This appeal is about how an Employment Tribunal should decide an application to restrict open justice by prohibiting the public disclosure of information deployed in the proceedings, where it is said that such disclosure would be contrary to the interests of justice, endanger personal safety, infringe human rights, and breach contractual rights to confidentiality.

Proceedings in the Employment Tribunal (“ET”) are subject to the strong common law principle that justice should be administered in public and fully reportable save in certain limited circumstances. The circumstances that are relevant for present purposes are where restrictions on transparency are necessary to secure the proper administration of justice or are provided for by statute.

Litigants in ET proceedings also enjoy rights under Article 6 of the European Convention on Human Rights (“the Convention”). Article 6 entitles a person whose rights are at issue in civil proceedings to “a fair and public hearing” from which the press and public can only be excluded in certain prescribed and limited circumstances. These rights sit alongside those conferred by Article 10(1) of the Convention. This guarantees freedom of expression: relevantly, the right to impart and receive information without state interference unless that interference is necessary in pursuit of one of the legitimate aims identified in Article 10(2). Article 10 rights are enjoyed not only by litigants but also by those who wish to observe legal proceedings. The Convention rights, including those under Articles 6 and 10, are enforceable domestically by virtue of s 6 of the Human Rights Act 1998 (“HRA”), which prohibits a court or other public authority from acting incompatibly with the Convention.

Restrictions on open and public justice in particular cases have come to be known as “derogations”.

The application for derogations in this case relied on rule 50 of Schedule 1 to the Employment Tribunals (Constitution and Rules of Procedure) Regulations 2013 (“Rule 50”). Rule 50 provides that an ET:-

The Convention rights that are relevant for this purpose in this case are Articles 2 (right to life), 3 (prohibition on inhuman and degrading treatment), 5 (right to liberty and security of person), 6 (right to a fair trial) and 8 (right to respect for private life).

Section 10A to the 1996 Act provides that:-

Rule 50(1) therefore identifies three grounds on which a derogation from open and public justice may be made: the interests of justice, the protection of a person’s Convention rights, and the protection of confidentiality. Clearly, more than one could apply in a particular case. All three arise for consideration in this appeal.

Rule 50(2) provides that an ET considering whether to make an order under the Rule “shall give full weight to the principle of open justice and to the Convention right to freedom of expression”.

Rule 50(3) contains illustrative examples of orders that can be made under the rule. One of these is “an order that the identities of specified parties, witnesses or other persons referred to in the proceedings should not be disclosed to the public, by the use of anonymisation or otherwise …” In this case the application was to anonymise “other persons” who were neither parties nor witnesses.

The first respondent (“Millicom”) is a member of a group of companies that provide digital services to emerging markets in Latin America and Africa. From 2017 Millicom employed the appellant (“Mr Clifford”) as a global investigations manager. His role was to conduct and oversee internal investigations into suspected wrongdoing in the group’s operations. In November 2019 Millicom dismissed Mr Clifford on grounds of redundancy. He then brought proceedings against Millicom in the ET complaining of ordinary unfair dismissal, automatic unfair dismissal, detriment on the grounds of protected disclosures, and disability discrimination. He also sued the three individual respondents on the grounds that they were fellow workers who were involved in subjecting him to the detriments of which he complained. I shall call the respondents collectively “the Millicom parties”.

For the purposes of this appeal Mr Clifford’s key allegations are that he was subjected to detriment and dismissed because of whistleblowing activity. In September 2017, he reported to Millicom that his investigations in a foreign country had revealed that staff of a Millicom subsidiary had tracked the mobile phones of a customer who was a prominent citizen in that country and disclosed their findings to a government agency there. The prominent citizen had later been the victim of a very serious criminal offence. Mr Clifford’s case is that all the Millicom parties treated him unfavourably and Millicom ultimately dismissed him for investigating these matters and reporting them to Millicom.

I have not named the customer or the subsidiary company or given details of the attack. Mr Clifford’s account of things is not public knowledge. But he set it all out in his claim documents. The Millicom parties then applied to the ET for an order under Rule 50 prohibiting the public disclosure or reporting of the identity of the customer, details of the attack, the alleged link between the attack and the Millicom company and its staff, or anything that was likely to lead to the identification of those matters. They proposed that any reference to any of those matters should be by way of a code, such that (for instance) the customer would be referred to as “Person X” the offence as “Event Y” and the foreign country as “Country Z”.

The Millicom parties’ case was that such an order was necessary in the interests of justice and/or to protect rights under Articles 2, 3, 5, 6 and 8 of the Convention and/or because Mr Clifford owed Millicom a contractual duty of confidence the breach of which would not be justified in the public interest. Evidence was filed in support of those contentions. This included assertions as to the risks to which Millicom employees would be exposed if the information was made public, and a statement by the second respondent (“Mr Frechette”) that those risks were such that if the ET declined to make the order sought, he would not be willing to give evidence or to permit Millicom to defend the proceedings. Mr Frechette is Vice President Legal – Corporate of the Millicom group. It has not been in issue that he has control over how Millicom deals with the litigation.

By the time the application was heard the Millicom parties had abandoned reliance on Article 2. But they maintained reliance on the interests of justice and the other Convention rights I have mentioned, contending that disclosure was “likely to put at serious risk the safety and security of current and former Millicom employees” including the individual respondents and Mr Clifford himself. At a late stage the Millicom parties added reliance on confidentiality, submitting a copy of Mr Clifford’s contract of employment containing a confidentiality clause.

On 23 October 2020 the ET dismissed the application. Employment Judge Henderson (“the EJ”) held that (1) the ET had no jurisdiction under Rule 50 to protect the Convention rights of individuals who are outside the jurisdiction of the signatory states; (2) if and to the extent that the Millicom parties were entitled to rely on the rights under Articles 3, 5 and 8 of the Convention their case did not satisfy the applicable threshold tests as there was no “objective evidence” to support it; (3) Mr Frechette’s evidence that he would not be a witness or allow Millicom to defend the proceedings if no derogations were granted was legally irrelevant, and there was nothing else that outweighed the open justice principle or the Article 6 right; and (4) although Mr Clifford owed Millicom a contractual duty of confidence this could not outweigh the open justice principle.

The Millicom parties obtained permission to appeal. On 11 May 2022 Mrs Justice Eady, the President of the Employment Appeal Tribunal (“EAT”), allowed the appeal in part. Eady P upheld the EJ’s conclusions that the foreign staff did not enjoy Convention rights, and that the evidence did not demonstrate a sufficiently verified objective basis for the alleged risks to safety and security. But she concluded that the EJ had erred by (1) confining her analysis of the case to consideration of rights under Articles 3, 5 and 8 of the Convention without considering whether the evidence justified an order in the interests of justice at common law and/or under Article 6; (2) failing to consider whether the respondents’ subjective fears might be enough to engage Article 8; (3) failing to conduct a proper fact-specific balancing exercise; and (4) failing to address the question of whether it was in the public interest for the duty of confidence to be breached by disclosure within the proceedings.

By paragraph 2 of her order, Eady P directed that the application be remitted for re-determination by a differently constituted ET in accordance with the principles laid down in her judgment. By paragraph 3, she ordered that the ET determining the remitted application should consider in particular, the following questions at the date of its determination:

Mr Clifford appeals to this court by permission of Simler LJ on the grounds that Eady P was wrong (1) to interfere with the ET’s properly reasoned conclusion that Article 8 was not “engaged” so that there was no need for any balancing exercise; and (2) to find that the ET should have undertaken any further or different balancing exercise in relation to the confidentiality clause. Mr Clifford’s case is that these points would be enough to dispose of the appeal because on a proper analysis everything turns on the question of whether Article 8 is “engaged”; if not, there is no separate “interests of justice” issue to be considered.

The Millicom parties take issue with that suggestion, maintaining that the interests of justice are a distinct ground for derogation. They support Eady P’s decision to remit the case for a decision on the “interests of justice” issue both for the reasons she gave and on the additional ground that (contrary to the EJ’s view) a party’s willingness or otherwise to give evidence or take part in proceedings is a legally relevant factor. The respondents were also given permission to cross-appeal. They argue that Eady P should have held that the Millicom parties’ fears, “which the ET accepted were genuinely held”, (a) had such a clear objective basis that the ET was perverse to find otherwise and (b) were in any event enough to show that Article 8 was “engaged”. The Millicom parties maintain that the right order is to remit the application for re-determination on this wider basis.

These points and others have been skilfully and helpfully developed in written and oral argument by Mr Callus on behalf of Mr Clifford and by Mr Hickman KC for the Millicom parties. Having reflected on their submissions, I have concluded that Eady P was right to order the redetermination of all three issues, that is to say whether the derogations sought were necessary (a) in the interests of justice and/or (b) to protect Mr Frechette’s Article 8 rights and/or (c) to protect rights of confidence. I would therefore dismiss the appeal. I have however concluded that the evidence placed before the EJ provided a sufficient objective basis for the fears expressed by the Millicom parties and the ET and EAT both erred in finding otherwise. I would allow the cross-appeal.

The first issue raised by the Rule 50 application was whether the derogations sought were “necessary in the interests of justice” within the meaning of the rule. The ET was required to consider whether that was so either because publicity would place the individual respondents and/or other Millicom employees at risk of serious harm or because the Millicom parties were so concerned about the risk of such harm that if no order was made the company would concede the claim or its position would be weakened by Mr Frechette’s refusal to give evidence.

The EJ began her consideration of this part of the case under the heading “Common law”. And she started (at [103]) by identifying open justice as a “fundamental principle” of the common law to be departed from only in cases of “strict necessity”, the burden of proof being on the Millicom parties. But her next step was to conduct an analysis of the case under Articles 3, 5 and 8 of the Convention. She identified the threshold tests for proving a risk of interference with the absolute rights under Articles 3 and 5. She held that those tests were not satisfied. There was, said the EJ, “no objective evidence presented to the Tribunal to support the fears expressed by Mr Frechette” that disclosure would place Millicom staff at risk of harm within the scope of Articles 3 and 5. She went on to hold that Mr Frechette’s subjective fears were insufficient to engage Article 8 so that she did not need to consider “the balancing test of proportionality under Article 8(2)”. She moved from this to the conclusion (at [110]) that “both under the common law and under Convention rights, if applicable, the respondent has not demonstrated the clear and cogent evidence necessary to establish a departure from the principle of open justice.”

The next section of the EJ’s judgment was headed “Article 6”. The EJ referred back to what she had said earlier about Mr Frechette’s evidence that in the absence of the derogations sought he would not give evidence or allow Millicom to defend the case. She had said (at [96]): “I accept Mr Callus’s submission that I should not take [that evidence] into account as part of my rule 50 deliberations … I do not intend to allow his intentions with regard to continuing with the proceedings, to influence me in reaching my decision.” The EJ noted that Article 6 allows for the exclusion of the press and the public “where the protection of the private life of the party so requires or to the extent strictly necessary in the opinion of the court in special circumstances where publicity would prejudice the interests of justice”. But she said that a refusal to make the order sought would not deprive the Millicom parties of their rights to a fair trial under Article 6. They would “still have access to a fair hearing but it would be their choice whether to proceed with it.” She concluded that the evidence had failed to establish any “special circumstances or risk to the private life of the parties” such as “to outweigh the principle of open justice”.

In my judgment, there are five flaws in this approach.

The first and main difficulty with the EJ’s approach is that she gave no separate and distinct consideration to the relevant language of Rule 50 or to the common law open justice principle. She wove these together with issues arising under the Convention and treated her Convention analysis as decisive of the application under the “interests of justice” limb of Rule 50. Eady P held (at [104]) that the EJ had erred by failing to apply common law principles in its consideration of the respondents’ evidence. I agree, although I would put it slightly differently.

The EJ was required to apply a rule made under a statutory power. The wording of that rule was recited but there was no consideration of the meaning or ambit of the “interests of justice” provision. Further, whilst the basic common law principles of open justice were identified there was no detailed discussion of those principles or their application, separately from the issues arising under the Convention. There was no acknowledgement that analysis by reference to the domestic Rule and/or the common law might yield a different answer.

Mr Callus does not dispute this. The answer he offers is that the absence of a separate common law discussion is immaterial. He argues that this aspect of Rule 50 represents a codification of the common law; that the answer under the common law could never be different from the answer arrived at by reference to the Convention rights; and that it was therefore unnecessary and wrong for Eady P to remit this aspect of the application for redetermination. Mr Callus submits that the common law and the Convention “walk in step” with one another; Articles 3, 5 and 8 have common law “analogues” which leave no room for an order which would not be justified by the Convention. He says that so far as risks to safety and security are concerned, there is no “balancing” process to be undertaken: if the Convention thresholds for intervention under Articles 3 and 5 are met an order would be merited. But if (as the EJ held) those thresholds are not met there can be no room for intervention under the common law. As for Article 8, Mr Callus accepts that if that provision is “engaged” the court must undertake a balancing process, as it would do at common law; but if (as the EJ held) Article 8 is not engaged, and no question of balancing arises, the same would obtain at common law.

I do not agree with this approach. The effect of the HRA is not that the Convention supplants or replaces domestic statutory or common law rules; rather it provides certain guarantees against the enforcement of those rules to the extent that would be incompatible with fundamental human rights. As Mr Callus eventually conceded, it is not necessarily the case that the answer given by the common law will be the same as that arrived at through a Convention analysis. And if the two are different, that does not necessarily mean the common law answer is incompatible with the Convention.

For his submission that the two strands of law are “in step”, Mr Callus draws on what Lord Reed said in A v BBC [2014] UKSC 25, [2015] AC 588 at [57]. But Lord Reed was not saying that the two are interchangeable or in all circumstances identical. The words relied on by Mr Callus need to be seen in their wider context. They were part of a discussion of “The relationship between the Convention and domestic law” in which Lord Reed rejected the proposition advanced by the BBC that “the source of the court’s power to restrict public disclosure of a party’s identity, in a situation where Convention rights are engaged, is to be found in the Convention rights themselves, rather than the common law”: see [55]. At [56], Lord Reed emphasised that “the common law principle of open justice remains in vigour, even when Convention rights are also applicable …” At [57] he said this:

In this case too, the appropriate starting point is the common law. This holds that open justice is a fundamental principle. But it also contains a key qualification: that every court or tribunal has an inherent power to withhold information where it is necessary in the interests of justice to do so: see Khuja v Times Newspapers Ltd [2017] UKSC 49, [2019] AC 161 [14] (Lord Sumption), citing the foundational common law authority of Scott v Scott [1913] AC 417, 446. I see nothing in Rule 50 or the context to suggest that when enacting the “interests of justice” limb of the Rule the draftsman intended to extend or to restrict the scope of the common law principle. On the contrary, the language of Rule 50(1) coupled with that of Rule 50(2) suggests that the principal intention was to reflect the common law, expressly authorising the ET to derogate from open justice to the extent that would in any event be permitted at common law, whilst emphasising the strength of the open justice principle.

The EJ should therefore have begun by asking herself whether the derogations sought were justified by the common law exception to open justice. This has been put in various ways in the authorities. In Scott v Scott at 439 Lord Haldane spoke of the need to show “that the paramount object of securing that justice is done would be rendered doubtful of attainment if the order were not made”. Earl Loreburn said, at 446, that the underlying principle that justified the exclusion of the public was “that the administration of justice would be rendered impracticable by their presence”. In Attorney General v Leveller Magazine Ltd [1979] ACT 44, 550 Lord Diplock spoke of the need to depart from the general rule “where the nature or circumstances … are such that the application of the general rule in its entirety would frustrate or render impracticable the administration of justice”. Usually, the court’s concern will be with the requirements of the due administration of justice in the proceedings before it. That is the focus of attention in the present case.

The qualification is certainly of wider application, as Eady P noted at [69]. It certainly permits derogations that are required for the protection of the administration of justice in other legal proceedings or even to secure the general effectiveness of law enforcement authorities: see Lord Reed’s discussion of the point in A v BBC at [38]-[41]. It may go further. But this appeal does not require us to identify the boundaries of the common law exception to open justice.

In Libyan Investment Authority (No 2) [2016] EWHC 375 (Comm) [26] Teare J held that the common law’s protection of the fairness of the court’s proceedings “extends to ensuring that those proceedings do not risk life and limb whether within the jurisdiction or without” (my emphasis). Eady J adopted that proposition, which is not in dispute on this appeal and must be right. I can see no reason why the court should be prohibited from doing what it considers necessary to ensure fairness in the proceedings just because the risk is to someone outside the jurisdiction. The fact that, as will often be the position in such a case, the person said to be at risk does not enjoy the protection of the European Convention cannot be a decisive factor.

In this case most of the persons who were said to be at risk abroad are not parties or witnesses but foreign-based employees of Millicom. But Eady J held that there is no reason in principle why the court should ignore risks of harm to non-participants. That conclusion is not challenged either. In my view it was clearly right, for two reasons. The first is that fairness to parties and witnesses may require the court to take account of risks to related persons, such as those with whom they live or work, wherever those persons may be, and whatever their status under the Convention. The second is that the issue is not just one of fairness to parties and witnesses. The aim is to protect and further the interests of justice more generally. Eady P put it this way (at [75]):-

As Eady P noted (at [70]), the common law “interests of justice” qualification requires or authorises derogations from open justice to protect “the right to life and security of a person” and to ensure that the proceedings “do not risk life and limb”: see Libyan Investment Authority v Societe Generale (No 1) [2015] EWHC 550 (QB) [31] (Hamblen J), and Libyan Investment Authority v Societe Generale (No. 2) [26] (Teare J). In such cases, the court will look for evidence of a real and immediate risk of harm. The EJ did this, although she did it as part of her Convention analysis. She concluded that there was a lack of “objective evidence” that publicity for the facts at stake would place the life and limb of Mr Frechette or the foreign-based employees at real and immediate risk. The assessment of a first-instance tribunal on a question of this kind deserves respect. But in my opinion, so far as it concerned the foreign non-participants, the assessment was clearly wrong. It was not open to the EJ and should have been overturned by Eady J. This was the second main flaw in the EJ’s approach. I shall explain my reasons for taking that view when I come to deal with the second limb of the application.

Moreover, and in any event, as Eady P also noted at [70] the domestic authorities show that derogations from open justice may be granted under common law principles in circumstances where the evidence does not meet the high threshold for interference on the grounds that there would be a risk to life, limb or security.

The most significant case is In Re Officer L [2007] UKHL 36, [2007] 1 WLR 2135. The case was about applications for anonymity by serving or former police officers who were due to be called as witnesses at a public inquiry into a death caused during an affray in Portadown, County Armagh. The applications were put on the twin bases that compelling the officers to give evidence without anonymity would be a breach of Article 2 and a breach of the common law duty of fairness to the witness. The tribunal of inquiry considered the case separately under both heads, noting that “the protection sought under article 2 … is also available at common law, but the common law goes further in providing protection to witnesses in appropriate cases which is not available under Article 2”. The House endorsed this approach.

The House of Lords held that under the common law and under Article 2 the tribunal had to ask itself whether there was a “real and immediate” risk to life which would be “materially increased” by compelling the witnesses to give evidence without anonymity. For this purpose, a “real risk” is one that is “objectively verified”. The subjective concerns of the applicants are not enough; fears of harm are only relevant if they are “objectively well-founded”. But the House went on to say that the principles that apply to the common law duty of fairness to the witnesses are “distinct and in some respects different from” the Convention requirements. At common law the tribunal could still take account of the witnesses’ subjective fears: R v Lord Savile of Newdigate, ex p A [2000] 1 WLR 1855. For these points see In re Officer L at [13], [14], [22] (Lord Carswell, with whom the other members of the House agreed). In the last of these paragraphs Lord Carswell observed that “It is unfair and wrong that witnesses should be avoidably subjected to fears arising from giving evidence …”

An order that is required to give effect to the court’s duty of fairness to a party or witness is one that is “necessary in the interests of justice". Here, Mr Frechette was a party and a potential witness who harboured fears about the consequences for himself and others of giving evidence and proceeding with the defence of the claim in the absence of derogations. But – and this is the third flaw in her reasoning – the EJ gave no consideration to those fears because she confined her analysis to the Convention and concluded that the fears were not supported by objective evidence and were therefore immaterial for that purpose.

For that reason, the EJ conducted no balancing process, which is the fourth flaw in her approach to the interests of justice limb of Article 50. If life and limb are shown to be at risk there is ordinarily no question of balancing competing rights. But in cases outside that category the common law requires a balancing process. Lord Carswell made this clear in In re Officer L at [22]:

As Eady P observed (at [37]-[38]) the factors that need to be weighed in the balance include (a) the extent to which the derogation sought would interfere with the principle of open justice; (b) the importance to the case of the information which the applicant seeks to protect; and (c) the role or status within the litigation of the person whose rights or interests are under consideration: see the decision of this court in R v Legal Aid Board, ex p Kaim Todner [1999] QB 966, at paras 6 and 8 (Lord Woolf) and Libyan Investment Authority (No 2) (above) at [34(3)].

I would add that the decision-maker should bear in mind the harm disclosure would cause and, conversely, the extent to which the order sought would compromise “the purpose of the open justice principle and the potential value of the information in advancing that purpose”: A v BBC [41] (Lord Reed). The main purposes of the open justice principle were identified by Baroness Hale in Dring v Cape Intermediate Holdings Ltd [2019] UKSC 429, [2020] AC 629 [42]-[43] : (1) “to enable public scrutiny of the way in which courts decide cases – to hold the judges to account for the decisions they make and to enable the public to have confidence that they are doing their job properly” and (2) “to enable the public to understand how the justice system works and why decisions are taken”.

As a general proposition, it may be said that the more remote an item of information is from the issues requiring resolution in the case the less likely it is that a restriction on its disclosure will offend the open justice principle or compromise its purposes. In this case, the ET will need to consider the Millicom parties’ contentions that the derogations they seek are “minor” and peripheral, relate to people who are not parties or witnesses, and concern information which has “no relevance” to the issues in dispute in the ET proceedings.

I agree with Eady P that on the facts of this case, the factors to be weighed in the balancing exercise also include Mr Frechette’s evidence that, because of his apprehensions about the risks of violence, he would not give evidence or allow Millicom to defend the proceedings if derogations were refused. The fifth flaw in the EJ’s reasoning was to leave this entirely out of account.

The EJ’s decision was based on a passage in Kaim Todner (above) at page 978 para 9 (Lord Woolf MR), later relied on in Moss v Information Commissioner [2020] EWCA Civ 580 [55] (Haddon-Cave LJ). The EJ accepted the submission of Mr Callus that the effect of those passages is that “a party insisting that it would not continue a claim if anonymity were not granted” should not “be a factor which has any purchase” in the decision whether to grant anonymity; “[t]he basis for any injunction must be made out objectively” (see [82] of the judgment). This is too stark a statement of the position. What Lord Woolf said in the passage cited is this:

I do not consider Kaim Todner or Moss to be authority for the bald proposition that a court or tribunal must always ignore an assertion that a party will abandon a claim or defence if the derogation sought is refused. That is not what Lord Woolf said: see the words I have emphasised in the quotation at [46] above. Of course, a threat to abandon a claim or defence or part of it if anonymity is not granted cannot be enough of itself to justify an application for that relief. And as Viscount Haldane LC emphasised in Scott v Scott (at p439), “A mere desire to consider feelings of delicacy or to exclude from publicity details which it would be desirable not to publish is not … enough”. There was thus no justification for a private hearing in that case where the issue was whether the marriage was void because the husband was impotent, “the petition was practically undefended and the evidence was very simple” (p431). But Viscount Haldane acknowledged (at p439) that a case might come within the exception to the open justice principle “[i]f the evidence to be given is of such a character that it would be impracticable to force an unwilling witness to give it in public”. And one of the illustrations which Earl Loreburn gave of the underlying principle (at p446) was a case in which “the administration of justice would be rendered impracticable” because “the parties entitled to justice would be reasonably deterred from seeking it at the hands of the court” Courts and tribunals must take a strict and disciplined approach to cases where this kind of assertion is made. But in my judgment, the question of whether publicity would affect the willingness of a party or witness to take part in a case is in principle a relevant factor. Deripaska v Cherney [2012] EWCA Civ 1235 [21] provides some modern support for that view.

In Kaim Todner the applicants did not want the name of their law firm to be associated with a decision of a court. In Moss, the court rejected the claimant’s case that publicity would risk a breach of his Article 8 rights. In other words, there was no reasonable basis for seeking anonymity in either case. Here, there were Mr Frechette’s concerns, and his statements as to what he would do if the case had to proceed with no restrictions on publicity. His concerns and his intentions were not just asserted. They were explained in some detail by him in two witness statements and supported by other evidence. The EJ made no finding that what he said was insincere or unreliable. She accepted (at [96]) that his stated intention “may well” be genuine. She was wrong to rule those matters out of consideration. She should have considered them and assessed their reasonableness. If she held them to be more than “mere feelings of delicacy” but to have some reasonable foundation she should have factored them into her consideration of whether the order sought was more than just desirable, but necessary in the interests of justice.

That is not the end of the decision-making process on the “interests of justice” limb of Rule 50. As indicated by Lord Reed in A v BBC at [57], a court or tribunal that has struck the common law balance will need to check its conclusions against relevant human rights. The ET will need to undertake this task when the Rule 50 application is remitted for redetermination. Here, there is an added reason for doing so. Rule 50 is delegated legislation which must be construed and given effect compatibly with Convention rights: ss 2, 3 and 6 of the HRA all apply. And for good measure Rule 50(2) expressly requires the tribunal to give effect to the Convention right to freedom of expression.

A decision to grant a derogation would therefore need to be reviewed for compatibility with the Article 6 and 10 rights of the parties and the Article 10 rights of the press and public. Under Article 6 the question would be whether the restrictions on disclosure are justified because “the protection of the private life of the parties so require” or whether they are “strictly necessary” because “publicity would prejudice the interests of justice”. On the facts of this case the relevant justifications under Article 10(2) would seem to be “for the protection of the … rights of others” and “for preventing the disclosure of information received in confidence”.

A decision to refuse a derogation would need review for compatibility with any other Convention rights that are properly relied on in favour of protecting the information in question, or which might support a decision to derogate. I shall deal with these next.

The second issue raised by the Rule 50 application was whether the derogations sought were “necessary … in order to protect the Convention rights of any person”. There is an overlap with the task I have just discussed. The Article 6 fair trial rights of all parties are, for instance, relevant here as well. But the two tasks are not the same. For the reasons given by the EJ, the rights of the foreign staff of Millicom to protection of their physical safety and security are not Convention rights. The individual respondents’ cases under Articles 3 and 5 have been rejected on the evidence. The issues raised by the appeal and cross-appeal under this limb of the rule relate only to the rights which Mr Frechette asserts under Article 8.

There are two issues: was Eady P wrong (1) to uphold the EJ’s conclusion that there was no objective evidence to support Mr Frechette’s fears for the safety of others? (2) to interfere with the EJ’s conclusion that those fears were insufficient to “engage” his Article 8 rights?

I have referred here and above to Article 8 being “engaged” because that is the language of the appeal documents. But I have put the word in quotation marks because, as Lewis LJ pointed out at the hearing, it is potentially confusing to speak of whether Article 8 is “engaged” in a case of this kind. Consideration of Article 8 in this case requires a two-stage process. The first question is whether the conduct under consideration (public disclosure of information by the state in legal proceedings) would involve an “interference” with a person’s Article 8 rights. If so, the second question arises: would that interference be justified as necessary in pursuit of one of the legitimate aims identified in Article 8(2)?

The Millicom parties adduced evidence from five witnesses: the three individual respondents and two independent professionals, a Mr Mifsud-Bonnici, and Adrian Stones.

Mr Frechette’s evidence was based on his “years of experience with working in [Country Z]”. The relevant features of his evidence were, in summary, that:-

This evidence was endorsed by the third and fourth respondents in short witness statements which said they shared and agreed with his concerns. Those two respondents were among the foreign Millicom workers about whose safety Mr Frechette was concerned.

Mr Mifsud-Bonnici was a consultant in environmental, social and governance matters who was currently advising executives on strategy relating to investments in Country Z and on cross-industry coalition building between the continent concerned and the EU. He gave evidence of the general background in Country Z. His view was that the government had a persistent lack of respect for the rule of law and fundamental rights such that there was a significant risk to businesses in Country Z and their employees. He too said that the authorities arbitrarily arrested and prosecuted government critics, including businesses, on trumped up charges and engaged in extortion, and that the government was implicated in abductions. He gave details, including a list of examples of foreign and Country Z workers charged with economic crimes. He said that Event Y (the offence against Person X) was widely believed to have been politically motivated.

Mr Stones was a former MI6 intelligence officer now working as a business intelligence consultant. His expertise was in security risk assessment. His MI6 role had required him to assess physical security threats to UK citizens. He had not been to Country Z but had extensive experience of working in the continent of which it formed part and was broadly familiar with the political and cultural situation there. He endorsed what Mr Frechette and Mr Mifsud-Bonnici had said about the absence of the rule of law in Country Z. His evidence was that Event Y was carried out by organs of the state or by individuals believing they were acting in accordance with the government’s wishes. His opinion was that the disclosure of a possible link between Millicom and Event Y would be a “game changer”. The story would be “incendiary” and would be widely reported in the press. Readers would be likely to conclude that Millicom was implicated in the serious criminal offence. His view was that this was likely to result in “a material increase” in two kinds of “real and immediate” risk to Millicom employees in Country Z: (1) serious violence from vigilantes, which could take the form of action against the offices, homes or persons of employees, who could be beaten up; (2) arrest and detention by the police, for questioning without access to telephones or lawyers, because the government would have to be seen to be doing something.

The EJ found Mr Frechette’s evidence that he and others might be at risk outside Country Z to be implausible. She held that the individual respondents were not at risk as they were not based in Country Z and had no need or reason to travel there. She concluded that much of Mr Mifsud-Bonnici’s evidence did not have “any direct relevance to the factual issues before me”. But the evidence of Messrs Frechette and Mifsud-Bonnici as to the general level or risk in Country Z was not in dispute, and the EJ accepted it. As for Mr Stones, the EJ rejected Mr Callus’s invitation to disregard his evidence as non-expert opinion from someone who had never been to the country in question. She acknowledged Mr Stones’ knowledge of the political situation in Country Z, concluding that the fact he had not been there only detracted “to some extent” from the weight to be given to his “expertise”.

The EJ found however (at [66]) that “Mr Stones cannot know for sure that the results he predicts will happen” and was “speculating on the possible outcome” of publicity. She held (at [67]) that the Millicom parties “have not shown that disclosure of the information …. would place Mr Frechette or the employees in [Country Z] at real or immediate risk of harm which would engage Articles 2, 3 or 5 ECHR.” She went on (at [104]) to summarise her conclusions in this way:

The EJ’s approach was, therefore, to treat the evidence of risk as insufficient on the grounds that it was “speculative” and not “objective”. Mr Hickman submits that this was perverse. The question was whether the employees would be exposed to a material increase in a real and immediate risk of harm. The background evidence, and (in particular) the evidence of Mr Frechette and Mr Stones was plainly relevant to that issue. The EJ was clearly not rejecting the evidence as inadmissible or lacking in credibility. Nor was she evaluating its weight against the imperatives of open justice; she never reached any kind of balancing process. Mr Hickman points to the EJ’s observation that Mr Stones could not know “for sure” what “will happen”. He points out that an applicant does not need to make the court or tribunal sure that the consequences at issue would happen, or even that this is probable; what must be proved is a real possibility of significant harm – a possibility that cannot be ignored having regard to the nature and gravity of the harm: Rabone v Pennine Care NHS Foundation Trust [2012] 2 AC 72 [35]-[38]; RXG v Ministry of Justice [2019] EWHC 2026 (QB), [2020] QB 703 [35(iii)-(iv)]. On a proper analysis, submits Mr Hickman, the EJ was departing from this approach, and treating the evidence as speculative and not objective because it involved an assessment of the risk that some event would happen. That was a clear error of law. The whole exercise was and had to be one of prediction and estimation.

Reflecting on those submissions I have been alert to the risk of placing undue weight on individual sentences or phrases in a judgment. It is necessary to look at the judgment as a whole. But the EJ nowhere explained what she meant by “objective evidence”. I have found myself driven to the conclusion that the EJ did indeed misdirect herself on this aspect of the case in the way contended for by the Millicom parties. The authorities show that a court or tribunal cannot find that a real risk that some event will occur just because a witness fears this will be so. But this evidence went beyond mere subjective fears. Mr Frechette and Mr Stones both gave reasoned evidence setting out facts to explain why, in their opinion, the disclosure which the Millicom parties seek to prevent would materially increase the existing background risk of harm from vigilantes and a corrupt and unreliable public sector in Country Z. This was on any reasonable view “objective” evidence, whatever view might have been taken of the weight to be given to it.

For these reasons I cannot agree with the view of Eady P, that the EJ “did not lose sight of the fact that it was being asked to assess a potential future risk”. The President took a benevolent approach to the EJ’s language, interpreting it as meaning that the evidence was not “objectively verified” in the sense that it did not convincingly demonstrate the seriousness of the risk. This is not how the EJ explained her conclusions. The President observed that Mr Frechette’s views were “inevitably subjective” and Mr Stones had not been to Country Z. That is not the way the EJ approached the evidence of these two witnesses. She accepted that Mr Frechette’s background evidence was reliable and that Mr Stones had relevant expertise. It was not open to the EAT to substitute its own reasoning on this point. In any event, the former point is unsound; Mr Frechette had considerable direct experience of Country Z. And although Mr Stones did not, the fact he had not visited the country could not of itself justify the outright rejection of his evidence as lacking any “objective” quality.

On this topic the EJ said only this (at [109]): “I do not accept that the subjective fears raised by Mr Frechette are sufficient to engage Article 8 and therefore I do not need to go on to consider the balancing test of proportionality under Article 8(2).” This is ambiguous, as Eady P recognised. She read it as a statement that, as a matter of law, Mr Frechette’s fears for himself and his colleagues could not be enough to call for a balancing exercise. She held that was an “unduly restrictive” approach and an error of law. Her reasoning was that the authorities show that causing a person to live in fear of a physical or verbal attack upon them may amount to an interference with their Article 8(1) right to respect for private and family life which calls for justification under Article 8(2); and the right to respect for private and family life can extend to workplace relationships. For the former proposition, Eady P cited Abbasi v Newcastle [2021] EWHC 1699 (Fam) [105]-[107] (Sir Andrew McFarlane P). For the latter, she cited Niemitz v Germany (1993) 16 EHRR 97 [29]-[31].

In essence, the President’s reasoning was that the EJ should have considered Mr Frechette’s case under Article 8 on the footing that his subjective fears about what might happen to him and his work colleagues if the pieces of information at issue were publicly disclosed in the course of the litigation were relevant. The EJ should have considered whether a decision to allow publicity for the relevant information, could amount to an interference with his Article 8 rights on that basis.

Mr Callus does not take issue with Eady P’s legal analysis. His argument is that she misinterpreted the ambiguous statement in paragraph [109] of the EJ’s judgment. He says that it was never in dispute that Article 8 is theoretically capable of extending to workplace relationships; the EJ well understood that Mr Frechette had “subjective fears … on behalf of his colleagues”; what happened here is that the EJ made a finding that “the evidence provided was insufficient to engage Article 8”, and the EAT should not have interfered with that assessment. Mr Callus further submits that in deciding to remit the case on this issue Eady P wrongly elided the threshold test for deciding whether particular conduct is an interference with Article 8 rights with the balancing test that has to be undertaken if that is so.

I am not sure it is necessary to resolve these questions, given what I have said about the previous issue. Even if Mr Callus was right, the EJ’s conclusion would still be flawed by reason of her mistaken approach to the question of “objective evidence”. For that reason, the case has to go back for redetermination on the Article 8 issue anyway. But insofar as it matters, I would reject Mr Callus’s criticisms of Eady P’s approach. I think her interpretation of paragraph [109] of the EJ’s decision was not just legitimate but correct. And I do not accept that Eady P elided the threshold and balancing tests as Mr Callus submits. That argument depends upon a misreading of the relevant section of her judgment.

The third issue raised by the Rule 50 application was whether the derogations sought fell within the “the circumstances identified in section 10A of the [1996 Act]”. Read literally, the question posed is whether the hearing would involve taking “evidence from any person which in the opinion of the tribunal is likely to consist of … information which has been communicated to him in confidence or which he has otherwise obtained in consequence of the confidence reposed in him by another person.” But the wording of Rule 50(1) suggests that the test is whether a restriction on open justice is necessary in order to protect information of these kinds. That would be the test at common law. For both these reasons I think that is the right basis on which to approach this limb of Rule 50.

Mr Clifford and his representatives were not happy with the late introduction of a copy of his employment contract, but they did not seek an adjournment. Nor did they dispute the authenticity of the document. In the relatively informal procedural context of ET proceedings the EJ was entitled to treat it as a sufficient evidential basis for findings of fact. Clause 15 of the contract, headed “Confidentiality” contained an undertaking by Mr Clifford that he would not during or after his service under the agreement disclose or communicate, make use of or divulge to any person, directly or indirectly “confidential information of any kind whatsoever of the Company or its associated companies or its respective clients which he may acquire in the course of his service hereunder…”

The EJ held (at [112]) that “the late introduction of the claimant’s contract of employment did establish a duty of confidentiality as regards matters undertaken in his employment duties. …” But she went on to decide that this did not necessitate an order under Rule 50 because

There are several obvious problems with this approach. Eady P identified two. First, she observed that review of HRH the Prince of Wales case shows that it is inaccurate to suggest that enforcement of a contractual duty of confidence depends on proof that “it is legitimate for the owner … to seek to keep [the information] confidential.” The overall question is “not simply whether the information is a matter of public interest but whether, in all the circumstances, it is in the public interest that the duty of confidence should be breached” (my emphasis). Further, the relevant circumstances include the nature of the information and the nature of the relationship that gives rise to the duty of confidentiality. (See HRH the Prince of Wales at [68]-[69].) The EJ failed to address these aspects of the test. She thereby failed to have regard to relevant factors. I agree. Indeed, I would characterise this as a misdirection in law.

Secondly, Eady P pointed out that the EJ’s floodgates argument was simplistic and flawed. Not all whistleblowing claims will involve information protected by a contractual duty of confidence. And in those that do, the ET will have to conduct a fact-specific analysis, to answer the questions just identified, and to determine whether restrictions on disclosure are compatible with open justice. Again, Eady J was clearly correct on this point.

To these points, I would add the following.

Mr Callus argues, however, that none of this matters. He says that the EJ did not need to consider whether there was any public interest justification for disclosure. Indeed, he would say she was wrong to do so. He submits that the EJ made no finding of fact that the duty of confidentiality applied to the information in question. There was thus no obligation to undertake any evaluation or balancing of competing factors.

In my opinion this is a misreading of the judgment, or a point that goes nowhere, or both. My own interpretation of the EJ’s decision, read as a whole, is that she held (a) that there was a contractual duty of confidence as regards “matters undertaken in the course of employment” and (b) that this applied to the information that was the subject of the Rule 50 application. That is why she went on to consider the further questions. Mr Callus’s submission that there was no evidence to support a finding to the latter effect cannot be upheld. It is not open to Mr Clifford. It was and remains an essential element of his own case that he had acquired the relevant information in the course of his employment. Indeed, he has filed evidence verifying that proposition.

I have sympathy with the EJ. She was faced at short notice with what she rightly described as a “complex and difficult application” of an unusual nature, the grounds of which were shifting and developing as the hearing proceeded. It also seems that both sides focused most of their attention on the human rights arguments, devoting rather less of it to the first and third limbs of the application. But regrettable as it is that this issue has to be considered on four separate occasions, I think the application must be sent back to the ET.

If my Lord and my Lady agree with that the consequence will be an order that the application be remitted for redetermination by a differently constituted ET in accordance with the judgments of the EAT and this Court. Paragraph 3 of the order of Eady P needs no amendment except to remove the words in brackets, for the reasons given at [54] above.

I agree.

I also agree.