Harris, R (on the application of) v Secretary of State for Justice

The appellant, Lorraine Allen, formerly Lorraine Harris, had her conviction for the manslaughter of her 4-month old son quashed by the Court of Appeal (Criminal Division) (“CACD”), which had heard fresh evidence. She applied to the Secretary of State for compensation for miscarriage of justice but his decision was that she was not eligible. She sought judicial review of that decision but failed before Mitting J, who rejected the specific reasons of the Secretary of State, but held that the appellant’s application for compensation could not in any event succeed. She now appeals against Mitting J’s refusal to quash the decision. The Secretary of State’s original reasons are no longer in issue. Three issues are potentially raised by the appeal:

the proper meaning of section 133 Criminal Justice Act 1988 and of the expression ‘miscarriage of justice’ contained in it; in particular which of the two interpretations advanced respectively by Lord Bingham and Lord Steyn in R (Mullen) v Secretary of State for the Home Department[2004] UKHL 18; [2005] 1 AC 1, is correct;

whether, if Lord Bingham’s construction be correct, this appellant was treated by the CACD as someone against whom there was, on the evidence as now understood, no basis on which a jury could properly convict, and should for that reason be considered to be a person who has suffered a miscarriage of justice;

whether, even if this is not so, the presumption of innocence, enshrined in Article 6(2) of the European Convention on Human Rights (“ECHR”) has the effect that the appellant, having had her conviction quashed, is entitled to be treated as having had her innocence established and therefore ipso facto to have suffered a miscarriage of justice.

The material parts of section 133 provide as follows:

It is common ground that section 133 was enacted in order to make the clearer this country’s compliance with its treaty obligations under Article 14(6) of the International Covenant on Civil and Political Rights 1966 (Cmnd 6702; 1977) (‘the ICCPR’). Except that the statute substitutes the expression “beyond reasonable doubt” for the covenant word “conclusively”, section 133(1) is in terms identical to Article 14(6).

There existed before the enactment of section 133 a discretionary ex gratia scheme for the payment of compensation; it has now been withdrawn, and there may be outstanding litigation concerning the withdrawal. We are, however, concerned in this case only with the statutory provision for compensation contained in section 133.

The facts of the appellant’s case are set out very clearly in the judgment of Gage LJ in the CACD: R v Harris, Rock, Cherry and Faulder[2005] EWCA Crim 1980. She was convicted of the manslaughter of her 4-month old son. He died of closed head injuries of the kind often referred to as “shaken baby syndrome” or, as the judgment of the CACD prefers to term it, “non-accidental head injury”. The Crown’s case was that she had shaken him in a violent manner beyond any proper handling, and/or thrown him down in his cot or otherwise struck his head against some object. Her case was that she had discovered him in the middle of the night floppy and unable to wake up; she had done no more than at most to bounce him properly on her knee in panic.

The household consisted of the appellant, the father of her son, and two daughters of hers by a previous relationship. There were no financial difficulties and the health professionals who had dealt with the family said that it gave the impression of perfection. On the night in question the boy was in the sole care of the appellant; father being away working a night shift.

At just before 1 a.m. the appellant telephoned her general practitioner and reported that she had woken to feed her son and found that he was not breathing. Her later evidence was that he was having difficulty breathing and she overstated his condition to get the doctor to hurry. The doctor visited. After examining the child he concluded that apart from his being a little snuffly there was nothing wrong with him. His eyes were normal and there were no signs of injury or abuse. He left the house at 0130.

At 0234 the appellant made a 999 call for an ambulance. She said that her son would not wake up. The ambulance crew, which arrived seven minutes later, found no pulse and could not revive him. He had fixed and dilated pupils and both then and at hospital very obvious retinal haemorrhages were apparent. Despite the best efforts of the medical teams at two hospitals, he could not be saved. The pathologist found no external injuries of any kind, but subdural and subarachnoid haemorrhages, marked swelling and softening of the brain associated with widespread hypoxic changes, and extensive bleeding in the retinae and vitreous of the eyes. These three elements of brain damage, subdural bleeding and retinal haemorrhages have frequently been called ‘the triad’ and often treated by medical opinion as strongly suggestive, or by some doctors as diagnostic, of non-accidental injury. That was the basis of the medical evidence called by the Crown.

Extensive medical evidence was also called for the appellant at the trial. The medical case for her accepted that the findings were consistent with violent shaking, with or without impact, but disputed that they necessarily demonstrated that this had happened. Alternative causes of the injuries found which were canvassed were bleeding from a blood disorder and re-bleeding of an old blood clot.

The appellant was convicted by the jury of manslaughter. Thus, the jury found that it was sure, to the criminal standard of proof, that she had used unlawful force on her son and thereby caused his death.

Her case was one of four referred to the CACD after well known doubts had been expressed about expert evidence in so-called shaken baby syndrome cases. In particular a team led by Dr Geddes had postulated a hypothesis (‘Geddes III’) that hypoxia (lack of oxygen), both alone and in combination with infection, was capable of giving rise to the triad of symptoms, without there being any unlawful force and indeed possibly as the result of an incident of apnoea (temporarily ceasing to breathe). Dr Geddes was one of many medical experts who gave evidence before the CACD. In the course of her evidence, however, she acknowledged that her team’s hypothesis did not stand up to critical analysis and the court rejected it as a credible cause of the triad of injuries.

There remained a substantial divergence of medical opinion amongst the expert witnesses called in the CACD. I borrow extensively, and with gratitude, much of Mitting J’s analysis of it. In particular, there was a “head-on collision” between two consultant experts in neuro-pathology. Dr Waney Squier, who had at the time of trial advised the appellant that the injuries found were non-accidental, had changed her mind and before the CACD firmly expressed her view that there was no incontrovertible evidence justifying this conclusion. Dr Rorke-Adams, by contrast, was firmly of the view that the medical findings did indeed lead to this conclusion. The two doctors disagreed not only as to their interpretation of the findings, but as to what clinical signs could be seen from the photographs; neither had examined the body personally. As to these two experts, Gage LJ expressed the court’s conclusions in this way:

Almost as sharp a disagreement existed between the evidence of the ophthalmic surgeons called in the CACD. Dr Gilliam Adams expressed the opinion that the fact that the ambulance crew noted fixed and dilated pupils at 0241 was a sign that the brain was swollen at that stage, which could in turn give rise to a possibility that there was a cause for retinal haemorrhages other than shaking. By contrast a radiologist and a neuro-surgeon called for the Crown did not accept that there could have been brain swelling by 0241 and Mr Richards was in no doubt that the force used to cause the triad of injuries must have been more than rough handling.

One matter which was resolved in the CACD was the theory, espoused by Dr Squier amongst others, that the cause of brain swelling might have been an infection. The end result of the evidence called on that issue was that leading counsel for the appellant accepted that infection as a possible cause of collapse was effectively disproved.

The CACD expressed its conclusions on the appellant’s case initially as follows:

The court concluded its judgment as follows:

The test in fresh evidence cases referred to in paragraph 70, and demonstrably applied in paragraph 153, was that advanced by Lord Bingham in R v Pendleton[2001] UKHL 66; [2002] 1 WLR 72 at paragraph 19, as explained by the Privy Council in Dial v State of Trinidad and Tobago[2005] 1 WLR 1660 at paragraphs 31 and 32. The decision as to the safety of a conviction in a fresh evidence case is for the court itself and is not what effect the fresh evidence would have on the mind of a jury, but in a difficult case the court may find it helpful to test its provisional view by asking whether the evidence now available might reasonably have affected the decision of the trial jury to convict. In the present case it is clear that the CACD adopted this latter approach and relied significantly for its decision on what might have been the impact of the medical evidence which it had heard if such evidence had been available to the jury. From the passages quoted above, and particularly those which I have emphasised, there can be no doubt that the court regarded the proper interpretation of the clinical findings in this case as a matter which it ought not itself to resolve, but rather as one which could and should be resolved by a jury on hearing the competing expert opinions. Adopting that approach, it decided that the evidence which was now available might, if it had been heard by the jury, have led to a different result.

The events leading to the death of the appellant’s son took place in the early hours of 5 December 1998. Her trial was concluded on 7 September 2000. She was sentenced to three years imprisonment. The appeal in the CACD was heard in July 2005, nearly five years out of the normal time for appeal and as a result of a general review of similar cases initiated by the Law Officers of the Crown. It followed that by the time of the appeal the appellant had served her sentence and a great deal of time had passed. Understandably, in those circumstances, there was no application by the Crown for a re-trial, as there would no doubt have been had the conviction been quashed for these reasons shortly after trial.

By section 2 of the Criminal Appeal Act 1968, the CACD must quash a conviction if it is shown to be unsafe. In one sense, therefore, everyone whose conviction is thus quashed on appeal might be said to have suffered a miscarriage of justice, because they have been unsafely convicted. It is quite clear that the right to compensation under article 14(6) and section 133 is much more restricted than that. Firstly, the right to compensation only arises in those cases of quashed convictions where the appeal succeeds on the basis of ‘a new or newly discovered fact’. The common cases of error of law at trial, or misdirection of the jury, or procedural irregularity are thus excluded. Secondly, it arises only where the miscarriage of justice is established by such new fact ‘conclusively’ or ‘beyond reasonable doubt’. The common case in which a conviction is quashed because there exist some doubts about its safety is thus also excluded. Thirdly, section 133 gives effect to the requirement that there shall have been a new or newly discovered fact by restricting compensation to cases other than appeals allowed in ordinary course, namely to those either heard out of time or referred by the Criminal Cases Review Commission (s 133(5)).

In R (Mullen) v Secretary of State for the Home Department[2004] UKHL 18, [2005] 1 AC 1, the House of Lords held unanimously that the expression ‘miscarriage of justice’ in article 14(6) and section 133 did not extend to every case of conviction quashed, even after the discovery of new facts. Mullen had been convicted of conspiracy to cause explosions. His trial in England had been possible only because the British authorities had arranged his deportation from Zimbabwe in flagrant breach of local and international law. That that was so emerged only after his trial and conviction. His case reached the CACD approximately seven years after his conviction. His conviction was quashed on the ground that his deportation had involved abuse of process, namely a gross abuse of executive power. R v Horseferry Road Magistrates’ Court, ex p Bennett[1994] 1 AC 42 had established the jurisdiction to stay a criminal prosecution on such grounds, but had done so only after Mullen’s trial. It followed that the circumstances of his deportation became relevant to any power of the criminal court only after trial and there had been no material failure of disclosure at the time of trial. He did not contend on appeal that he was other than guilty of the offence; he relied exclusively on the abuse of process. His subsequent claim for compensation under section 133 and/or the ex gratia scheme was refused. His challenge to that refusal failed in the House of Lords. The unanimous decision was that section 133 did not extend to the case where the conviction was quashed on grounds entirely unconnected with the guilt or innocence of the defendant and where there had been no failure in the trial process.

In the course of the speeches in that case a well-known difference as to the interpretation of section 133 emerged. Lord Steyn held (at paragraphs 34 - 56) that in this context ‘miscarriage of justice’ means that the innocence of the defendant is acknowledged. Lord Bingham (at paragraphs 4 - 9) expressed no concluded opinion on this question, but made it clear that he “hesitated to accept” this interpretation. For his part, he was ready to accept that ‘miscarriage of justice’ extended in this context to serious failures of the trial process, whether or not innocence was demonstrated. Of the other members of the House, Lord Scott expressed no view on this topic, Lord Rodger agreed with Lord Steyn, and Lord Walker, whilst holding that it was unnecessary to go further than the limited unanimous ground for refusing Mullen’s application, was of opinion that Lord Steyn had set out powerful reasons for his interpretation.

Given the unanimous view of the House that Mullen’s claim failed in any event, these different interpretations were not strictly necessary to the decision. That is no doubt why Lord Bingham expressed himself in the cautious manner in which he did. The speeches, however, followed sustained and detailed submissions on both sides on this topic.

For the appellant, Mr Owen QC accepts that her innocence has not been demonstrated beyond reasonable doubt, or conclusively, by the decision of the CACD. Accordingly, if the correct interpretation of section 133 is Lord Steyn’s her claim for compensation must fail. But his first submission is that Lord Bingham’s interpretation should be adopted and that on this interpretation the appellant succeeds because something went seriously wrong with the trial process. That is so, he contends, because the effect of the decision of the CACD is that the trial court was presented with evidence which significantly overstated the conclusions which could safely be drawn from the triad of injuries. He says that it can now be seen that the expert evidence at trial was flawed, not because it was other than conscientious and experienced, but because advances in medical thinking show that the triad does not alone bear the weight which was ascribed to it by the evidence of the doctors called for the Crown at trial. He further contends that the effect of the decision of the CACD is to hold that where the prosecution depends on the triad alone there is no case which can properly be left to the jury, because it cannot properly convict.

At this stage, I take, for the purposes of argument, Lord Bingham’s interpretation of section 133 (see, however, below). At its most favourable to Miss Harris, it contains this passage in paragraph 4 of his speech:

As the opening sentence makes clear, Lord Bingham was there considering the use of the term ‘wrongful convictions’ by successive Home Secretaries in the context of the ex gratia scheme. It must be remembered that in Mullen, both statutory and ex gratia schemes were under consideration. However, when he came to consider section 133 and “very briefly” to indicate why he hesitated to accept the interpretation which Lord Steyn espoused, Lord Bingham went on at paragraph 9(1) to say this:

Whatever arguments may exist about the inter-relation of Lord Bingham’s paragraphs 4 and 9, it is plain that the critical feature of the extended interpretation of ‘miscarriage of justice’ which he was prepared to contemplate is that “something has gone seriously wrong in…..the conduct of the trial”: see the concluding words of paragraph 4. That is made the plainer by his references to a defendant who “should clearly not have been convicted” (paragraph 4) and “certainly should not have been convicted” (paragraph 9(1)) [my emphasis].

In the present case there was nothing which went wrong with the conduct of the trial, whether seriously or otherwise. In speaking of ‘flawed expert evidence’ it is clear that Lord Bingham cannot have been contemplating evidence which was conscientiously given and based upon sound expertise at the time of trial. The most that could be said against the expert evidence given at this trial is that it might need adjustment in the light of new medical research and/or thinking. In any event, the medical evidence given at time of trial has not been demonstrated to be flawed, even in this limited sense. As the passages from the judgment of the CACD which I have cited show, this court’s decision went no further than to say that the differences of medical opinion needed to be resolved by a jury. Nor was this a case in which the jury was presented with a medical consensus that the triad was diagnostic of unlawful killing. The medical evidence called for the appellant accepted that it was consistent with unlawful killing but disputed that it necessarily led to that conclusion. The appeal was allowed because over the intervening years more possible force had emerged for the opinion voiced on the appellant’s behalf and now supported by Dr Squier’s evidence, which the jury had not heard and which the CACD, despite plain doubts about it, was not in a position wholly to dismiss.

For the same reasons, I have no doubt that the decision of the CACD does not begin to carry the implication that there was no case for the appellant to answer once the fresh evidence was available. Mr Owen points to paragraphs 70 and 152, cited above, and to the statements that the court did not feel able to say that the triad on its own must automatically and necessarily lead to a diagnosis of non-accidental injury. He further reminds us that in R v Cannings[2004] EWCA Crim 1; [2004] 1 WLR 2607, in the context of repeated and wholly unexplained sudden infant deaths where no injuries could be found, the CACD warned (at paragraph 178) that in such cases, if the outcome depends exclusively or almost exclusively on a serious disagreement between distinguished and reputable experts, it will often be unwise and therefore unsafe to proceed. But in the subsequent case of R v Kai-Whitewind[2005] EWCA Crim 1092, Judge LJ, who was the author of the judgment in Cannings, made it absolutely clear (at paragraphs 82 – 90) that the proposition cited was confined to cases where there was no evidence except such conclusions as could be drawn from repetition or coincidence. Cannings was held not to be any authority for saying that when experts disagree as to the conclusions which can be drawn from injuries discovered, the case should be withdrawn from the jury; on the contrary, the resolution of such disagreements, bearing in mind the criminal standard of proof, is an important part of the functions of a jury.

In the present case, there was no basis for saying that, on the new evidence, there was no case to go to a jury. Moreover, if the court had meant to say that there was (now) no case to answer, it would have said so in plain terms. On the contrary, its oft-repeated statements that the evaluation of the rival medical opinions would be a matter for the jury are wholly inconsistent with a finding that there was no case to answer on the new state of medical evidence. Likewise, the posing of the Pendleton question by way of check is inconsistent with a finding that the case should never have reached the jury if the fresh evidence had been known.

In those circumstances, I reach the clear conclusion that, even on the interpretation of section 133 which Lord Bingham favoured, this case cannot succeed. That means that it is unnecessary to resolve the difference of interpretation between Lord Bingham and Lord Steyn.

That accords with the conclusion recently arrived at by the Divisional Court in R (Clibery) v Secretary of State for the Home Department[2007] EWHC 1855 (Admin). Mr Clibery had been convicted in 1998 of rape of his ex-wife. Subsequently, material became known which would have been usable to cast doubt upon the complainant’s general credibility. On a reference to it by the CCRC, the CACD quashed the conviction on the grounds that it was unsafe. The defendant’s subsequent application for compensation under section 133 was refused by the Home Secretary and the refusal was challenged by application for judicial review. Dismissing that application, Lord Phillips CJ said this:

Mr Owen asks us to say that the decision in Clibery was wrong. He contends that upon Lord Bingham’s interpretation all cases where the conviction is quashed on the basis of new or newly discovered fact are within section 133 except for the Mullen type of case in which the appeal has been allowed for reasons which are unconnected with the guilt or innocence of the defendant. For the reasons already given, this contention cannot stand with Lord Bingham’s emphasis that article 14(6) is directed to ensuring that defendants are fairly tried and to the need even on the extended meaning of miscarriage of justice for there to have been something go seriously wrong with the investigation or conduct of the trial. In my respectful opinion, Clibery is not wrong but right.

Mr Owen advances alternatively what he describes as a free-standing ground of appeal. It is that the appellant is entitled to compensation under section 133 because once her conviction was quashed she reverts to the position of an unconvicted person who can rely on the presumption of innocence contained in article 6(2) ECHR. For the state to voice any qualification to her innocence is, Mr Owen submits, to infringe the appellant’s rights under that article. To deny her compensation on the grounds that she has not been demonstrated beyond reasonable doubt (or conclusively) to be innocent, or, as the case may be, to have suffered a miscarriage of justice by reason of serious error in the investigation or conduct of the trial, is to voice such qualification and thus to infringe article 6(2).

This argument is founded upon a number of decisions of the European Court of Human Rights when considering particular national provisions for compensation payable to those who are acquitted at trial. In a series of cases beginning with Sekanina v Austria(1994) 17 EHRR 221 the Court considered an Austrian statutory provision which enabled a person acquitted at trial to claim compensation if time had been spent in custody on remand but permitted the court to refuse it unless any suspicion that he had committed the offence had been dispelled. Compensation had been refused by the criminal court, albeit a different branch from the court of trial, on the grounds that “the jury took the view that the suspicion was not sufficient to reach a guilty verdict; there was, however, no question of the suspicion being dispelled”. The Strasbourg court held that that amounted to voicing a suspicion which undermined the acquittal and thus constituted an infringement of article 6(2). Other cases following Sekanina in identical circumstances are Rushiti v Austria(2001) 33 EHRR 56 and Weixelbraun v Austria(2003) 36 EHRR 45. Similarly, in O v Norway (application 29327/95) and Hammern v Norway (application 30287/96), decided on the same day, the Court reached a like conclusion where a Norwegian court, sitting with a jury, had acquitted the defendant but the Judges of that same court went on to refuse compensation on the ground that the statute required the defendant to show on the balance of probabilities that he had not committed the offence of which he had been acquitted. Mr Owen’s submission is that this reasoning applies equally to the case of a person whose conviction has been quashed, because there is a final determination in her favour. Moreover, he says that it would be anomalous to hold otherwise since a person such as this appellant may have been in prison for much longer than the remand period spent there by persons in the Austrian or Norwegian cases.

In my view these cases do not lead to the conclusion which Mr Owen seeks to derive from them, for the following reasons:

None of these cases considered the ICCPR scheme for payment of compensation for conclusively proved miscarriage of justice, which is what is in issue here.

Article 14 of the ICCPR juxtaposes within it both the provision for compensation in article 14(6), now under consideration, and, in article 14(2), a provision in terms identical to article 6(2) ECHR. Yet by article 14(6) it plainly requires something more than the quashing of the conviction before the right to compensation arises, namely that a miscarriage of justice be conclusively demonstrated by new or newly discovered facts. It does not seem to me that these provisions could co-exist in these terms if the consequence of article 14(2) was that nothing more could be required for compensation beyond the quashing of the conviction on the basis of new fact. Moreover, in WJH v The Netherlands (Communication No 408/1990; 31 July 1992) the Human Rights Committee established under the ICCPR decided that article 14(2) applied only to criminal proceedings and not to an application for compensation.

Whilst the ICCPR is a treaty independent of the European Convention, provisions identical to article 14(6) are to be found in Protocol 7 to the ECHR, article 3. For the same reasons, it is inconceivable that article 3 could be in the terms it is if article 6(2) of the main Convention meant that compensation necessarily followed the quashing of a conviction on the basis of fresh evidence.

As Lord Steyn pointed out in Mullen (at paragraph 41), the distinction between the Austrian domestic scheme then under consideration and the international scheme under Protocol 7 article 3 was one to which the Strasbourg Court carefully drew attention in Sekanina. It said, at paragraph 25:

It is plain from the Austrian and Norwegian cases that the line between the application and non-application of article 6(2) is frequently a fine one. In Sekanina the Commission (paragraph 37) expressly stated that article 6(2) “naturally” does not prevent the same facts being relied upon, post acquittal on the merits, to found a civil claim against the defendant, and this must occur routinely, as also must subsequent child care cases. Yet in Orr v Norway (31283/04; 15 May 2008) the Court held that article 6(2) disabled the complainant in a rape case from recovering compensation post acquittal notwithstanding the different standard of proof attributable to the civil claim; the decision was grounded upon the manner in which the court expressed itself in dealing with the latter question.

The basis for the decisions in the Austrian and Norwegian cases was the closeness of the link between the decision to acquit on the merits and the decision as to compensation. In the Austrian cases the compensation decision was within the jurisdiction of the criminal court, albeit it was usually made by a differently constituted criminal court some time after the acquittal, as for example a confiscation order may be in England. Moreover, the court proceeded in part by analysing the decision of the trial jury. In the Norwegian cases the acquittal was made by a court composed of judges and jury, and the same judges went on more or less immediately to consider compensation. In Sekanina the court said at paragraph 22, after referring to the time lag before the compensation claim was decided:

By contrast, compensation in a fresh evidence case under article 14(6) and section 133 is not linked to any acquittal on the merits. Rather, it is to be paid when not only has there been a reversal of the conviction but also where the additional factor exists of a miscarriage of justice demonstrated beyond reasonable doubt, or conclusively, to have taken place.

It can no doubt be said, with Mr Owen, that just as compensation for acquittal under the Norwegian scheme was described by the Court as a procedure whose object was “to establish whether the State had a financial obligation to compensate the burden it had created for the …person by the proceedings it had instituted against him” (see O v Norway paragraph 38), so too is the scheme for compensation for miscarriage of justice under article 14(6). But that is to beg the question when the scheme in question creates such an obligation. If article 6(2) were to apply to claims under the scheme here under consideration, there would be no reason in logic or fairness to distinguish between those whose convictions are quashed on grounds of fresh evidence and those whose convictions are quashed on other grounds; each would be in the position of being able to rely on the presumption of innocence. Indeed, there would be no obvious reason for distinguishing between those who are convicted but whose convictions are quashed, and those who are acquitted at trial. But it is clear that article 14(6) does not provide for compensation to be paid except in the limited circumstances to which it refers.

In Mullen, Lord Steyn held (at paragraph 44) that article 6(2) ECHR did not apply to the special rules created by article 14(6) ICCPR. Lord Bingham’s decision was that, even if the Austrian and Norwegian cases here reviewed might differ from the conclusion reached in WJH v The Netherlands, they could not assist Mullen since his ‘acquittal’ was unrelated to the merits of the accusation against him. However, he also said this (at paragraph 10):

For these reasons I would reject the alternative submission of the appellant based upon the presumption of innocence as established by article 6(2) ECHR.

On that analysis it is not necessary for our decision to resolve the difference of construction articulated by Lord Bingham and Lord Steyn.

I would venture to suggest that debate on the rival interpretations of section 133 is not advanced by use of the expressions ‘would not’ or ‘should not’ or ‘could not’ have been convicted, for all that they do is to beg the question what is the reason why conviction would not, or should not, or could not, have followed. Lord Bingham’s reference in the last sentence of paragraph 4 to the defendant who ‘should not’ have been convicted is in terms confined to the defendant in whose case something has gone seriously wrong with the investigation or conduct of the trial.

For my part, I was initially attracted by Lord Bingham’s proposition that in ordinary parlance the expression ‘miscarriage of justice’ is applied to cases beyond those where innocence is plainly established. It was certainly so used by Sir John May in his report on the Guildford bombings (1994) (HC 449, paragraphs 21.3-21.4), although it is perhaps significant that Sir John believed that the general public understanding of the expression meant the conviction of the innocent or the acquittal of the guilty. I would be inclined to accept, for example, that many would, in ordinary parlance, describe a conviction achieved by police corruption or by falsified police evidence, as a miscarriage of justice.

However, on reflection, and after hearing the arguments deployed before us, I have become convinced that ordinary parlance cannot yield the correct construction of section 133. The reasons which lead me to that conclusion are largely those very clearly set out in the opinion of Lord Steyn in Mullen and there is no occasion for me to attempt to re-state them. I mention as briefly as I can some additional factors which have thus persuaded me:

I do not myself think that ordinary parlance can be a reliable guide to the meaning of an expression found in an international treaty, at least when it is an expression which can mean a great variety of things to different people in different circumstances. For some, it would embrace every defendant whose conviction is quashed, on whatever grounds, who may be said to have suffered a miscarriage of justice in the sense of an unsafe conviction, but it is clear that section 133 cannot mean this. Lord Bingham’s reference in paragraph 4 of his speech in Mullen (cited above at paragraph 24) to judicial misdirection provides an example of exactly this difficulty. One would not ordinarily, as it seems to me, categorise as a miscarriage of justice every case in which a conviction is quashed by the Court of Appeal on grounds of some misdirection by the trial Judge. Nor, as it seems to me, would one speak of every such case as involving something going seriously wrong with the conduct of the trial. Unless, therefore, some method is to be found of distinguishing those misdirections which amount to something going seriously wrong with the trial from those which merely justify a conclusion that the conviction is unsafe, resort to ordinary parlance cannot provide the necessary test. Moreover, a misdirection case would not ordinarily be within section 133 in any event, because the appeal would not depend on any new or newly discovered fact.

It follows that unless the correct construction is Lord Steyn’s I cannot see where the meaning of ‘miscarriage of justice’ can logically, or even sensibly, stop short of every case of appeal allowed on grounds of new or newly discovered fact.

Whilst I agree of course that the CACD does not ordinarily address the question of guilt or innocence, but only the safety of the conviction, those cases where the innocence of the convicted defendant is genuinely demonstrated beyond reasonable doubt by new or newly discovered fact will be identifiable in that court and the judgment will, in virtually every case, make plain that this is so. Accordingly I respectfully part company from with the view expressed by Schiemann LJ, when Mullen was in the Court of Appeal, that the approach of the CACD makes the operation of section 133 difficult or unworkable unless a broader definition of miscarriage of justice is adopted. On the contrary, as it seems to me, the operation of the section poses very real difficulties if the broader definition is adopted, for then it becomes necessary to ask in every case of conviction quashed on grounds of fresh evidence whether it satisfies the section 133 criterion of miscarriage proved beyond reasonable doubt or is merely a case of doubt raised to the extent that the conviction is unsafe. If, however, miscarriage of justice means the establishment of innocence beyond reasonable doubt, there will usually be no difficulty in those cases being apparent from the judgments of the CACD.

Cases in which police or similar malpractice is established beyond reasonable doubt will ordinarily give rise to irresistible claims for compensation under the ordinary law of tort, even if innocence is not conclusively demonstrated.

For those reasons, I would if necessary also dismiss this appeal on the grounds that the correct construction of section 133 is that explained by Lord Steyn.

I agree that this appeal should be dismissed for the reasons given by Hughes LJ.

I also agree and for the same reasons.