R (on the application of Sturnham) v The Parole Board of England and Wales and another (No 2)

From 4 April 2005 until 3 December 2012, English law provided for the imposition of sentences of imprisonment for public protection (“IPP”). This is another case in which courts have had to address the practical and legal issues resulting from this innovation.

To impose a sentence of IPP the court had (inter alia) to be of “the opinion that there is a significant risk to members of the public of serious harm occasioned by the commission by [the offender] of further specified offences” (Criminal Justice Act 2003, section 225(1)(b)). When imposing it, the court was required to specify a minimum period (the “tariff” period) after the expiry of which the prisoner was eligible for review by the Parole Board who could direct his release on licence (Powers of Criminal Courts (Sentencing) Act 2000, section 82A). The Parole Board was required not to direct release unless “satisfied that it is no longer necessary for the protection of the public that the prisoner should be confined” (Crime (Sentences) Act 1997, section 28(6)(b)).

The case is before the Supreme Court as an application for permission to appeal, with the appeal to follow if permission be granted. Having heard the submissions, I consider that it raises issues of importance which merit the Court’s consideration and would therefore grant permission. On that basis, there are two grounds of appeal before the Supreme Court. The first focuses on the relationship between the criteria for the court to impose a sentence of IPP and for the Parole Board to direct release on licence. The appellant submits that they must, though differently worded, be read as involving the same substantive test. The Parole Board and the Secretary of State submit that the difference in wording represents a difference in substance. The second ground of appeal is that, even if the criteria differ in substance, the Parole Board in fact applied a wrong test when deciding whether to order the appellant’s release. Although Mr Hugh Southey QC for the appellant accepts that this ground is now largely, if not entirely, academic in view of the appellant’s release, he submits that the court should address it to clarify the test for release.

For reasons which follow, I would dismiss the appeal on both grounds.

During an altercation outside a public house on 19 May 2006 the appellant, then aged 28, punched a man, who fell backwards, struck his head on the ground and died on the next day. The appellant was convicted of manslaughter. The judge, HHJ Findlay Baker, concluded that the appellant was dangerous. He was forceful and physically very strong, had indicated that he regarded it as his right to respond with violence to any tendered or threatened towards him, had uncontrolled heavy drinking and cocaine taking problems, and had shown no commitment to change these. Accordingly, on 31 January 2007 the judge imposed a sentence of IPP, with a tariff of 2 years 108 days, which expired on 19 May 2009. He said that, had it not been appropriate to impose IPP, a sentence of six years’ imprisonment would have been appropriate. Deducting half of that, and the time spent on remand, gave the tariff.

A Parole Board review took place only on 10 May 2010, when the Board concluded that Mr Sturnham had made significant progress, but still presented a low risk of re-offending and a medium risk of serious harm. It declined to order release, but recommended transfer to open conditions which took place on 12 August 2010.

Mr Sturnham issued proceedings for judicial review, claiming that the Parole Board had applied the wrong test and also claiming damages for the delay in holding the review. These proceedings were the subject of judgments given by Mitting J on 14 March 2011, [2011] EWHC 938 (Admin), and the Court of Appeal on 23 February 2012, [2012] EWCA Civ 452; [2012] 3 WLR 476. The claim for damages was ultimately disposed of in the Supreme Court by decision on 1 May 2013, [2013] UKSC 23; [2013] 2 WLR 1157, restoring Mitting J’s award at first instance of £300 damages for six months undue delay. The former claim is now before the court, having been dismissed by both Mitting J and the Court of Appeal.

The result of the present appeal no longer has direct significance for Mr Sturnham’s detention. He was released on licence pursuant to a Parole Board decision dated 7 September 2011. But it had a live general significance for the Parole Board at the date when the appeal was considered by the Court of Appeal (23 February 2012) and it may have a continuing significance in other cases, including potentially for prisoners serving life sentences.

IPP was a child of the Criminal Justice Act 2003. Section 225(1) to (3) of that Act identified the circumstances in which IPP was appropriate (until 13 July 2008 without, but thereafter subject to, any residual discretion on the sentencer’s part) by differentiating those in which it required a discretionary life sentence to be imposed. Release after the tariff period was provided for by the insertion into the Crime (Sentences) Act 1997 of a reference to IPP so as to make IPP subject to the same statutory regime of review by and release by direction of the Parole Board as applicable to mandatory and discretionary life sentences.

As enacted and in force when Mr Sturnham was sentenced, section 225 read, so far as material:

Section 225 falls to be read with section 224 and Schedule 15. Section 224 provided, so far as material:

Schedule 15 contained a very substantial list of over 150 different offences, starting with manslaughter, kidnapping, false imprisonment, threats to kill and malicious wounding. Notably, however, it did not include murder, for the obvious reason that murder would carry a mandatory life sentence.

Section 229 explains the concept of risk (or dangerousness) relevant under section 225:

As to release by the Parole Board on licence, sections 28 and 34 of the Crime (Sentences) Act 1997 read, so far as material:

Section 225 accordingly applied whenever the sentencing court was of the opinion that “there is a significant risk to members of the public of serious harm occasioned by the commission by [the offender] of further specified offences”. It offered two mutually exclusive courses in such a case. Under section 225(2) the court was required to pass a sentence of imprisonment for life if (a) the offence was one in respect of which the offender would apart from section 225 be liable to imprisonment for life, and (b) it considered “that the seriousness of the offence, or of the offence and one or more offences associated with it, is such as to justify the imposition of a sentence of imprisonment for life”. Alternatively, if either of (a) or (b) was not satisfied, the court was obliged to impose a sentence of IPP.

Prior to the 2003 Act, the criteria for imposition of a discretionary life sentence consisted, broadly, of the commission of a very serious offence and a conclusion that the offender was a serious danger to the public and likely to remain so for an indeterminate period. In R v Hodgson (1967) 52 Cr App R 113, 114 the Court of Appeal put the matter as follows:

Further guidance was given by Lord Lane CJ in R v Wilkinson (1983) 5 Cr App R (S) 105, 108-109. He said that a discretionary life sentence should be reserved for the most exceptional circumstances, and for the most part for offenders who were incapable of being dealt with under the Mental Health Act 1959, “yet who are in a mental state which makes them dangerous to the life or limb of members of the public” and in respect of whom “It is sometimes impossible to say when that danger will subside”.

The reference in Hodgson to future offending being “likely” was read in a mathematical sense of “more probable than not” by counsel for the appellant and it seems the Divisional Court in R v Parole Board, Ex p Bradley [1991] 1 WLR 134, 143F, 144H, 145F-G and (in particular) 146A-C. But “likely” is a word of open meaning, and I regard any attempt to state or apply a test of mathematical probability in this context as inappropriate. The formulation “good grounds for believing that the offender may remain a serious danger to the public for a period which cannot be reliably estimated at the date of sentence” was, in my view rightly, preferred in the later case of Attorney General’s Reference No 32 of 1996 (R v Whittaker) [1997] 1 Cr App R(S) 261, 264. By the same token, the Court of Appeal was right, in the context of IPP, to regard it as “wholly unhelpful to attempt to redefine ‘significant risk’ in terms of numerical probability, whether as ‘more probable than not’ or by any other percentage of likelihood”: R v Pedley [2009] EWCA Crim 840, [2009] 1 WLR 2517, para 19.

Other authority also indicates that the criteria for imposing a discretionary life sentence are in some measure inter-related. In R v Chapman [2000] 1 Cr App R 77, 85 Lord Bingham CJ approved Whittaker and said:

That being the common law position relating to the imposition of a discretionary life sentence, the next question is the effect of section 225(2) and (3). In R vLang [2005] EWCA Crim 2864; [2006] 1 WLR 2509, para 8 Rose LJ said:

The point did not however arise for decision. Subsequently in R v Kehoe [2008] EWCA Crim 819; [2009] 1 Cr App R(S) 41, para 17 the Court of Appeal expressed the view that:

To the extent that this passage in Kehoe suggests that section 225(2) involves a raising of the threshold for imposition of a discretionary life sentence, it is difficult to see in what that can really have consisted. But it is on any view clear that, in cases of significant risk to the public of serious harm by the commission of further specified offences, section 225(3) introduced a new form of indeterminate sentence, based on actual offending which was either of a kind for which a life sentence was not available or not of such seriousness as to justify the imposition of a life sentence.

The amendments made to the Crime (Sentences) Act 1997 had the effect of fitting the new form of sentence of IPP into a pre-existing framework established for mandatory or discretionary life sentences. In the light of decisions such as Thynne, Wilson and Gunnell v United Kingdom (1990) 13 EHRR 666 and Stafford v United Kingdom (2002) 35 EHRR 1121 and R (Noorkoiv) v Secretary of State for the Home Department [2002] 1 WLR 3284, the role of considering whether it was “no longer necessary for the protection of the public that the prisoner should be confined” after the expiry of the tariff period (or “minimum term”) fell upon the Parole Board under section 28(6)(b) in relation to all these types of sentence. Two clear differences did however exist between sentences of IPP and life sentences: the licence period in respect of an IPP could be brought to an end ten years after release (section 31A of the Crime (Sentences) Act 1997, as inserted by Schedule 18, paragraph 2, to the Criminal Justice Act 2003); and a minimum term of whole life could not be imposed in respect of IPP (Schedule 18, paragraph 4, to the 2003 Act).

Since sentences of IPP were fitted into a larger framework, it is relevant to consider how that framework operated. The Criminal Justice Act 1967, section 61(1) provided merely that the Secretary of State “may, if recommended to do so by the Parole Board, release on licence a person serving a sentence of imprisonment for life”. In R v Secretary of State for the Home Department, Ex p Benson (The Independent, 16 November 1988) and in Ex p Bradley, the Divisional Court considered the nature and level of risk by reference to which the Parole Board should measure whether continuing detention was justified. In Benson Lloyd LJ said that the decision whether to recommend or release on licence depended on whether there was a risk of repetition of “the sort of offence for which the life sentence was originally imposed, in other words risk to life or limb”, including in that category a non-violent rape.

In Bradley the court was concerned with the extent of the risk relevant to sentencing and to release on licence. In rejecting Mr Edward Fitzgerald’s submission that the two must equate, the court said (p 145F-H):

The court went on to say of the level of risk relevant at the release stage, that it “must indeed be ‘substantial’ …, but this can mean no more than that it is not merely perceptible or minimal”, that it must be unacceptable in the subjective judgment of the Parole Board and that the Parole Board must have in mind all material considerations, scrutinising ever more anxiously whether the level of risk is unacceptable, the longer the time the offender has spent in prison post-tariff (p 146D-F, and see R v Parole Board, Ex p Wilson [1992] QB 740, 747E-G).

In Ex p Wilson Mr Fitzgerald sought initially to renew the submission, rejected in Ex p Bradley, that the test which the Parole Board should apply for continuing detention should be likelihood of re-offending. On reflection, he abandoned the challenge and accepted the correctness of Ex p Bradley, and the Court of Appeal commented that in its judgment he was right to do so (p 747A).

Section 34(4)(b) of the Criminal Justice Act 1991 introduced a statutory definition of the Parole Board’s function in relation to release of a discretionary life prisoner, using language subsequently reproduced in section 28(6)(b) of the Crime (Sentences) Act 1997. The Parole Board was not to give a direction with respect to a discretionary life prisoner unless “satisfied that it is no longer necessary for the protection of the public that the prisoner should be confined”. The approach established at common law by Ex p Bradley and Ex p Wilson was, in R v Parole Board, Ex p Lodomez (1994) 26 BMLR 162, held applicable under the statutory test introduced by section 28(6)(b). Leggatt LJ in the Divisional Court in Ex p Lodomez recorded his view that it was “unhelpful to invent alternative versions of the statutory test” (p 184).

In the present case, the Court of Appeal applied to sentences of IPP the approach taken in Ex p Bradley with regard to discretionary life sentences. The appellant’s first ground of appeal is that IPP is distinguishable in principle from a life sentence and that it was wrong to apply Ex p Bradley to IPP. Alternatively, if that submission be rejected, then it is submitted Ex p Bradley was wrong and should be over-ruled.

The appeal has been conducted on both sides on the basis that a substantial distinction exists between the test of “significant risk to members of the public” applicable under section 225(1)(b) and the test “no longer necessary for the protection of the public” as understood and applied to discretionary life sentences by Ex p Bradley. But I have already indicated my view that Ex p Bradley went too far in equating significant risk with mathematical likelihood and that “good grounds” would represent a more acceptable elaboration. As to the phrase “no longer necessary for the protection of the public”, the Divisional Court considered that the level of risk which it involved must remain undefined, but offered two observations. First, it seemed inevitable that the risk must be “substantial”, which the Divisional Court thought “can mean no more than that it is not merely perceptible or minimal” (p 146). Second, it must be sufficient to be judged unacceptable in all the circumstances in the subjective judgment of the Parole Board, and, third, in exercising their judgment as to the level of risk acceptable the Parole Board must have in mind all relevant considerations.

A possible difficulty about the Divisional Court’s first observation is that the equation of “substantial” risk with any risk that is “not merely perceptible or minimal” tends to change the focus or starting point in a way which may influence the conclusion. It is preferable to concentrate on the statutory language and not to paraphrase. This applies to the assessment both of “significant risk” under section 225(1)(b) of the 2003 Act and of whether detention is “no longer necessary for the protection of the public” under section 28(6)(b) of the 1997 Act.

In relation to the first ground of appeal, Mr Southey relies upon the exceptional nature of a life sentence, imposed, the Divisional Court said in Ex p Bradley, only where there was “compelling justification [consisting in] the perception of grave future risk amounting to an actual likelihood of dangerousness” (p 145). However, as indicated in paragraphs 14 and 21 above, the distinction under section 225 between circumstances calling for a discretionary life sentence and for IPP may depend more upon the seriousness of the offence actually committed than upon any difference in the offender’s assessed dangerousness. The threshold criterion of dangerousness for the purposes of applying section 225 (“significant risk to members of the public of serious harm occasioned by the commission by him of further specified offences”) was the same in each case.

Mr Southey argues that discretionary life sentences and IPP are or may also be different in two other respects. One is that it was wrong to speak of an onus on an offender serving IPP to disprove his dangerousness. Although the default position is that detention will continue “unless … the Board is satisfied that it is no longer necessary for the protection of the public that the prisoner should be confined”, the Parole Board is an investigative body which will make up its own mind on all the material before it: see R v Lichniak [2002] UKHL 47, [2003] 1 AC 903, para 16, per Lord Bingham, and In re McClean [2005] UKHL 46; [2005] UKHRR 826, paras 74-78, per Lord Carswell. Mr Southey suggests that the position may be different in relation to an offender serving a life sentence, while at the same time wishing to reserve the argument that it should not be. If there is in this respect a difference (which I doubt), I regard it as immaterial to the issues which the Supreme Court has to decide on this appeal.

The other respect turns on reasoning of the Supreme Court in R v Smith (Nicholas) [2011] UKSC 37, [2011] 1 WLR 1795. This is relied upon as establishing that, when considering whether to impose a sentence of IPP, the sentencing court was not making a predictive judgment of risk at the expiry of the tariff period. If that is right, then there was a marked distinction between the criteria governing imposition of a discretionary life sentence (as hitherto understood) and a sentence of IPP. The reasoning in Ex p Bradley relies at least in part upon the predictive assessment in relation to the post-tariff period which a sentencing court makes when considering whether to impose a discretionary life sentence, and the distinction between that assessment and the contemporary evaluation of the Parole Board at the post-tariff review stage.

Counsel for the Parole Board and for the Secretary of State did not in their written cases or oral submissions take issue with the reasoning in R v Smith. But I am far from satisfied that it can be regarded as the last word. There is nothing in the language of section 225(1)(b) to suggest any distinction between the nature of the assessment required for the purposes of considering whether to impose a discretionary life sentence and a sentence of IPP. On the contrary, the same kind of assessment of risk was on the face of it required for both, with the distinction between them for the purposes of the section focusing, according to section 225(2) and (3), on the seriousness of the actual offence committed.

Further, in a number of other cases, the predictive approach appears to have been assumed to be correct in relation to the imposition of a sentence of IPP. In R v Johnson [2006] EWCA Crim 2486, [2007] 1 WLR 585, Sir Igor Judge P said (para 10) that “It does not automatically follow from the absence of actual harm caused by the offender to date [i.e. to the date of sentencing], that the risk that he will cause serious harm in the future is negligible”. More clearly, in R (Walker) v Secretary of State for Justice [2009] UKHL 22, [2010] 1 AC 553, Lord Judge CJ referred repeatedly to the “predictive” assessment to be made when sentencing as to the risk of dangerousness at the expiry of the tariff period: see paras 102, 103 and 108. Citing Sir Igor Judge’s words in para 10 in Johnson,Hughes LJ in R v Pedley [2009] EWCA Crim 840, [2009] 1 WLR 2517 also referred to the imposition of a sentence of IPP as depending upon an assessment of future risk: see paras 16, 20 and 21.

In R v Smith (Nicholas), the primary issue was whether it was legitimate to pass a sentence of IPP for armed robbery and possession of a firearm on a career criminal who had already been recalled to prison to serve the remainder of a previous life sentence also imposed for armed robbery and having a firearm with intent. The submission was that, because he could not be released from that life sentence unless and until the Parole Board was satisfied that it was no longer necessary for the protection of the public that he should be detained, there could not be said to be any “significant risk to members of the public of serious harm occasioned by the commission by him of further specified offences” so as to justify IPP. Counsel supported this submission by arguing that the sentencing judge’s duty was to “consider whether he [the offender] will pose a significant risk when he has served his sentence” (para 14). Lord Phillips, giving the judgment of the court, addressed the submission on the same basis, and rejected it, saying:

It is notable that, although Pedley was cited in argument (though not in the judgment), Walker does not appear to have been referred to in Smith at all. Nor does the reasoning in Smith address the relationship between discretionary life sentences and IPP or consider what basis there could be for requiring a different approach to the assessment of risk under the latter, when compared with the former. An important part of the rationale of a discretionary life sentence is, on the authorities, an assessment of future risk; such a sentence may be appropriate because “It is sometimes impossible to say when that danger will subside”; or “the offender may remain a serious danger to the public for a period which cannot be reliably estimated at the date of sentence" (see Wilkinson and Whittaker, cited in paras 16 and 17 above). While the Divisional Court in Ex p Bradley placed too much emphasis in my view on mathematical probability, it was unquestionably right to consider that a sentencing judge deciding whether to impose a discretionary life sentence was required to assess risk on a predictive basis. The natural inference would be that a parallel approach was expected to be applied to the new sentence of IPP, when fitted into the pre-existing framework governing discretionary life sentences. There is nothing unrealistic about asking a sentencing judge to assess whether an offender presents a risk for a period which cannot reliably be estimated and may well continue after the tariff period.

Logically, it is also difficult to see why it was necessary at all in Smith to address the question whether the sentencing judge’s assessment was of present risk or predictive. If the fact that the offender was in prison was relevant at all, it would exclude any present as much as any future risk of the offences to which he was evidently prone. The point which required decision was that, when deciding whether to order IPP, any concurrent prison sentence was to be ignored and the offender was to be assumed to be at liberty. More generally, unless the judgment required in the case of IPP is predictive, it must logically follow that, even though the fixed (tariff) period would in the judge’s view be sufficient to eliminate any further future risk before the tariff expired, the judge would still be required (even after the time when the imposition of IPP became discretionary) to impose a sentence of IPP, although convinced that there was no point in doing so. The concept of a long determinate sentence sufficient to eliminate future risk would be largely superseded.

In these circumstances, I have grave reservations about the reasoning in para 15 in Smith even in relation to sentences of IPP. But, since it was not challenged on this appeal and is not in my opinion ultimately decisive, I say no more on this.

In support of the appellant’s case on the first ground, Mr Southey is able to point to a number of statements in the cases. In R (Bayliss) v Parole Board [2008] EWHC 3127 (Admin); [2009] EWCA Civ 1016, Cranston J and the Court of Appeal were content to proceed on the basis accepted by counsel that the test for release from IPP mirrored the test for imposition of a sentence of IPP. In Pedley Hughes LJ held that a sentence of IPP was Convention compatible, because inter alia it was “proportionate to the risk of serious harm, particularly since when the tariff sentence attributable to the instant offence has been served, the system provides for release once that significant risk no longer exists” (para 22). In Ex p Walker when that case was in the Court of Appeal, [2008] EWCA Civ 30; [2008] 1 WLR 1977, Lord Phillips CJ described the primary object of IPP as being “to detain in prison serious offenders who pose a significant risk to members of the public of causing serious harm by further serious offences until they no longer pose such a risk” (para 35).

None of these statements was however based on any detailed examination of the present issue, and I have come to the conclusion that they are wrong, so far as they suggest that the test which the Parole Board must apply when considering whether to direct release from IPP is precisely the same as that which the sentencing judge had to apply in order to pass a sentence of IPP in the first place. I set out my reasons in the following paragraphs. On the same basis, as well as in the light of what I have said in para 36 above, I also reject the submission that Ex p Bradley was wrongly decided.

First, the two tests are, both in their terms and in their default position, substantially different. Imposition depends upon the court being positively satisfied of “a significant risk to members of the public of serious harm occasioned by the commission of further specified offences”. Release depends upon the Parole Board being “satisfied that it is no longer necessary for the protection of the public that the prisoner should be confined”.

Second, the test for release applied under the 2003 Act to a sentence of IPP was the test for discretionary life sentences encapsulated in statutory form first in section 34(4)(b) of the 1991 Act, and later in section 28(6)(b) of the 1997 Act, and since also applied to mandatory life sentences. Those drafting and enacting the 1991 Act must be taken to have been aware of the decision in Ex p Bradley (decided on 4 April 1990). Those drafting and enacting the 1997 Act must be taken to have been aware of and accepted the line of authority consisting of Ex p Bradley, Ex p Wilson and Ex p Lodomez. Parliament therefore accepted a difference in the tests for imposing and for release from a discretionary life sentence. In introducing a sentence of IPP into the same framework for release as applies to discretionary life sentences, Parliament must on the face of it have intended to apply to sentences of IPP the same test for release as for discretionary life sentences, again even though that differed from the test for imposition.

Third, the phrase “no longer necessary for the protection of the public” in the test for release does not import any reference to the threshold risk justifying the imposition of the sentence. The sentence imposed will itself operate as a complete protection of the public against any real risk during the tariff period. The phrase does no more than raise the question whether continued detention, after the tariff period, is any longer necessary to achieve that protection.

Fourth, I see no inconsistency or incongruity in a scheme involving a higher initial threshold of risk for the imposition of a life sentence or a sentence of IPP, but requiring a somewhat lower risk to be established in order for the convicted offender to be eligible for release. This is so even if a sentencing judge deciding whether to impose a sentence of IPP was not engaged in the predictive exercise held in Ex p Bradley to be required when a court considers whether to impose a discretionary life sentence. Those who cross the initial threshold have notice from the case-law that they are at peril of being held to protect the public against a more general and lesser level of risk. The threshold consists of the commission of a serious offence coupled with the existence of a significant risk of the commission of further specified offences. A person who has not committed a serious offence cannot be detained, even if he presents a significant risk of the commission of specified offences: that is because the threshold has not been crossed. But where the threshold is crossed, it does not follow that the objective of detention beyond the tariff period is confined to the elimination of any significant risk (whether that means whatever significant risk was identified when the sentence of IPP was imposed or any significant risk which may at the end of the tariff period be thought to exist). The objective may well be the more general protection of the public for as long as necessary. This, on the face of it, is also what the statutory test for release under section 28(6)(b) states.

Fifth, the appellant’s case is that an offender serving a sentence of IPP should not continue to be detained after the tariff period, if he can show that there is no significant risk to members of the public of serious harm occasioned by the commission by him of further specified offences. The appellant does not carry the logic of his case to the point of suggesting that the risk must be of the commission of the same kind of specified offences as those for which he was originally sentenced. But, assuming that the test for release were to be whether there was or was not a risk of “further specified offences of whatever kind, even though not of the same kind as those for which he was originally sentenced”, still it would appear to follow that an offender who was known to present a risk to his partner’s life because of her unfaithfulness during his imprisonment could not continue to be detained. This is because murder is not a specified offence, which is because it carries a mandatory life sentence. That in turn reflects the fact that the schedule of specified offences was designed to meet the requirements for imposition of a sentence of IPP. It was not intended to operate as part of the test for release from IPP. What matters when release is being considered is whether the prisoner presents a continuing risk to life or limb, including a non-violent rape, from which the public needs protection by way of his continuing detention (see para 23 above).

Sixth, it would seem logically also to follow by extension, on the appellant’s case, that an offender serving a life sentence should not continue to be detained after the tariff period, if he can show that there no longer exists the same severe risk of serious re-offending that justified the life sentence in the first place. That would mean that an offender serving a life sentence would be entitled to release despite a risk of re-offending that would justify an offender serving IPP continuing to be detained. If this difficulty is sought to be avoided by distinguishing between offenders serving life sentences and those serving sentences of IPP, that brings the argument back to the second point. It is implausible to think that any such distinction was intended.

Seventh, I do not accept Mr Southey’s submission that the provisions of the European Convention on Human Rights require a different interpretation to be put on the statutory language to that which would otherwise apply. The submission is that the Convention requires that any decision to maintain detention becomes illegitimate if “based on grounds that had no connection with the objectives of the legislature and the court or on an assessment that was unreasonable in terms of those objectives”: Van Droogenbroeck v Belgium (1982) 4 EHRR 443, para 40, M v Germany (2009) 51 EHRR 976, para 88 and James v United Kingdom (2012) 56 EHRR 399, para 195.

In the last case, the European Court accepted in para 198 that there was a sufficient causal connection between the imposition of a sentence of IPP, because the offender was “considered … to pose a risk to the public”, and his deprivation of liberty in the post-tariff period, when his “release was contingent on [him] demonstrating to the Parole Board’s satisfaction that [he] no longer posed such a risk”. It was not concerned with the question whether the minimum level of risk at each stage was required to be identical in order to comply with the Convention. In principle, I see no reason why it should be. There was and is no reason why the objectives of the statutory scheme should not involve a high threshold for imposition of a discretionary life sentence or sentence of IPP, but a high level of security against risk for release (in other words, a lower threshold for continuing detention) after the tariff period. The provisions whereby IPP was introduced into law and related to or assimilated with the position regarding discretionary life sentences must be read as a whole. So read, in my view, the natural conclusion is that this is what Parliament intended. I see nothing in the Convention inconsistent with such an approach.

I would therefore dismiss this appeal on the first ground.

The second ground arises from a complaint by Mr Sturnham that the Parole Board in determining whether to direct his release took into account a direction by the Secretary of State to the effect that the appropriate test was whether the Board was satisfied that his level of risk was “no more than minimal”. In R (Girling) v Parole Board [2006] EWCA Civ 1779; [2007] QB 783 the Court of Appeal held that the Secretary of State had no power to give such a direction, and accordingly it was wrong of the Parole Board to act on it.

Mitting J was shown a copy of an internal Parole Board document dated July 2010, in which the Parole Board had itself adopted the “more than minimal” test internally in guidance issued to its members. He recorded that there was no direct challenge to that guidance bearing in mind its date. But he added that, if the Board had followed it, he “would not be prepared to say that it was an unlawful test”, but that “Beyond that it would not be wise for me to go, given that the issue has not been fully ventilated” (para 32). He dismissed the actual ground of appeal on the basis that there was nothing to show that the Board had taken that part of the Secretary of State’s direction into account. The Court of Appeal upheld the judge’s decision on the latter ground.

Before the Supreme Court, Mr Southey barely addressed the complaint that the Parole Board wrongly guided itself by reference to directions which the Secretary of State had no power to give, and has not established any reason for this court to do other than uphold the decisions below on that aspect. The appeal on the second ground should therefore also be dismissed.

In so far as Mr Southey invited us to try to define more closely the meaning of section 28(6)(b), I would decline the invitation and repeat that it is preferable to concentrate on the statutory language, rather than to seek to paraphrase it.

Having granted permission to appeal as stated in paragraph 3, I would therefore dismiss the appellant’s appeal on both grounds as stated in paragraphs 49 and 52.