Case Nos: CA-2024-000713 - [2025] EWCA Civ 1271

The Revenue appealed on four grounds, namely:

To the extent that any challenge is made to the Judge’s fact-findings based on his evaluation of the evidence, as Lewison LJ said in Volpi v Volpi [2022] EWCA Civ 464 at [2]: “the approach of an appeal court to that kind of appeal is a well-trodden path.” He then set out a series of well-settled principles derived from numerous authorities of the Supreme Court and Court of Appeal, including Henderson v Foxworth Investments Ltd [2014] UKSC 41; [2014] 1 WLR 2600 and Perry v Raleys Solicitors [2019] UKSC 5; [2020] AC 352, which are worth repeating in the light of the nature of the Revenue’s case, particularly on Grounds 3 and 4:

Lewison LJ went on to say at [4]:

“4.

Similar caution applies to appeals against a trial judge’s evaluation of expert evidence: Byers v Saudi National Bank [2022] EWCA Civ 43, [2022] 4 WLR 22. It is also pertinent to recall that where facts are disputed it is for the judge, not the expert, to decide those facts. Even where expert evidence is uncontroverted, a trial judge is not bound to accept it: see, most recently, Griffiths v TUI (UK) Ltd [2021] EWCA Civ 1442, [2022] 1 WLR 973 (although the court was divided over whether it was necessary to cross-examine an expert before challenging their evidence).”

Grounds 1 and 2 are inextricably linked. It was accepted by Mr Ewart KC, on behalf of the Revenue, that the Judge correctly set out the test articulated by the Supreme Court, but he contended that in substance he failed to apply it. Four points were made in support of that submission, namely:

There is no need to dwell on criticisms (iii) and (iv) of Ground 1, which were not developed by Mr Ewart in his oral submissions. The Judge used “constructive discovery” as a shorthand for the point at which the claimants could with reasonable diligence have discovered their mistake, as he stated at [18] (i.e. what is referred to elsewhere as “the date of discoverability”). He made it clear at [21] vi) that the question was not whether the mistake should have been discovered earlier, and he did nothing in his judgment that would indicate that he had treated that as the test. Whilst it might have been better if the Judge had continued to use “date of discoverability” instead of choosing terminology which means something different in another context, there was no confusion about what he meant. Indeed, there are examples of similar language being used within FII SC 2 itself, see e.g. [195] and [236]. The Judge was entitled to approach the matter in the way in which he did.

As to the burden of proof, the Judge correctly directed himself at [21] ii) and applied that direction consistently throughout the judgment, see for example [117] and [129].

I shall therefore turn to consider the remaining facets of Ground 1, which were developed orally by Mr Ewart in conjunction with Ground 2.

Having set out at [21] iv) a) and b) the two ways in which the Supreme Court had put the test at FII SC 2 [193], the Judge stated in terms that he was not going to give either any greater significance than the other. He was as good as his word (as demonstrated for example at [139] and [140]). There was no need for the Judge to carry out a detailed analysis of what was meant by a “worthwhile” claim, and there would have been obvious dangers in seeking to paraphrase a test that has been laid down by the Supreme Court. All the sub-paragraphs of [21] indicate that he understood very well what the Supreme Court meant, and that he was not equating “worthwhile” with having a good prospect of success (let alone with thinking you were going to win).

Perhaps the simplest and clearest explanation of what the Supreme Court meant by a “worthwhile claim” appears in FII SC 2 at [178]:

“It is therefore possible to investigate how legal thinking on a particular question (for example, in the present case, whether the UK tax treatment of dividends received by UK-resident companies from non-resident subsidiaries was compatible with EU law) developed over time, and to ascertain, by means of evidence, the time by which a reasonably diligent person in the position of the claimant (such as, in the present case, a UK-based multi-national company) could have known of a previous mistake of law, to the extent of knowing that there was a real possibility that such a mistake had been made, and that a worthwhile claim could therefore be made on that basis…” [emphasis added].

That is precisely the exercise which the Judge carried out.

At one stage the Revenue sought to equate a “worthwhile claim” with “a claim that is worth investigating,” though Mr Ewart resiled from that position to an extent in his oral submissions. He was right to do so, as that is plainly not what the Supreme Court was referring to. They meant a claim which is worth raising/pursuing (in the sense of being legally, as opposed to economically, viable). The mere identification of a possible legal argument that the payment had been made under a mistake of law would not suffice. A claimant would not have sufficient confidence to embark on the preliminaries to litigation or to submit a claim to HMRC for repayment of the tax, if he or his advisers had not yet taken any steps to investigate whether the argument would even get off the ground.

Any scintilla of doubt about this is dispelled by the express espousal by the Supreme Court in FII SC 2 of what was said by Lord Walker at an earlier stage of this litigation, when he referred to the time when “a well-advised multi-national group based in the UK would have had good grounds for supposing that it hada valid claim to recover ACT levied contrary to EU law” (see FII SC 2 at [211] and [255]) [Emphasis added]. It would not have good grounds for supposing that it had a valid claim if it was told by an Appropriate Adviser that the argument was hopeless or that such a claim was vulnerable to being struck out. To the extent that the Revenue’s arguments under either Grounds 1 or 2 involved the proposition that the Appropriate Adviser would be tasked with giving advice upon whether a claim was worth investigating to see if it was worthwhile pursuing or raising with the Revenue, they were based on a fallacious premise.

Although the Supreme Court referred to a “well-advised multi-national group” without spelling out what was meant by “well-advised”, it is self-evident that in order to discover whether there was a “real possibility” that a payment of tax was made under a mistake of law, (or that the basis for the payment was “legally questionable”) and thus recognise that a claim would be worthwhile pursuing, or to feel confident enough about it to embark upon the preliminaries to litigation, one would need to take legal advice from a suitably qualified lawyer or lawyers. That would be part and parcel of the exercise of reasonable diligence.

Consequently, when in FII SC 2 Lord Reed and Lord Hodge referred to the time at which a claimant, having exercised reasonable diligence, would have had sufficient confidence to justify embarking on the preliminaries to the issue of proceedings such as submitting a claim to HMRC, taking advice and collecting evidence, the “advice” to which they were referring must have been something other than the legal advice that is a necessary ingredient of the exercise of reasonable diligence. The advice at the “preliminaries” stage would no doubt include advice as to the prospects of success, possibly involving some form of risk/benefit analysis, which would require a more in-depth analysis and consideration of the merits of the legal argument, as the Judge envisaged at [32].

Limb (ii) of the Revenue’s criticism under Ground 1 involved the type of pedantic textual analysis of the judgment that was rightly deprecated in Volpi v Volpi. Upon carrying out that analysis it was also demonstrably unjustified. There are references in the judgment to what the claimants “would” have known or “would” have been advised, but it is clear from reading the judgment as a whole that the Judge applied the correct test, and that there was no confusion about what the claimants had to prove, namely, that they could not have discovered (by exercising due diligence) that they had a worthwhile claim earlier than a particular date. He did not require the claimants to have some greater degree of certainty about the truth of the law than the Supreme Court did.

If one asks, what could have been found out by a claimant exercising reasonable diligence? the answer is: what an Appropriate Adviser would have told them (and therefore what they would have discovered about the viability of the claim in consequence of that advice). Lord Reed and Lord Hodge themselves refer to “the point in time when the test claimants could with reasonable diligence have discovered, to the standard of knowing that they had a worthwhile claim, that they had paid tax under a mistaken understanding that they were liable to do so” (FII SC 2 at [255]). [Emphasis added].

In determining whether the well-advised multi-national could have discovered with reasonable diligence that there was a worthwhile claim, or (to use the alternative formulation) whether it would have had sufficient confidence in a prospective ACT Challenge or DV Challenge to embark upon the preliminaries to litigation, it was necessary to investigate what (initial) legal advice it would have received at the relevant time from an appropriately qualified adviser or advisers who were sufficiently expert and competent to be able to provide it – to use the Judge’s shorthand, an Appropriate Adviser – and therefore what the claimant would have known in consequence of taking that advice.

As the Supreme Court recognised, that in turn would depend on what such an adviser would have known and believed about the state of the law, and what the state of professional thinking (both among academics and practitioners) was at that time, without using hindsight based on later developments in the law and in legal thinking. The Judge was astute to avoid this trap. As he said at [28]:

“It is all too easy to make assertions as to what could have happened in the light of knowledge of the ultimate outcome of the litigation. Assertions as to what a well-advised multi-national could have discovered between 20 and 30 years ago need to be tested carefully against evidence of what people were thinking and doing at the time.”

How legal thinking developed over time, and therefore what the law was thought to be at a particular time, are questions of fact. One difficulty in the present case was that it was established on the evidence that nobody in the period with which the court was concerned actually considered the specific legal issue which, for these purposes, the court had to assume the multi-national company would have asked the Appropriate Adviser to advise it upon.

If such advice had been sought, in order to answer the question, a competent adviser would have needed to carry out the type of research that all lawyers would when asked to give an initial opinion on a point of law. They would no doubt have looked at the leading textbook, Farmer & Lyal on EC Tax Law, at academic writings, and at such of the case law as they considered to be relevant, to ascertain whether there was a viable argument that the provisions under which the charges to tax were made were incompatible with EU law, or whether the carve-out which gave Member States a wide degree of autonomy over matters of this nature would apply. On the Judge’s findings, they would have found nothing in any commentaries or textbooks to support the contention that the provisions in question were incompatible with EU law, and some limited evidence pointing the other way. He found that, as at July 1996 there was no obvious “general direction” in the CJEU’s jurisprudence [80] – [90] (an assessment with which I respectfully agree). Against that background, the question of what the Appropriate Adviser would have told the multi-national company is a question of fact.

The Judge accepted the evidence of Mr Gammie that there was a professional consensus that there were three fundamental difficulties facing any argument of this nature, which he identified at [71]. Based on that evidence the Judge found as a fact at [70], reiterated at [85] and [106], that in July 1996 an Appropriate Adviser would have advised that there was no worthwhile claim. He was entitled to do so, based on the evidence of the state of legal thinking at that time. Mr Gammie was particularly well-qualified to give direct evidence on that subject and about how (if at all) legal thinking developed in the light of the various relevant decisions of the CJEU. He had been a partner in Linklaters & Paines for ten years, specialising in tax matters (particularly cross-border corporate taxation) before leaving in 1997 to join the tax bar. He also had a distinguished academic career in the UK and abroad, with expertise in tax law and European law. He had taught and written on those subjects during his period as Professor at the University of Leiden between 1998 and 2019. He also sat as a fee-paid judge in both the First-Tier Tribunal and Upper Tribunal (Tax Chamber).

The Judge did not just take Mr Gammie’s evidence about the professional consensus at face value, but tested it rigorously against all the contra-indications argued for by the Revenue. He found support for it in a passage in Chapter 16 of the contemporaneous edition of Farmer & Lyal (a Chapter entitled “The Prohibition of Discrimination and the Fundamental Freedoms: Their Impact in the Sphere of Direct Taxation”) which he quoted at [81]; and in the Ruding Committee’s report, as well as to some extent in the Parent/Subsidiary Directive, with which the UK’s system was compliant, and which he addressed at [84].

All these sources supported Mr Gammie’s evidence of a general perception (even among the distinguished members of the Ruding Committee) that the problem of double taxation arising from conflicts of tax jurisdiction must be resolved by convention or harmonisation. The remit of the Ruding Committee was to consider every means by which the frictions (i.e. detrimental effects on business) of cross-border flows could be reduced, and yet there was no suggestion in its report that the answer lay in the Treaty freedoms, but rather, an assumption that there would have to be Community-wide measures. Whilst it is true that case-by-case challenges based on the Treaty freedoms would not give rise to a Community-wide solution to the problem, that does not blunt the force of the point that the Ruding Committee report does not suggest that the discriminatory taxation of dividends from profits earned in another Member State was contrary to EU law. Insofar as the Revenue sought to suggest otherwise, I respectfully disagree. What the Judge found was also consistent with what was said at [34] of FII SC 2.

In those circumstances, the Revenue faced an uphill struggle in establishing that the Judge was “plainly wrong” to make the fact-findings that he did. They sought to get around this difficulty by criticising both the Judge’s use of the concept of an Appropriate Adviser and his findings as to the approach that the Appropriate Adviser would have taken. It is convenient to consider these aspects of Ground 1 in conjunction with Ground 2 of the Grounds of Appeal, which contends that the Judge adopted a “flawed approach” to how an Appropriate Adviser would have behaved and that this led to a mistake in his analysis of the date of discoverability.

At one stage in his oral submissions Mr Ewart went so far as to describe the Judge’s findings at [58] to [65] as to what the Appropriate Adviser would know or believe, as “legally irrelevant” to the question whether the mistake of law could have been discovered with reasonable diligence. He submitted that the Court was in as good a position as the Judge to evaluate the case law of the CJEU and decide, based on the principles established at the time, and such contemporaneous commentaries as existed, whether there was a viable argument that the Relevant UK Law was incompatible with EU law. He asserted that there should be no difference between what a judge would conclude was the legal position at the time and what the well-advised multi-national would have been advised at that time.

Quite apart from the fact that this argument appears to me to go well beyond the scope of the Grounds for which permission to appeal was granted, it flies in the face of the evidence-based evaluative exercise which the Supreme Court had in mind at FII SC 2 [178] and [255] which, as I have said, was the exercise which the Judge carried out. If Mr Ewart’s submission were right, the Supreme Court would have carried out the exercise for itself. Instead, it made it clear that this was a matter which could not be decided in the abstract and which required evidence of the very nature that Mr Gammie gave, and was pre-eminently qualified to give.

Mr Ewart took issue with the Judge’s finding at [32] that an Appropriate Adviser would have been “guided by a high-level appreciation of relevant legal principles” and argued that any such adviser would have conducted an analysis based upon the cases already decided by the CJEU at the relevant time, particularly Commission v France, Case C-270/83 [1986] ECR 273 (“Avoir Fiscal”), R v Commissioners of Inland Revenue, ex parte Daily Mail, Case C-81/87 [1989] QB 446 (“Daily Mail”); R v Inland Revenue Commissioners, ex parte Commerzbank,Case C-330/91 [1994] QB 219 (“Commerzbank”) and Bachmann. But the Judge was not suggesting that the Appropriate Adviser would not have considered the case law. The point he was making was that at the stage with which he was concerned - i.e. before one got to the preliminaries to litigation - where the advice which was being sought was designed to enable the claimant to decide if it was worthwhile embarking on those preliminaries, the Appropriate Adviser would not have carried out the same in-depth analysis of vast quantities of wide-ranging materials that the experts carried out. The Adviser would have endeavoured to extract the relevant principles from the decided cases and apply those principles in order to decide whether the basis of the payment was legally questionable. There is nothing objectionable about that approach.

The Revenue also contended, as they did below, that the consensus referred to by Mr Gammie was only in the UK and that it was relevant to have regard to thinking elsewhere in the EU. Quite apart from the fact that the members of the Ruding Committee were not all from the UK, and the editors of Farmer & Lyal are distinguished EU lawyers, the Judge answered that point definitively at [86]. He found that even if the well-advised multi-national based in the UK would have sought advice from EU lawyers as well as UK tax specialists (albeit, I interpolate, that they could have approached Mr Gammie, who was qualified to give advice both on tax and on relevant EU law) the overall advice it received would inevitably have to take into consideration the specific detail of the UK tax regime in order to determine the question of comparability, and that would engage the professional consensus. He was entitled to make those fact-findings, which cannot be disturbed on appeal.

In any event, the Judge did address all of the cases relied upon by the Revenue in his judgment. He referred to Avoir Fiscal, Daily Mail and Commerzbank at [63] and Bachmann at [71], and he reached a view that was properly open to him as to what could be usefully extracted from them. He had well in mind the evidence of the Revenue’s expert Professor Barnard, which sought to draw certain conclusions from those and other authorities, which he expressly rejected at [67] and [68] and [89] to [93], preferring the evidence of Mr Gammie. He found nothing in the authorities relied on by the Revenue which would have supported the argument that the payments of tax by these claimants were made under a mistake of law. It was not good enough to say that the authorities demonstrated that direct taxation might in some circumstances have an impact on Treaty freedoms. That, in itself, would not be enough to establish the incompatibility of the Relevant UK Law with EU law.

The fundamental problem with the analysis of the cases upon which the Revenue relied was that there was nothing in any of those cases which suggested that in the context with which we are concerned, the situation of a non-UK EU resident subsidiary paying a dividend to a UK parent was comparable with or might be compared to that of a UK subsidiary paying a similar dividend (as the Judge pointed out at [90]). In the DV cases the foreign dividend income had been subject to overseas tax, not to UK tax. In the ACT cases the income in the UK had been subject to ACT but the non-resident income had not been.

The court was referred to passages in Mr Gammie’s cross-examination in which Mr Gammie not only took the clear view that the situations were not comparable but stated that he was not just reflecting his personal view but reflecting his understanding from all his contacts (both in practice and other contacts) as to what the general view would have been at the time. Mr Gammie’s evidence was that if someone hypothetically asked the Appropriate Adviser: “have I paid tax in circumstances where I shouldn’t have paid tax?” the answer would be “no, because I don’t think the situation is comparable.” In other words, an argument that the Relevant UK Law was incompatible with EU law was not regarded at the time by the body of qualified professionals who would have been approached to give advice to a UK based multi-national group as a valid basis for making a claim or embarking on the preliminaries to litigation.

Moreover, Mr Gammie’s evidence did not focus on the UK to the exclusion of other jurisdictions. The body of materials appended to his report covers an extensive and impressive range of European jurisprudence. Whilst he accepted that legal thinking in terms of the potential impact of EU law on matters of direct taxation was more advanced in Europe than it was in the UK, Mr Gammie said in his report that he had “not detected in my papers or in the literature at the time any specific suggestion that the ECJ might eventually decide this is a breach of the Treaty”.

Having weighed up the competing expert evidence, considered the documentary evidence, and considered the case law relied on, the Judge was entitled to conclude as he did. He was entitled to find that neither of the ways in which the test was formulated by the Supreme Court was satisfied, at a time when the professional consensus was that a claim of that nature was not viable, for the clear and cogent reasons given by Mr Gammie.

The conclusion I have reached above is not just enough to warrant dismissing this appeal on Grounds 1 and 2 but also provides a complete answer to Grounds 3 and 4.

Ground 3, the head-on rationality challenge, was always ambitious, and on application of the principles adumbrated in VolpivVolpi it is simply unarguable. There was a rational basis for the Judge’s findings about the existence of the consensus, namely, the evidence of Mr Gammie. I do not accept Mr Ewart’s submissions to the effect that the Judge misunderstood or mischaracterised Mr Gammie’s evidence. The fact that the professional consensus, to which Mr Gammie himself subscribed, was not recorded in contemporaneous documentation did not mean that it did not exist. The weight to be ascribed to Mr Gammie’s evidence, and which of Mr Gammie’s views he accepted, were matters for the Judge. The Revenue’s submissions were tantamount to a contention that he should have preferred the evidence of Professor Barnard; but that ship has sailed.

There was also a rational basis for the Judge’s fact-finding that the professional consensus was undermined by Verkooijen, namely, that (regardless of whether any commentators said so at the time) the reasoning of the CJEU struck at the heart of the three limbs of the consensus the Judge had described at [71] for the reasons he set out at [133]. In simple terms, the CJEU had dismissed (or brushed aside) all the legal obstacles to making such a claim which previously the Appropriate Adviser would have regarded as fatal – perhaps most significantly, comparability. It had also put very stringent limits on the extent to which the principle of cohesiveness applied in Bachmann would afford a justification for any impediment to the Treaty freedoms caused by the differences in treatment. That would have been enough, on the Judge’s findings, to embolden a well-advised multi-national sufficiently to embark upon the preliminaries. It (and its advisers) would now be aware that the claim was unlikely to fall at the first hurdle. Thus it did not matter that Mr Gammie did not claim that the decision in Verkooijen dismantled the consensus, nor that the situation with which Verkooijen was concerned was factually very different. I will consider those differences in more detail when I address FCE’s cross-appeal.

The Judge was entitled to conclude that the CJEU judgment in Verkoojen meant that there was no longer a basis for regarding the three pillars of the consensus described by Mr Gammie as insurmountable obstacles to a claim. His finding that this was the date of discoverability was also consistent with the evidence about the time when the possibility of making a challenge to the lawfulness of these tax payments began to be considered by well-advised multi-nationals based in the UK.

In a passage in FII SC 2 dealing with Henderson J’s judgment in FII HC 1, the Supreme Court referred at [33] to evidence from a tax manager at BAT, Mr Cohn, that “the first time we considered that the denial of [FII] treatment of foreign dividends might be a breach of EC law was when we discussed internally the Verkooijen judgment shortly after it was published on 6 June 2000.” As Mr Bremner KC accepted when it was put to him by my Lord, Lord Justice Nugee, in the course of oral argument, that was a reference to an ACT Challenge, although he pointed out that the head of taxation at BAT, Mr Hardman, focused more on a DV Challenge in his evidence. Lord Reed and Lord Hodge went on to describe the evidence of the BAT witnesses, which Henderson J accepted in FII HC 1, as consistent with other evidence. All that evidence supports the Judge’s conclusion as to the date of discoverability, at the very least so far as a DV Challenge is concerned.

However, the reasoning which the CJEU deployed in Verkooijen also emerges clearly from the Advocate General’s second Opinion in that case. Whilst the CJEU does not always accept the views of the Advocate General, it often does. It is important to bear in mind that the Supreme Court in FII SC 2 has made it clear that the date of discoverability is not the date on which a claimant knows he will win, or even that he is likely to win the argument, it is the date on which, if he takes the right advice, he would become aware that the payment is legally questionable, in the sense that there is a viable argument that the Relevant UK Law contravened EU law, (and would have sufficient confidence to embark upon the preliminaries to litigation).

If the Appropriate Adviser had read that Opinion in conjunction with the Advocate General’s first Opinion, they must have appreciated that there was a viable argument that, at least so far as differential taxation of dividend income was concerned, the positions of a foreign subsidiary and a domestic subsidiary could be regarded as comparable and that the principle of cohesiveness would provide no justification for the difference in treatment. It was viable because the Advocate General, far from dismissing it out of hand, had espoused it. It was foreseeable that the CJEU might agree with him, and on that basis a challenge to the discriminatory tax treatment of dividend income from an EU subsidiary would have been worth pursuing. That is why I consider that on the application of the Judge’s own reasoning, the date of discoverability was December 1999 rather than June 2000.

Ground 4 can be dealt with even more shortly. One starts with the assumption that an appellate court must make that the judge in the lower court has taken the whole of the evidence into consideration, absent compelling evidence to the contrary. Here, there is no need to rely on that assumption because the Judge said at [31] that he had had full regard to all the evidence and submissions of the parties but that he was not going to deal in any great detail with evidence and submissions that appeared to him to be unduly influenced by hindsight. There is no reason to disbelieve him.

In the Revenue’s written submissions, this criticism was developed into a complaint that the Judge did not consider the state of the case law in 1996, in particular Avoir Fiscal, Daily Mail, Commerzbank and Bachmann. That is incorrect; the Judge did consider those cases, but he preferred Mr Gammie’s evidence about what lawyers thought about them, to that of Professor Barnard.He rejected, as unduly influenced by hindsight, her argument that it would have been apparent to an Appropriate Adviser at the time that those cases would support a claim that these payments were legally questionable. He based his view, as he was entitled to, on Mr Gammie’s evidence as to what the professionals who would have been approached to give such advice to a UK based multi-national were actually thinking at the time. There is no substance in the complaint that the Judge failed to have regard to all the evidence.

For those reasons each of the challenges to the Judge’s judgment fails, and I would therefore dismiss the appeal.