Roberts, R (on the application of) v Secretary Of State For Home Department

MR JUSTICE ELIAS: The claimant is a serving prisoner. In 1991 he was convicted and sentenced to two concurrent life sentences for murder. He carried out a frenzied knife attack on an elderly couple and there was evidence that the wife may have been sexually assaulted. He has always denied the offences. He accepted that he was in the house having attempted a burglary when drunk, but he claimed that he had left the house by a downstairs window when he saw the figure on the garden path and had no contact with the couple at all.

His tariff was set at 17 years by the Home Secretary. He will not, therefore, be considered for parole until that period has elapsed. Since his incarceration he has been a category A prisoner. By this application for judicial review he seeks to challenge the decision taken by the Category A Review Team on 20th May 2003, when it determined that he should remain a category A prisoner and not be downgraded to category B. In fact the review team initially made its determination on 29th April 2003, but that was without considering representations drafted by the claimant's solicitors. Following receipt of those representations the team reconsidered its decision and confirmed the original determination.

There are four categories of security, A to D, of which only the first two are relevant in this case. They are as follows:

The principle is:

There are various ways in which the regime for prisoners in category A will differ from that of prisoners in category B. For example, category A prisoners are subject to closer monitoring within the prison, certain jobs are unavailable to them because they increase the risk of escape, and they are more closely managed and supervised on those occasions when they leave the prison. In addition, in practice it is likely to take longer for a prisoner in category A to achieve parole than someone in category B. Whether that is a relevant consideration for the review team to take into account is one of the issues raised in this case and I return to consider it below.

Until a recent change to the definition of category A, it was expressly provided that prisoners would be placed in category A if they posed the requisite degree of danger, however unlikely their escape may be. However, in William Pate v Secretary of State for the Home Department [2002] EWHC 1018 (Admin), Turner J held that whilst it was legitimate to have a policy that aimed to make escape impossible for such persons, it was not lawful to exclude the possibility that this objective could be achieved even in category B. In that case the claimant was a physically frail prisoner who contended that it was unlawful for the relevant committee to place him in category A without any reference at all to the fact that he was not realistically in a position to escape. Turner J said this (paragraph 38):

The response to this judgment was that the definition of category A was altered, as I have indicated. Prison governors were notified by the director of high security prisons, that the new definition should be accompanied by the following statement:

Until this amendment the issue of escape risk was considered independently of the issue of danger by placing all category A prisoners on three escape risk classifications. These remain in force and are not affected by the fact that there may now be a small number of cases where prisoners who pose the risks referred to in category A may nonetheless be catered for in category B. The three categories of escape risk are determined, as one might expect, according to the risk of an escape attempt. The lowest of these is the standard escape risk on which most category A prisoners are classified. This is the category in which the claimant falls. The other two categories are "high escape risk" and "exceptional escape risk".

All category A prisoners must have their security category reviewed at least annually. A number of bodies may become involved in the process. Most of the decisions are made by the Category A Review Team, as in this case. Membership of that team is made up of officials from Prison Service headquarters. The team cannot determine that there should be a re-categorisation into category B, but if it makes that recommendation then the matter will be considered by the Category A Committee. This is a more senior committee comprising the Deputy Director General, the head of Police Advisers to the Prison Service, the head of the Resettlement and Programmes Group, the head of the Category A Review Team, and representatives from the high security prisons.

The Category A Committee will also consider every case where five years has elapsed since the Committee last considered the prisoner's case, and it will consider cases where the prison itself has recommended downgrading. The final decision to downgrade is, however, always made by the Deputy Director General of the Prison Service.

When determining the issue of the appropriate category, the Category A Review Team will have before them reports from various individuals who have responsibility for the prisoner when he is in prison. At the time when the decisions were taken in this case, the prisoner was entitled to have the gist of these reports communicated to him so that he could then make observations and representations upon them, but not the full reports.

This was in accordance with the established case law (see for example R v Secretary of State for the Home Department ex p Duggan [1994] 3 A.E.R. 277, approved by the Court of Appeal in R v Secretary of State for the Home Department ex p McAvoy [1998] 1 W.L.R. 790.

However, following the decision of Munby J in R (on the application of Alan Lord) v Secretary of State for the Home Department [2003] EWHC 2073 (Admin), the full reports now have to be provided, save for those parts which can properly be withheld under the Data Protection Act. Following that decision the reports were in fact provided to the claimant and the court in the context of these judicial review proceedings, but obviously only after the decision had been taken.

In this case the prisoner was sent a note on 10th February 2003 setting out the gist of the reports which were being considered by the panel. The reports confirmed that the prisoner's custodial behaviour was acceptable and was compliant with the regime. It noted that there were no control problems and that he had received no adjudications for several years. In addition, he was polite and cooperative in dealing with the staff. It also noted that he continued to deny guilt for the offences and as a consequence refused to participate in any offence-related work. He had taken part in a reasoning and rehabilitation (R & R) programme and also an oral communication course, and had expressed a willingness to take part in the CALM management programme, which addresses anger management, and an alcohol awareness course.

The paper informed the claimant that the reports noted that the risk factors relating to the specific offences remained largely unaddressed, and that the recommendation of the prison was that the claimant should remain within category A.

On the basis of these reviews the review team concluded on 29th April 2003 that the prisoner should remain in category A. It recognised the satisfactory custodial behaviour but commented that:

It then noted the fact that the prisoner had participated in certain programmes and courses and that benefit had been gained in particular from the R & R programme. However, it observed that that programme did not discuss prisoners' specific offences but rather offending behaviour more generally, and commented:

It concluded as follows:

Finally, the review team observed that the index offences and also the previous offending history evidenced a high level of potential dangerousness, and that downgrading could not be justified until there was clear and cogent evidence that the risk of re-offending in a similar way had significantly diminished. The team was not satisfied that there was such evidence, "either through offence-related work or otherwise."

The claimant's solicitors then sent detailed representations to the panel. Included in that were the gist of reports that had been provided when earlier categorisation reviews had taken place. These included the concerns expressed by two prison chaplains that the prisoner may have been the victim of a miscarriage of justice. It was also noted that the prisoner had been told that in 1994 a majority of the prison staff had supported a downgrading of his security category, although that recommendation was not of course acted upon.

It was submitted to the review team that denial of guilt should not be a bar to re-categorisation. Plainly, as the solicitors recognised, it was a relevant factor, but one which was overridden in this case, it was submitted, by very persuasive counter-balancing features. His solicitors urged upon the review team the fact that the claimant had spent a long period in prison and had become involved in very few confrontational situations and that this, together with a lack of any adjudication for a period of some 4 years, demonstrated a sufficiently diminished risk of danger in the future. They pointed out that although he would not become involved with programmes which require him to deny his guilt, he would participate in any other general offence-related work which was designed to address previous offending behaviour.

In substance, they were contending that the position had now been reached where there was a stalemate. He was and will continue to deny his guilt, and yet the attitude appeared to be that without participating in programmes which involved recognition of guilt, he would never be able to satisfy the review team that the risk of similar offending had diminished significantly enough to allow a re-categorisation.

The review team found the representations unpersuasive. They considered the original decision was both reasonable and rational. It also rejected a submission that he should be put into category B on the grounds that he posed no real risk of escape. This has given rise to a distinct ground of challenge which I consider below.

The decision is challenged on four grounds, although the first two are, to some extent, interrelated. They were formulated by Mr Huseyn, counsel for the claimant, identifying four questions as follows:

1. Was the Category A decision unreasonably reliant on the Claimant's denial of guilt and consequent failure or inability to participate in offence focused courses?

2. Did the decision fail to give due weight to other factors such as good custodial behaviour in assessing the risk of harm to the public represented by the Claimant?

3. Is the Respondent correct and acting rationally in excluding considerations of the effect of the categorisation decision on 'progression through the system' and progress towards parole from categorisation assessments?

4. Did the Respondent apply the right test in excluding from consideration any assessment of the actual escape potential of the claimant on the basis that this need only be done where there are exceptional circumstances?

Ground 1

The claimant recognises that in formal terms the report of the review team recognises that the lack of participation in offence related work did not itself bar downgrading. Indeed, it stated so in terms. The team also left open the possibility that the prisoner would be able to demonstrate a sufficiently significant diminution of risk, not only through offence based work, but also in other ways.

Mr Husyen for the claimant submits that this is, in essence, window dressing. The reality is that there is no way in which risk reduction can properly be established to the satisfaction of the authorities, save by admitting guilt and participating in the relevant courses. He submits that failure to participate in such courses is plainly the overwhelming factor in the review team's determination that there should be no re-categorisation. He referred me to the reports before the review team in which various members of the prison staff had expressed their opinion on the question of downgrading. All these reports were opposed to a re-categorisation.

Mr Huseyn submitted, and I accept, that looking at these various reports they virtually all conclude either that his failure to participate in specific offence focused work means that there can be no proper assessment of the extent to which risk is being diminished or, alternatively, that the fact of non-participation demonstrates that it has not significantly diminished. He says that even if the review team is conscious of the fact that the requisite reduction in risk may be demonstrated in other ways, even for those on denial, the prison staff do not appear to do so. Moreover, the tone of their reports must have influenced the decision of the review team.

Mr Kovats, for the Secretary of State, recognises and accepts that as far as the Parole Board is concerned, when it is exercising its functions, it may not refuse parole simply because the prisoner denies his guilt. That is established by the decision of the Court of Appeal in R v The Parole Board and Secretary of State for the Home Department ex p Oyston, (unreported, 1st March 2000). At paragraph 43 of that decision Lord Bingham of Cornhill, then the Lord Chief Justice, summarised the difficulties faced by the Parole Board when dealing with prisoners in denial. He said this:

The court also approved the following principles enunciated by Laws J, as he was, in an unreported decision in 1997, R v Secretary of State for the Home Department, ex p Hepworth and Others. The four propositions were these:

Mr Kovats accepts that these principles apply equally to categorisation. That seems to me to be correct given that both decisions are concerned with the assessment of risk. That is not to say, however, that identical considerations operate in each case. As the Court of Appeal noted in R (Williams) v Secretary of State for the Home Department [2002] EWCA Civ 498, [2002] 1 W.L.R. 2264, the Parole Board is concerned with assessing risk in the context of someone who is lawfully released and subject to continuing monitoring and control. Furthermore, there are incentives to behave, since in the event of non-compliance the licence is revocable.

By contrast, those determining whether a prisoner should remain in category A are concerned with risks posed to the public by someone who, they must anticipate, may unlawfully be at large. Plainly such persons would not be supervised or monitored. Although both decisions address public safety, they perforce do so in different ways and for different purposes. Even so, as the parole cases show, it is not the denial itself which is relevant, but the effect which this has on the ability of the prisoner to come to terms with his offending behaviour and to demonstrate the necessary reduction in risk. This must equally be so in the categorisation process.

Mr Kovats says that it is wrong to say that the review team has made the error attributed to it by the claimant. It would be unjust to assume that it has merely paid lip service to the possibility that there may be other ways of demonstrating a reduction of risk. He referred me to a statement in these proceedings from Clare Lewis, who is the Operations Manager and Director of High Security Operations Unit and the current temporary head of the Category A Review Team. She gave examples of the way in which the diminution of risk may be established without doing courses focused on the specific offence, such as by one-to-one work with specialist staff, increasing maturity and other evidence of a sustained and material change in attitude towards offending behaviour gathered by staff through regular contact.

Moreover, in this case Mr Kovats submits that the team was placed in a particularly difficult dilemma. In some cases it may be possible to identify relatively easily what has motivated the prisoner to commit the crime and to assess risk even for those in denial. Here, however, it is not clear whether the motive was sexual, financial, the fact that the prisoner was in drink, or perhaps is inherently violent, or even a combination of some of these. Without a recognition of guilt and some understanding as to why the crime was committed, it is difficult to be satisfied that there is a sufficient reduction in the risk that it might occur again. He submits that this is effectively one of those cases which is identified in the fourth proposition set out by Laws J, where without participation it may be extremely difficult objectively to assess the risk.

I accept Mr Kovats' submissions. There is a very real difficulty facing the review team in cases of this nature. The guilt of the prisoner must be assumed. That is what the review team properly did here. The review team must then assess the nature of the risk in the event of an escape. Where the index offences are so grave, as they will inevitably be in category A cases, the review team can justifiably require cogent evidence that that risk has diminished.

That evidence will, in the normal way, be most cogently demonstrated by the prisoner participating in courses and programmes which are directed to the specific offences, so that there can be some self-awareness into the gravity and consequences of his conduct. However, it is a condition of a number of these courses that the prisoner must admit his guilt. That is so, I am informed, for the Sex Offences Treatment course, the CALM course (controlling anger and learning to manage), and the CSCP course (cognitive self change programme). By not participating in such courses or programmes the prisoner inevitably makes the task of the review team more difficult, and in some cases practically impossible.

It must be recognised that this compounds the injustice for anyone who has suffered the grave misfortune to be wrongly committed of such terrible crimes, and there will inevitably be such people. It puts pressure on the innocent to admit guilt in order to facilitate release, or, alternatively, to serve a longer sentence than they would have had to do had they committed the crime and felt properly able to admit guilt. But that seems to me to be inevitable, the system cannot operate unless the verdict of the jury is respected.

Moreover, on very, very, many more occasions defendants deny guilt for offences which they have in fact committed, for a whole variety of reasons. Given that the danger must be presumed from the nature of the index offence, it is plainly a proper requirement that there should be cogent evidence in the diminution of risk if the safety of the public is to be secured. No doubt to those in denial the recitation by a review team that being in denial does not of itself preclude re-categorisation may appear to have something like a mantra-like quality. There is no doubt that if they disqualify themselves from the courses which address their specific offending, it will be considerably more difficult than to be able to satisfy the review team that re-categorisation is justified. This is not, however, a punishment meted out to them because they have not admitted the offences, but it is because by being in denial they limit - and in many cases severely limit - the practical opportunity of demonstrating that the risk has diminished. Indeed, their denial demonstrates that they have not accepted that the risk was ever present. In the circumstances, therefore, I do not consider that the review team can be criticised on this ground.

I would, however, make two observations before leaving this particular aspect of the appeal. First, it is not difficult to understand the frustration of someone like the claimant who was informed in 1994 that there was significant support for a re-categorisation, and who should discover some 9 years later that there is now no support, notwithstanding the fact that he would, on the face of it, appear to be better placed now than he was then. Plainly the observations made in 1994 do not begin to give rise to a legitimate expectation, or anything of that nature, and it has not been suggested that they did. But the disappointment of someone in the claimant's position is readily understandable.

The reason for the change in attitude, I am told, is that the assessment of risk, and its reduction by specialist courses, has developed significantly since 1994. Whereas formerly the determination of whether or not there was a reduction was carried out, at least in significant part, by focusing on behaviour in prison, it is now a much more sophisticated process. Behaviour in prison figures as one factor in the assessment, but the successful participation in offender focused courses is now seen as very important. These changes operate to the detriment of those who will not render themselves eligible for these courses.

The second point, which I make with considerable diffidence, is this. It is plainly going to be extremely difficult for some prisoners to satisfy the authorities that the risk has reduced if they fail to do the specific offence directed courses. In the long term this effects their chances of parole, which is not only to their detriment but also that of the public, not least because any incarceration of category A prisoners is extremely costly to the public purse.

Of the three ways identified by Miss Lewis to demonstrate reduction in risk (and I am not suggesting she intended these to be exhaustive), it seems that good behaviour over a period and growing maturity would not, in the vast majority of cases, be likely to be considered enough to demonstrate reduction of risk for understandable reasons. Indeed this case itself suggests that this is the position.

Whether there is scope for more one-to-one work with specialist staff than currently exist, and whether this is a satisfactory alternative, I do not know; but the tenor of the reports I have seen in this case suggests to me, maybe mistakenly, that this possibility may not be widely appreciated by some staff in the prisons. I do no more than raise the question whether it is explored as an alternative as often as it might be.

The second ground is essentially directed to the weighing process carried out by the team. Mr Huseyn submitted that where the review team does not have the benefit of successful participation in offence-directed work, it must give more weight to those factors which it can assess. Here the review team, he submitted, gave virtually no weight either to the good behaviour and constructive attitude, or to the successful participation in the R & R course.

This argument is doomed to fail, not least because the weight to be given to those factors is plainly for the review team. Furthermore, as Mr Kovats pointed out, if the argument were correct, then it might be beneficial for prisoners to choose not to participate in courses and rely on a sustained period of good behaviour.

The review team plainly took these matters into account and referred to them in the decision. It could not adopt different principles of assessing risk for those in denial than for the other prisoners.

In order to succeed on this ground it seems to me that the claimant would have to show that the review team had acted in a perverse way in reaching its conclusion. In this case that would involve the court concluding that a reasonable team would have been compelled, because of the good behaviour of the claimant in prison, and the other matters relied upon, to have concluded that this significantly reduced the risk of his committing like offences in the future. That argument plainly has no prospect whatsoever of success.

There was a more specific argument directed at the particular wording of the decision. The Tribunal did not say that involvement in the R & R programme could not be sufficient evidence on its own for reduction of risk; rather, it said that it could not be evidence of such a reduction. I accept that this cannot be right. Plainly it will be evidence and should be taken into account, but I am satisfied that, read sensibly, that is what the decision is saying. It could hardly be intending to say that participation in the course would not be evidence at all, if considered on its own, but would somehow magically become relevant evidence if it can be combined with some other evidence.

The argument here essentially is as follows. Before the prisoner can be considered for parole it is necessary that he be progressively de-categorised until he is able to spend a period in prison in open conditions. All this is necessary because the Parole Board will not, in practice, permit long-term prisoners to be released on parole until they have demonstrated, in conditions closer to those to the community, that they will act responsibly and can be trusted in the community. It is perfectly true that in legal theory the factors weighing with the review team are different from those which a parole board has to consider, as I have indicated above. But in reality they are interlinked. The interrelationship between these two decisions was considered by Judge LJ giving the judgment of the Court of Appeal (Lord Phillips of Worth Matravers MR, Judge and Carnwarth LJJ) in the Williams case. He said this:

The danger is that a decision by a categorisation team can frustrate the operation of a parole process. If a category A prisoner does not have access to opportunities to demonstrate that he can be trusted in a lower category, then he will have an almost impossible task in persuading the Parole Board that he should be released.

The claimant's contention here is that the potential implications of its decision to the parole process have not been addressed or considered by the team. The Williams case makes plain, as the prison authorities accepted in that case, once there is a recommendation from the Parole Board then that is a relevant consideration for a review team to take into account when considering the question of re-categorisation. It was not made plain in Williams precisely how it is to be taken into consideration, but I assume that it is then incumbent on the review team or committee, whichever is charged with a duty to reconsider the security category, to have regard in particular to the Parole Board's views about risk.

The court in Williams recognised that there may be real difficulties facing a prisoner in circumstances where the Parole Board recommends a period in a lower category, but the category review team rejects that approach. In Williams the court held that the risk of different assessments would be mitigated if the prisoner were provided with the full reports rather than their gist, as is now the case following the Law decision, and if the prisoner were permitted an oral hearing. Ultimately, however, the question of re-categorisation is still a matter for the review team and its function cannot be fettered or controlled by the Parole Board. The procedural safeguards identified in Williams therefore do not eradicate the possibility that a decision of the prison authorities may still effectively frustrate the wishes of the Parole Board.

Mr Huseyn submits that just as a review team must have regard to any decision of the Parole Board in exercising its discretion, so it must have regard to the potential implications of its decision for parole even though no parole decision has been made. In this case the first parole consideration will, I am told, be in a few months, which is some two years before the claimant can in fact be released on parole.

Mr Kovats pointed out that the problem facing the prisoner in Williams was that he could not demonstrate that he was a significantly lower risk because he was not able, at that time, to do all the relevant courses as a category A prisoner. The claimant in that case was caught in what was described as a Catch 22 situation. That is no longer the case. Here, the claimant is able to do all the courses, although not necessarily in the current prison, and therefore he is not disabled from demonstrating the reduction in risk, although, of course, there is still the possibility of conflict between the review team and the Parole Board.

There is clearly force in the submission that the claimant here is the author of his own misfortune. But those observations do not wholly meet Mr Huseyn's point that the review decision may still impact adversely upon the claimant's future parole prospects, and that this is a matter which should be taken into account.

I do not accept Mr Huseyn's argument. There has, as yet, been no consideration of the case by the Parole Board at all. I do not accept that it is either feasible or desirable that the prison authorities ought to have half an eye on what attitude might be adopted by the Parole Board to the question of parole. In my judgment, the review team must focus on the question it has to determine, which is whether the prisoner continues to constitute a serious danger such that escape must be rendered impossible. Once there is evidence from a decision of the Parole Board itself that the category A status may be defeating or undermining the prisoners' opportunity for parole, then the opinion of the Board is a factor which the committee must bear in mind, particularly insofar as it focuses on danger in the community. But it cannot sensibly be required to anticipate what approach a parole board might take and adjust its determination accordingly.

There are plainly potential tensions in the operation of the two systems as the Johnson case indicates, but I do not think that a review team is required to anticipate those potential difficulties. Indeed, I think it ought not to do so.

The final ground is directed to the review team's implementation of a policy adopted by the Prison Service after the decision of Turner J in the Pate case. Mr Huseyn accepted, as I understand it, that the statement accompanying the category A definition, which I have set out above, is not, in principle, objectionable. It recognises that exceptionally a prisoner may be in category B and yet escape may still be impossible. However, he criticises the way in which the review panel approached the matter. It said this:

Mr Huseyn accepts that it is going to be a rare case where the situation will arise that somebody can be placed in category B because the risk of escape is so minimal. But he says that nonetheless the matter must be properly considered. His criticism here is that the review team has effectively adopted a two-stage approach. First they ask themselves whether there are any exceptional circumstances, and then they are indicating that they will only consider whether the prisoner might be accommodated within category B if they have found that there is something exceptional in his case.

Mr Kovats submits that the passage I have quoted does not demonstrate that the review team looked at the matter in this two-stage way, but I recognise the force of Mr Huseyn's observations that the drafting of the paragraph suggests that they may well have done that. Even if they did, I must confess that it seems to me that it is a legitimate approach. The distinction between a category A and a category B prisoner is that there is greater security for the former. The prison authorities have taken the view that something more is required to ensure their incarceration than is necessary for those who fall into category B.

The definition of standard escape risk prisoners recognises that many prisoners will have few resources and little inclination to escape, but nonetheless would take an opportunity to do so if they could. If it is a legitimate policy, as it surely is, to seek to eliminate any risk of escape for category A prisoners, then the starting point must be that category A safeguards are required. They are fuller and are considered more appropriate than the category B safeguards.

The Pate case recognises that this cannot be a universal rule, but, as I think everybody can see, it is going to be an exceptional case where that principle does not apply. As Miss Lewis points out in her witness statement, even for those with very limited resources to escape there is a possibility of human error, or corruption, or they may tag along perhaps in the course of mass escape with other prisoners.

Accordingly, it seems to me that the review team might perfectly properly in effect ask itself if there is anything that takes this particular case out of the ordinary. In other words, is there something on the face of it which makes this prisoner's situation exceptional? That is not to say that they are entitled to limit in advance those factors which may be considered exceptional, but at a practical level they are entitled to say that if there is nothing unusual in a category A prisoner's case then there is no purpose in analysing in detail the risk of escape.

If, on the other hand, there are considerations which point to it being a special case, and one which might fall into the exceptional category, then that is a matter which they must carefully attend to before reaching a final determination. I do not think the panel in this decision has sought to limit what might potentially constitute reasons for concluding that escape would not be a real possibility even if the prisoner were put into category B.

Furthermore, in this particular case, I find it impossible to say that they could not properly reach that conclusion. It is true that there are some factors which might suggest that this particular prisoner would be reluctant to escape: for example, he seems to have no links with anybody outside the prison. But that is very far removed from saying that he would not seek to escape if the opportunity presented itself.

As Mr Kovats points out, this is not a balancing question where one weighs up different considerations, rather it is a threshold question, and that is a conceptually different exercise.

Finally, even if the review team did approach the matter in the wrong way, which I do not accept, it cannot possibly be said that it made any difference whatsoever to the outcome of this particular case.

For these various reasons therefore, I reject this application.

Can I thank you both very much.