CA-2024-001499 - [2025] EWCA Civ 1321

My conclusions on this part of the case can be summarised as follows. Many of Mr Fox’s criticisms of the judge’s approach are ill-founded. But three of them have merit. Those three criticisms are not sufficient to undermine the judge’s overall conclusion on the threshold issue of serious harm, which I think was correct. But they do have resonance and weight when it comes to the judge’s assessment of the extent of the reputational harm that was caused and the scale of her award. Her judgment on damages was careful and thorough, but in my view it was flawed. And although this court is slow to interfere with assessments of damages in this field, the awards in this case go beyond the range properly available to the trial judge. For these reasons, on which I shall now expand, I would dismiss the appeal on liability but allow the appeal on damages, reducing each award by half to £45,000.

In any defamation case the single meaning of the statement complained of plays an important role. The single meaning identifies the “charges” or imputations against the claimant which the statement would convey to the ordinary reasonable reader. The claimant must always satisfy the court that each such imputation has a defamatory tendency according to the common law tests identified at the start of this judgment. The claimant must also show, in relation to each imputation, that “its publication caused, or is likely to cause, serious harm to the reputation of the claimant”. That is one effect of s 1 of the 2013 Act (see Sube v News Group Newspapers Ltd [2018] EWHC 1961 (QB), [2018] 1 WLR 5767; Amersi v Leslie [2023] EWCA Civ 1468 [55]). The single meaning identifies the target for any defence of truth. It is relevant, albeit not decisive, when assessing a defence of reasonable belief that publication was in the public interest (see Economou v de Freitas [2018] EWCA Civ 2591, [2019] EMLR 7 [92]-[95]) or issues of malice (see Loveless v Earl [1999] EMLR 530, 538). In logic and in fairness the single meaning must also govern the nature and scope of the remedies to which the claimant is entitled if liability is established. Damages should be assessed on the basis that the claimant is entitled to be compensated for the consequences of the publication of the defamatory single meaning, no more and no less. Any other approach would lack coherence and risk injustice.

So the court will not assume, nor will it necessarily find, that the publication of the single meaning caused reputational harm commensurate with the gravity of that meaning and the extent of its publication. If the evidence showed that in fact many people interpreted the statement complained of in a sense that was not injurious, or less injurious than the single meaning, that could have a significant bearing on serious harm or on damages. Nicklin J recognised this when giving directions for the preliminary issue trial in this case ([2022] EWHC 2726 (QB) [21], [24]) and I agree. To this extent I accept the submissions on behalf of Mr Fox. I do not, however, agree that the application of these principles to this case has the consequences that have been suggested.

Although Mr Fox was trying to make a rhetorical point the High Court held that this was not clear to the reader from the words he chose or the context in which all readers viewed his posts. The claimants established that, to the ordinary reasonable reader, Mr Fox’s tweets about them bore the same single meaning. Each tweet conveyed a single factual imputation: that the claimant in question was a paedophile. This court affirmed those conclusions. The appropriate starting point for an enquiry into the reputational harm caused by the tweets was that the imputation of paedophilia was an exceptionally damaging thing to say, and its mass publication was highly likely in the ordinary course of things to cause serious reputational injury. It was not incumbent on the claimants to adduce additional evidence to establish that such injury was probably caused. It was, in principle, for Mr Fox to identify reasons and supporting evidence for reaching a different conclusion. That is the established practice in the rare case where a defendant seeks to establish a “reverse innuendo” meaning. I think the same must apply in the present situation, where the defence does not rely on innuendo facts but on additional context which is said to affect the meaning which the statement would convey to the reader. Mr Fox sought to make an inferential case to that effect. Although the judge did not address that aspect of his case expressly, she plainly did not consider that it had any material bearing on the issue of serious harm. I have not been persuaded that she was wrong in that assessment.

The judge expressly rejected the secondary submission for Mr Fox, that the claimants bore the additional burden of proving that people took the single meaning to be true. In my opinion she was right to do so. A claimant might, perhaps, bear an evidential burden of this kind in the exceptional case where the allegation is inherently likely to be discounted because it was plainly outlandish and its source lacked any credibility (as in Riches v News Group Newspapers Ltd [1986] QB 256 and Oriental Daily Publisher Ltd v Ming Pao Holdings Ltd [2013] EMLR 7). But this was not such a case. As the judge put it, this was a “not incredible” allegation. And I agree with the approach taken by the judge in her remedies judgment. To prove that a serious allegation caused serious reputational harm it is not necessary to persuade the court that it was believed to be true. A reader of the statement who takes from it the single meaning may, for instance, think “I am not sure that is true but it may be, and I am going to avoid the claimant in case it is”. That would be evidence of serious harm.

There was some direct evidence before the judge that some readers had not taken the allegation of paedophilia literally, or seriously, and that some did not believe it. The Stonewall statement, referred to by the judge, is evidence that the allegation was taken literally and seriously but rejected by that organisation. There was some other, similar evidence. This evidence was admissible and relevant. But I think Mr Fox overstates his case very considerably when he maintains that the evidence as a whole showed that all or the majority of readers fell into one or other of the categories I have mentioned, so that no serious harm was caused. The judge took Mr Fox’s points into account and, in my opinion, her treatment of this aspect of the evidence was legitimate.

I can deal quite shortly with some of the other points made by Mr Fox.

The three criticisms with which I agree are these. First, it was wrong effectively to discount all the mitigating steps taken by Mr Fox after publication. The judge should have made a substantial allowance for their cumulative impact. Mr Fox’s “language is powerful” tweet of 4 October 2020, his prompt deletion of the offending tweets, and his “baseless insults” tweet of 5 October were all steps taken shortly after the tweets complained of. All were aimed directly at the relevant constituency: Mr Fox’s own followers. They were not unequivocal retractions, and they did not contain an apology. But these steps were inherently likely to result in significant mitigation of the harm initially caused. It will have been clear to any reasonable reader that Mr Fox was saying that his earlier paedophile tweets were “baseless insults”. That message was repeated in Mr Fox’s tweet of 13 October 2020. Listeners to the interviews with Julia Hartley-Brewer on 5 October and Nick Ferrari on 13 October 2020 received the same message. Although there is no evidence as to the overlap between the audiences for those radio shows and Mr Fox’s followers, the existence of some overlap can be inferred.

Secondly, it was wrong to treat mainstream media reporting of the parties’ exchanges as evidence of additional reputational harm. A person who publishes a defamatory statement is responsible for any republication that is foreseeable as a natural and probable consequence of what they do. But republication of a statement cannot count towards reputational harm if the republication takes place in a context that draws the defamatory sting of the original statement; in such a case “the claimant must take the meaning as it emerges from the entire publication”: Economou v de Freitas [2016] EWHC 1852 (QB), [2017] EMLR 4, [17]. Here, there was no question of media reports adopting or endorsing the allegation of paedophilia. As the judge said more than once, what happened was that the media “discussed” the allegation. They did so in a way that “looked askance” at it and encouraged the public to think that Mr Fox had made an error of judgment or a poor joke. To these points can be added that the media reported what Mr Fox had said in his mitigating tweets of 4 and 5 October. On that basis, so far from counting towards reputational harm, mainstream media reporting should have been assessed as a mitigating factor, to the extent it came to the attention of those who read the tweet complained of.

Thirdly, and less importantly, it was wrong in principle for the judge to treat the Stonewall statement as evidence of reputational harm. It cannot be legitimate to reason, as she did, from the proposition that the statement showed that in Stonewall’s opinion a defence to the charge was necessary, to a finding of fact that its publication had caused serious reputational harm.

These three points do not, individually or collectively, lead me to conclude that serious harm was not made out. On the contrary, my conclusions on this part of the case reflect much of what I have said about Mr Fox’s claim. This was an imputation of criminal misconduct of a kind that revolts ordinary people. On Mr Fox’s own case he published the allegation to some 250,000 followers of his. That in itself was mass publication. Mr Fox failed to show that his tweets were not in fact understood in the way complained of, or that they were disbelieved. Mr Blake and Mr Seymour enjoyed a presumption that their pre-publication reputation in this sector of their lives was good. No attempt was made to rebut that presumption. For these reasons, each claimant had a strong inferential case that Mr Fox’s own publication of the allegation of paedophilia probably caused serious harm to the reputation of that claimant. Most of the factors relied on by Mr Fox as mitigating harm are ill-founded. The three that I have identified as well-founded are not sufficient to displace the inference of serious harm. The judge’s ultimate conclusion on the threshold is unaffected. But the three points are important when it comes to damages.

None of the three errors I have identified has any major bearing on the judge’s evaluation of the compensable distress suffered by the claimants. For the most part that stands. It does however follow from my earlier conclusions that the award should not have compensated the claimants for the emotional consequences of having to face Mr Fox’s claims. And the three points, and especially the first two, have a direct and obviously material impact on the key issue, namely the extent of the reputational harm for which the claimants are entitled to compensation.

For the reasons I have given, damages should have been assessed in respect of Mr Fox’s publication to his own followers and not any media republication. The assessment should have been carried out on the footing that upon first reading his tweets Mr Fox’s followers generally understood them to bear the single meaning identified by the court; that many took them seriously; and that they therefore caused serious harm at the time of publication. But that initial harm was substantially reduced by the steps that Mr Fox himself took on 4, 5 and 13 October 2020, as well as by the way the matter was reported in the media over that same period. These conclusions, critically, reduce the size of the award required to achieve full and adequate vindication of the claimants’ reputations.

Reputational injury and emotional harm are not easily measured in money. As the judge observed, there are no comparable defamation cases to assist in identifying the appropriate award. We do however have the conventional figures for personal injury damages. These cannot be read across directly, but they do provide relevant context. Awards for libel should not be, or appear to be, wholly out of kilter. The conventional or “tariff” awards for pain, suffering and loss of amenity are set out in the Judicial College Guidelines (17^th ed). The maximum is £493,000. For moderate brain injury, with permanent effects on concentration, memory, and the ability to work, the range is £52,550 to £110,720. For total loss of sight in one eye with reduced vision in the other the range is £78,040 to £129,330. Viewed in this light, awards of £90,000 are considerably in excess of what was necessary in this case. Substantial sums were appropriate. But awards of £45,000 each would be sufficient to achieve proper compensation and to show the world that there was nothing in the allegations.

I agree with both judgments.