SERAFIN v Malkiewicz began as an everyday action for libel. A Polish-language magazine edited and published by the defendants had alleged disreputable conduct on the claimant's part, towards London's Polish community. Jay J.'s decision dismissing Mr. Serafin's libel claim ([2017] EWHC 2992 (Q.B.)) triggered a remarkable sequence of appellate consequences.

The Court of Appeal held the judge had been wrong to uphold the defence of publication on a matter of public interest (Defamation Act 2013, s. 4). There may be nothing especially unusual in that. But it further held that the conduct of the trial had been unfair to the claimant, who had represented himself. The learned judge had “descended to the arena, cast off the mantle of impartiality and taken up the cudgels of cross-examination”: [2019] EWCA Civ 852, at [114]. The Supreme Court upheld this extraordinary finding. Their Lordships concluded, “with profound regret”, that Jay J.'s “barrage of hostility towards [the claimant], fired by the judge in immoderate, ill-tempered and at times offensive language … did not allow the claim to be properly presented”: [2020] UKSC 23, [2020] 1 W.L.R. 2455, at [48]. The logical result had to be a retrial before a different judge. Yet the Court of Appeal “did not in its judgment proceed to address the consequences that should flow from its conclusion that the trial had been unfair”: ibid., at [33]. It had ordered only an assessment of damages. The Supreme Court could not understand what precisely this required (was the damages inquiry to assume that all the alleged libels were actionable? Or only the ones held actionable by Jay J.?). The Court of Appeal's “unexplained order” could not stand and a new trial was required: [2020] UKSC 23, at [49]. To round things off, that court's detailed consideration of the public-interest defence had been so flawed that “the new judge should determine [its] availability … without reference to the reasoning which led the Court of Appeal to conclude that the defendants had [failed to meet its requirements]”: [2020] UKSC 23, at [78]. Lord Wilson, giving judgment for a unanimous Supreme Court, evidently found the case embarrassing. No wonder.

Serafin v Malkiewicz will be a leading case on unfair trials – since they are fortunately rare. But our concern here is the public-interest defence in defamation. This is no less constitutionally important. Serafin saw the first consideration of the statutory defence at the highest level. While the Supreme Court's reasoning was technically obiter, it naturally has great authority.

The House of Lords first derived a public-interest defence out of qualified privilege in Reynolds v Times Newspapers Ltd. [2001] 2 A.C. 127. The Defamation Act 2013 states: “The common law defence known as the Reynolds defence is abolished” (s. 4(6)). Section 4 enacts its replacement. The key requirements are “a statement on a matter of public interest” (s. 4(1)(a)) and that “the defendant reasonably believed that publishing [it] was in the public interest” (s. 4(1)(b)). About these the statute gives little guidance save that “the court must make such allowance for editorial judgement as it considers appropriate” (s. 4(4)) and “must have regard to all the circumstances of the case” (s. 4(2)). Unsurprisingly section 4's meaning has been contentious.

Low Kee Yang (2014) 130 L.Q.R. 24 thought “Reynolds privilege transformed”. Section 4 was “far more favourable to the defendant” than the common law defence. Specifically compared with the inquiry, founded upon Lord Nicholls’ speech in Reynolds, into whether the defendant's behaviour met standards of “responsible journalism” (see [2001] 2 A.C. 127, 202). Yang argued that Lord Nicholls’ celebrated checklist had become irrelevant: “Nowhere in s.4 is there any express indication that the defendant must have acted responsibly and it would be an unbearable stretch of interpretation to argue that responsible conduct is embedded within the concept of reasonable belief”. Yang supported this view by noting that whereas the Defamation Bill, as introduced, had originally contained a list of eight behavioural factors derived from Lord Nicholls’ Reynolds speech, this subclause was removed during parliamentary consideration and is absent from section 4 as enacted. That amendment, Yang contended, indicated a decisive break with the Reynolds approach – an “astonishing” recalibration of defamation law towards freedom of speech.

This argument has not been accepted in the courts. Economou v De Freitas [2018] EWCA Civ 2591, [2019] E.M.L.R. 7 confirms the continuing relevance of the Reynolds factors. At [101] Sharp L.J. approvingly paraphrased Warby J.'s reasoning below: “it would be hard to describe a belief as reasonable if it has been arrived at without care, without any examination of the relevant factors or without engaging in appropriate enquiries.” Despite the focus in Reynolds on the defendant's behaviour, contrasted with section 4(1)(b)'s requirement of “reasonable belief” Sharp L.J. said “it could not sensibly be suggested that the rationale for the Reynolds defence and for the public interest defence are materially different”: [2018] EWCA Civ 2591, at [86]. Accepting that the 2013 Act could have expressly incorporated the Reynolds behavioural factors but did not, those factors might still be relevant when taking “all the circumstances” into account, and potentially decisive: ibid., at [110]. After all (as noted at [77]) according to the Act's Explanatory Notes, section 4 is “intended to reflect the principles in Reynolds” (Explanatory Notes at [29]); indeed “intended essentially to codify the common law defence” (Explanatory Notes at [37]).

The Court of Appeal's decision in Serafin v Malkiewicz marks the high tide of “Reynolds continuity”. Enthusiastically approving Economou the court thought the Reynolds factors relevant to establishing the “public interest”, not just the defendant's reasonable belief about it: [2019] EWCA Civ 852, at [48]. Overruling Jay J., it held the defendants’ “standards of journalism plainly left much to be desired” (at [82]). In particular they had not put their allegations to the claimant prior to publication. The “tenacious” requirement that claimants should be given the opportunity to respond was “the core Reynolds factor” (at [66]). There were other journalistic failings too. Having analysed these, the court finally thought it “useful to consider the Reynolds factors seriatim” and did so (see [83]).

With this Lord Wilson was unimpressed. The kernel of His Lordship's rebuke ([2020] UKSC 23, at [72]) was that the Court of Appeal had misunderstood Sharp L.J.'s reasoning in Economou. Her Ladyship had correctly observed there that the rationale of the statutory defence was not materially different (see above) – however, continued Lord Wilson: “It is wrong to consider that the elements of the statutory defence can be equiparated with those of the Reynolds defence”. It followed that the Court of Appeal had been wrong in Serafin to apply Lord Nicholls’ Reynolds factors as a “checklist”. In particular, section 4 must not be “glossed” by elevating the solicitation of prior comment from the claimant into a strict “requirement”. Lord Wilson also discouraged future courts from investigating ECHR jurisprudence insofar as the very purpose of the defence is to balance claimants’ (art. 8) rights to reputation against defendants’ (art. 10) rights to freedom of speech. For good measure he noted that the Court of Appeal had misattributed both of the Strasbourg authorities it cited to the European (Union) Court of Justice.

Lord Wilson did not formulate any elaborate positive account of the defence. His judgment focused on counsel's criticisms of the two Court of Appeal authorities (on behalf of the defendants and the Media Lawyers Association, interveners). As seen, His Lordship accepted criticisms of the decision below in Serafin v Malkiewicz yet his view of Economou v De Freitas was broadly positive. He did however identify an understatement in Sharp L.J.'s remark (above) that Parliament could have incorporated the Reynolds list but did not. As L.K. Yang previously emphasised (also above), such a list had actually appeared in the original Bill. It was deleted from section 4's final form. Lord Wilson traced the parliamentary history (quoting extensively from Hansard). Given the list's removal Lord Wilson lamented that the Act's Explanatory Notes describe the “codification” of Reynolds (“a strong word”). He traced this mischaracterisation to the “unfortunate” failure to update a passage which had referred to the original Bill, with its Reynolds-inspired list of factors.

The resulting law sits between two extremes. The public-interest defence cannot be analysed as if Reynolds still held sway – the Court of Appeal's apparent error in Serafin. But nor has the Supreme Court approved the converse suggestion that the behavioural factors identified in Reynolds are irrelevant. The middle course is surely the correct interpretation of the statutory text. Since a defendant's belief in public-interest publication must be reasonable, her conduct (e.g. in verifying the information) has to be relevant. It is manifestly part of “all the circumstances” (s. 4(2)). But all depends on the facts of each case. How the test is framed, in the abstract, is important. Ultimately however, its application to particular fact situations is even more so. Everyone can agree that “responsible journalism” is essential and “fake news” deplorable (see Economou [2018] EWCA Civ 2591, at [109]); but free speech is equally vital. Section 4, like Reynolds before it, merely restates the intractable conflict. Its resolution requires sound judgement, and judgments.