R (on the application of Pearce and another) v Parole Board of England and Wales

The issue in this appeal is what approach the Parole Board (“the Board”) may properly take, when deciding whether or not to direct the release of a prisoner on licence, to potentially relevant assertions or allegations made about the prisoner which have not been determined, either by the Board or some other body, to be either proved or disproved on the balance of probabilities. After the decision of the Divisional Court in R (D) v Parole Board; [2019] QB 285 (the Worboys case; hereinafter “R(D)”) the Board published Guidance directed to this issue. The lawfulness of that Guidance is in issue in these proceedings.

The claimant, Dean Pearce (“the claimant”), is a prisoner whose release the Board declined to direct. He brought the present proceedings for judicial review to challenge that decision and in doing so contended that the published Guidance is unlawful. That contention was upheld by the Court of Appeal ([2022] 1 WLR 2216), although the decision of the Board on the particular facts of the claimant’s case was held to have been proper and justified. There is no longer any dispute about the Board’s decision in relation to the claimant, but the challenge to the lawfulness of the general Guidance remains.

The Parole Board

The Board is a statutory body, in being since 1967 and presently established under section 239 of the Criminal Justice Act 2003 (“CJA 2003”). Although in the past its functions were to advise the Home Secretary on the exercise of the Royal prerogative power to release prisoners before the end of their sentence, it now has statutory responsibilities for itself making the decision about early release, that is to say release on licence sooner than the end of the court’s sentence. The Secretary of State (now of Justice) is obliged to follow any directions for release which it may give. In so doing, the Board acts judicially and as a body independent of the executive. It is properly treated as a court for the purposes of the European Convention on Human Rights. In Weeks v United Kingdom (A/11) (1987) 10 EHRR 293 the Strasbourg Court explained that the relevant attributes of a court are that it is independent and impartial and that its procedures are fair, which includes the requirement that the prisoner is able properly to participate in the proceedings of the Board: paras 61–65.

Sentencing provisions have been extensively and frequently amended in the past 25 years or so, and early release schemes more than most. There are as a result several different permutations of sentencing and early release provisions. At present, there are six principal categories of prisoner in respect of which the Board has the function of considering whether to direct early release. They are:

It is not necessary here to set out the complex statutory provisions which contain the rules for these various categories of prisoner, because, convoluted as they are, they all involve the common feature that release on licence is made conditional upon the decision of the Board, and the statutory test to be applied by the Board is expressed in the same terms in each case. In all these cases, the effect of the statutes is that the prisoner will remain in prison under the sentence of the court unless the Board directs his earlier release.

In some (but by no means all) circumstances, the Board also has a separate function in relation to the question whether the prisoner merits transfer to a less severe regime, especially, but not only, to open conditions. In this case the Board’s role is to advise the Secretary of State, who makes the decision. In this role also the Board fulfils a judicial function and the principles on which it acts are in many respects similar to those applicable to directions to release. In many cases where release is considered but refused, the Board may move on to consider a recommendation for transfer. This separate function is not, however, the subject of the present appeal, which concerns only the decision-making powers of the Board in relation to release on licence.

The Statutory Test for Release

For all the cases where the Board has the duty to make the decision whether to direct release, the statutory test which it must apply is in effect the same, albeit written in several different places in the statute book. It is enough to set out just one of the provisions, found in section 28(6)(b) of the Crime (Sentences) Act 1997 (“CSA 1997”) and applicable to the first category of prisoner mentioned, ie those on life and IPP sentences:

“28 (6) The Parole Board shall not give a direction under subsection (5) above with respect to a life prisoner to whom this section applies unless—

(a)…….; and

(b)

the Board is satisfied that it is no longer necessary for the protection of the public that the prisoner should be confined.”

It is convenient to note that in R (Sim) v Parole Board [2004] QB 1288 (“Sim”) it was held that in the specific case of an early model of extended sentence, created by the Crime and Disorder Act 1998, article 5(4) of the European Convention on Human Rights (“ECHR”) required the Board positively to be satisfied that it remained necessary for the protection of the public that the prisoner remain in custody, as distinct from permitting release only when the Board was satisfied that it was no longer necessary for him to remain confined. This inversion of the natural meaning of the words of the universally formulated statutory test was held to result from the fact that for that particular form of extended sentence the sentencing court would not necessarily have found the prisoner to pose any risk of causing serious harm to the public, and would not, as a result, have had it in mind that he should be detained unless it was shown that he no longer presented a danger to the public of such harm: see para 47 in the judgment of Keene LJ in the Court of Appeal. The Board appears since to have taken the view that this approach is also required in the case of extended sentences of two different models, created by the CJA 2003 (the Extended Sentence for Public Protection (“EPP”) and the LASPO Extended Determinate Sentence (“EDS”)). The correctness of this view has not been debated in the present case, and it does not affect the question which is raised before us. It ought, however, to be observed that there is an arguably significant difference between the 1998 model of extended sentence and the two later models. Both later models are available to the sentencing court where, and only where, the prisoner is found to be dangerous, that is to say presents a danger to the public of serious harm through the commission of further offences — see sections 227(1)(b) CJA 2003 and s 226A(1)(b), now section 280(1)(c) Sentencing Act 2020. In this respect, both later model extended sentence prisoners are in a similar category to those made subject to life or IPP indeterminate sentences. In all such cases, it might be said that the sentencing court did indeed have it in mind that continued detention after the end of the custodial element of the sentence would be necessary unless it be shown that the danger to the public was not present and that that was the occasion for the extension of the licence period. If that were correct, the necessary link between the sentence of the court and the continued detention of the prisoner would be established and article 5(4) would not require any inversion of the natural meaning of the statutory language (see the reasoned speeches of the House of Lords in R (Walker) v Secretary of State for Justice [2010] 1 AC 553, especially per Lord Brown, at paras 49–50 and 60, and Lord Judge CJ, at paras 103, 108 and 123). It is, however, not necessary to resolve this question for the purposes of the present case.

It is important to recall that a life sentence is not quantified by, or limited to, the minimum term. It is an indefinite sentence, which lasts until it is safe to release the prisoner. Similarly, in the case of extended sentences (ie the EPP or EDS) or prisoners who would otherwise be on licence but who have been recalled to prison, the original sentence was (i) for a fixed or minimum term and also (ii) for a licence period, to be served at large if the public was safe but not if it was not.

The Board proceeds by decision of a panel of members, either on paper or by oral hearing. Whichever the form of procedure, the prisoner is entitled to be represented before it.

By section 239(3) and (4) of CJA 2003 the Board must consider any documents provided by the Secretary of State and also “any other oral or written information” obtained by it. It is well established that it has control over the material at which it looks in what is essentially an inquisitorial process in which interested parties, notably the prisoner and the Secretary of State, are entitled to participate. It will typically have a good deal of material, such as reports of behaviour in prison, statements of attitude to offences, records of rehabilitation courses or work undertaken, assessments of relationships in the world outside prison (whether supportive, potentially criminal or likely to de-stabilise), and health reports, particularly covering drug or alcohol use. It will normally have a dossier containing information about the prisoner which is prepared by the Public Protection Casework Section (“PPCS”) (part of the Prisons and Probation Service in the Ministry of Justice). It will have the details of his past criminal record, his index offence and conviction, and the sentencing remarks of the judge. It will generally have assessments by psychologists and/or offender managers of the risk which he presents to the public, whether upon release or upon movement to a lower category of prison regime. The prisoner is entitled to put material before the panel. The Board considers something of the order of 25,000 cases per year.

The Issue in this appeal

It may sometimes happen that the Board is told that complaints or allegations have been made about the prisoner. Such might relate to his conduct in prison, but they might also arise from events either before he started his sentence or after he left prison, for example if he has been allowed out on temporary licence release or has been in an open prison. Such a complaint or allegation may be of conduct which, if true, would amount to a criminal offence, or may be of some other behaviour which might affect the statutory question whether the Board can be satisfied that it is no longer necessary for the protection of the public for the prisoner to remain confined. The particular question raised by the present appeal is what limitations, if any, there are upon the way the Board can approach such a complaint or allegation.

This court held in R (Osborn) v Parole Board [2014] AC 1115 that one of the circumstances in which it is likely that the common law duty of procedural fairness will require an oral hearing is where there are disputed facts which are likely to be important to the decision. There is no doubt that it is within the competence of the Board to enquire into such an allegation, and indeed to make findings as to what were the facts. But the Board’s powers to gather information are limited. While its process is essentially inquisitorial, it is not itself an investigatory body and it cannot command an investigation by, for example, the police. In R v Vowles [2015] 1 WLR 5131, the Court of Appeal (Criminal Division) noted, at para 42 per Lord Thomas CJ, that the Board further lacks any compulsive power to enforce directions for case management which it may give, and expressed the hope that the Ministry of Justice or its agency, now the PPCS, might give an undertaking to enforce compliance at least by its officers. The Board has a residual power to ask the High Court to issue witness summonses if it is satisfied that the situation calls for it, but that is not a procedure which it uses lightly or often. Recently a report of the charity, JUSTICE (“A Parole Board fit for purpose” January 2022) also drew attention to the absence of compulsive powers and suggested that this hampers the Board’s effectiveness; in particular that the residual and rarely used power to seek a High Court order for the attendance of witnesses was said to be costly and impractical. The ability of the Board to find facts depends on the material before it, whether in the form of documentation from the Secretary of State or oral evidence from witnesses. It is, moreover, settled law and was common ground before this court, that the Board is not, in making its decisions, in any sense trying a criminal case. If it does find facts, the prisoner stands in no sense convicted of any offence, and no kind of punishment is imposed. It is equally settled law and common ground that, if it sets out to make findings of fact, the Board applies the civil standard of proof, the balance of probabilities, rather than the criminal standard of proof. Thirdly, it is equally settled law and common ground that the Board is not bound by the rules of admissibility which apply to a criminal trial. For example, it is entitled to take hearsay material into account. These propositions have been established since at least the decision of the Court of Appeal in R (Brooks) v The Parole Board [2004] EWCA Civ 80; [2004] Prison LR 324 (“Brooks”).

The central issue in this appeal is whether the Board, in making its decision, is confined to acting only on facts which it (or some other competent body) has found to be proved on the balance of probabilities, or whether there can be circumstances in which despite the absence of fact-finding, it is entitled to take a complaint or allegation into account when confronting the statutory question.

The published “Guidance on Allegations”

The Board initially published Guidance on the treatment of “allegations” in March 2019. It was somewhat amended in July 2021 to take account of the decision of the Divisional Court in R (Morris) v The Parole Board [2020] EWHC 711 (Admin) and of the first instance judgment of Bourne J in the present case ([2020] EWHC 3437 (Admin)). The Guidance thus fell for consideration by the Court of Appeal in the present case in its amended form, and that court held that it was in some respects unlawful. The question for this court is whether that decision is correct. (The Guidance has been suspended pending the decision of this court.)

The Guidance defines the “allegations” to which it is designed to relate as follows in paras 2 and 3:

“2.

The term 'Allegation' refers to conduct alleged to have occurred which has not been adjudicated upon. Adjudications can include a finding by a criminal or a civil court or a prison adjudication. Allegations which are relevant are those which, if true, could affect the panel's risk analysis. Sometimes these allegations are currently being investigated by the police or others and may be disposed of or adjudicated on in the future.

3.

Allegations may be of harmful behaviour and/or 'risky' behaviour.”

The ensuing paragraphs 4–6 contain the principal advice on how a panel of the Board should deal with such an allegation:

“4.

Once it is established that an allegation has been made, panels will need to consider the allegation when making a parole decision.

5.

Panel decisions must be made objectively, based on (a) the information and evidence provided to the panel and (b) information and evidence obtained as a result of the panel's inquiries and (c) what can properly be inferred from that information and evidence.

6.

Panels faced with information regarding an allegation, will have to assess the relevance and weight of the allegation and either:

a.

choose to disregard it; or,

b.

make a finding of fact; or

c.

make an assessment of the allegation to decide whether and how to take it into account as part of the parole review.”(Emphasis added.)

As will be seen, objection is particularly taken on behalf of the claimant to para 6(c). A similar form of words occurs later in the Guidance at para 9(3). The claimant’s contention is that in the absence of findings of fact, either by the Board or by some other competent body, an allegation is simply a “non-fact”, and as such it is not permissible for the Board to pay any attention to it.

Subsequent paragraphs of the Guidance expand somewhat on the advice as to the correct approach. It is not necessary to set out the whole of a discursive document, but attention should be drawn to the following paragraphs.

“16.

Panels should be very careful about making findings of fact in relation to allegations that are being investigated and may result in further enforcement action, such as a prosecution. It is not the panel's role to pre-judge any future case that may be brought against the prisoner. Prisoners and their representatives may claim that it is unfair that a finding of fact is made to a lower standard of proof than the criminal standard (beyond reasonable doubt) and in circumstances where the procedural safeguards of a criminal trial do not apply. Panels should be clear on what they are making findings of fact about and why.”

“18.

Panels may need to make an assessment of an allegation when the allegation is capable of being relevant to the parole review, but the panel is not in a position to make a finding of fact either because there is insufficient material available to make such a finding on the balance of probabilities, or because it would not be fair to do so. This most often arises when there is information regarding an allegation, but, critically important aspects of the evidence cannot fairly be tested. The allegation and the circumstances around it can form a basis for testing the reliability of the prisoner's evidence. It can be material on which an expert's evidence can be tested. The wider circumstances of the allegation might also give rise to areas of concern. However, in cases where there is a mere allegation without any factual basis that can be identified by the panel or the allegation is not relevant to the question of risk before the panel. the allegation should be disregarded and no weight placed on it.” (Emphasis added.)

20.

If an allegation is relevant to the parole review, the panel will need to form a judgement as to what weight to give the allegation. This will require an examination of the allegation and any underlying facts that the panel can find (on the balance of probabilities). The following factors can be considered when judging what weight to give an allegation:

a.

Source: can the credibility and reliability of the source be assessed and, if so, what is their credibility as a source; were the actions of the source consistent with the allegation; does the source have a motive to act against the prisoner; how contemporaneously (sic) was the making of the allegation with the events concerned; has the source's account been consistent? Allegations from a credible source are likely to be given greater weight than allegations from a less credible source.

b.

Supporting information: is there other evidence that supports the specific allegation whether from other sources and/or documentary evidence that record the allegation? Allegations that are supported by other information will normally have more weight than allegations that come from a single source.

c.

Nature of the allegation: an allegation that is of more serious misconduct is capable of having a greater effect on the panel's risk assessment.

d.

Contemporaneity: is the allegation relating to events in recent times or at some time in the distant past? Allegations that relate to more recent times are likely to be more relevant than allegations relating to events in the distant past.

e.

Context: does the allegation fit with other information known about the prisoner (which could include convictions or known behaviour including patterns of behaviour or other known allegations) in which case it may have more weight than an allegation that does not fit; and

f.

The prisoner's evidence: panels should take account of the prisoner's denial or limited admissions/minimisation of the allegation, and, in doing so, make an assessment of the prisoner's credibility and reliability as a witness.”

“24.

An allegation that is only marginally relevant, or is relevant but which carries little weight, is likely to be of little concern to the panel and therefore have little to no impact on the parole decision. Mere allegations without any underlying factual basis or irrelevant allegations should be disregarded. The panel's risk assessment should always be based on found facts even if they are unable to make a finding of fact about all the matters raised.” (Emphasis added.)

The immediate occasion for the issuance of this Guidance was the decision of the Divisional Court in R (D). In that case the Board, following a practice which it had developed over time, had taken the view that it was disabled from giving any consideration at all to additional criminal allegations against the prisoner, beyond those for which he had been convicted. On the very striking facts of that case, that practice had led the Board to rule altogether out of consideration the fact that some 80 or so additional complainants had recorded allegations of rape against the prisoner, beyond the 12 counts on which he had been convicted, and moreover had alleged, apparently in many cases independently, an unusual and strikingly similar modus operandi. The decision of the court was that there was no bar to the Board considering the additional allegations, in particular as a means of testing the truthfulness of the circumstantial account of his offending and its origin which the prisoner was advancing. That was potentially relevant to the question whether he was now repentant and thus to the risk of further offending and harm which he did or did not present. The court in that case did not, however, have to address the proper approach to the facts of the additional allegations, or the question here in issue of whether the allegations could be relied upon in the absence of findings of fact about them.

The case of Pearce

In 2009 the claimant, born in 1982 and then aged 27, committed two separate offences of sexual assault by penetration of women. In each case the woman was unknown to him and was walking home alone late at night. The method of offending was similar. He affected to befriend the victim but, when she resisted his advances, he took her to a secluded area and there used violence and threats to achieve the offence. He had previous convictions which included one for a sexual assault on a 13–year–old girl in 2005. On his plea of guilty to the index offences, he was sentenced on 13 October 2010 to imprisonment for public protection with a specified minimum term of three years and 23 days. The minimum term expired in November 2013. The present Board review, in May 2019, was his fourth.

The panel making the decision had a dossier of some 357 pages to consider. It heard the evidence of the prisoner’s Offender Supervisor and Offender Manager, from the Prison Psychologist, from the PIPE Key Worker and from the claimant, who was represented at the hearing. No view had been expressed by the Secretary of State who was not represented at the hearing. The claimant had a history of failed community sentencing, and of failing to honour bail. The panel identified a number of risk factors which he presented. They included a pre-occupation with sex, beliefs and attitudes which supported the abuse of the vulnerable, including children, a lack of stable relationships, isolation and poor emotional self-management and drug and alcohol misuse. He had, however, completed a number of work programmes and from them emerged strong evidence of signs of developing maturity and the hope that he might be able to manage himself better in future.

In addition to the professional assessments of the claimant, the Board was made aware of a number of allegations against him beyond those comprised in his index offences and previous convictions. They were as follows.

The Board’s panel questioned the prisoner about six of these complaints—not, so it would appear, incidents (i) or (vi). It recorded that he replied that he had often been arrested when young and mostly could not remember the details. As to incident (iii) he denied any recollection of it, or of arrest. He said that he remembered being arrested in respect of matter (iv) but denied that there had been any sexual activity. As to incident (vii) he said that the woman’s father had seemed cross with him for being in a taxi with his daughter. Of incident (viii) he asserted that he had been helping a victim of crime. The panel concluded that his denial of recollection of several of the arrests was not plausible. It noted and made a finding of fact that there must have been sexual contact in incident (iii) as shown by the DNA evidence. It observed that in relation to incident (iv) his present denial of any sexual activity was inconsistent with his claim at the time that it had been consensual. As to four other complaints, apparently matters (ii), (v), (vii) and (viii), it recorded a finding that they were “of concern”. It drew attention to the similarities between several of the complaints, which had in common encounters with lone females late at night and initial efforts to befriend them. Elsewhere in its decision, the panel disagreed with the prison psychologist’s contention that there was no substantial evidence relating to any of the six allegations. Its view was that there was witness evidence from some alleged victims and the DNA evidence in addition. It concluded that all six allegations considered by it were relevant to the prisoner’s risk of sexual offending and of causing serious harm if at liberty in the community.

The Board also considered the evidence of the professionals as to the extent of the prisoner’s progress, and evidence relating to the stability or otherwise of his plans for how he would live, and who with, if released. On all the material before it the Board’s panel declined to direct release but recommended transfer to an open prison as a means of testing the risk which the prisoner posed.

The proceedings in the courts below and the issues in this appeal

The claimant challenged the Board’s determination by judicial review. He did not succeed at first instance before Bourne J ([2020] EWHC 3437 (Admin)), who held (at paras 80, 82) that the procedure adopted by the Board had not been procedurally unfair to the claimant and rejected the challenge that the Board had failed to carry out a proper enquiry as to whether the allegations were or were not made out. He also held (at paras 56, 58, 60), and this alone is relevant to the appeal, that the Guidance was lawful and rejected the challenge that the Board could take into account, in assessing risk, only those matters on which it had been able to make findings of fact on the balance of probabilities.

The claimant failed on appeal in his challenge to the merits of the Board’s determination. The Court of Appeal (Lewison, Macur and Snowden LJJ) [2022] 1 WLR 2216) held (para 56) that there could be no reasonable challenge to Bourne J’s analysis (at paras 74-82) of the Board’s decision letter which included findings of facts on which to base an assessment of future risk. Macur LJ, who delivered the leading judgment, agreed with Bourne J that “the Board was entitled to question the reliability of Dean Pearce’s evidence as to his recollection of the incidents and gauge his reaction to the allegations, as contra-indicating his ability to self-manage sexual thoughts and behaviour”. The Court of Appeal therefore dismissed the appeal. But it also found that the Guidance was unlawful, declaring (in its Order of 18 January 2022, point 1) that “The April 2019 and July 2021 versions of the [Board’s] Guidance on Allegations contain unlawful misstatements of the law regarding the use of unproven allegations in the assessment of risk, to the extent set out in the judgment of the court”.

The essence of the reasoning underlying the declaration was that the Board’s process of risk assessment must involve a two-stage process by which the Board must first make findings of fact in relation to the truth of an allegation, or as to the facts which are a constituent part of or consequential to the allegation, and which may be indicative of harmful behaviour and then have regard only to the facts so found in its assessment of risk (see paras 27, 35, 43, 46–47 and 51 of the judgment). In para 43 of the judgment Macur LJ stated that she could not conceive “how the touchstone of “public law fairness” can operate in the circumstances in which an allegation which is not proved on the balance of probabilities is taken into consideration in the assessment of risk”. The Court of Appeal’s analysis is perhaps most clearly stated in that paragraph and in para 35 of the judgment in which Macur LJ said:

“… what is clear to me is that the panel must conscientiously evaluate the information before it to make findings of fact upon which to make the assessment of the prisoner’s risk … Established or undisputed constituent or consequential facts to an overarching allegation may provide compelling and convincing indications of risk in themselves, whereas simply to assess the seriousness of the nature of an allegation, provided there is some evidential basis for it is to embark down the route of ‘no smoke without fire’”.

The Court of Appeal accordingly held that the paragraphs of the Guidance which countenanced and invited the carrying out of a risk assessment by reference to unproven allegations misstated the law and were therefore unlawful.

The Board challenges this conclusion on appeal to this court. Ben Collins KC submits that the Board’s task is to assess the risk of releasing a prisoner; that this assessment is an evaluative exercise of judgment in which the concept of a standard of proof has no role; that the court should be slow to interfere with the Board’s decisions which were such judgments informed by its specialist skills and experience; and that there is no requirement for the Board to base its assessment of risk only on facts which it has found proved on the balance of probabilities.

The Board is a statutory body with a statutory remit. It is treated as a judicial body in certain respects: para 3 above. How it is to carry out its remit is determined as a matter of statutory interpretation; and the statute is to be interpreted against the backdrop of the general law. The statutory remit is stated succinctly in section 28(6) of the CSA 1997: para 7 above. The appeal has focussed on how that statutory provision operates in the context of the general law. The issues in this appeal are therefore: (i) whether there is a general legal principle that the law adopts a binary approach of fact or non-fact which the Board must adopt; (ii) whether there is an analogy between the Board’s risk assessment and the court’s assessment of risk when considering a care order under the Children Act 1989, to which reference is made in the relevant case law; (iii) what the concept of fair proceedings entails; (iv) the use of allegations in the Board’s risk assessments; and (v) the terms of the Guidance.

Fact or non-fact; a binary principle?

The central part of Mr Rule KC’s argument on behalf of the claimant is:

A general principle of law?

The first question to be considered is whether there is such a general rule of law.

Mr Rule KC relies on the observations of Lord Hoffmann in In Re B [2009] 1 AC 11. He said, at para 2:

“If a legal rule requires a fact to be proved (a ‘fact in issue’), a judge or jury must decide whether or not it happened. There is no room for a finding that it might have happened. The law operates a binary system in which the only values are zero and one. The fact either happened or it did not. If the tribunal is left in doubt, the doubt is resolved by a rule that one party or the other carries the burden of proof. If the party who bears the burden of proof fails to discharge it, a value of zero is returned and the fact is treated as not having happened. If he does discharge it, a value of one is returned and the fact is treated as having happened.”

In that case, the question arose in the context of court proceedings. Lord Hoffmann was discussing the standard of proof required in the context of section 31(2) Children Act 1989, when the making of a care order in relation to a child is under consideration. He took as read, in reliance on the decision of the House of Lords in In re H, the rule that in that context the law required that any facts used as the basis of a prediction that a child is "likely to suffer significant harm" must be proved to have happened; every such fact was, in other words, to be treated as a fact in issue (see para 3).

But Lord Hoffmann made it clear that he was not articulating any universal rule for all legal proceedings that everything could be separated out into proven facts and non-facts. On the contrary, he went on immediately at para 3 to add that in In re H

“The majority of the House rejected the analogy with facts which merely form part of the material from which a fact in issue may be inferred, which need not each be proved to have happened. There is of course no conceptual reason for rejecting this analogy, which in the context of some predictions (such as Lord Browne-Wilkinson's example of air raid warnings) might be prudent and appropriate. But the House decided that it was inappropriate for the purposes of section 31(2)(a).”

In other words, what Lord Hoffmann was describing as a binary approach to facts and non-facts only applied to facts in issue, that is to say to facts which were required by the law to be proved in order to lead to a legal entitlement. He readily accepted that there were many cases where the law proceeds by analysing or assessing allegations or asserted facts which are not proved. He gave as one example facts which are part of the material from which a fact in issue may be inferred, and as another the assessment of predicted risk, as described by Lord Browne-Wilkinson in In re H, dissenting in the result in that appeal but, as Lord Hoffmann confirmed, giving a valid example of how risk may be assessed without being able to find proven all the foundation facts. Critically, Lord Hoffmann looked at the particular requirements of the statute in question.

Shagang Shipping Co Ltd v HNA Group Co Ltd [2020] 1 WLR 3549 provides another example of a situation in which a court may perfectly properly have regard to a possibility which it has not been able to find proved. The claimant made a demand under a guarantee which supported a charterer’s obligations under a charterparty. The defendant guarantor asserted that the charterparty had been procured by the payment of a bribe to one of the charterer’s senior employees. The allegation of bribery was based solely on confessions made by persons when detained by the Chinese Public Security Bureau in the course of a criminal investigation. The claimant asserted that the confessions had been obtained by torture. The trial judge did not make any finding on the balance of probabilities that torture had been used to obtain the confessions. Instead, he held that he could not rule out the possibility that torture had been used but did not need to decide the matter as other surrounding circumstances, which he found proven, persuaded him that no bribe had been paid. The Court of Appeal overturned the trial judge’s ruling. It held (paras 62, 63, 65) that the judge had failed to give proper consideration to the probative value of the confessions; he ought to have considered first the question of torture, and, if that was not established on the balance of probabilities, to have disregarded any possibility of torture when considering what weight to give to the confessions.

This court allowed the claimant’s appeal, holding that the trial judge had not decided and did not need to decide whether or not torture had been used because he had already concluded that, notwithstanding the confession evidence, no bribe had been paid. The court stated (para 94): “The absence of a finding on that question is not the same as a finding that torture had not been proved on the balance of probabilities. Even if the binary principle operated in this context, therefore, the judge could not be treated as having, in law, made a finding that there was no torture.”

It is of relevance to this appeal to quote this court’s subsequent reasoning:

“95.

Even if, however, the judge had reached a definite conclusion that the use of torture had not been proved on the balance of probabilities, there would have been no inconsistency between that conclusion and the judge’s finding that torture was a real possibility which affected the reliance that should be placed on the confessions.

96.

It is of course true that, as Lord Hoffmann observed in In re B, if a legal rule requires a fact to be proved, the law operates a binary system. So where it is necessary to prove a fact for the purpose of a rule governing the admissibility of evidence, there are only two possibilities: either the evidence is admissible or it is not, which depends on whether the fact has been proved or not. There is no room for a finding that the fact might have happened. But not all legal rules do require relevant facts to be proved in this binary way. In particular, the rule governing the assessment of the weight to be given to hearsay evidence in civil proceedings does not. It requires the court to have regard to ‘any circumstance from which any inference can reasonably be drawn as to the reliability or otherwise of the evidence’: see section 4(1) of the Civil Evidence Act 1995. Such circumstances are not limited to facts which have been proved to the civil standard of proof.

97.

HNA’s argument depends on the assertion that, if failure to prove a fact to the requisite standard of proof requires a value of zero to be returned for the purpose of a particular legal rule, then that fact must be treated as not having happened for the purpose of other legal rules as well. There is no logical reason why that should be so. Nor is there anything in In re B (or any other authority cited in these proceedings) which lends the notion any support. What was decided in In re B was that section 31(2)(a) of the Children Act 1989 requires any facts used as the basis of a prediction that a child is ‘likely to suffer significant harm’ to be proved on the balance of probabilities, and that the assessment of the child’s welfare required in care proceedings once the threshold in section 31(2) has been crossed must be conducted on the same factual basis as the determination of whether that threshold has been crossed. Hence, if a particular fact (in that case an allegation of sexual abuse) has not been proved, it must be treated as not having happened for the purposes of both section 31(2) and the assessment of the child’s welfare. That is a decision about the meaning and effect of particular provisions of the Children Act. It does not establish any general principle that failure to prove that a fact happened for the purpose of a particular legal rule has the legal consequence that the fact must be treated as not having happened for all other purposes in the litigation. …”

In Shagang Shipping this court went on to discuss the concept of a “fact in issue” which Lord Hoffmann used in In re B, stating:

“98.

… This phrase commonly — and in our view most usefully — refers to those facts which as a matter of law it is necessary to prove in order to establish a claim or a defence: see eg Phipson on Evidence, 19^th ed (2018), para 7-02; Cross and Tapper on Evidence, 13^th ed (2018), p 30. …

99.

The requirement to discharge the legal burden of proof, which operates in a binary way, applies to facts in issue at a trial, but it does not apply to facts which make a fact in issue more or less probable. Lord Hoffmann was alert to this point in In re B as, immediately after [the passage we have quoted in para 24 above] he contrasted facts in issue with ‘facts which merely form part of the material from which a fact in issue may be inferred, which need not each be proved to have happened’ (para 3). …”

This court concluded (para 112) that it was contrary to principle to hold that if the use of torture had not been proved on the balance of probabilities, the court when assessing the weight to be given to a statement must ignore a serious possibility that the statement had been obtained by torture.

There is some assistance as to the correct approach to a risk assessment to be found in Secretary of State for the Home Department v Rehman [2003] 1 AC 153 which concerned the statutory power of the Secretary of State to deport a person on the ground that it would be conducive to the public good in the interests of national security. The principal issue in the appeal before the House of Lords was whether the Special Immigration Appeals Commission (“SIAC”), hearing the appeal against the deportation order, had misinterpreted the concept of national security when it held that for there to be a threat to national security, the Secretary of State had to establish that the subject person had to be shown to engage in, promote or encourage violent activity targeted at the United Kingdom, its system of government or its people. That misinterpretation had led SIAC to enquire into whether specific acts of encouragement of terrorism, so targeted, had been proved against the deportee, and it had held that they had not. Both the Court of Appeal and the House of Lords held that this requirement of targeting was a misinterpretation, since national security could easily be endangered by activity which was aimed at others, such as a friendly country. A second issue in the appeal related to the appellate role of SIAC. At that time, it was empowered to allow an appeal in such a case if it concluded that the Secretary of State should have exercised his or her discretion differently. That was an unusually broad basis of appellate jurisdiction that required specific statutory authority which was removed by amendment in 2003. This appeal is not concerned with those matters; what is relevant is the courts’ discussion of the correct approach to the Secretary of State’s decision-making.

On that question, all their Lordships approved the following statement of law by Lord Woolf CJ in the Court of Appeal in the same case ([2003] 1 AC 153). He had said this at para 44:

“However, in any national security case the Secretary of State is entitled to make a decision to deport not only on the basis that the individual has in fact endangered national security but that he is a danger to national security. When the case is being put in this way, it is necessary not to look only at the individual allegations and ask whether they have been proved. It is also necessary to examine the case as a whole against an individual and then ask whether on a global approach that individual is a danger to national security, taking into account the executive's policy with regard to national security. When this is done, the cumulative effect may establish that the individual is to be treated as a danger, although it cannot be proved to a high degree of probability that he has performed any individual act which would justify this conclusion.” (Emphasis added.)

In the House of Lords Lord Slynn of Hadley, after quoting Lord Woolf and adopting this approach, added this (para 22):

“Here the liberty of the person and the opportunity of his family to remain in this country is at stake, and when specific acts which have already occurred are relied on, fairness requires that they should be proved to the civil standard of proof. But that is not the whole exercise. The Secretary of State, in deciding whether it is conducive to the public good that a person should be deported, is entitled to have regard to all the information in his possession about the actual and potential activities and the connections of the person concerned. He is entitled to have regard to precautionary and preventative principles rather than wait until directly harmful activities have taken place, the individual in the meantime remaining in this country. In so doing he is not merely finding facts but forming an executive judgment or assessment. … There must be material on which proportionately and reasonably he can conclude that there is a real possibility of activities harmful to national security …” (emphasis added).

Lord Hoffmann, in arriving at the same conclusion, said this (para 56):

“In any case, I agree with the Court of Appeal that the whole concept of a standard of proof is not particularly helpful in a case such as the present. In a criminal or civil trial in which the issue is whether a given event happened, it is sensible to say that one is sure that it did, or that one thinks it is more likely than not that it did. But the question in the present case is not whether a given event happened but the extent of future risk. This depends upon an evaluation of the evidence of the appellant’s conduct against a broad range of facts with which they may interact. The question of whether the risk to national security is sufficient to justify the appellant’s deportation cannot be answered by taking each allegation seriatim and deciding whether it has been established to some standard of proof. It is a question of evaluation and judgment, in which it is necessary to take into account not only the degree of probability of prejudice to national security but also the importance of the security interest at stake and the serious consequences of deportation for the deportee.” (Emphasis added.)

In the recent case of R (Begum) v Special Immigration Appeals Commission [2021] AC 765, which concerned the exact nature of the appellate role of SIAC in relation to different decisions under legislation which had changed from time–to–time, Lord Reed (at para 59) helpfully identified a difference of emphasis between Lord Slynn and Lord Hoffmann and observed that the latter was more consistent with the orthodox modern approach of public law. But that difference related to the proper appellate approach. It did not affect the reasoning of Lords Slynn and Hoffmann as to the original decision-making process, which both agreed was not confined to determining whether specific allegations of past conduct were or were not proved. In particular, Lord Slynn, in referring to the requirement of fairness that past acts should be proved, was saying no more than if, on such an exercise, specific past activity was to be relied on as having occurred, that activity must be proved to have happened. He was, explicitly, not saying that this was the only basis on which a decision as to future risk could be arrived at. That would have been entirely inconsistent both with his adoption of Lord Woolf’s formulation at para 44 in the Court of Appeal and with his own statement that such fact finding, if undertaken, was “not the whole exercise” (per Lord Reed, para 53).

There are clear differences between the SIAC cases and the role of the Board with which this appeal is concerned. First, in the SIAC cases the Secretary of State is the primary decision-maker and he or she is a member of the executive who is not acting in a judicial capacity. Secondly, before April 2003, SIAC performed a fact-finding role but it did so in an appellate capacity. By contrast, the Board acts as a judicial body (see para 3 above) and, in the circumstances with which this appeal is concerned, makes the primary decision on whether to direct that the prisoner be released. Notwithstanding those differences, we consider that the observations of Lord Slynn and Lord Hoffmann, which we have quoted in paras 41 and 42 above, support the view that a decision-maker, whether a member of the executive branch of government or a judicial body, when assessing future risk, is not as a matter of law compelled to have regard only to those facts which individually have been established on the balance of probabilities; the decision-maker, from the assessment of the evidence as a whole, can take into account, alongside the facts which have been so established, the possibility that allegations, which have not been so established, may be true.

An example of the process of risk assessment by a judicial officer, which may be a closer analogy, is the consideration by a magistrate of an application for bail where someone is brought before the justice after arrest for suspected breach of his bail conditions. Under section 7(5) of the Bail Act 1976 the justice has to grant bail unless of the opinion that the person is not likely to surrender to custody or has broken or is likely to break a bail condition. In R (Director of Public Prosecutions) v Havering Magistrates Court [2002] 1 WLR 805, Latham LJ expressed the view that in carrying out the assessment of the relevant risk the magistrate was not restricted to admissible evidence in the strict sense. He stated (para 41):

“What undoubtedly is necessary, is that the justice, when forming his opinion, takes proper account of the quality of the material upon which he is asked to adjudicate. This material is likely to range from mere assertion at the one end of the spectrum which is unlikely to have any probative effect, to documentary proof at the other end of the spectrum. The procedural task of the justice is to ensure that the defendant has a full and fair opportunity to comment on and answer that material. If that material includes evidence from a witness who gives oral testimony clearly the defendant must be given an opportunity to cross-examine. Likewise, if he wishes to give oral evidence he should be entitled to. The ultimate obligation of the justice is to evaluate the material in the light of the serious potential consequences to the defendant, having regard to the matters to which I have referred, and the particular nature of the material, that is to say taking into account, if hearsay is relied upon by either side, the fact that it is hearsay and has not been the subject of cross-examination, and form an honest and rational opinion.”

In Sim [2004] QB 1288, which we discussed in para 8 above, Keene LJ, delivering the leading judgment with which Munby J and Ward LJ agreed, expressed approval of this analysis and suggested (para 57) that it was generally applicable to proceedings before the Board when it was assessing risks. He went on to state that the Board could take into account hearsay evidence on a disputed factual matter and should normally bear in mind that the evidence was hearsay when attaching weight to it. He envisaged the possibility that the evidence in question might be so fundamental to the decision that fairness required that the offender had an opportunity to test it by cross-examination before it was taken into account at all. He did not suggest that in assessing risk, the Board could have regard only to matters which it had found as fact on a balance of probabilities but instead appears to have envisaged a holistic exercise in which the Board evaluates all the material before it in its assessment of risk. This approach is consistent with the judgment of Elias J at first instance in Sim [2004] QB 1288 in which he emphasised (para 60) that the Board was “considering a question of judgment which raises many considerations” in reaching a view on whether the conduct of the prisoner was such that the risks could not be controlled in the community. This was, he stated, “not simply a factual issue”.

The suggested Children Act analogy

We turn then to address s 31(2) of the Children Act 1989, on which Mr Rule KC relies, and conclude that even in that specific context there is no general rule that an unproven fact can never be considered by way of an assessment of risk.

By section 31(2) of the Children Act 1989 a court may only make a care order “if it is satisfied—(a) that the child concerned is suffering, or is likely to suffer, significant harm” attributable to the care given, or likely to be given, not being what it would be reasonable to expect a parent to provide. The conditions for the consideration of a care order are known in family law as “the threshold criteria”, or sometimes “the threshold conditions”. Only if the preliminary hurdle they raise is surmounted can the court move on to the evaluative judgment whether a care order is the right solution in the case before it, often termed the “welfare” or “disposal” stage of the process. It is established law that a child is “likely to suffer significant harm” if there is a real possibility or a real risk that such may happen to him; it is not necessary that harm is more likely than not to occur to him: In re H.

Sometimes, although not all that commonly, the contention that a child is likely to suffer significant harm is based squarely upon an allegation that a previous child was harmed in a specific way by one or both parents. Where this is the contention, In re H holds that unless that allegation is an established fact found by a court, it cannot be taken into account. It is not permissible to take into account the possibility that it may be true and thus that the parent(s) in question present a risk to the subject child whose future is now under consideration. This is now very well settled. The decision to that effect in In Re H has been specifically endorsed by the House of Lords and later by this court in at least six subsequent cases including: Lancashire County Council v B [2000] 2 AC 147, In re O (Minors) (Care: Preliminary Hearing) [2004] 1 AC 523, In re B (children) [2009] 1 AC 11(above), In re B (A Child) (Care Proceedings: Threshold Criteria) [2013] 1 WLR 1911. There is no occasion to begin to question it. This is the basis of Mr Rule KC’s submission in the present case that the Board cannot consider at all any allegation which is not proved on the balance of probabilities to be true.

The origin of the suggestion that there is an analogy between the Children Act cases and the Board’s statutory test may lie partly in some observations of Wall LJ, concurring in the outcome, in Brooks. It is true that he, speaking from a wealth of experience of child care cases, pointed out in Brooks that the Board panel appeared to have proceeded by way of finding the facts in relation to a disputed allegation (there of rape) and then answering the statutory question informed by the facts which it had found. So it had, but the question which arises in the present case, namely whether that is the only and mandatory way in which the Board can answer its statutory question, simply did not arise in Brooks. The panel had found the rape established on the balance of probabilities. The issue before the court was whether it was entitled to do so on the basis of the hearsay account of the complainant, and the answer to that question was “yes”. The Board had not purported to rely on the decision in In re H, except in relation to the standard of proof—see para 22 in the judgment of Kennedy LJ, and that was a topic which remained misunderstood until resolved by this court in In re B (children) in 2009([2009] 1 AC 11). Wall LJ certainly commended the In re H approach to fact finding, but the possible analogy was not addressed by either Kennedy or Clarke LJJ and does not appear to have been argued; it cannot be said that Brooks represents any decision that an analogy applies and ought to be applied as a mandatory principle of procedure.

It is no doubt true that the concept of risk is common to the two legal exercises undertaken by the family court and the Board. The family court cannot consider a care order unless, in the type of case postulated, there is a real risk that the child may suffer significant harm in future. The Board, in addressing the statutory question, is inevitably concerned with assessing what risk the prisoner would present to the public if released. It does not, however, follow that the two legal exercises are truly analogous, or that the rule in In re H is capable of transplantation into the decision-making process of the Parole Board.

An examination of the decision in In re H demonstrates clearly that there are two separate reasons which underlie it which have no application to the Board process.

First, there is a powerful legal policy reason for a restrictive construction of the threshold criteria. They were specifically designed as a brake on the potentially very intrusive powers of the state to break up families by removing children from their homes. This was part of the reasoning of Lord Nicholls in In re H, and has been a recurrent theme since. At para 80 (p 588 ) Lord Nicholls prefaced his discussion of the question whether proof of the earlier allegation was a sine qua non for passing the threshold criteria with the heading “Suspicion and the threshold conditions”. After discussion of the practice of courts generally, and of the statutory context afforded by neighbouring provisions of the Children Act, he concluded by reverting to the impropriety of proceeding to a care order on the basis of suspicion.

“100.

…. Before the section 1 welfare test and the welfare “checklist” can be applied, the threshold has to be crossed. Therein lies the protection for parents. They are not to be at risk of having their child taken from them and removed into the care of the local authority on the basis only of suspicions, whether of the judge or of the local authority or anyone else. A conclusion that the child is suffering or is likely to suffer harm must be based on facts, not just suspicion.”

In the years since In re H the concern of the family courts to ensure that there is no scope for unwarranted social engineering by way of the removal of children on grounds simply that they would be better off in care has received repeated emphasis. Parents may well be feckless, dishonest, aggressive or otherwise thoroughly unsatisfactory, but their children do not fall to be removed unless there is a real risk of significant harm befalling them. As Lady Hale, who was concerned with the framing of the Act, has several times reminded the profession, the threshold criteria were put in place as a protection for parents. In In Re B [2009] 1 AC 11 (above) at para 54 she said this:

“The reasons given by Lord Nicholls for adopting the approach which he did in In Re H remain thoroughly convincing. The threshold is there to protect both the children and their parents from unjustified intervention in their lives. It would provide no protection at all if it could be established on the basis of unsubstantiated suspicions … ”

She said the same in In re J [2013] 1 AC 680 when this court rejected an attempt to reverse In re H. At para 44:

“Time and again, the cases have stressed that the threshold conditions are there to protect both the child and his family from unwarranted interference by the state. There must be a clearly established objective basis for such interference. Without it, there would be no ‘pressing social need’ for the state to interfere in the family life enjoyed by the child and his parents which is protected by article 8 of the ECHR. Reasonable suspicion is a sufficient basis for the authorities to investigate and even to take interim protective measures, but it cannot be a sufficient basis for the long term intervention, frequently involving permanent placement outside the family, which is entailed in a care order.”

Lord Wilson gave similar reasons in In re J for affirming In re H. At para 75 he said:

“My view remains that the need for the local authority to prove the facts which give rise to a real possibility of significant harm in the future is a bulwark against too ready an interference with family life on the part of the state. And, subject to the caveat that the court received no argument on the impact of article 8 of the European Convention on Human Rights, I incline to the view that nothing less than a factual foundation would justify such grave interference with the rights of the child and the parents thereunder to respect for their family life: see Olsson v Sweden (1988) 11 EHRR 259, in which, at paras 67 and 68 (which it has cited with approval on many subsequent occasions), the European Court of Human Rights stressed that a child’s removal into care was justified only if it was necessary in a democratic society in the sense that it corresponded to a pressing social need and was based on reasons which were relevant and sufficient.”

The several judgments of this court in In re B (A Child) (Care Proceedings: Threshold Criteria) [2013] 1 WLR 1911, which is principally concerned with the proper approach on appeal to decisions in care cases, repeat yet again the concerns of family courts that the intrusion involved in a care order mandate an especially careful approach to making such an order. Moreover, these powerful statements of legal policy mirror several others made by senior family judges to the effect that the greatest care must be taken to justify care orders (see for example the cases summarised by this court in In re H-W (Children) [2022] 1 WLR 3243, at para 47).

There is no equivalent policy consideration in the statutory question which the Board must address. Of course, if a prisoner is safe to release, he should be released, but the decision the Board makes whether to bring his imprisonment, under sentence of the court, to an end is not comparable to the intrusion involved in the removal by the state of children from their parents, siblings and homes. The process of the family court hearing a care application is necessarily adversarial; the local authority makes an application for an order and it is for it to justify it. The process of the Board is essentially inquisitorial, albeit with the participation of interested parties such as the prisoner and the Secretary of State if he wishes.

The second factor in In re H,which is not mirrored in the legislation governing the Board, was found by Lord Nicholls in the legislative context. The Children Act 1989 contains, as well as section 32 relating to the making of indefinite care orders, several provisions for interim or emergency measures when children may be at risk. Lord Nicholls relied heavily on the contrast with section 31(2).

“90.

There are several indications in the Act that when considering the threshold conditions the court is to apply the ordinary approach, of founding its conclusion on facts, and that nothing less will do. The first pointer is the difference in the statutory language when dealing with earlier stages in the procedures which may culminate in a care order. Under Part V of the Act a local authority are under a duty to investigate where they have "reasonable cause to suspect" that a child is suffering or is likely to suffer harm. The court may make a child assessment order if satisfied that the applicant has ‘reasonable cause to suspect’ that the child is suffering or is likely to suffer harm. The police may take steps to remove or prevent the removal of a child where a constable has "reasonable cause to believe" that the child would otherwise be likely to suffer harm. The court may make an emergency protection order only if satisfied there is ‘reasonable cause to believe’ that the child is likely to suffer harm in certain eventualities. Under section 38 the court may make an interim care order or an interim supervision order if satisfied there are ‘reasonable grounds for believing’ that the section 31(2) circumstances exist.

91.

In marked contrast is the wording of section 31(2). The earlier stages are concerned with preliminary or interim steps or orders. Reasonable cause to believe or suspect provides the test. At those stages, as in my example of an application for an interlocutory injunction, there will usually not have been a full court hearing. But when the stage is reached of making a care order, with the far-reaching consequences this may have for the child and the parents, Parliament prescribed a different and higher test: ‘a court may only make a care or supervision order if it is satisfied … that … the child . . . is suffering, or is likely to suffer, significant harm … ’ This is the language of proof, not suspicion. At this stage more is required than suspicion, however reasonably based.”

There is no equivalent pointer from statutory context in the case of the Board’s test.

These two differences apart, it is apparent from the Children Act cases that neither they generally, nor Lord Nicholls in particular, recognise any overriding legal principle that courts can only act on proven facts. After In re H the family courts had to confront the cases where an allegation of past fact was made in the course of care proceedings, and the trial court was unable to reach a conclusion, even on the balance of probabilities, about it. There are two principal possible cases in which this may happen. The first is where there is a proven allegation that the subject child sustained significant harm (for example shaking injuries or sexual abuse) but the court is unable to say which of the two adults in his family was responsible. This is likely to present acute questions when the parents have since separated and thus the subject child is now living with only one of the possible perpetrators of the harm. The second case is where the allegation of past significant harm is one on which the court cannot reach a conclusion as to whether it happened or did not — this may occur, for example, where the allegation is of sexual abuse.

These two cases were considered by the House of Lords in in In re O (Minors) (Care: Preliminary Hearing) [2004] 1 AC 523, where the principal speech was again delivered by Lord Nicholls. Largely by applying the policy considerations mentioned above, the House concluded that the two cases receive different answers. In the first case (“the uncertain perpetrator”) the threshold criteria are passed because the child incontrovertibly did suffer harm, and it is not necessary to enquire whether he is likely to suffer it in future. In that case, the court moves on to the “welfare” or “disposal” stage and considers whether a care order ought to be made or not. In doing so it is entitled to take into account the fact that the remaining parent, with whom the subject child still lives, is a possible (but unproven) perpetrator of the harm. Lord Nicholls expressed the conclusion in strong terms:

“27.

Here, as a matter of legal policy, the position seems to me straightforward. Quite simply, it would be grotesque if such a case had to proceed at the welfare stage on the footing that, because neither parent, considered individually, has been proved to be the perpetrator, therefore the child is not at risk from either of them. This would be grotesque because it would mean the court would proceed on the footing that neither parent represents a risk even though one or other of them was the perpetrator of the harm in question.”

This passage was expressly approved by this court in Inre S-B at para 24 per Lady Hale.

Conversely, in the second case, where previous harmful injury to the child is not proven, Lord Nicholls held in In re O that the rule is the opposite:

“37 … At the welfare stage, to what extent may the court take into account the possibility that the non-proven allegation might, after all, be true?

38.

This raises a question of legal policy. On the one hand there is the family protection purpose of the threshold criteria. On the other hand there is the general principle that at the welfare stage the court has regard to all the circumstances. On balance, I consider that to have regard at the welfare stage to allegations of harm rejected at the threshold stage would have the effect of depriving the child and the family of the protection intended to be afforded by the threshold criteria. Accordingly, at the welfare stage in this type of case the court should proceed on the footing that the unproven allegations are no more than that.”

These clearly expressed decisions demonstrate, beyond argument, first that In Re H cannot be an application of a universal legal principle that an unproven fact can never figure in legal reasoning, and second that, to the contrary, when one comes to assessing risk the unproven possibility that a fact may be true is a factor which can in some circumstances properly be taken into account.

The plain fact is that there is no sound analogy between the Board’s process and the passing of the threshold criteria in a care case. Quite apart from anything else, the exercise is not the same because the default position is different. In the context of section 31(2) of the Children Act, the default position is that if the threshold conditions are not met the subject child must stay with his family, whatever their inadequacies, peculiarities or faults. By contrast, the prisoner is already subject to sentence. The default position is that if the statutory test for release is not met, he must remain in custody.

The reasoning from policy in In re H is conveniently summarised in Lord Nicholls’ proposition at para 100, set out at para 53 above. It is not enough to pass the threshold criteria for the making of a care order that there is a “real possibility” that on a previous occasion a child was caused significant harm. But there is neither logical nor principled occasion to carry this reasoning over from the case of state social engineering by breaking up a family to the case of whether a prisoner subject to a sentence of imprisonment which is not yet expired should or should not be released early on the grounds that it is no longer necessary that he should be confined. Indeed, one would positively expect that a “real possibility” that such a prisoner would cause further significant harm to the public would mean that it is impossible to be satisfied that it is no longer necessary for him to be confined.

Summary on the binary approach

Having set out at some length our reasons for rejecting the suggested existence of a general principle of law that the law knows only the binary concept of fact and non-fact, we summarise our conclusions from the cases which we have discussed:

The concept of fair proceedings

In the conduct of its proceedings the Board must comply with the requirements of procedural fairness, which is the modern term for the rules of natural justice. Those requirements are, as Singh LJ explained in R (Talpada) v Secretary of State for the Home Department [2018] EWCA Civ 841 (“Talpada”) para 57, first that the decision-making body is impartial, and must not be, or appear to be, biased, and, secondly, that the decision-making body is under a duty to hear the other side: ie that the person whose legally protected interests may be affected by its decision must be given the opportunity to make representations to the decision-maker before the decision is taken. The duty of procedural fairness is a flexible obligation to adopt fair procedures which are appropriate and adapted to the circumstances of the particular case. In Kioa v West (1985) 60 ALJR 113,127 Mason J stated:

“In this respect the expression ‘procedural fairness’ more aptly conveys the notion of a flexible obligation to adopt fair procedures which are appropriate and adapted to the circumstances of the particular case. The statutory power must be exercised fairly, that is, in accordance with procedures that are fair to the individual concerned in the light of the statutory requirements, the interests of the individual and the interests and purposes, whether public or private, which the statute seeks to advance or protect or permits to be taken into account as legitimate considerations …”

This statement was cited with approval by Lord Bingham of Cornhill in R (West) v Parole Board [2005] 1 WLR 350, para 28, in which he held that in considering what procedural fairness requires, account must be taken of the interests at stake: para 30.

This court has held that, given the importance of the issues at stake for a prisoner, and particularly for a prisoner subject to an indeterminate sentence who remains in prison following the expiry of his minimum or tariff term, the Board has to hold an oral hearing before making its decision, for example where important facts were in dispute or significant explanations or mitigations were advanced which needed to be heard orally: R (Osborn) v Parole Board (above) [2014] AC 1115, paras 2, 83-85 per Lord Reed.

The rules of substantive fairness in public law are, as Singh LJ explained in Talpada paras 59–61, closely related to the doctrine of substantive legitimate expectation and substantive unfairness entails behaviour on the part of a public authority that amounts to an abuse of power (see also R (Gallaher Group Ltd) v Competition and Markets Authority [2019] AC 96, paras 40–41 per Lord Carnwath, para 50 per Lord Sumption). While Singh LJ’s and this court’s comments were made in the context of decisions by the executive branch of government, it is important in the context of a judicial review of the decisions of a judicial body to adopt an analogous rigour in relation to the requirements of substantive fairness. In particular, we do not see any basis for an assertion that substantive fairness requires a judicial body to treat as if they were “facts in issue” information which, on a proper analysis, is not of that nature. In so far as Macur LJ may have been suggesting otherwise in para 43 of her judgment, we would respectfully disagree. The question therefore is one of procedural fairness. We do not understand Mr Rule KC in his submissions on the requirements of fairness to argue otherwise; he squarely presents the case as one of procedural fairness. But, as explained above, we do not accept his submission that procedural fairness requires the Board to apply the binary concept of fact and non-fact to all allegations.

When making a risk assessment in relation to a person which might result in seriously adverse consequences to him, a decision-maker must have in mind the consequences to that person of an adverse decision against him. This can be seen as a question of fairness: see for example Lord Slynn in Rehman (above) at para 22. It can also be seen as a relevant and necessary component of the evaluation which the decision-maker is mandated to make: see Lord Hoffmann in Rehman (above) para 56. In R v Parole Board, Ex p Watson [1996] 1 WLR 906 Sir Thomas Bingham MR took the latter view, describing the role of the Board in making its decision whether to direct the release of a prisoner serving a discretionary life sentence as a balancing exercise in which preponderant weight is to be given to the need to protect members of the public. He stated (p 916–917):

“In exercising its practical judgment the board is bound to approach its task … balancing the hardship and injustice of continuing to imprison a man who is unlikely to cause serious injury to the public against the need to protect the public against a man who is not unlikely to cause such injury. In other than a clear case this is bound to be a difficult and very anxious judgment. But in the final balance the board is bound to give preponderant weight to the need to protect innocent members of the public against any significant risk of serious injury.”

We view the prisoner’s interest as a component in the evaluation but would question whether it is correct to describe the task as a balancing exercise. In our view the Board’s task is to apply the statutory test of asking whether it is satisfied that it is no longer necessary for the protection of the public that the prisoner should be confined. In carrying out that difficult task the Board must bear in mind the risk of injustice to the prisoner of which Sir Thomas Bingham spoke.

A failure by the Board to give due consideration to the adverse consequences of a decision not to direct release would amount to a failure to take into account a relevant consideration. The matter is, as Lord Bingham stated, often one involving a difficult and anxious judgment.

In R v Johnson (Practice Note) (Attorney General’s Reference (No 64 of 2006) [2007] 1 WLR 585, a case giving guidance on the imposition of a sentence of imprisonment for public protection, the Court of Appeal, in a judgment handed down by Sir Igor Judge P, stated (para 10):

“To the extent that a judge is minded to rely upon a disputed fact in reaching a finding of dangerousness, he should not rely on that fact unless the dispute can fairly be resolved adversely to the defendant.”

In R v Considine [2008] 1 WLR 414, in a judgment also delivered by Sir Igor Judge P concerning the assessment of dangerousness in the context of sentencing under section 229 of the CJA 2003, the Court of Appeal endorsed (at para 35) the comment in R v Johnson which we have quoted above. The Court of Appeal went on to state (para 37):

“We have deliberately declined to lay down any hard and fast rules about how the court should approach the resolution of disputed facts when making the section 229 assessment. In reality there will be very few cases in which a fair analysis of all the information in the papers prepared by the prosecution, events at the trial, if there has been one, the judicial assessment of the defendant’s character and personality (always a critical feature in the assessment), the material in mitigation drawn to the attention of the court by the defendant’s advocate, the contents of the pre-sentencing report, and any psychiatric or psychological assessment prepared on behalf of the defendant, or at the behest of the court itself, should not provide the judge with sufficient appropriate information on which to form the necessary judgment in relation to dangerousness.”

The assessment of (future) dangerousness at the stage of sentencing has much in common with the statutory question which the Board is required to answer when deciding whether or not to release. There is, however, a difference: at the sentencing stage the court is deciding what sentence to impose, whereas the Board is dealing with a prisoner whose sentence is determined, and who remains subject to it, with the proviso that he may be released on licence if the Board is satisfied that it is no longer necessary in the public interest for him to remain confined. The endorsement by the Court of Appeal in Considine of a holistic exercise in the court’s assessment of dangerousness when sentencing does not support by analogy the contention that fairness requires the Board to have regard only to facts established by admission or on the balance of probabilities in making its assessment. Procedural fairness in this context will involve giving the defendant the opportunity to state his position in relation to the disputed fact and to argue that no reliance should be placed on it.

The use of allegations in the board’s risk assessment

The central question is whether there is anything in the legal context of the Board’s role, when it addresses allegations of past criminal or otherwise risky behaviour, which confines the matters which the Board may take into account to proven facts of past behaviour while excluding from consideration in any circumstance the possibility that the unproven allegations might be true.

In our view there is not. The question at all times is one of statutory interpretation. The statutory remit of the Board (para 7 above) is that it may not direct the release of the prisoner unless it is satisfied that it is no longer necessary for the protection of the public that he should be confined. There is no further express statutory specification as to how the Board satisfies itself of that test. Nonetheless, the statutory provision does not stand alone; it is to be interpreted in the context of the general law. As we have reasoned above, there is no general legal principle that a court of law or the Board in making a risk assessment can have regard to evidence or information only if it is established as a fact by admission or on the balance of probabilities. Further, the public law concept of fair proceedings does not require that only facts so established are to be considered in a risk assessment. There is therefore no requirement to be implied into the statute that the Board must disregard the possibility that an allegation, which has not been established as true on the balance of probabilities, may be true.

That is not the end of the matter for it does not follow from that conclusion that the Board should not seek to resolve disputed facts by making findings of fact on the balance of probabilities where it is reasonably practicable to do so. In the Guidance under challenge (para 2) the term “allegation” is said to refer to conduct alleged to have occurred which has not been adjudicated upon. A prisoner with a history of past offending is vulnerable to such allegations. If an allegation could, if true, affect the Board’s risk assessment, the Board’s task, so far as it can on the information which has been made available to it or which it is able to obtain, is to explore the nature of that allegation and its surrounding circumstances in order to make such findings of fact as it can about either or both on the balance of probabilities. By this means the Board gives due consideration in its assessment both to the public interest and to the prisoner’s interests and acts with procedural fairness.

In doing so, the Board must as a matter of procedural fairness give the prisoner the opportunity to challenge the relevant evidence or information. If the allegation is a disputed issue of fact which is likely to be material to the outcome of the risk assessment or if issues of explanation or mitigation of accepted facts are likely to arise, the Board may, if it is reasonably practicable to do so, in compliance with its duty of procedural fairness, have to hold an oral hearing to receive oral evidence and allow cross-examination and oral submissions, before reaching a conclusion as to the truth of the allegation: R (West) v Parole Board (above), Lord Bingham, para 35, Lord Slynn of Hadley para 50, Lord Hope of Craighead paras 63 and 67.

In other circumstances, the Board may not be able to reach a conclusion that the relevant allegation is true or is untrue. It may be impossible to do so where the Board has not got the relevant evidence. As we have said, the Board cannot itself compel a reluctant witness to attend a Board hearing. Unless the Board or the prisoner takes the exceptional course of obtaining an order of the High Court to compel the attendance of a witness, attendance in response to a letter from the Board or a party is a voluntary act. In Brooks, Kennedy LJ recorded the observations of Judge Inigo Bing, the panel chairman and a Circuit Judge sitting regularly in the Crown Court, who gave three broad categories of reluctant witnesses. They were (para 24): (i) people who are in fear of the prisoner, (ii) children who are understandably unwilling to attend a Board hearing, which generally takes place in a prison establishment, and (iii) ex or current prisoners, or those with close relationships with a prisoner, who do not want to be responsible, and to be seen to be responsible, for putting someone back in prison. The statement was made in the context of a case involving the recall of an offender to prison. In our view Judge Bing’s third category would also extend to those who do not wish to be responsible, and seen to be responsible, for keeping a prisoner in prison. Further, people complaining of sexual offending may often be reluctant for reasons of privacy to become involved in a criminal process or in hearings before the Board. Where the Board does not have a witness statement from the relevant person, to which it is prepared to attach sufficient weight, after any challenge by the prisoner, it may not be possible to make findings as to the truth of the allegation. Similar problems may arise if the dossier provided to the Board by the Secretary of State does not enable such findings to be made.

It may also in certain circumstances be procedurally unfair for the Board to make findings of fact as to the truth of an allegation. For example, even if there is information available to the Board in support of an allegation or allegations, circumstances may prevent the prisoner from responding to or challenging the allegation in an effective manner. Such circumstances may arise if the allegation is or may be the subject of a criminal prosecution or prison disciplinary proceedings which have yet to be determined. In such circumstances the Board might be in danger of pre-empting the decision of a subsequent court or tribunal, and moreover on different evidence and to a different standard of proof. Further for the prisoner to answer questions before the Board about such an allegation may risk damaging his right to silence or otherwise prejudicing his defence in those other proceedings.

It does not, however, follow from the Board’s inability to make a finding as to the truth of an allegation in a particular case that the allegation is irrelevant and must be disregarded. Where the prisoner can comment on an allegation, for example in the absence of pending criminal or disciplinary proceedings relating to that allegation, the Board may use an allegation or allegations to test the credibility of the prisoner’s account of his behaviour and, as a result of his responses, reach conclusions favourable or adverse to the prisoner without reaching a concluded view on the veracity of those allegations (see for example R (D) [2019] QB 285, para 155). As we have discussed in para 23 above, in the present case the claimant’s responses to questions about the allegations caused the Board concern and were properly facts which were relevant to the assessment of the risk which he presented, as the judge and the Court of Appeal held. In particular, it was implausible that he could not remember details of the incidents for which he had been arrested and the Board made a finding that he must have had sexual contact with a 12-year-old girl.

In some cases, the Board may not be able to make a finding that allegations of criminal behaviour are true but nonetheless be able to make findings concerning the surrounding facts which persuade it that it is not safe to release the prisoner. In the present case, Macur LJ stated (para 35): “[e]stablished or undisputed constituent or consequential facts to an overarching allegation may provide compelling and convincing indications of risk in themselves”. We agree.

In R (Morris) v Parole Board [2020] EWHC 711 (Admin) the Divisional Court considered a challenge to the Board’s decision not to direct the prisoner’s release, based on its findings in relation to two incidents. The first, in 2014, concerned an allegation by the prisoner’s ex-partner that he had attended her home, been verbally abusive and had assaulted her. The ex-partner gave a witness statement to the police but later retracted her statement. The second incident, in 2017, concerned an allegation that he had harassed another woman with whom he may have had an intimate relationship by persistently trying to contact her or her friends. The police did not have sufficient evidence to charge him with harassment but issued him with a “prevention of harassment letter”, warning him that harassment was a criminal offence. The Board attempted unsuccessfully to obtain further information about the 2014 and 2017 incidents and had to make its determination in the absence of that information. The Divisional Court rejected the challenge that the Board was obliged to ignore the two incidents. McGowan J, delivering the judgement, at para 56 stated that there was sufficient evidential material to allow the Board to make some findings of fact. In relation to the 2014 incident the Board was entitled to rely on the fact that the prisoner had been arrested, his admission of a heated argument at his ex-partner’s house, and the existence of the retracted statement and the relevant police reports. In relation to the 2017 incident, the existence of the police statements and the “prevention of harassment letter”, which the Board was entitled to treat as having been issued in good faith and with a proper basis of fact, showed that the police had perceived a risk and so had warned him. The police statements and the prevention of harassment letter were information available to the Board, which indicated a risk that the allegation might be true. McGowan J concluded that the Board was entitled to consider both the 2014 and the 2017 allegations “in the context of well-established concerns about [the prisoner’s] relationships or friendships with women”. We agree with that approach, although we question below, when we discuss the Guidance, the utility of the concept of a “mere allegation” which the Divisional Court used in Morris and which found its way into the Guidance.

In Morris McGowan J at para 34 quoted the Divisional Court in R(D) where it stated that the Board’s evaluation of risk was in part inquisitorial and that its risk assessment was quintessentially a matter for the panel’s judgment. We agree. She also referred in para 35 to the undisputed point made by Kennedy LJ in Brooks, at para 28, that the Board is not determining a criminal charge but is concerned with the assessment of a more than minimal risk of further serious offences being committed in the future. Again, we agree. Whatever the Board’s conclusions, no conviction ensues and no punishment is imposed. The Board may sometimes have to address the question whether the making of an unresolved allegation, which the prisoner has not been able to challenge by cross-examination, causes it concern relevant to the statutory test where it does not treat the allegation as established to be true.

The observations of Stanley Burnton LJ in R (McGetrick) v Parole Board [2012] 1 WLR 2488 are in point. He stated at para 33:

“It is essential to bear in mind that it is not the function of the Board to find a prisoner guilty or innocent of any offence or other misconduct. Its function is to assess the risk that would be created if the prisoner is released on licence. For that purpose, the Board must take into account hearsay and other evidence of misconduct or criminal offences on the part of the prisoner, whether the misconduct or offence took place before or after or at the same time as the offending for which he was sentenced. Similarly, the Board must take into account evidence as to the relevant good conduct of the prisoner, whenever it took place. The weight, if any, to be given to that evidence is a matter for the Board.” (Emphasis added.)

In some cases, the number and nature of multiple allegations of a similar nature from independent sources might justify the Board in concluding that the prisoner had engaged in a course of conduct giving rise to risk to the public even if no single one of the allegations resulted in a finding of fact that it was true. This is for two reasons. First, the Board would be entitled to consider that it was improbable that a number of similar incidents alleged against the prisoner were false. Secondly, obvious similarities in various incidents may constitute mutual corroboration of those incidents. Whether the remarkable facts of R(D) might have presented such an example did not arise in either of the Board decisions relating to him. In other cases, where the Board has to address only one allegation of criminal behaviour or other conduct which could affect the analysis of risk, and that allegation cannot be proved, a holistic assessment of all the circumstances may persuade the Board that there is a significant chance, short of the balance of probability, that the allegation is true. We agree with McGowan J in Morris (para 53) that “a consideration of allegations which have not been established is not itself intrinsically unfair”.Procedural fairness would, nonetheless, require the Board to give the prisoner the opportunity to make submissions about how the Board should proceed.

It is necessary also to address R (Delaney) v Parole Board [2019] EWHC 779 (Admin) (“Delaney”) on which the Court of Appeal relied in this case in support of the proposition that the Board had to proceed by a two-stage process of making findings of fact and then making a risk assessment in reliance only on those findings. In that case a prisoner faced allegations of several incidents of domestic violence, which were said to have caused a fractured wrist and a perforated eardrum. The police evidence to the Board was that there were reasonable grounds for the belief that the prisoner had been guilty of violent offending but, despite efforts to do so, the police had not been able to obtain any evidence which it was satisfied would stand up in court. The prisoner accepted that he and the complainant had been drinking and had had heated arguments but denied any violence. The panel declined to make any findings of fact as to what had happened between the prisoner and the complainant but considered that the original complaints may well have been true and that there was therefore a real risk if the prisoner was released. The panel also recorded that the prisoner had overreacted angrily to an unrelated matter at his hearing. Andrew Baker J upheld the prisoner’s challenge that the Board’s decision was irrational. He founded on the erroneous approach of the panel in its decision letter, which was (i) the decision of the panel not to make any relevant findings of fact, and instead to conclude that the allegations may be true, and (ii) the flawed logic that the allegations led to an increase in risks. He was not in a position to consider whether there was evidence before the panel sufficient to enable it to make findings about the prisoner’s behaviour which might have justified its conclusions as to risk.

We do not read his judgment as holding that the Board could have regard only to found facts when making a risk assessment. Counsel’s proposition, which the judge accepted, was that an allegation of violence cannot, in itself, found a conclusion that a prisoner presents a particular risk of violence. We emphasise the words “in itself” because the authorities which counsel cited in support of the proposition, which the judge accepted, support the view that an allegation, in the form of a police charge or investigation, may in certain circumstances provide powerful evidence from which risk may be inferred when one has regard to the circumstances alleged in the charge: R (Broadbent) v Parole Board [2005] EWHC 1207 (Admin) , paras 26–29 per Stanley Burnton J as further explained in R (J) v Parole Board [2010] EWHC 919 (Admin) , para 48 per Irwin J; and McHale v Secretary of State for Justice [2010] EWHC 3657 (Admin), para 16 per Langstaff J. What appears to have been decisive in Delaney’s challenge is the express decision of the panel not to make any findings of fact as to the prisoner’s behaviour when it was open to it to do so.

For completeness, we consider briefly the decision of the House of Lords to which Mr Rule KC referred the court: In re D (Secretary of State for Northern Ireland intervening) [2008] 1 WLR 1499. In that case the Life Sentence Review Commissioners, acting under Northern Irish legislation which was analogous to that with which this case is concerned, decided on the balance of probabilities that the prisoner had been guilty of sexual assaults on children of which he had not been convicted because the charges had been withdrawn in the interests of the complainants and no prosecution had ensued. The relevant challenge by the prisoner, which the House of Lords rejected, was that the panel had erred in not applying a standard higher than the balance of probabilities. The House did not consider any argument as to what the panel could have done in its assessment of risk if it had not been able to make the findings relevant to risk on a balance of probabilities. It may fairly be said that their Lordships proceeded on the assumption that the allegations had to be proved on the balance of probabilities, but nobody argued otherwise. The argument, which this court addresses, has arisen only recently, since R(D) was decided, and the court does not derive any assistance in relation to that question from this decision.

We summarise our conclusions as follows:

The terms of the Guidance

We have set out in paras 16–18 above the relevant paragraphs of the Guidance. The Court of Appeal at para 51 of its judgment held that paras 6(c), 9(3), and 18-24 were inconsistent with correct legal principles. This finding was based on the view that the Board in law could have regard in its risk assessment only to those facts which it held to be established on the balance of probabilities. For the reasons which we have set out above we do not agree with that analysis.

In our view para 6 of the Guidance is not unlawful but could be more clearly rephrased to reflect the analysis that the Board should, if it can, make relevant findings of fact. The same point applies to para 9(3) of the Guidance.

Para 18 of the Guidance is in our view a correct statement of the law subject to our comments below on the final sentence which appears to have adopted the concept of a “mere allegation” from Delaney and Morris. Para 18 of the Guidance is expressly stated to apply where “the panel is not in a position to make a finding of fact either because there is insufficient material available to make such a finding on the balance of probabilities, or because it would not be fair to do so.” In our view it is only in such circumstances that the Board should rely on its assessment of concerns about an unproven or disputed allegation. In that context, procedural fairness would require the Board to give the prisoner the opportunity to argue, for example by having regard to the considerations listed in para 20 of the Guidance, that no account should be placed on unproven allegations, including because there was no serious possibility that the allegations were true, or that only very limited account should be so placed.

What the Guidance goes on to set out in paras 19–24 falls to be read against that background and in that context is in our view unobjectionable, but again subject to our comments below on the use of the “mere allegation” which reappears in para 24.

We are not persuaded that the Guidance is unlawful but we are satisfied that it could be redrafted to emphasise the importance of making proper findings of fact when it is possible for the Board to do so and the requirements of procedural fairness which we have addressed above. We also note the criticisms of the drafting of the Guidance by both Bourne J and Macur LJ. It is not the task of this court to rewrite the Board’s Guidance. Our principal concern is that paras 6, 9, and the opening words of para 11 (“Panels may need to make a finding of fact regarding the allegation …” (emphasis added)) can be read as giving insufficient emphasis to the importance of making relevant findings of fact when it is possible to do so. Further, the succinct statements in paras 6 and 9 appear in contrast to the more discursive presentation of the assessment of levels of concern in paras 18–24 and may benefit from greater emphasis. It should be straightforward to provide greater clarity and to put paras 18–24 in their proper context.

Finally, as we have trailed above, we do not find the concept of a “mere allegation” and the contrast between that and an allegation for which there is some factual basis to be helpful. There may be some cases where the fact that an allegation has been made should be treated as carrying no weight. For example, if the police interviewed a person in relation to an offence solely on the basis that he had a previous conviction for an analogous offence and without anything otherwise to suggest he might have been the perpetrator of the offence in question, the fact of such an interview could carry no weight. We agree that, generally, a simple report of an accusation, without more, is likely to be incapable of informing the Board’s decision on the statutory question, unless there is material indicating the source and circumstances of the complaint. But such circumstances are likely to be rare. There will often be a police report or a witness statement by the complainant which provides some evidential basis. Creating a dichotomy between a “mere allegation” and an allegation for which there is some factual basis unwarrantably shifts the focus from where it should be, which is on assessing the quality of the evidence of the circumstances surrounding the allegation to determine what, if anything, can be established as relevant facts either as to the truth of the allegation or as to the surrounding circumstances. In the Court of Appeal’s judgment in this case, Macur LJ gives a helpful example of the latter where she speaks of a dossier of a prisoner, who has been convicted of sexual assaults against children, which contains evidence of his being arrested or questioned as a person of interest on another occasion regarding a sexual assault. If the prisoner had been arrested on an occasion other than those which led to his convictions because he was often seen in the children’s playground in which the child had been assaulted, the Board would properly question him on that matter. If his frequent presence in proximity to the playground was established or undisputed, then in the absence of a plausible explanation for his presence, it would point towards risky behaviour which was relevant to the risk assessment. We note that in each of Delaney, Morris and this case there was evidence of facts connected with the allegation of criminal behaviour which the Board could have used or did use as a basis for assessing that the prisoner posed a significant risk to the public. As has often been stated, the weight to be attached to evidence is a matter for the Board subject always to a challenge on the ground of public law irrationality.

Conclusion

We would allow the appeal and hold that the Board’s Guidance on Allegations is lawful. We would nevertheless invite the Board to review the terms of the Guidance in the light of this judgment.