Danny Osborne, R (on the application of) v Parole Board for England & Wales

Mr Osborne was sentenced to an indefinite term of imprisonment for public protection in 2008, for an offence of robbery. His minimum period having expired, he was released on licence in 2015 but recalled later that year. A similar thing happened in 2016. He was released again in June 2019.

On 21^st October 2020, he was arrested for allegedly stalking and assaulting a former partner. His licence was revoked on 22^nd October. He returned to custody the following day.

His case was referred by the Secretary of State for Justice to the Parole Board on 23^rd November 2020, and heard by a panel on 8^th November 2021. The panel comprised two independent members (including the chair) and a psychiatrist; there was no psychologist member. It received written and oral evidence, hearing from Dr Bromley (forensic psychologist), Ms Brownsey-Joyce (senior prison probation officer) and Ms Lynn (community offender manager). Mr Osborne gave evidence and was legally represented by Mr Coningham, who continues to represent him in these proceedings. The panel adjourned for further written evidence and submissions.

By a decision dated 6^th January 2022, the panel declined to direct Mr Osborne’s release. He submitted an application for reconsideration on 30^th January. That was rejected by a decision dated 23^rd February.

By these judicial review proceedings, and with permission granted on all grounds, Mr Osborne challenges the decision of the Parole Board not to direct his release, and the rejection of his application for reconsideration. The Parole Board maintains a neutral stance, as is its common practice.

The refusal to release

The Parole Board panel decision of 6^th January sets out its reasoned analysis in three sections. First, there is an analysis Mr Osborne’s offending behaviour (‘the past’). This considered his criminal record, as well as further ‘unproven allegations’ of violence. The decision states (at [1.5]) that:

The panel places only minimal weight on those unproven allegations, which are nevertheless indicative to that extent of the possibility that Mr Osborne’s violent offending has been more extensive than the offending for which he has been convicted.

It recorded Dr Bromley’s analysis of his history and ‘precipitators of violence’.

Second, the decision analyses evidence that Mr Osborne had changed (‘the present’). It considered the programmes he had completed, the progress he had made in addressing his offending behaviour, and the history of his recalls, and in particular the events of October 2020. It notes Mr Osborne’s denial of the allegations of stalking and assaulting his former partner and that he had not been charged in this connection, the complainant having withdrawn her original statement to the police. It considered the allegations nevertheless relevant to its assessment of the justification of Mr Osborne’s recall and his present risk. So it directed itself to whether it was fair and appropriate to make findings of fact regarding this particular alleged behaviour.

It noted that Mr Osborne’s convicted violent offending included assaulting a different intimate partner, and his own account that the relationship with the later complainant had been violent (on her part). It noted the matter had not been proceeded with by the police, in view of the complainant’s retraction, but also that she had withdrawn support for further proceedings, not retracted the allegations. It noted there were witness statements on the police file, and a photograph of the complainant at the time with facial bruising consistent with this evidence. It considered Mr Osborne’s own account of the events, which was that he himself had been a victim of assault by third parties. It rejected his account as ‘not credible’. It concluded (at [2.29]):

Considering the allegations in the round, the panel considers that it is in a position to find, as a fact, that Mr Osborne was violent and abusive towards Ms G on the 21/10/2020, including causing a bruise to her right eye. The panel notes that Mr Osborne was, by his own admission, intoxicated with benzodiazepine at the time of the violent and abusive behaviour towards Ms G.

Third, the decision analyses the manageability of risk (‘the future’). It summarises and considers the evidence of Ms Lynn and Dr Bromley about this. It records that during the hearing the panel had explored with these witnesses the interplay between Mr Osborne’s underlying personality disorder (antisocial personality disorder – APSD) and his attention deficit hyperactivity disorder (ADHD). Both witnesses had said that while ADHD might be relevant, it was not a primary driver of his offending. The panel noted its agreement with that analysis, adding that it considered ‘the primary drivers of Mr Osborne’s offending are more likely to be his underlying traits of ASPD’. It noted that Ms Lynn and Dr Bromley considered Mr Osborne suitable for re-release, because of his willingness to work with Dr Bromley’s assessment of his risk and recommendations for managing it in the community. This was also supported by Mr Osborne’s prison offender manager Mr Bennett.

In reaching its final conclusions, the panel took into account Mr Osborne’s convicted violent offending (including towards a former partner); his own account of the violence in his relationship with the complainant Ms G (albeit he attributed that to her); its finding of fact that he had been violent and abusive to her in October 2020; the fact that he had failed to report to Ms Lynn that he had recommenced this relationship; as well as his largely positive conduct in prison, and concluded that recall was justified.

It considered (at [4.7]) that:

underlying traits of ASPD are likely to be the primary drivers of Mr Osborne’s violent and abusive behaviour in intimate relationships, and of his inability or unwillingness to be open and honest with those who are tasked with assessing his risk.

It went on to conclude as follows (at [4.9]-[4.12]):

the panel considers that Mr Osborne’s repeating pattern of emotional mismanagement, violence, and his inability or unwillingness to be open and honest is indicative of a need for intervention to address his underlying ASPD traits. The panel considers that there is insufficient evidence that Mr Osborne would be likely to engage with any voluntary measures targeting ASPD traits in the community. The proposed BBR [building better relationships] programme would not address the ASPD traits and the risk management plan does not include any mandatory intervention in the community that would do so.

The panel also notes the context of Mr Osborne’s history in which he has committed further offences while on licence.

It is for those reasons that the panel considers that the proposed risk management plan is unlikely to be effective in protecting the public. The panel considers that Mr Osborne’s risk of serious harm will be imminent when he enters into or resumes an intimate relationship and openness and honesty, which was lacking when he was in the community most recently, is essential to effective risk management of Mr Osborne outside of custody.

The panel therefore considers that it remains necessary that Mr Osborne remains confined in prison.

The decision turns finally to the question of suitability for open conditions. It noted past absconding from open conditions, and that allegations had been made that he had stabbed someone while unlawfully at large – ‘although the panel has given minimal weight to those allegations’. It also noted that ‘the higher intensity intervention targeting underlying traits of ASPD that is required in Mr Osborne’s case prior to his release could not be made available to him in open conditions’. It concluded he was unsuitable for open conditions.

The reconsideration decision

Mr Osborne applied under Rule 28(1) of the Parole Board Rules 2019 for reconsideration of the decision to refuse release. His grounds were:

Irrationality

a)

The panel was not entitled to find (as it did) that the Applicant had assaulted a former partner.

b)

The panel was not entitled to place weight (as it did) on unproven allegations.

c)

The panel wrongly rejected the recommendations of the professional witnesses.

Procedural unfairness

d)

The panel had a duty (which it ignored) to inform the parties that it considered the release plan inadequate.

The decision of Mr James Orrell for the Parole Board of 23^rd February rejected all of these grounds. He concluded:

Mr Osborne’s challenge in these judicial review proceedings is to both decisions.

In relation to the panel decision, he says the finding of fact that Mr Osborne assaulted his former partner, and the placing of weight on unproven allegations, were irrational and procedurally unfair (Grounds 1 and 2). He says the decision to reject the release recommendation of the professional witnesses was irrational and procedurally unfair (Ground 3). And he says the failure to notify him, his community offender manager Ms Lynn, and Dr Bromley, that it considered the release plan inadequate – so that no opportunity was given to provide further evidence – was procedurally unfair (Ground 4).

In relation to the reconsideration decision, he says the Parole Board failed to take into account findings of the panel relating to whether Mr Osborne’s recall on licence was justified (Ground 5).

Giving permission for judicial review on the papers, Morris J observed that grounds 3 to 5 seemed to be the stronger grounds, but that permission was being granted on all five grounds. Mr Coningham confirmed to me that Mr Osborne’s case rested on the existing documentation: there was no new information to be considered.

Mr Coningham directed my attention to the discussion of irrationality in R (Bousfield) v Parole Board [2021] EWHC 3160 (Admin) at [53]-[58], and asked me to bear in mind that to be rational and fair, the panel’s decision-making needed to take the right matters into account, and to give clear and adequate reasons, as well as being substantively within the range of outcomes properly available to it in all the circumstances.

In his closing submissions to the panel, Mr Coningham had acknowledged the panel was required to consider the assault allegation, and to make a finding of fact if appropriate; but that it had to act fairly and not make a finding unless it could be satisfied that the civil standard of proof had been met. He had drawn the panel’s attention to the Parole Board’s Guidance on Allegations. Paragraph 11(c) of the Guidance says this:

Panels may need to make a finding of fact regarding the allegation when…

(c)

the prisoner’s case can be fairly considered. The prisoner must have a fair opportunity to contest the allegations. This may be achieved through oral evidence, written submissions or in interview with an offender manager, depending what is fair in the case. Fairness may be particularly difficult to achieve as panel hearings do not have the safeguards that are present in criminal proceedings. Fairness may be particularly difficult with allegations arising out of events which happened a long time ago.

Mr Coningham says the evidence of the assault was so unsatisfactory no reasonable panel could have found it proved to the civil standard; and that it was made in a way which was unfair to Mr Osborne and in breach of the Guidance. He makes a number of specific points.

First, he says the panel was obliged to have made a finding about the complainant’s credibility, but there is no sufficient assessment or finding of that on the face of the decision. In her initial police statement she had alleged Mr Osborne had been violent to her throughout their relationship, but she later resiled from that, maintaining only the alleged violence of the October 2020 incident. He says this was a significant lie which fundamentally undermined her credibility and damaged Mr Osborne; but the panel did not deal with it.

Second, he says the panel was entitled to look at the contemporaneous photo of the complainant, showing her injury, but was wrong to treat her contemporary social media posts as capable of being corroborative.

Third, he says that although the decision records the panel directing itself to the question of making a finding about the stalking allegation as well as the assault allegation, it did not do so and did not explain why not.

Fourth, the panel failed to take proper account of the fact that the police investigation stopped at an early stage and no prosecution file was available; there was no complete set of evidence before the court and such evidence as there was contained allusions to other material suggesting its incompleteness.

Fifth, he points out that Ms Lynn’s evidence had contained indications that she had reservations about the complainant’s reliability and character and that there was undisclosed material potentially relevant.

Sixth, the panel insufficiently addressed the reasons for the prosecution not having been proceeded with.

And finally, he says the panel took a defective approach to Mr Osborne’s credibility. It had found him not to have given a truthful account of the October incident, particularly as regards his account that, before going to see the complainant, he himself had been assaulted. Mr Coningham says the panel did not give itself a Lucas-style direction, to the effect that not having told the truth about the antecedent matter did not mean he had not told the truth about not assaulting the complainant. He drew my attention in this connection to what is said about this in A, B & C (Children)[2021] EWCA Civ 451 starting at [54], and I have read that. The panel, he says, made an impermissible or generalised inference of Mr Osborne’s untruthfulness which was not properly relevant to its finding about the assault – or at least, inadequately explained its reasoning in this respect.

The underlying substance of most, if not all, of these points were put to the panel in Mr Coningham’s closing submissions, so Ground 1 is effectively a challenge to the way the panel responded, or failed to respond, to them. I have returned to the panel’s analysis in its 6^th January 2022 decision to consider that.

I can see that the panel’s route to its finding that Mr Osborne assaulted the complainant included a number of steps. It appears Mr Osborne acknowledged he had been at the complainant’s home overnight at the relevant time (in breach of his licence conditions). It is a matter of record that Mr Osborne was taken to hospital on 22^nd October with facial injuries, and lacerations and human bite marks on his body. It noted the photo of the complainant taken at the time showing her facial injuries.

The panel noted that the police dossier included a statement from an apparently independent witness who testified to seeing a man fitting Mr Osborne’s description, seeming intoxicated but not smelling of alcohol, being verbally abusive and aggressive to the complainant in the street, including screaming at her where’s my xani?. Xanax is a benzodiazepine, and Mr Osborne accepted he was intoxicated with benzodiazepine at the relevant time. The witness described a large bruise to the complainant’s right eye, consistent with the photo taken of her at the time.

So aside from the evidence of Mr Osborne and the complainant, the panel was starting from undisputed evidence they were both together on the night in question, and was satisfied that both were physically injured. That raised an obvious question about whether or not that was a coincidence. That was a starting point the panel was entitled to take on the materials before it, and, from my reading of the decision, clearly did take.

Apart from the witness statement in the police dossier, to which the panel was entitled to have at least some regard, the only other accounts capable of being considered relevant by the panel were those of Mr Osborne and the complainant.

The panel makes clear the principal reason it rejected Mr Osborne’s account was that it did not provide a reasonable, or indeed any, explanation for his injuries. Not only was it extremely vague, but the story of a random attack by strangers (or possibly at the instigation of the complainant) did not make sense on its own terms, and did not account for there being bite marks on his torso. That, together with there being no sign of any alternative explanation for the complainant’s simultaneous injury, is readily graspable as the core of the panel’s negative answer, on the balance of probabilities, to the question of whether the injuries were coincidence.

Similarly, the panel’s principal reason for accepting the complainant’s account was that it was a more probable explanation, objectively speaking, for the two sets of injuries. From that starting point, it was able to note (a) that the retraction of her original decision was fully explained on the basis that Mr Osborne was no longer at large and the complainant could not face further proceedings; (b) her account had been detailed and consistent; and (c) it was supported by the witness statement in the police dossier.

I am not persuaded in these circumstances that the panel’s decision, on its face, was anything other than entirely rational and adequately supported and explained. The more substantial question is whether it was vitiated by unfairness to Mr Osborne. I have thought about all the points of criticism Mr Coningham makes, but I am not persuaded that it is, for the following reasons.

As a general point, I accept that although Parole Board proceedings are civil in nature, and the standard of proof is the civil standard, it is right that where allegations amounting to criminal misconduct are concerned, and of course where individual liberty is at stake, proceedings must be conducted mindfully of that and with a suitable degree of circumspection and anxiety. Having said that, what is fair in a hearing of this sort has to be considered in the round with due sensitivity to the facts.

Starting with the panel’s treatment of the complainant’s credibility, I do not agree that anything more specific or detailed was required of the panel in this case, or that the claimant’s withdrawn account of earlier violence from Mr Osborne or Ms Lynn’s unspecified reservations needed any more anxious probing, in order for the fact-finding to be fair to Mr Osborne. That is because the panel was entitled to accept, and begin with, the evidence of the injuries. The claimant’s evidence was relevant not principally to the fact of the synchronous injuries, but to the explanation for them. The panel was entitled to consider that that evidence largely spoke for itself, and that the witnesses’ accounts were by way of testing the hypothesis that coincidence was an unlikely explanation. So in preferring the complainant’s account, the panel was not accepting a bare assertion, where credibility would have been wholly determinative. Nor do I agree that the panel treated the complainant’s contemporaneous social media messaging as corroborative, rather than simply as indications of consistency. They were entitled to do that.

And I do not agree that the panel was required to regard the fact that the CPS declined to support an evidence-led prosecution without the co-operation of the complainant as tending against the allegation being credible. The panel accepted the logic of the complainant’s reasons for not being willing to support prosecution – and they are indeed plainly rational – and the CPS would have been entitled to a pragmatic decision as to the prospects of a prosecution without her support, simply on grounds of evidential deficit.

The panel made no finding of fact about the withdrawn allegation of historic violence, and in the circumstances was entitled to treat it as irrelevant, both to the objective probability of the assault and to the credibility of the complainant’s account of it. The panel had accepted that the withdrawal of her complaint had been multi-factorial and complex, and her withdrawal of the historical account was capable of being viewed in a similar light. The matter of the complainant’s interaction with the police was properly capable of being considered in the context of the interpersonal dynamic between the two individuals seen in the round.

Turning to Mr Osborne’s credibility, I do not agree, for the reasons I have set out, that an explicit Lucas self-direction was necessary for fairness in this case. His denial of the assault was considered incredible not because he was seen as an established liar but because his story simply did not account for his injuries. Although Mr Osborne had not held back from alleging violence against himself by the complainant in the past or in general, it does not appear that he put this forward as an alternative explanation in the present instance. His explanation instead was violence at the hands of third parties. Rejecting that account did not raise a danger of importing the infection of a lie about one thing into a fair assessment of the truth of something else, since both were in effect sides of the same coin. The panel was not inferring from a lie about one assault another lie about not assaulting the claimant; nor was it evaluating a bare assertion, or two competing bare accounts. It was looking for an explanation for the bite marks and the black eye happening at about the same time. It did not hear anything from Mr Osborne reasonably capable of accounting for that. So the panel did not need to engage in any anxious evaluation of the truth of his account much beyond that.

The panel dealt with Mr Coningham’s points about there being possible uncertainty as to the existence of other relevant material at [2.28]. It said this:

Mr Osborne also asserts in his written submission that there was no apparent recognition on the part of the panel ‘that there may be considerations of disclosure or trial strategy which are not within the knowledge of the panel’. It is further asserted that, as no prosecution file was produced for a criminal trial, neither the panel nor Mr Osborne has any idea as to whether the totality of the evidence has been produced or whether there was unused or undiscovered material. However, the evidence was provided by police in response to a direction by the Board dated 11/02/2021 for ‘MG5, key witness statements, interview summaries and any additional evidence in relation to the allegations leading to recall’. The panel therefore assumes that all the available material has been provided. The panel also notes that Mr Osborne has at no stage requested that additional directions were made to procure further materials or witnesses.

That seems to me to be a reasonable answer, and I did not hear Mr Coningham engage any more directly with it.

Nor do I consider the fairness and/or rationality of the decision to be vitiated by the panel not having made findings of fact about the stalking allegations. It did not need to do so, and would apparently have had no evidence other than the accounts of Mr Osborne and the complainant relevant to doing so. It is interesting, perhaps, that in this respect – which would then have turned on a pure competing credibility evaluation – the panel took the more cautious route.

Mr Osborne was ably represented at his panel hearing and he and the panel had the benefit of Mr Coningham’s submissions on these and other matters. I can see no reason to hesitate over the fairness of the panel’s ostensibly rational findings on the question of assault (bearing in mind the civil standard, albeit elevated), or to interfere with its decision on this ground.

In reviewing Mr Osborne’s offending history, the panel had noted the details of the index robbery in 2007 (planned and committed with an accomplice) and that the trial judge then had described his offending history as ‘dreadful’, comprising 14 convictions for a total of 24 offences since he was around 16 in 1999. The relevant convictions for the dangerousness assessment included (a) an extremely violent attack on two men in 2000, with the same accomplice, causing traumatic brain injury, for which he was convicted of causing grievous bodily harm with intent and received six years’ imprisonment, and (b) the assault (four days before the index robbery) in which he punched an ex-partner in the face, causing her to lose consciousness, when her children had been present.

The panel also noted from the file three further unproven allegations of violence. Two were against his accomplice (respectively, punching him in the face including breaking his jaw, and a knife attack), but it seems the accomplice refused to give evidence. The third was alleged to have happened while Mr Osborne had absconded from open conditions, and related to stabbing a man who declined to make a statement. The panel said it was placing ‘only minimal weight on those unproven allegations, which are nevertheless indicative to the extent of the possibility that Mr Osborne’s violent offending has been more extensive than the offending for which he had been convicted’.

The panel also made mention subsequently of allegations surrounding his recall in 2016, after four months on licence; these allegations included threats to kill a former partner and her unborn baby.

Mr Coningham says it was irrational and unfair for the panel to place even ‘minimal’ weight on the historic allegations, and that it is unclear what weight was in fact placed on this material (especially the last mentioned) or why they were mentioned at all.

The reconsideration decision had concluded that, despite these mentions of unproven allegations and the reference to ‘minimal weight’, a fair and comprehensive reading of the decision indicates the panel actually put no weight on them. So I have looked again at the panel’s decision.

I can see only one reference to the unproven allegations, after their mention in the panel’s historical reviews. That comes at the very end of the decision, where the panel was considering the suitability of open conditions. It mentions that he had previously absconded from open conditions, and that allegations were made of a stabbing while he was unlawfully at large ‘although the panel has given minimal weight to those allegations’.

The decision against open conditions seems to me to have rested principally on the fact of the absconding, concerns about Mr Osborne’s ability or willingness to comply with the conditions of any temporary release, and the unavailability of the risk-management interventions it considered necessary. The unproven allegations seem to me on a fair reading to have been mentioned only as a footnote detail to the fact of the absconding. I cannot see that the panel placed any discernible weight on them in that context, or that they could have made any difference in themselves to its conclusions about the unsuitability of open conditions. That decision was – at least on its own terms – within the range properly available to the panel on the facts before it and for the reasons it expressed itself to be relying on.

So I am inclined to agree with the reconsideration decision that, on a fair reading of the panel decision, it is not apparent that the panel placed any weight on these unproven allegations in reaching this decision – or any other of the decisions it reached. It plainly did not need to – and I do not agree with Mr Coningham that it was unfair and prejudicial to Mr Osborne in effect even to have mentioned them at all – for the following reasons.

The arguable potential relevance of these allegations could be said to have been to Mr Osborne’s propensity for violence in general, and specifically for violence (a) involving facial injury or (b) involving a former partner. That propensity in turn could have been relevant to two issues – the fact-finding about the October assault, and Mr Osborne’s future risk. But his propensity for violence, in both of these respects, was amply demonstrated by his conviction history and his indeterminate sentence following a dangerousness assessment. He had been convicted of causing a disastrous head injury (which had not inhibited him from further violence thereafter) and of assaulting a former partner by punching her in the face. In any event, for the reasons already given, I cannot see that general issues of credibility and/or propensity weighed significantly in the fact-finding exercise itself.

So far as risk assessment is concerned, Mr Coningham reminds me of the guidance of the Court of Appeal in R (Pearce) v Parole Board [2022] EWCA Civ 4 that for present purposes ‘an assessment of risk can only be made upon undisputed or established facts’. I have looked at the risk assessment the panel made and cannot find that it did otherwise here, by reference to historic allegations or otherwise. I cannot find that the allegations did or could have made any difference to the panel’s analysis or conclusions. The risk assessment was based on consideration of Mr Osborne’s conviction history, his established history of conduct in prison and on licence, his established recall history, his association with drug misuse, and his psychological profiling and interventions. These alone were a clearly sufficient basis for the assessment made and I cannot, in other words, see any respect in which the unproven allegations affected or were capable of affecting the panel’s decisions in this respect.

My conclusion on this ground, therefore, is that there is no sign in the panel’s decision that they did in fact take the unproven allegations into account, even to the extent of ‘minimal weight’, nor that there is any material unfairness in the mere mention of them in the course of the panel’s historical scene setting. So I find no basis for interfering with its decision on this ground.

Dr Bromley, the expert psychologist witness, in both her written and oral evidence, had considered that Mr Osborne’s risk could be managed in the community and that the ‘building better relationships’ (BBR) programme was the best of the available interventions for him. The prison and community offender managers agreed.

The panel, however, considered that Mr Osborne’s underlying traits of ASPD were likely to be the primary drivers of two matters of concern in his conduct: abusive and violent behaviour in intimate relationships, and failure to be open and honest with those charged with assessing and managing his risk. It considered his ‘repeating pattern of emotional mismanagement, violence, and his inability or unwillingness to be open and honest is indicative of a need for intervention to address his underlying ASPD traits’. It considered there was no evidence he would be likely to engage with voluntary measures targeting his ASPD in the community. And it concluded ‘[t]he proposed BBR programme would not address the ASPD traits and the risk management plan does not include any mandatory intervention in the community that would do so’.

Mr Coningham accepts in the first place that a panel is in principle entitled to reject professional evidence. He accepts that the witnesses in this case did not give a ‘ringing endorsement’ to the prospects for Mr Osborne’s risk management to be successfully undertaken in the community in any event. But he says it was irrational and unfair to reject their recommendation here, when there was no psychologist on the panel, no alternative psychological evidence before it, no lines of questioning of the witnesses raising reservations and/or an alternative view of the psychological evidence, no advance indication of the reservations about the suitability of BBR programme, and no opportunity for Mr Osborne or the professional witnesses to address these key points on which the panel’s decision turned. He points out that the opportunity of the adjournment of the panel’s proceeding, and the further materials it asked for on the papers, was not taken to raise these matters when it could, and should, have been.

He also directs my attention to a number of decided authorities in which he says challenges on similar grounds have been successful. We looked again at R (Bousfield) v Parole Board [2021] EWHC 3160. In that case, the court noted that, until receipt of the panel’s decision, neither the claimant nor those advising him could reasonably have anticipated that the decision would turn on an apparent conclusion there was a real risk of the offender returning to his drug habit if living at the approved premises indicated in the recommended risk management plan. The plan had indicated that the offender’s risk could be safely managed in the community and the panel’s rejection of that on this basis could not have been anticipated.

We looked at R (O’Sullivan) v Parole Board [2009] EWHC 2370 (Admin). The court accepted the panel was fully entitled in principle to reject the professional evidence. But here, the panel had had reservations about the offender’s ‘internal coping strategies’. The court was unable to find a good reason in the panel’s decision as to why the necessary further work on these could not and should not be done in open conditions. There was a failure to explain the disagreement with the professionals.

And we looked at R (Wells) v Parole Board [2019] EWHC 2710 (Admin). Again, the court accepted that the panel was not bound by the professional evidence. But it considered there to be a heightened duty to give reasons where it rejects that evidence, and the court found the panel to have fallen short of that elevated standard. The court applied a test of rationality formulated in this way (at [32] and [34]):

A more nuanced approach in modern public law is to test the decision-maker’s ultimate conclusion against the evidence before it and to ask whether the conclusion can (with due deference and with regard to the Panel’s expertise) be safely justified on the basis of that evidence, particularly in a context where anxious scrutiny needs to be applied.

…

This may in certain respects also be seen as an aspect of the duty to give reasons which engage with the evidence before the decision-maker. An unreasonable decision is also often a decision which fails to provide reasons justifying the conclusion.

I find support in these decisions for an approach which accepts that a panel is not bound by professional evidence as to risk and the manageability of risk, even where it is unanimous and uncontroverted. But where it is minded to depart from unanimous professional evidence, it may be necessary for a panel to take a number of careful steps. It may be necessary for it to consider testing with those professionals any doubts it has at the time, in order to be sufficiently confident those doubts are well-founded. It may be necessary for it to seek further evidence. And/or it may be necessary for it to explain particularly clearly its reasons for ultimately coming to a different view from the professionals. These steps may be necessary in order to ensure that a decision to reject unanimous professional evidence is both securely well-founded in substance, and also fairly reached and fully articulated.

Whether some or all of these steps are necessary or desirable is likely to be fact-sensitive. On the facts of the present case, I note both that the professional evidence was unanimous, but also that the panel’s disagreement with it was specific. It seems to turn on the panel’s primary conclusion that the principal drivers, both of Mr Osborne’s offending and of his failure of engagement with professional risk-management, were likely to be his underlying traits of ASPD. Its secondary conclusion was that there was therefore a need for intervention to address these underlying traits. Then third, it concluded that this intervention needed to be mandatory, not voluntary. Fourth, there was nothing of that nature in the community risk management plan. And fifth, therefore, Mr Osborne needed to be kept in prison, in closed conditions.

I am troubled as from the first step in this analysis. The decision records that the panel explored the operative effect of ADHD with the professional witnesses, but does not record that it explored the operative effect of ASPD with them. It records agreement with the professionals’ opinion on ADHD, but does not record agreement or disagreement on ASPD, neither citing evidence in support nor reasons for departing from any other evidence. I can see that Dr Bromley considered ASPD to be a factor ‘exacerbating’ Mr Osborne’s risk. But I cannot see in the professional evidence an indication that ASPD traits were ‘the primary drivers’ of his offending. It appears to have been the panel’s own conclusion on a matter of psychological expertise (and there was no psychologist on the panel).

That does not necessarily make it irrational in the sense of being a conclusion no reasonable panel could have reached. But it does seem to me that a panel minded to reach such a view needed to apply its mind to taking some or all of the ‘careful steps’ identified above: testing its thesis with the witnesses, seeking more evidence, and/or articulating its thinking in some detail so as to explain either how it is supported by the evidence or why the panel is departing from that evidence. But I cannot see that it thought about, or took, any of those steps here.

That is particularly troubling where the assessment that ASPD was a primary driver of Mr Osborne’s concerning behaviours was the root reason for the decision that Mr Osborne should remain in (closed conditions in) prison. But the second step in the panel’s analysis – that it was necessary for there to be ‘intervention to address his underlying ASPD traits’ – is also of concern in itself. There are implicit in that some conclusions or assumptions about the potential nature, effectiveness and availability of ‘intervention’ of this sort which are not anywhere articulated. There was an evidential basis in Mr Osborne’s history for the third step – that reliance could not be placed on him actively and honestly engaging voluntarily with (any) professional interventions. And the fact that ASPD interventions were not articulated as such in the risk management plan was no doubt plain on its face. But it is not possible to understand from this decision what ‘intervention’ the panel had in mind as being necessary, or why the BBR programme did not furnish it. Further, the fact that the plan was rejected for not including ‘any mandatory intervention in the community that would do so’ does not enable the reader to understand whether either mandatory community interventions could and should have been put in place (and if so why that solution was not explored) or, if not, why ASPD intervention was possible only in closed prison conditions.

So returning to the ‘careful steps’ set out above, it seems to me that, if the panel was minded to conclude ASPD was a, or the, ‘primary driver’ of Mr Osborne’s risk profile, it needed to focus on the evidential basis for that conclusion, test it, and either augment it – by exploring it further with the existing witnesses or directing further evidence about it – or clearly articulate it by reference to the detail of the existing evidence. I cannot see that it did either here. It needed to do a similar exercise with the conclusion it was minded to reach on intervention to address/manage ASPD – but again, I cannot see that it did. So I do not see how, allowing all due deference to the panel, I can be satisfied that its conclusion can be ‘safely justified’ on the evidence, in a context where anxious scrutiny needs to be applied – nor indeed that the panel had properly so satisfied itself in the first place.

Whether approached from the angle of rationality or fairness, that is not satisfactory. I cannot agree with the reconsideration decision that, having cited the guidance in Benson [2019] PBRA 46, the panel’s decision was one in which ‘everyone, but particularly the prisoner, can understand the reasons for the decision’, or, it might be added, one which can be safely justified as being soundly structured and supported on its own merits. No reasons are given for the conclusion that ASPD was a primary driver of Mr Osborne’s behaviour and no evidence is cited or apparent to support it. It seems to be an unexplained inference of the panel which does not speak for itself.

The conclusion that the plan for managing risk in the community was deficient for want of mandatory interventions directed to ASPD is, firstly, based on that unsatisfactory premise, and, secondly, also problematic in having itself been untested with the existing witnesses, unsupported by any other evidence, and insufficiently articulated. It is also an unexplained inference of the panel or an insufficiently explained extrapolation from its first inference. And what is certainly unapparent is why, if the BBR programme was deficient, the panel concluded that was an irremediable deficiency and there was no practical alternative to closed conditions. That is a considerable jump of logic, and the intermediate steps can only be guessed at.

It is not easy to understand this decision, and it is consistent with operative assumptions or inferences which are not visibly tested, articulated or explained, and which may or may not be misconceived. So I do not think this decision can be safely upheld.

In these circumstances, I would uphold grounds 3 and 4 and set aside the panel’s decision. That is effectively dispositive of this case.