is why the
election has to be made jointly. The
advantage to the paying subsidiary comes at a price to the recipient parent.
[20] Accordingly the loss sustained by the
subsidiary cannot fairly be assessed in isolation. The commercial reality is that by not having
the opportunity to make a group income election the group lost the opportunity
to take advantage of a fiscal package: a package which affected the parent in
one way and its subsidiary in a different way.
In calculating the loss suffered by the group, that is, the parent and
the subsidiary, regard must be had to both elements in the
90
package. The
effect on the parent must be considered as well as the effect on the
subsidiary. The subsidiary lost the use
of the money paid as ACT. In some
instances, where the ACT was not set off against its mainstream corporation
tax, the subsidiary lost the money altogether.
But this cannot be treated as the amount of compensation payable by the
Commissioners of Inland Revenue without also taking into account any adverse
consequence a group income election would have had on the parent. By the same token, assessment of the
compensation must take into account any countervailing fiscal benefit received
by the parent which would not have been available had a group income election
been made.
[21] Pirelli sought to side-step this
difficulty by presenting their claim as a claim by the subsidiary alone. But this difference in presentation cannot
make any material difference in the outcome.
A group income election cannot be made, or continued in force, by a
subsidiary acting alone. Pirelli cannot,
by presenting their claim in this way, alter the fundamental nature of the
wrong for which compensation has become payable, namely, the loss of an
opportunity for the parent and the subsidiary jointly to take advantage of a
single fiscal package having different effects on the parent and the
subsidiary.
[22] It is for this reason that assessment of
the group’s overall loss, rather than the loss of the subsidiary alone, does
not involve departure from the basic principle of company law that a parent
company and its subsidiary are separate legal entities. Assessment of the overall loss represents the
only fair way to assess the amount of loss suffered where a subsidiary and its
parent have been denied the opportunity jointly to obtain a single package of
this nature.
[23] Assessment of the compensation now in
issue is not exclusively a matter for the domestic law of the United
Kingdom. Assessment of the compensation
has a Community law dimension because the compensation is payable for breach of
an EC Treaty obligation. But this
feature does not disturb the conclusion expressed above. I can see nothing in the above analysis which
offends the Community law principles of equivalence and effectiveness.
[24] Pirelli, however, raised a different
point of Community law. The European
Court of Justice has held that in certain circumstances the need to ensure
cohesion of a member state’s tax system may justify rules liable to restrict
fundamental freedoms. This principle has
been applied where a direct link existed, in the case of one and the same
taxpayer, between the grant of a tax advantage and the offsetting of that
advantage by a fiscal levy, both of which related to the same tax (see
Bachmann v Belgium Case C-204/90 [1994] STC 855, [1992] ECR I-249 and EC
Commission v Belgium Case C-300/90 [1992] ECR I-305) [1993] 1 CMLR
785. The precise limits of this
principle remain to be explored. But the
European Court has declined to apply the principle where the link between the
taxes in question comprised the grant of a tax exemption to shareholders resident
in the Netherlands in respect of dividends received by them and taxation of the
profits of companies in another member state.
They were ‘two separate taxes levied on different taxpayers’ (see
Staatssecretaris van Financiën v Verkooijen Case C-35/98 [2002] STC 654 at
695, [2000] ECR I-4071 at 4132–4133 (paras 56–58)).
[25] Pirelli sought to apply this limitation
on the scope of the ‘tax cohesion’ principle by analogy in the present
case. They sought, further, a reference
to Luxembourg if their submission on this point did not find favour. I am unable to accept either submission. Pirelli’s argument breaks down at the very
outset. The
91
suggested analogy is misconceived. The grounds on which a restriction on a
fundamental freedom may be justified say nothing about the principles applicable
in assessing compensation for breach of a Treaty freedom. Assessment of compensation is primarily a
matter for the domestic legal system of each member state, provided always that
the principles of equivalence and effectiveness are duly observed.
[26] Accordingly, I would hold in answer to
the second question that the amount of the convention tax credits received by
the Dutch and Italian parent companies is to be taken into account in
calculating the compensation payable for the breach of art 43 EC. Essentially what is to be taken into account
is the value of the benefits thus obtained.
Depending on the circumstances a notional rate of interest on the
amounts paid may properly be regarded as an element of the value obtained. But this is a matter to be evaluated in the
context of the facts of the particular claims.
THE THIRD QUESTION
[27] The third question is of a different
character. It arises out of the terms of
Council Directive (EC) 90/435 of 23 July 1990 (OJ L225 20.8.90 p 6), known as
the ‘Parent/Subsidiary Directive’. On
this I have nothing to add to the observations of Lord Scott. The terms of arts 5(1) and 7(1) of the
directive, coupled with the decision of the European Court of Justice in Océ
van der Grinten NV v IRC Case C-58/01 [2003] STC 1248, [2003] ECR I-9809,
make plain that ACT is not prohibited by the directive. This is acte clair. I would not refer a question to the European
Court of Justice as sought by Pirelli.
[28] For these reasons, and those set out in
the speeches of my noble and learned friends Lord Hope of Craighead, Lord Scott
and Lord Walker of Gestingthorpe, I would allow this appeal and remit the
proceedings to Park J to decide the unresolved factual question whether, had
group income election been available to the Pirelli group, the group would have
elected to have the United Kingdom subsidiaries pay the dividends in question
free of ACT or, instead, would have chosen that the United Kingdom subsidiaries
should pay the dividends outside group income elections, thus enabling the
overseas parents to receive convention tax credits.
LORD HOPE OF CRAIGHEAD.
[29] My Lords, I have had the advantage of
reading in draft the speeches of my noble and learned friends Lord Nicholls of
Birkenhead, Lord Scott of Foscote and Lord Walker of Gestingthorpe. I agree with them, and for substantially the
same reasons as they have given I too would allow the appeal and make the order
which has been proposed by Lord Nicholls.
But we are differing from judges in the courts below whose opinions in
revenue matters have always commanded the greatest respect. So I should like to add these brief remarks
to explain why I have reached the same conclusion as my noble and learned
friends have done.
[30] The issues which we have been asked to
decide are (1) whether, if a group income election had been made under s 247(1)
of the Income and Corporation Taxes Act 1988, the three European Union (EU)
parent companies would have been entitled to a tax credit under the relevant
Double Taxation Agreement (DTA) (the election issue) on the dividends paid to
them as group income by their United Kingdom subsidiaries; (2) whether, if that
question is answered in the negative, the tax credits which were received by
the EU parent companies under the relevant DTA should be brought into account
in assessing the compensation
92
payable to the United Kingdom subsidiaries for the
breach of art 52 of the EC Treaty (now art 43 EC) because the tax advantage of
a group income election was not made available where the parent company was not
resident in the United Kingdom (the assessment issue); and (3) whether ACT is a
withholding tax within the meaning of art 5(1) of Council Directive (EC) 90/435
(the Parent/Subsidiary Directive) (OJ L225 20.8.90 p 6). I would give the answers ‘No’ to the first
question, ‘Yes’ to the second question and ‘No’ to the third question, holding
the answer to the third question to be acte clair.
THE ELECTION ISSUE
[31] The answer to the question raised by the
election issue is to be found in arrangements for double taxation relief that
were created for companies not resident in the United Kingdom by ss 788(3)(d),
231(1) and 247(2) of the 1988 Act. These
arrangements have to be seen in the context of the system which Lord Scott has
described for giving tax credits to a parent company resident in the United
Kingdom in respect of distributions made to it on which its United Kingdom
subsidiary had paid advance corporation tax (ACT) under s 14(1) of the 1988
Act. I have listed the sections of the
1988 Act in that order because taking them in that order reveals most clearly
the question of statutory construction on which the answer to the issue
depends. In my opinion it is to the
domestic legislation, and not only to the relevant DTA, that one must look for
the answer. For the reasons that I shall
give, and in agreement with Lord Scott, I think that the legislation provides
the answer which is contended for by the Revenue.
[32] Section 788(3)(d) of the 1988 Act gave
domestic effect to the arrangements specified in a DTA for conferring on
non-resident companies such as the EU parent companies ‘the right to a tax
credit under section 231 in respect of qualifying distributions made to them’
by companies which were resident in the United Kingdom. The expression ‘tax credit’ is defined in s
832(1) as meaning a tax credit under s 231.
So it is to the provisions of that section that one must go in the first
instance to see what the right was that was being referred to in s 788(3)(d).
[33] Section 231(1) provided that, where a
company resident in the United Kingdom made a qualifying distribution and the
person receiving the distribution was another such company, the recipient of
the distribution was to be entitled to a tax credit equal to such proportion of
the amount or value of that distribution as corresponded to the rate of ACT in
force for the financial year in which the distribution was made. But the opening words of that subsection
stated that it was subject to ss 95(1)(b) and 247. Section 95(1)(b), which applied where a
company purchased its own shares from a dealer, is not in point in this
case. But s 247 is very much in
point. It is that section, and in
particular s 247(2), that creates the difficulty for the Pirelli companies.
[34] Section 247(1) of the 1988 Act enabled an
election (a group income election) to be made not to account for ACT in two
situations that might arise within a group of companies all of whose members
were resident in the United Kingdom. The
first is the one that would have applied to this case, had all the companies
within the Pirelli group had been so resident.
It was where the paying company was a 51% subsidiary of the receiving
company or of another United Kingdom resident company of which the receiving
company was a 51% subsidiary. In that
situation the receiving company and the paying company could jointly elect that
ACT was not to be payable on the dividends which the parent received from the
paying company.
93
[35] The consequences of such an election, so
long as it was in force, were twofold.
First, ACT was not payable under s 14(1) of the 1988 Act on the election
dividends. Second, the income
represented by the amount of the election dividends was not franked investment
income (that is to say, franked as having borne ACT) in the hands of the
recipient company, but was to be treated in the ordinary way as group income of
the receiving company. This meant that
income of this description did not qualify for a tax credit under s 231. As s 247(2) puts it, the election dividends
‘shall be excluded from sections 14(1) and 231’. The payment of ACT was not enacted as a
condition that had to be fulfilled before a shareholder could become entitled
to a tax credit, as Park J was right to point out (see [2003] EWHC 32 (Ch) at
[36], [2003] STC 250 at [36]). But the
link between these two provisions—imposing the liability to ACT and giving the
right to the tax credit—could not have been more clearly expressed.
[36] The reason for the exclusion, in the
domestic context, of the right to a tax credit under s 231 was simple. A group income election dividend was a
dividend on which no ACT was payable.
The purpose of s 231 was to provide the receiving company with a tax
credit equivalent in amount to the ACT payable on the dividend. The tax credit was designed to avoid tax
having to be paid twice on the same dividend when tax was paid on its profits
by the parent company. As no ACT was
payable in the case of a group income election dividend, there was nothing
against which credit needed to be given to the recipient company. There was no risk of tax being paid twice in
respect of the amount distributed to the parent, so there was no need of a tax
credit to avoid that result.
[37] It has to be borne in mind, of course,
that ACT was a tax payable by the paying company. It was not a tax on profits in the hands of
the recipient. Receipt of distributions
that had borne ACT could arise in circumstances where, due to such features as
capital allowances, the paying company did not have sufficient profits at the
end of its relevant accounting period to give rise to a liability to mainstream
corporation tax. This was frequently the
case, as Park J explains in his judgment (at [60]). But the tax credit was always given, and
given only, where the company making the qualifying distribution in respect of
which it was given was liable under s 14(1) of the 1988 Act to payment of ACT
on an amount equal to the amount or value of the distribution. As Mr Glick QC for the Revenue put it, it was
the payment of the ACT that made the giving of the tax credit to the recipient
necessary.
[38] Park J said (at [38]–[41]) that the
answer to the election issue was to be found in the DTA and that s 788(3)(d),
which referred only to s 231 and made no mention of s 247(2), had to be read
consonant to the Treaty. The Court of
Appeal adopted the same approach (see [2003] EWCA Civ 1849 at [46], [2004] STC
130 at [46]). Mr Aaronson QC for the Pirelli
companies urged your Lordships to take the same view. If this argument was right its effect would
be that s 247(2) had the effect already mentioned in the domestic context but
did not qualify the entitlement to relief from double taxation provided for by
the DTA. But I do not think that this
approach fits either with the clear language of s 231 itself, or with the
system which the tax credit mentioned in that section was designed for.
[39] It seems to me that there is no escape
from the fact that it is s 231 that s 788(3)(d) uses to identify the relief
that is to be given in accordance with the DTA by way of a relief under the
domestic system to the non-resident companies.
It was the domestic system, not the Treaty, that defined the extent of
that relief. According to its own terms
s 231 had to be read subject to s 247.
And the giving
94
of a tax credit for qualifying distributions only
became necessary because qualifying distributions were distributions on the
making of which the paying company was liable to ACT. Reading these two provisions together, it is
clear that the prerequisite for the giving of a tax credit was the making of a
qualifying distribution which was liable to ACT. A group income election extinguished that
liability and with it the right to the tax credit that was the counterpart of
the liability. It follows that, if the
same system had been available to them and a group election had been made, no
ACT would have been payable on the distributions to the EU parent
companies. So there would have been no
entitlement to a tax credit with respect to those distributions under s
788(3)(d). I would answer the question
raised by this issue in favour of the Revenue.
THE ASSESSMENT ISSUE
[40] The assessment issue raises an entirely
different point. The argument for the
Pirelli companies is that the harm of having to pay the ACT and of losing the
cash flow advantage that a group income election would have given them was
suffered by the United Kingdom subsidiary only, not by the parent which
received the advantage of the tax credit paid to it under s 788(3)(d). So the amount of the compensation to be paid
to the subsidiary should be assessed without bringing the tax credit received
by the parent companies into account. To
this the answer for the Revenue is that s 247(1) gave the right to make a group
income election to both the parent and to the subsidiary jointly. It was an election which was available to
them which would only have been made where it was in their joint interests to
make it, and in the event of a joint election the tax credit would not have
been payable. So the group should be
treated as a single unit for the purpose of assessing the amount of the
compensation payable in order to give effect to the judgment of the European
Court of Justice. Park J said ([2003]
STC 250 at [48]) that this was a huge and unjustifiable step, as it ignored the
separate corporate identities of the respective companies.
[41] I do not think that the Revenue’s
approach falls into the trap of ignoring the companies’ separate
identities. What it does is to look at
them instead as members of the group. I
agree with Lord Scott that the relevant harm was the harm that the group
suffered by reason of the breach of the parent companies’ right to freedom of
establishment under art 52 of the Treaty.
The breach deprived the group of the benefit of a joint election which,
if there had been no breach, would have been available under s 247(1) of the
1988 Act for the benefit of the group as a whole. It would have been exercisable by the paying
and the recipient members of the group jointly.
It is the joint nature of the exercise that makes it appropriate to look
at the group as a whole when the compensation is being assessed rather than the
effect of the breach on each company taken in isolation.
[42] In Metallgesellschaft Ltd v IRC, Hoechst
AG v IRC Joined cases C-397/98 and C-410/98 [2001] All ER (EC) 496 at 539,
[2001] Ch 620 at 664 (para 88) the European Court stressed that in an action
for restitution the principal sum due was none other than the amount of
interest which would have been generated by the sum use of which was lost as a
result of the premature levy of the tax.
In para 89 it said that art 52 of the Treaty entitled a subsidiary
resident in the United Kingdom and/or its parent company having its seat in
another member state to obtain interest on the ACT paid by the subsidiary
during the period between the payment of the ACT and the date on which the
mainstream corporation tax became payable.
Elaborating on this point (at para 96), the court said that art 52
95
of the Treaty required that resident subsidiaries and
their non-resident parent companies should have an effective legal remedy in
order to obtain reimbursement or reparation of the financial loss which they
had sustained as a result of being deprived of the entitlement to a tax regime
that allowed them to avoid the obligation on the subsidiary to pay ACT in
respect of dividends paid to the parent company.
[43] The language that the European Court uses
in these and the other passages in the judgment referred to by Lord Scott shows
that the loss for which an effective legal remedy is to be made available is
the loss which was sustained by the subsidiaries and their parent companies as
members of a group. This is because the
taxation regime which was denied to them was a regime that was available to a
group under the domestic system. The
effect of the breach of art 52 which denied the opportunity to opt for that
regime to the Pirelli companies cannot be assessed without looking at both
sides of that regime—at the differences that it would have made to the
receiving companies within the same group as well as to those companies that
were paying the dividends. Decisions as
to whether to opt for that regime, which had to be made by these companies
jointly, were always made by reference to what was in the best interests of the
group looked at as a whole. It would be
artificial, and therefore wrong, to look only to the effect on the subsidiaries
that opting for that regime would have had without having regard to its effect
on the parent companies within the same group also.
[44] Mr Glick QC for the Revenue accepted that
a common law claim against the parent companies for repayment of the tax
credits would fail because they were not paid to the parent companies under a
mistake. But he submitted that it was
nevertheless open to the Revenue to rely on the amount of the tax credits given
to the parents to reduce or extinguish its liability to the subsidiaries in the
same group. In my opinion reparation for
the breach of art 52 of the Treaty for the loss that was sustained by the group
in this case, as explained by the European Court’s judgment, permits this
approach. I would answer this question
too in favour of the Revenue.
THE WITHHOLDING TAX ISSUE
[45] The third issue is whether ACT was
contrary to EC law on the ground that it was a withholding tax within the
meaning of the Parent/Subsidiary Directive.
As Park J pointed out ([2003] STC 250 at [61]), it only needs to be
addressed if both of the two previous questions are answered in favour of the
Revenue. Both of those questions were
answered both by him and by the Court of Appeal against the Revenue, so it was
not necessary for them to deal with it.
As we are in favour of the Revenue on both points we must do so.
[46] The wording of art 5(1) does not provide
a clear answer to the problem. But it
comes close to doing this, as it provides that profits which a subsidiary
distributes to its parent shall be exempt from withholding tax. The wording of this paragraph suggests that
the exemption applies where the taxable person from whom the tax is being
withheld is the parent company to whom the profits are being distributed. That is how the concept of withholding tax
was explained in Ministério Público, Fazenda Pública v Epson Europe BV
Case C-375/98 [2002] STC 739, [2000] ECR I-4243, where the European Court said
that a tax was a withholding tax where it was a tax on the parent company’s
dividend income, not a tax on the profits of the subsidiary (see [2002] STC 739
at 756, [2000] ECR I-4243 at 4274–4275 (paras 23, 24)). ACT did not fall within that description, as
under
96
s 239 of the 1988 Act it was a tax payable by the
subsidiary which was to be set against the subsidiary’s liability to mainstream
corporation tax. So it was a tax on the
profits, if any, of the subsidiary, not those of the parent company. This is made clear by the fact that income of
the parent by way of distributions made by its subsidiary under deduction of
ACT entitled the parent to a tax credit under s 231 equal in amount to the tax
payable by the subsidiary as ACT.
Article 7(1) however seems to me to put the matter beyond doubt. It provides that the term ‘withholding tax’
does not cover an advance payment of corporation tax to the member state of the
subsidiary which is made in connection with a distribution of profits to its
parent company. ACT answers to this
description.
[47] In Athinaïki Zithopiia AE v Greece
Case C-294/99 [2002] STC 559, [2001] ECR I-6797 it was argued to the European
Court that art 5 did not apply because the tax in question was a tax on the
profits of the subsidiary to its parent company within the meaning of art
7(1). In para 29 of its judgment (see
[2002] STC 559 at 572, [2001] ECR I-6797 at 6826) this argument was rejected on
the ground that it related to income which was taxed only in the event of a
distribution of dividends and up to the limit of the dividends paid. This description seems to me to indicate that
the reason why the taxation was held not to fall within art 7(1) was that it
was a tax on the parent’s income as holder of the shares which entitled it to
payment of the dividends, not on the income of the subsidiary. This interpretation of that decision is
confirmed by what the European Court said in Océ van der Grinten NV v IRC
Case C-58/01 [2003] STC 1248 at 1264–1265 (para 47), [2003] ECR I-9809 where,
treating the matter as having been settled by its decisions in Ministério
Público, Fazenda Pública v Epson Europe BV and Athinaïki Zithopiia AE v
Greece, the court made it clear that a withholding tax is a tax on the
income from the shares on which the dividend is paid and where the holder of
the shares is the taxable person from whom the tax is withheld.
[48] It is true, as Park J pointed out ([2003]
STC 250 at [60]), that ACT could only be said to be an advance payment of tax
payable by the subsidiary if its profits were sufficiently large at the end of
the accounting period to attract mainstream corporation tax. If this was not the case—and it often was not
the case—it could not be said to be an advance payment of tax that was
ultimately to be borne by the subsidiary.
But this not the criterion to which art 7(1) as interpreted by the
European Court addresses itself. The
essence of a withholding tax within the meaning of art 5(1) as explained by art
7(1) is that it is a tax on the income of the parent company. The court’s own jurisprudence makes it clear
that ACT was not a withholding tax for the purposes of the Parent/Subsidiary
Directive.
LORD SCOTT OF FOSCOTE.
INTRODUCTION
[49] My Lords, the problem that your Lordships
must resolve on this appeal is produced by a combination of four things. First, the Finance Act 1972 introduced
advance corporation tax (ACT) to our tax system. The charging provision, originally s 84(1) of
the 1972 Act, became s 14(1) of the Income and Corporation Taxes Act 1988. It required a company which made a
‘qualifying distribution’ (typically the payment of a dividend) to its
shareholders to pay ACT to the Revenue.
The ACT (described in s 14(1) as ‘an amount of corporation tax’) was
made payable on an amount equal to the amount or value of the distribution and
at the rate at which income tax at the basic rate was charged. The ACT was not deducted from the dividend
paid to the shareholder (as had been the case with
97
Sch F income tax: see the explanation given by Peter
Gibson LJ in the judgment of the Court of Appeal (see [2003] EWCA Civ 1849 at
[5], 6], [2004] STC 130 at [5], [6])) but instead the ACT attributable to each
dividend was ‘imputed’ to the shareholder to whom the dividend was payable and,
if the shareholder was resident in the United Kingdom, the shareholder
generally became entitled to receive from the Revenue a tax credit equal to the
amount of the ACT (see s 231(1) of the 1988 Act). The shareholder was then taxable on the
aggregate of the dividend and the tax credit.
As for the company that had paid the ACT, that company was entitled to set
the ACT against its liability to mainstream corporation tax.
[50] The second contributory factor was the
availability of group income elections.
Provision for these elections was made by s 247(1) of the 1988 Act. Where both a subsidiary company and its
parent company were resident in the United Kingdom they were permitted by s 247(1)
to make a joint election that the sub-section should apply to the dividends
paid by the subsidiary to the parent.
Then, so long as the election remained in force, ACT would not be
payable in respect of those dividends and the parent company would not receive
a corresponding tax credit.
[51] The right to receive tax credits pursuant
to s 231(1) of the 1988 Act did not apply to recipients of dividends who were
not resident in the United Kingdom. But
the liability of companies to pay ACT when dividends were paid arose regardless
of the place of residence of the shareholder.
The position of non-resident shareholders was catered for, to the extent
that it was catered for at all, under double taxation agreements entered into
between the United Kingdom and other states.
Some of these double taxation agreements contained provision for tax
credits to be given by the Revenue to foreign shareholders in a United Kingdom
resident company which had paid dividends and, accordingly, also paid ACT. The amount of the tax credits to which the
foreign shareholders were to become entitled and the conditions on which
entitlement to the tax credits depended were agreed by negotiation between the
United Kingdom and the state in question and were spelt out in the double taxation
agreements. Not all double taxation
agreements contained provisions enabling the foreign shareholders to claim tax
credits but, relevantly for present purposes, the double taxation agreements
between the United Kingdom and the Netherlands and the United Kingdom and Italy
did. The agreements themselves did not
ipso facto become enforceable in our domestic law but s 788 of the 1988 Act
gave them that enforceability. Subsection
(1) of s 788 declared that:
‘If Her Majesty by Order
in Council declares that arrangements specified in the Order have been made
with the government of any territory outside the United Kingdom with a view to
affording relief from double taxation … and that it is expedient that those
arrangements should have effect, then those arrangements shall have effect in
accordance with sub-section (3) below.’
Subsection (3) says that:
‘Subject to the
provisions of this Part, the arrangements shall, notwithstanding anything in
any enactment, have effect in relation to income tax and corporation tax in so
far as they provide … (d) for conferring on persons not resident in the United
Kingdom the right to a tax credit under section 231 in respect of qualifying
distributions made to them by companies which are so resident.’
98
[52] The third factor contributing to the
problem which your Lordships must resolve is that the double taxation agreement
(the DTA) between the United Kingdom and the Netherlands (the Netherlands
Convention of 7 November 1980) (see the Double Taxation Relief (Taxes on Income)
(Netherlands) Order 1980, SI 1980/1961), and that between the United Kingdom
and Italy (the Italian Convention of 21 October 1988) (see the Double Taxation
Relief (Taxes on Income) (Italy) Order 1990, SI 1990/2590), provided, under art
10(3)(c), that a Netherlands/Italian company that had received a dividend from
a United Kingdom subsidiary should (subject to certain qualifications which are
irrelevant for present purposes)—
‘be entitled to a tax
credit equal to one half of the tax credit to which an individual resident in
the United Kingdom would have been entitled had he received those dividends,
and to the payment of any excess of that tax credit over its liability to tax
in the United Kingdom.’
For the sake of completeness I should add that art
10(3)(b) of each DTA provided for tax credits to be given to individual
shareholders resident in the Netherlands or Italy.
[53] The final event which has created the
problem is the judgment given on 8 March 2001 by the European Court of Justice
in two cases, heard together, Metallgesellschaft Ltd v IRC, Hoechst AG v IRC
Joined cases C-397/98 and C-410/98 [2001] All ER (EC) 496, [2001] Ch 620. The European Court of Justice was responding
to requests for rulings on the question whether the denial by s 247(1) of the
1988 Act to companies of other member states and their United Kingdom
subsidiaries of the right to make a group income election, and thereby to
enable the subsidiaries to avoid payment of ACT in respect of dividends paid to
their parents, while affording that right to United Kingdom companies and their
United Kingdom subsidiaries, was consistent with Community law (see the
formulation of the issue in the court’s judgment ([2001] All ER (EC) 496 at
531, [2001] Ch 620 at 655–656 (para 35))).
[54] The court noted ([2001] All ER (EC) 496
at 533, [2001] Ch 620 at 657 (para 44)) that the ACT/group income election
statutory provisions gave the subsidiary of a United Kingdom parent—
‘a cash flow advantage
inasmuch as it retains the sums which it would otherwise have had to pay by way
of advance corporation tax until such time as mainstream corporation tax
becomes payable …’ (See also [2001] All
ER (EC) 496 at 534, [2001] Ch 620 at 659 (para 54).)
The court concluded that it was contrary to art 52 of the
Treaty (now art 43 EC) for this tax advantage to be made available where the
parent company was resident in the United Kingdom but to be denied where the
parent company was resident in some other member state.
[55] Having reached the conclusion to which I
have referred, the court then had to consider what remedy the companies which
had been denied the right to make group income elections should be entitled
to. The court’s conclusion on this
question is set out in para 96 of the judgment ([2001] All ER (EC) 496 at 540,
[2001] Ch 620 at 665) and, having regard to some of the submissions addressed
to your Lordships, I think it important to set out in terms what the court
said:
‘Where a subsidiary
resident in one member state has been obliged to pay advance corporation tax in
respect of dividends paid to its parent company
99
having its seat in
another member state even though, in similar circumstances, the subsidiaries of
parent companies resident in the first member state were entitled to opt for a
taxation regime that allowed them to avoid that obligation, art 52 of the EC
Treaty requires that resident subsidiaries and their non-resident parent
companies should have an effective legal remedy in order to obtain
reimbursement or reparation of the financial loss which they have sustained and
from which the authorities of the member state concerned have benefited as a
result of the advance payment of tax by the subsidiaries.’
It is to be noticed that the European Court referred to
the financial loss sustained by and the legal remedy that had to be afforded to
‘resident subsidiaries and their non-resident parent companies’.
THE ISSUES IN THIS APPEAL
[56] Following the European Court’s judgment
there were, not surprisingly, claims for compensation by a number of foreign
companies and their United Kingdom subsidiaries. Having regard to the number of cases and to
the many common issues that were raised by these cases a group litigation order
(GLO) was made. The GLO identified a
number of common issues, one of which was the effect, if any, that the receipt
by a foreign parent company of a tax credit should have on the compensation to
which its United Kingdom subsidiary was entitled. That is the core issue which your Lordships must
now resolve. There are, however, sub-issues.
[57] Some of the facts that give rise to the
issues are actual. Some must be
assumed. The actual facts are as
follows: (i) Two of the respondents, Pirelli Cable Holding NV (Pirelli Cable)
and Pirelli Tyre Holding NV (Pirelli Tyre) were resident in the Netherlands and
one of the respondents, Pirelli SpA, was resident in Italy. (ii) The respondents Pirelli UK plc (the UK
holding company) and Pirelli General plc (the UK subsidiary) were resident in
the United Kingdom. (iii) The UK holding
company owned all the shares in the UK subsidiary and Pirelli Tyre, Pirelli
Cable and Pirelli SpA each owned 33
1/3% of the issued ordinary shares of the UK holding company. (iv) Over the period 14 July 1995 to about 20
April 1999 Pirelli Tyre, Pirelli Cable and Pirelli SpA could, if they had been
United Kingdom-resident companies, have made a group income election in respect
of all dividends declared by the UK holding company. (v) The UK holding company therefore became
liable to pay ACT on each occasion in the relevant period that it paid
dividends to its three parent companies and it paid ACT accordingly. On two occasions ACT was paid by the UK
subsidiary in respect of dividends it paid to the UK holding company. The UK holding company subsequently paid
dividends of the same amount to its three parent companies but in doing so
incurred no additional ACT liability because the dividends it paid were
‘franked’ payments equal to the ‘franked investment income’ it had received
from the UK subsidiary (see s 241(1) of the 1988 Act). Nothing turns on this and it is simpler, for
present purposes, to treat all ACT as if paid by the UK holding company. And (vi) in respect of all these dividends
the three parent companies received from the Revenue tax credits pursuant to
the DTAs between the United Kingdom and Netherlands or the United Kingdom and
Italy, the net amount of which was 6·875% of the value of the dividends. The assumed facts are that over the whole of
the relevant period a group income election would have been made by each
of the three parent companies if it had been
100
open to them to do so. This is an issue of fact yet to be resolved
if it becomes necessary to resolve it.
[58] The two main issues are, therefore,
these: (i) If the UK holding company had paid the dividends to its three parent
companies while a group income election was in place, and thereby avoided
liability to pay ACT in respect of those dividends, would the three parent
companies nonetheless have been entitled to the tax credits that they received
pursuant to the relevant DTA? (ii) If
the parent companies would not have been entitled to tax credits had the
dividends been paid by the UK holding Company without incurring any liability
to ACT, should the tax credits in fact received by the parents be brought into
account in calculating the compensation payable for the breach of art 52 that,
the European Court held, had been committed?
[59] There is another issue, namely, whether
ACT constituted a withholding tax within the meaning of art 5(1) of Council
Directive (EC) 90/435 of 23 July 1990 (the Parent/Subsidiary Directive) (OJ
L225 20.8.90 p 6) and, if so, whether it fell within the exception set out in
art 7(1) of that directive, and whether that question should be referred to the
European Court for a ruling. I will
leave any further explanation of this issue until later.
THE FIRST MAIN ISSUE
[60] Both Park J (see [2003] EWHC 32 (Ch),
[2003] STC 250) and the Court of Appeal (Peter Gibson and Laws LJJ and Sir
Martin Nourse) (see [2004] STC 130) concluded, for substantially the same
reasons, that even on the hypothesis that the Pirelli respondents had been
entitled to make and had jointly made a group income election under s 247(1),
so that the UK holding company in paying their dividends would have incurred no
liability to pay and would not have paid ACT, nonetheless the parent companies
would still have been entitled to tax credits pursuant to art 10(3)(c) of the
respective DTAs. The essential reasoning
behind this conclusion, strongly supported by Mr Aaronson QC in his written and
oral submissions to your Lordships, was that each DTA was the product of
negotiation between the United Kingdom and the Netherlands or Italy (as the
case may be), represented a give-and-take bargain between the negotiating
states and should be applied according to its own internal provisions without
regard to what might be thought to be the policy underlying the tax regime
regarding ACT and tax credits.
[61] Thus, Park J said (at [35]) that ‘the
question ultimately turns on the construction of the relevant provisions of art
10 in the two DTAs …’ (at [37]) that
‘Article 10, read with s 788(3) … lays down the conditions which have to
be fulfilled …’ for entitlement to tax credits and (at [38]) that the Pirelli
parent companies ‘complied with every specific condition [for entitlement to
tax credits], and would have done whether the Pirelli UK dividends were paid to
it as group income or not’.
[62] And the Court of Appeal judgment said
([2004] STC 130 at [45]):
‘The impression that the
provisions of the DTAs are intended to be exhaustive for prescribing the
circumstances in which the right to a tax credit under s 231 in respect of
qualifying distributions made by a UK-resident company is conferred on a
non-resident is strongly confirmed by consideration of the Italy and
Netherlands DTAs themselves. They are
extremely detailed, and although they are substantially similar they are not
identical … The DTAs contain their own formulas for calculating the
101
amount of tax credit
that is allowed, different from the amount of the tax credit available to a
UK-resident shareholder, albeit expressed by reference to what an individual
UK-resident shareholder would have received by way of tax credit. They contain their own conditions governing
entitlement to the tax credit in particular amounts.’
[63] My Lords, I am in respectful agreement
with Park J and the Court of Appeal that the issue depends upon the
construction of art 10(3)(c) of the DTAs but I do not agree that that provision,
or the DTA provisions as a whole, are ‘exhaustive’, if the use of that
adjective is intended to mean that the correct meaning and effect of the DTA
provisions can be ascertained simply by reading the provisions themselves.
[64] Let me stay with art 10(3)(c), under
which a Netherlands, or Italian, resident company can claim ‘one half of the
tax credit to which an individual resident in the United Kingdom would have
been entitled had he received those dividends’.
There are two points to be made about this language. First, there is no definition anywhere in
these DTAs of the expression ‘tax credit’.
But art 3(2) of each DTA says that:
‘As regards the
application of the Convention by one of the States any term not defined therein
shall, unless the context otherwise requires, have the meaning which it has
under the law of that State concerning the taxes to which the Convention
applies.’
[65] The expression ‘tax credit’ is defined in
s 832(1) of the 1988 Act as ‘a tax credit under section 231’ and the same
wording is to be found in s 788(3)(d).
So that is the meaning to be attributed to the expression ‘tax credit’
in art 10(3)(c). Re-writing art
10(3)(c), the Netherlands or Italian company is to be entitled to ‘one half of
the tax credit under section 231 of the 1988 Act to which an individual
resident in the United Kingdom would have been entitled if he had received
those dividends’.
[66] Section 231(1) says that:
‘Subject to section …
247 … where a company resident in the United Kingdom makes a qualifying
distribution and the person receiving the distribution is another such company
or a person resident in the United Kingdom, not being a company, the recipient
of the distribution shall be entitled to a tax credit …’
And s 247(2) says that where a group income election under
sub-s (1) is in force—
‘the election dividends
shall be excluded from sections 14(1) and 231 and are accordingly not included
in references to franked payments made by the paying company or the franked
investment income of the receiving company but are in the Corporation Tax Acts
referred to as “group income” of the receiving company.’
[67] It follows, in my opinion, from ss 231(1)
and 247(2) that where a group income election is in force and dividends are
paid by a United Kingdom company a s 231 tax credit cannot be claimed by the
recipient of the dividends. The language
of ss 231(1) and 247(2) bars such a claim.
The language of art 10(3)(c) involves, therefore, a contradiction. The entitlement to a s 231 tax credit of a Pirelli
parent company that received dividends that had been paid under a group income
election is said by art 10(3)(c) to depend upon the entitlement to a s 231
102
tax credit of a United Kingdom-resident individual
‘had he received those dividends’ (my emphasis). But ‘those dividends’ would be dividends paid
under a group income election and, therefore, in respect of which no ACT was
payable or paid and no entitlement to a s 231 tax credit could arise. But an individual could not have received any
such dividends. Only a company could
receive dividends paid under a group income election. In order, therefore, to apply the hypothesis
prescribed by art 10(3)(c), either the individual must be taken to have
received dividends of a character different from those received by the Pirelli
parent company, ie the individual must be supposed to have received dividends
not paid under a group income election and in respect of which ACT had
been paid, or, alternatively, the individual must be attributed for the purpose
of the art 10(3)(c) hypothesis with the ability to receive dividends paid under
a s 247 group income election.
[68] Each of these solutions is somewhat
unsatisfactory because each involves an element of unreality. But some degree of unreality in trying to
apply the art 10(3)(c) condition is, perhaps, inevitable if one bears in mind
that those who negotiated the terms of the DTAs and those, if different, who
drafted art 10(3)(c) could not have had in mind that the provision would
subsequently need to be applied to a hypothetical situation in which it has to
be supposed that group income elections have been made by United Kingdom
subsidiaries and their other member state parent companies, that dividends have
been paid by the subsidiaries to the parent companies without incurring any ACT
liability and that the parent companies that have received these dividends have
then claimed s 231 tax credits. For my
part I prefer the second of the two possible solutions which seems to me to do
less violence than the other to the language of art 10(3)(c). There is no doubt about the character of the
dividends assumed to have been received by the Pirelli parent companies and in
respect of which it is to be assumed that they claimed tax credits. They are dividends assumed to have been paid
under a group income election jointly made by the subsidiaries and the
parents. The reference in art 10(3)(c)
to ‘those dividends’ that the hypothetical individual is to be taken to have
received should be, it seems to me, dividends of the same assumed
character. So the individual must be
attributed with the ability to receive dividends of that character. If that is the right approach to the art
10(3)(c) hypothesis, the individual who had received dividends of that
character would not have been entitled to a s 231 tax credit and, consequently,
nor would the Pirelli parent companies.
The alternative appears to me to involve overlooking the statutory
definition of ‘tax credit’ as a ‘tax credit under section 231’ and attributing
to the individual, and thence to the Pirelli companies, the right to a tax
credit that an application of s 231 would reject.
[69] I disagree, therefore, with the
conclusion reached by Park J and the Court of Appeal on the first issue, the
so-called ‘election’ issue. The error
which I think was made can be identified at [37] and [38] of Park J’s judgment
(see [2003] STC 250). At [37] he sets
out in sub-paras (i)–(v) the art 10(3)(c) conditions that, in his view, the
Pirelli parent companies had to satisfy in order to claim tax credits. I respectfully agree that these conditions
must all be satisfied and were satisfied.
The judge then moves to his conclusion which he sets out in sub-para (vi):
‘If the foregoing conditions were satisfied, as they were, Pirelli SpA was
entitled to tax credits equal to half the tax credits to which a United Kingdom
resident individual would have been entitled.’
103
In my opinion, however, the judge omitted an additional
condition that had to be satisfied, namely, that a United Kingdom-resident
individual who had received the dividends received by Pirelli SpA (ie ‘those
dividends’) would have been entitled to s 231 tax credits. This, in my opinion, is the critical
condition.
[70] At [38], second sentence, Park J said
that: ‘Pirelli SpA complied with every specific condition, and would have
done so whether the Pirelli UK dividends were paid to it as group income or not.’ (My emphasis.) Not so, in my respectful opinion. The judge’s conclusion overlooks the art
10(3)(c) requirement that the hypothetical United Kingdom resident individual
must be taken to have received ‘those dividends’, that is to say, the dividends
paid to Pirelli SpA under a group income election. In that hypothetical event the omitted
condition would not be satisfied. The
United Kingdom-resident individual would not have been entitled to a s 231 tax
credit; nor, therefore, would Pirelli SpA.
[71] The Court of Appeal apparently did not
accept that references to ‘tax credit’ in art 10 of the two double taxation
agreements meant ‘tax credit under section 231’
(see [2004] STC 130 at [46]–[48]).
But the s 832(1) definition is essential in order to give a meaning to
‘tax credit’ in art 10, for the term is nowhere else defined and art 3(2)
expressly imports domestic law definitions for terms not defined in the DTAs
themselves. The Court of Appeal say (at
[48]) that—
‘the reference [in s
788(3)(d)] to s 231 was necessary in order to cause the tax credit to be
aggregated with the distribution in respect of which the tax credit is
conferred and so to be rendered chargeable to tax under para 2 of Sch F.’
That is no doubt true but does not, in my opinion, justify
writing the definition out of the DTAs.
Article 10 in express terms hinged a Netherlands/Italian parent
company’s right to a tax credit to the entitlement that a United
Kingdom-resident individual would have had to a tax credit if he had received
the dividends that the foreign parent company had received. That being so I do not, for my part, find it
at all surprising that specific provisions in domestic legislation restricting
in specified circumstances the right to a tax credit should govern the
availability of a tax credit under art 10.
Be that as it may, the only tax credit available, at least in this area
of tax law, is a tax credit under s 231.
There is no such thing as an art 10(3)(c) tax credit that is not a ‘tax
credit under section 231’.
[72] In my opinion, therefore, and in
agreement with my noble and learned friend Lord Walker of Gestingthorpe, on the
true construction of art 10(3)(c) the Pirelli parent companies would not have
been entitled to tax credits in respect of dividends paid to them by the UK
holding company as group income under a s 247(1) group income election and,
therefore, without incurring any liability to pay ACT.
THE SECOND ISSUE
[73] So what are the consequences of that
conclusion for the compensation claims made by the Pirelli respondents? Both Park J and the Court of Appeal treated
this second issue as one which depended upon whether the separate corporate
identity of the UK holding company, which had paid the ACT, could be
ignored. There were references to
Salomon v A Salomon & Co Ltd [1897] AC 22, [1895–9] All ER Rep 33 and
to the remarks about a single economic unit or lifting the corporate veil in
such cases as DHN Food Distributors Ltd v Tower Hamlets London BC [1976]
3 All ER 462, [1976] 1 WLR 852, Adams v Cape Industries plc [1991]
104
1 All ER 929, [1990] Ch 433 and Ord v Belhaven
Pubs Ltd [1998] 2 BCLC 447. Park J
held, and the Court of Appeal agreed, that compensation payable to the Pirelli
subsidiaries was not to be reduced by reference to the value of the tax credits
received by the Pirelli parent companies.
[74] It must, for the purposes of this second
issue, be accepted that if s 247 group income elections had been able to be
made and had been made by the Pirelli companies and if dividends had been paid
to the Pirelli parent companies by the UK holding company while the elections
were in force, the UK holding company would not have paid ACT and the Pirelli
parent companies would not have been entitled to tax credits.
[75] The answer to this issue does not, in my
opinion, depend upon whether the corporate veil can be lifted or whether the
Pirelli companies should for compensation purposes be treated as a single
economic unit. The answer requires as a
start that the nature of the wrong for which compensation is payable be
identified.
[76] The European Court of Justice held that
the ACT/group income election tax regime imposed an unwarranted restriction on
the freedom of establishment guaranteed by art 52 of the Treaty and hence was a
breach of that article. The rights
conferred by art 52 are described in para 41 of the judgment (see [2001] All ER
(EC) 496 at 532, [2001] Ch 620 at 656–657):
‘Article 52 of the
Treaty constitutes one of the fundamental provisions of Community law and has
been directly applicable in the member states since the end of the transitional
period. Under that provision, freedom of
establishment for nationals of one member state within the territory of another
member state includes the right to take up and pursue activities as
self-employed persons and to set up and manage undertakings under the
conditions laid down for its own nationals by the law of the country where such
establishment is effected. The abolition
of restrictions on freedom of establishment also applies to restrictions on the
setting up of agencies, branches or subsidiaries by nationals of any member
state established in the territory of another member state.’
And in para 42:
‘Freedom of
establishment thus defined includes, pursuant to art 58 of the Treaty, the
right of companies or firms formed in accordance with the law of a member state
… to pursue their activities in the member state concerned through a branch or
agency … With regard to companies, it should be noted in this context that it
is their corporate seat … that serves as the connecting factor with the legal
system of a particular state, like nationality in the case of natural persons …
Acceptance of the proposition that the member state in which a company seeks to
establish itself may freely apply to it a different treatment solely by reason
of the fact that its registered office is situated in another member state
would thus deprive art 52 of all meaning …’
[77] It is important, in my opinion, to notice
two things from these passages. First,
the right to freedom of establishment conferred by art 52 is the right of a
company (or an individual) with its seat in one member state to establish
itself in another member state.
Unwarranted restrictions imposed by the latter member state on the
branch or agency or subsidiary company by means of which the parent company is
seeking to establish itself in that member state is plainly a
105
breach of the art 52 right to freedom of
establishment of that parent company. It
is more difficult to describe an infringement of art 52 rights brought about by
unwarranted restrictions being placed by a member state on the subsidiary of a
parent company that has its seat in another member state as a breach of art 52
rights of the subsidiary. It is, surely,
strictly speaking, the parent company’s right to freedom of establishment that
is interfered with.
[78] Second, the language of these passages
shows that a company in one member state and its branch, or its agency, or its
subsidiary, in another member state are regarded as a group. The subsidiary, unlike the branch or an
agency, will have separate legal identity from that of the parent company but
there is no suggestion that this makes any difference.
[79] It is entirely consistent and natural,
therefore, that when the European Court discusses the remedies to be made
available for an infringement of art 52 rights it treats the parent companies,
whose rights have been infringed, and the subsidiaries, on whom the unwarranted
restrictions have been placed, as a group.
Thus: (i) para 77 refers to the situation where—
‘a subsidiary resident
in the member state concerned and its parent company having its seat in another
member state have been wrongfully deprived of the benefit of a taxation regime
which would have enabled the subsidiary to pay dividends to its parent company
without having to pay advance corporation tax …’ (see [2001] All ER (EC) 496 at
537, [2001] Ch 620 at 662);
(ii) para 96 concludes that—
‘art 52 of the EC Treaty
requires that resident subsidiaries and their non-resident parent companies
should have an effective legal remedy in order to obtain reimbursement or
reparation of the financial loss which they have sustained and from which the
authorities of the member state concerned have benefited as a result of the
advance payment of tax by the subsidiaries’ (see [2001] All ER (EC) 496 at 540,
[2001] Ch 620 at 665);
(iii) paras 98 and 107 refer to a claim for compensation—
‘brought … by a resident
subsidiary and its non-resident parent company for reimbursement or reparation
of the financial loss which they have suffered as a consequence of the advance
payment of corporation tax by the subsidiary …’ (see [2001] All ER (EC) 496 at
540, 542, [2001] Ch 620 at 665–666, 667);
(iv) in para 2 of the court’s ruling ([2001] All ER (EC)
496 at 543, [2001] Ch 620 at 668), the court says that—
‘art 52 of the EC Treaty
requires that resident subsidiaries and their non-resident parent companies
should have an effective legal remedy in order to obtain reimbursement or
reparation of the financial loss which they have sustained …’;
and (v) in para 3 of the ruling there is a similar
reference. It is true that in the
passages to which I have referred the European Court was responding to
questions which had been referred to it by the national court and that the
formulation of the responses may have owed something to the nature of the
questions. But the questions, set out in
para 13 of the opinion of the Advocate
106
General (Fennelly) (see [2001] All ER (EC) 496 at 506–507
[2001] Ch 620 at 628–629), do not require the explicit responses that the
court’s responses appear to me to constitute in making clear that the
compensation claims could be put forward by the subsidiaries and parent
companies as a group to recover the loss that they, the group, had suffered in
consequence of the art 52 infringements.
[80] The correct analysis, in my opinion, is
that the art 52 infringements of which the Pirelli respondents complain were
infringements of the Pirelli parent companies’ rights to freedom of
establishment in the United Kingdom, that the primary measure of the loss thus
caused was the financial detriment to their UK subsidiary caused by its
liability to pay ACT but its inability to make group income elections, but that
the benefit of the tax credits that the Pirelli parent companies would not have
received but for the art 52 infringements should be brought into account. Logically, as it seems to me, the
compensation should be compensation to the group for the detriment suffered by
the group. Compensation that
concentrates on the financial detriment of the subsidiary and ignores the
financial benefit of the parent seems to me to overlook the nature of the art
52 infringement. The identity of the
recipient or recipients of the compensation payable for the loss suffered by
the group can be left to be decided by the group, but cannot, in my opinion,
affect the quantum of the compensation.
[81] Park J said ([2003] STC 250 at [45]) that
the Revenue’s problem, in arguing that countervailing advantages to the Pirelli
parent companies should be balanced against the disadvantage of the UK holding
company having had to pay ACT soon after paying dividends to the Pirelli parent
companies, was that the harm suffered in consequence of the breach of Community
law was suffered by the United Kingdom subsidiaries whereas the countervailing
advantages were enjoyed by the Pirelli parent companies. The Court of Appeal repeated with approval
Park J’s point (see [2004] STC 130 at [53]).
But the point overlooks, in my opinion, the nature of an art 52 infringement. ACT was not held to be an unlawful tax. What was unlawful was interfering with the
Pirelli parent companies’ art 52 freedom of establishment by denying their
subsidiaries a facility (the ability to avoid ACT by making a group income
election) available to United Kingdom-resident parent companies’ subsidiaries. The charge which was found proved was
discrimination against other member state parent companies which had
established United Kingdom subsidiaries and in favour of United Kingdom parent
companies with United Kingdom subsidiaries.
The relevant harm, which took the form of financial detriment to the UK
subsidiary, was harm to the group brought about by a breach of the parent
companies’ art 52 freedom of establishment.
[82] I would, therefore, allow the Revenue’s
appeal on this issue too.
THE WITHHOLDING TAX ISSUE
[83] The Parent and Subsidiary Directive
recited that it was necessary, in order to ensure fiscal neutrality, that the
profits which a subsidiary distributed to its parent company be exempt from
withholding tax. Accordingly, art 1(1)
of the Directive required each member state of the EU to apply the Directive to
‘… distributions of profits by companies of that State to companies of other
Member States of which they are subsidiaries’ and art 5(1) of the Directive
said that: ‘Profits which a subsidiary distributed to its parent company shall
… be exempt from withholding tax …’
107
[84] The Pirelli respondents contend that ACT
constituted a withholding tax. If that
is correct ACT would have been an unlawful tax, at least in its application to
United Kingdom subsidiaries of parent companies of other member states. It would then follow that exactions of ACT
from these subsidiaries would have been unlawful exactions and the subsidiaries
that had paid ACT would, prima facie, be entitled to reimbursement of that which
had been unlawfully demanded of them.
This, I think, is common ground.
No question would arise of any countervailing advantage obtained by the
parent companies unless in an action brought by the Revenue against the parent
companies for reimbursement of the value of the tax credits as having been
provided under a mistake of law. The
possibilities of such a claim need not be pursued here.
[85] However, art 7(1) of the directive says
that:
‘The term “withholding
tax” as used in this Directive shall not cover an advance payment or prepayment
(précompte) of corporation tax to the Member State of the subsidiary which is
made in connection with a distribution of profits to its parent company …’
Article 7(1) would appear to rule out any claim that ACT
constituted a withholding tax for the purposes of the directive. Moreover, the European Court and the Advocate
General in Metallgesellschaft Ltd v IRC, Hoechst AG v IRC made a number
of references to ACT in terms which bring ACT within the art 7(1) exemption. In para 22 of his opinion, Advocate General
Fennelly said that ‘It is clear that ACT is nothing other than an advance
payment of [mainstream corporation tax]’ (see [2001] All ER (EC) 496 at 510,
[2001] Ch 620 at 633) and the court said ([2001] All ER (EC) 496 at 527, [2001]
Ch 620 at 651 (para 6)):
‘It is important to bear
in mind that ACT is not a sum withheld on a dividend, which is paid in full,
but is rather corporation tax borne by the company distributing dividends, paid
in advance and set off against the mainstream corporation tax (MCT) payable in
respect of each accounting period.’
If the point rested there it would be plain, and acte
clair, that ACT was not a withholding tax.
[86] But the Pirelli respondents rely on
certain passages in other European Court judgments which they submit make it
arguable that ACT is indeed a withholding tax and that the point should be
referred to the European Court for a definitive ruling.
[87] Article 7(1) said that an advance payment
of corporation tax was not a withholding tax, but did not go on to say what
was. Conceptually a ‘withholding tax’ is
a tax levied on the recipient of a payment and is an amount withheld from the
recipient. Park J said ([2003] STC 250
at [59]) that he would not regard ACT ‘as being conceptually a withholding
tax.’ He explained: ‘If a United Kingdom
company declared a dividend of 100 to its shareholders it paid them 100, not
100 less an amount withheld out of it on account of the tax on the dividend.’
[88] The obvious conclusion that ACT was not a
withholding tax is said to be called into question by the European Court’s
judgment in Athinaïki Zithopiia AE v Greece Case C-294/99 [2002] STC
559, [2001] ECR I-6797. In its judgment
the court commented ([2002] STC 559 at 572, [2001] ECR I-6797 at 6826 (para
28)) that ‘the chargeable event for the taxation … is the payment of
dividends. In
108
addition, the amount of tax is directly related to
the size of the distribution’. The same
could be said of ACT. But in para 29 the
court said this:
‘Contrary to the
submissions of the Greek Government, the taxation cannot be treated like an
advance payment or prepayment (précompte) of corporation tax to the member
state of the subsidiary which is made in connection with a distribution of
profits to its parent company, within the meaning of art 7(1) of the
directive. The taxation relates to
income which is taxed only in the event of a distribution of dividends and up
to the limit of the dividends paid.’
It cannot be said of ACT that ACT ‘relates to income which
is taxed only in the event of a distribution of dividends’. ACT is treated as an advance payment of
corporation tax which is assessed on the company’s profits whether or not the
profits are distributed as dividends.
[89] However, following the judgment of Park
J, but before the hearing in the Court of Appeal, the European Court returned
to the question of what constituted a withholding tax. The case was Océ van der Grinten NV v IRC
Case C-58/01 [2003] STC 1248 at 1264–1265, [2003] ECR I-9809 where, at para 47
of its judgment, the court referred to the Athinaïki Zithopiia case (and
Ministério Público, Fazenda Pública v Epson Europe BV Case C-375/98 [2002]
STC 739, [2000] ECR I-4243) as establishing that—
‘any tax on income
received in the state in which dividends are distributed is a withholding tax
on distributed profits for the purposes of art 5(1) of the directive where the
chargeable event for the tax is the payment of dividends or of any other income
from shares, the taxable amount is the income from those shares and the taxable
person is the holder of the shares …’
ACT is not a tax charged on the shareholders. It does not qualify as a withholding tax
under the description of a withholding tax given in the Océ van der Grinten
case.
[90] In my opinion, the terms of art 7(1) of
the directive coupled with the explanation of the nature of a withholding tax
given in the Océ van der Grinten case and the remarks about ACT by the
Advocate General and the European Court in the Metallgesellschaft case
establish, so that the point is acte clair, that ACT falls within the
art 7(1) exception and is not a withholding tax.
CONCLUSION
[91] For these reasons, and those set out in
the opinions of my noble and learned friends Lord Nicholls of Birkenhead, Lord
Hope of Craighead and Lord Walker of Gestingthorpe, I would allow this appeal,
set aside paras 1 and 2 of the order made by Park J on 22 January 2003 and
remit the case to Park J to decide the unresolved factual question referred to
by Lord Nicholls at [28], above. The
respondents must pay the costs of the appeal to the Court of Appeal and to this
House.
LORD WALKER OF GESTINGTHORPE.
[92] My Lords, I have had the advantage of
reading in draft the opinions of my noble and learned friends Lord Nicholls of Birkenhead,
Lord Hope of Craighead and Lord Scott of Foscote and I gratefully adopt Lord
Scott’s summary of the facts. I agree
with the reasoning and conclusions in these opinions. But in view of the importance of the matter,
and because we are differing from the courts
109
below, I add some comments of my own on what has been
called the election issue.
[93] Your Lordships are concerned with two
double taxation agreements (DTAs) entered into by the United Kingdom, one in
1980 with the Netherlands (see the Double Taxation Relief (Taxes on Income)
(Netherlands) Order 1980, SI 1980/1961) and the other in 1988 with Italy (see
the Double Taxation Relief (Taxes on Income) (Italy) Order 1990, SI
1990/2590). On what has been called the
election issue the provisions of the DTAs can assist the respondents, at any
rate in the national courts, only if and so far as those provisions have been
incorporated into the domestic tax law of the United Kingdom. That is the effect of the words ‘in so far as
they provide’ in s 788(3) of the Income and Corporation Taxes Act 1988 (see
R v IRC, ex p Commerzbank AG [1991] STC 271 (Divisional Court); NEC
Semi-Conductors Ltd v IRC [2003] EWHC 2813 (Ch), [2004] STC 489 (Park
J). The correct starting point for any
examination of the position under domestic tax law is therefore the relevant
paragraph of s 788(3) following the words ‘in so far as they provide’, that is—
‘(d) for conferring on
persons not resident in the United Kingdom the right to a tax credit under
section 231 in respect of qualifying distributions made to them by companies
which are so resident.’
[94] It is a striking feature of s 788(3)(d)
that it is expressed in precise and technical language (in contrast to the
three preceding paragraphs, which are expressed in more general, and largely
non-technical, language). What is to be
conferred is ‘a tax credit under section 231’ (which is in any case, under s
832(1) of the 1988 Act, the meaning of ‘tax credit’) and it is to be conferred
‘in respect of qualifying distributions’ made by companies resident in the
United Kingdom. Any Dutch or Italian
businessman encountering these expressions in (say) 1995 would, unless he
happened to be a tax expert himself, have had to seek advice as to their
meaning from someone who was expert in the intricacies of United Kingdom
corporation tax. This need would not be
diminished by the strong overriding language used earlier in s 788(3)
(‘notwithstanding anything in any enactment’), since para (d) is spelling out
what the content of the overriding provisions is to be.
[95] The inquiring businessman would have been
told that a tax credit under s 231 was the only significant tax credit known to
the imputation system of corporation tax introduced into the United Kingdom in
1973. (His adviser might, at the risk of
confusion, have added that under s 239 of the 1988 Act payment of advance
corporation tax (ACT) gives the paying company a credit in respect of
mainstream corporation tax (MCT), but that this is normally spoken of in terms
of set-off rather than credit; and that there could also be a specialised type
of tax credit under s 441A of the 1988 Act, but that that complication could be
ignored.)
[96] The adviser would perhaps also have gone
on to say that the basic structure of s 231 (as it then stood) was to be found
in the first three subsections.
Subsection (1) laid down the basic condition for the grant of the tax
credit, that is, a qualifying distribution made (usually in the form of a
dividend) by a company resident in the United Kingdom to a shareholder resident
in the United Kingdom. It also laid down
the amount of the tax credit: it was to be a proportion of the amount of the
distribution corresponding to the rate of ACT (so that if the rate of ACT was
25%, the credit would be one-quarter of the amount of the distribution). Subsection (2) dealt with certain rare cases
in which a corporate
110
shareholder resident in the United Kingdom would be
entitled to payment of the tax credit; the adviser probably would have told his
client to disregard sub-s (2) as an unimportant exception to the general rules
about franked investment income (to be found in Ch V of Pt VI of the 1988
Act). Subsection (3) dealt with the much
more common situation in which there was a right to payment of the tax credit,
that is when the dividend was paid to a non-corporate shareholder resident in
the United Kingdom (typically an individual, but possibly a body of trustees of
a pension scheme, a charity or some other type of trust), and the shareholder
was not liable to pay income tax, or paid it at a relatively low rate. Those were the salient characteristics of a
tax credit under s 231.
[97] On being told this much about a tax
credit under s 231 the foreign client might have objected that it was dealing
with a purely domestic situation (a United Kingdom company paying a dividend to
a United Kingdom shareholder), so how could such a credit benefit him? The adviser would have replied with some
confidence that the whole point of s 788(3)(d) was to confer on a non-resident,
through a DTA, a right which was, under s 231, conferred on residents, and that
this residence requirement was necessarily and obviously overridden in the
parallel universe of the DTA. The
adviser might also have added, with little less confidence, that the
quantification of the tax credit in s 231(1) was also liable to be overridden,
and was in practice very likely to be overridden, by arrangements negotiated
under s 788 between HM Government and the Dutch or Italian government. Neither side would wish its fisc to have to
pay out to non-resident shareholders more cash than it had to, and the amount
of the credit under the DTA would depend on hard bargaining rather than on any
transparent principle. In the present
case the negotiated formula produces, on an assumed ACT rate of 25%, the rather
awkward figure of 6·875% (explained in the judgment of Park J [2003] EWHC 32
(Ch) at [24], [2003] STC 250 at [24]).
[98] So (the client might have said,
recapitulating his understanding of the
advice) the shareholder’s residence requirement in s 231(1) is to be
disregarded, and the quantum of the tax credit under s 231(1) cannot be relied
on; are there any other characteristics of ‘a tax credit under section 231’
about which advice is needed? The
adviser would then have drawn attention to the only other obvious technicality
in s 231(1), the need for the shareholder to receive a ‘qualifying
distribution.’ He would have explained
that that expression meant, in practice, a dividend, and that under the imputation
system the payment of a dividend was closely associated with the payment
(within a maximum period of 15 weeks) of an amount of ACT (see s 14 of and Sch
13 to the 1988 Act). He might have added
that this was however subject to s 247 (‘Dividends etc paid by one member of a
group to another’), to which both ss 14 and 231 are expressly made
subject. Where a group income election
(GIE) under s 247 was in force, the dividends paid under it (election
dividends) did not give rise to an obligation for payment of ACT but were
(under s 247(2)) to be ‘excluded from sections 14(1) and 231.’ In other words, they were not to count as
qualifying distributions triggering a liability to pay ACT, nor as qualifying
distributions giving the right to a tax credit.
[99] The Dutch or Italian client would no
doubt have evinced particular interest at this point; it was because his
company had a subsidiary resident in the United Kingdom that he was interested
in how the DTAs work. If this imaginary
consultation had taken place in 1995, the adviser (unless very prescient and
knowledgeable about the European Union treaties) would have hastened to
reassure his client that he need not worry about s 247, because a GIE was
111
available only between companies resident in the
United Kingdom. The question ‘What if a
Dutch (or Italian) parent company and its United Kingdom subsidiary joined in
making a GIE and dividends were paid under it, without ACT?’ would have been
regarded as a stupid question. But the
momentous decision of the Court of Justice of the European Communities in
Metallgesellschaft Ltd v IRC, Hoechst AG v IRC Joined cases C-397/98 and
C-410/98 [2001] All ER (EC) 496, [2001] Ch 620 has since made it a meaningful
and highly pertinent question.
[100] That is the heart of the so-called
election issue, which is the first issue in this appeal. Is it an essential characteristic of ‘a tax
credit under section 231’ that it is conditional on a qualifying distribution
in respect of which ACT is payable (so that if the company paying the dividend
is not liable to pay ACT, there is no tax credit for the shareholder)? Or is the liability to pay ACT simply a
practical necessity in the domestic context of a United Kingdom shareholder
obtaining a tax credit on the payment of a dividend by a United Kingdom
company, but not a prerequisite (because of the ‘notwithstanding anything in
any enactment’ rubric) when the concept of ‘a tax credit under section 231’ is
transferred to the international context (or parallel universe) of the DTAs?
[101] Park J had little difficulty in deciding
that there was no requirement for ACT to be payable. He said of s 231(1) ([2003] STC 250 at [36]):
‘There is no requirement
there or anywhere else that ACT must have been paid, or at least must be
payable. It is probably true that the
draftsman in 1972 assumed that, in any case where a recipient of a dividend
would be entitled to a tax credit under the predecessor of s 231(1), there
would have been a liability on the part of the paying company, or of one or more
lower tier companies, to pay ACT in amounts which would match the amount of the
tax credit. Further, in all
circumstances which the draftsman could realistically have anticipated at the
time, his assumption would have been correct.
It remains the case, however, that the policy considerations which
influenced the architects of the imputation system and the assumptions which
the draftsman probably made were nowhere enacted as conditions which had to be
fulfilled before a shareholder could be entitled to a tax credit.’
The judge repeated (at [38]) the same point in relation to
art 10 of the DTAs. After summarising
submissions made by Mr Glick QC (for the Revenue) as to ss 231 and 247 he set
out his comment very pithily (at [40]): ‘So what?’
[102] The Court of Appeal (Peter Gibson, Laws
LJJ and Sir Martin Nourse: [2003] EWCA Civ 1849, [2004] STC 130) upheld the
judge’s conclusion, and again seem to have had little difficulty in doing
so. The core reasoning is at [46] and
[47] of the judgment of the court delivered by Peter Gibson LJ:
‘[46] Given the
elaborately detailed nature of the DTAs and their purpose of relieving double
taxation, we would find it very surprising if specific provisions limiting the
conferring of tax credits in the domestic tax legislation so as to exclude it
in particular circumstances were intended to govern the availability of tax
credits to a non-resident. Is it really
to be supposed that every statutory qualification enacted from time to time in
the UK fiscal legislation to the availability of a tax credit under s 231
qualifies the entitlement conferred by the DTA to the limited relief of double
taxation by the tax credit as provided for in the DTA? We think not.
112
[47] Moreover, s 247(2),
which forms the linchpin of Mr Glick’s argument, is not in fact purporting to
affect the meaning of “tax credit” in UK tax law. It is merely limiting the circumstances in
which a tax credit under s 231 will be granted.’
[103] I have found this issue much less
easy. It is to my mind a short but very
difficult point of statutory construction.
The unanimous view of the very experienced judges in the courts below
commands great respect. But in the end I
have come to the conclusion, differing most reluctantly from the courts below,
that they reached the wrong conclusion because they did not give enough weight
to two factors. One is that in applying
the DTAs it is necessary to look, not only at their terms, but also at the
language of s 788(3)(d), which uses a technical expression of domestic tax law,
‘qualifying distribution.’ The other is
that the clear scheme of the 1988 Act is that the payment of a dividend should
be accompanied by a payment of ACT if a tax credit is to come into existence,
and if exceptionally (because of a GIE) the payment of a dividend is not
accompanied by a payment of ACT, the dividend would not give rise to a tax
credit, because of s 247(2). Section
247(2) does not directly affect the meaning of ‘tax credit’, but it does to my
mind affect the meaning of ‘qualifying distribution’; a dividend paid under a
GIE is in terms excluded from s 14(1), and s 231 is in terms made to take
effect subject to s 247.
[104] These processes of exclusion and
subjection are no doubt not strictly a matter of definition (although the types
and purposes of statutory definitions are manifold: see Bennion Statutory
Interpretation (4th edn, 2002) p 479ff).
But they are a fundamental element of how a tax credit under s 247 was
intended to work under the imputation system.
A thunderbolt from Luxembourg, in the form of the decision in
Metallgesellschaft Ltd v IRC, Hoechst AG v IRC, has shown that under EU law
the statutory scheme was flawed, and has been flawed since its inception in
1973. There is no answer which resolves
all the difficulties. But in those
circumstances your Lordships should in my opinion adopt a construction which
best accords with the original scheme of the 1988 Act, flawed though it is now
seen to be, rather than abandoning the attempt to find any sort of purposive
construction.
[105] In his printed case Mr Aaronson QC (for
the respondent Pirelli companies) acknowledged that if common sense or plain
justice required it, the Revenue might be able to justify what he has called a
selective reading of ss 231 and 247. The
requirements of justice are not easy to discern in the world of cross-border
taxation of multinationals, but I think that commonsense does point in favour
of the Revenue’s appeal. The evident
purpose of s 788(3)(d) and of art 10 of the DTAs is to give a tax credit (of a
certain sort) to a non-resident shareholder who receives a dividend from a
United Kingdom company. It is central to
the concept of the United Kingdom granting a tax credit to the shareholder in
respect of a dividend that some United Kingdom tax should have been paid (or at
least payable) in respect of that dividend.
It would be an abuse of language, and contrary to commonsense, to speak
of granting a tax credit when no such tax has been paid. Moreover the DTAs would not then, in this
sort of situation, be relieving double taxation. It is the respondents, to my mind, who seek
to adopt a selective reading of the relevant provisions of the 1988 Act by
ignoring a central feature of the statutory scheme, that liability to pay ACT
is a concomitant of a qualifying distribution.
113
[106] Mr Aaronson relied on the well-known
presumption that Parliament does not intend to act in breach of public
international law, including the treaty obligations of the United Kingdom. I would accept the submission (made by Mr
Glick in his reply) that that principle does not assist the respondents, since
s 788 (like its statutory antecedents) is a general enabling provision capable
of incorporating an indefinite number of bilateral arrangements (that is, DTAs)
into our domestic tax law, but only for the purposes specified in sub-s (3),
and within the scope of those purposes (see R v IRC, ex p Commerzbank AG [1991]
STC 271 and NEC Semi-Conductors Ltd v IRC [2004] STC 489, to which I
have already referred). If s 788(3)(d)
is clear and unambiguous, its meaning cannot be altered by the text of any
particular DTA (see Salomon v Comrs of Customs and Excise [1966]
3 All ER 871 at 875–876, [1967] 2 QB 116 at 143).
[107] For these reasons I would resolve the
election issue in favour of the Revenue.
In agreement with the reasoning of Lord Nicholls, Lord Hope and Lord
Scott I would also decide the assessment issue and the withholding tax issue in
favour of the Revenue. I would therefore
allow the appeal and make the order that Lord Scott proposes.
LORD BROWN OF EATON-UNDER-HEYWOOD.
[108] My Lords, I have had the advantage of
reading in draft the opinions of my noble and learned friends, Lord Nicholls of
Birkenhead, Lord Hope of Craighead, Lord Scott of Foscote and Lord Walker of
Gestingthorpe. For the reasons they
give, with which I entirely agree, I too would allow the appeal and make the
order proposed.
Appeal allowed.
KUJUA STYLE TAMU ZA KUMKUNA MSICHANA AKAKUPENDA DAIMA ASIKUSALITI BONYEZA HAPA CHINI
0 comments: