5.2.1 High-Profile Wrongful Convictions That Led to the Creation of the Court of Criminal Appeal

A number of well-publicized cases of suspected wrongful convictions played a key role in the eventual creation of the Court of Appeal in 1907. Sir Fitzroy Kelly KC argued before a 1846 Royal Commission on Capital Punishment that at least forty-two people later proved to be innocent had been sentenced to death between 1802 and 1840 (Poyser, Nurse and Milne, Reference Poyser and Milne2018: 14). Israel Lipski, a Jewish Polish immigrant, was convicted of murder by a jury in the Old Bailey in 1887 after eight minutes of deliberation. Despite concerns in the Yiddish press and the trial judge’s own concerns about his summing up to the jury, the Home Secretary did not commute Lipski’s death sentence (Friedland, Reference Friedland1984). The same trial judge, Justice James Fitzjames Stephen, also contributed to the 1889 conviction of Florence Maybrick for murdering her husband through arsenic poisoning. This case will be examined in Chapter 8.4 because Maybrick was convicted largely on the basis that she was having an extramarital affair and in the face of evidence that her husband took arsenic as medicine. The Home Secretary commuted her death sentence on the basis that the American woman only attempted to murder her husband by poisoning him (Pattenden, Reference Pattenden1996: 14–15). As will be seen in relation to death row exonerations in both the United States and China, concerns about wrongful executions lent urgency to intertwined debates about miscarriages of justice and capital punishment.

The two wrongful convictions that immediately led to the creation of the Court of Appeal did not, however, result in death sentences. Adolph Beck was convicted of fraud, and George Edalji was convicted of cruelty to animals. Beck was a Norwegian who spoke with a strong accent. After one of his wrong convictions for defrauding women, he declared: “I am absolutely innocent of every charge that has been brought against me…. I ask the Press to help me to get all the evidence in support of me…” (Irving, Reference Irving1912: 28). He petitioned the Home Office unsuccessfully sixteen times for relief. He was wrongfully convicted multiple times because of repeated false identifications by victims of financial crimes before the real perpetrator was discovered. An inquiry recommended that those who handled requests for mercy at the Home Office have legal training but did not recommend the creation of a Court of Appeal. Beck eventually received both free pardons and compensation (Pattenden, Reference Pattenden1996: 28–29). Writing in 1912, HR Irving observed that there could be no question that Beck was innocent, and “it is melancholy to think that the long-drawn-out agony of this unfortunate Norwegian should have been necessary to bring about a reform which logic and common sense demanded, and only legal conservatism and the unreasoning indifference of the public had so long postponed” (Reference Irving1912: 3). Beck was awarded £5000 in compensation, the equivalent of £90,000 today for each of his five years in prison (Naughton, Reference Naughton2013: 210).

George Edalji was a Brown solicitor whose family immigrated from India. He was unpopular in the village he lived in and was blamed for a series of mutilations of animals. He was convicted and sentenced to seven years’ imprisonment in part on the basis of evidence from the same handwriting analyst who had given faulty evidence in Beck’s wrongful conviction (Risinger, Reference Risinger2006: 58). His case was championed, among others, by Sir Arthur Conan Doyle, who created the fictional detective, Sherlock Holmes. An inquiry into his case recognized that his conviction was unsatisfactory but concluded that neither a pardon nor compensation was appropriate because Edalji could not prove his innocence. The inquiry found evidence of tunnel vision or confirmation bias when it concluded “the police commenced and carried out their investigations, not for the purpose of finding out who was the guilty party, but for the purpose of finding evidence against Edalji who they were already sure was the guilty man” (Wilson, Reference Wilson1907: 3). Edalji’s statement that he had expected to be arrested was interpreted at his trial as a statement not that he had been framed but that he was guilty (Risinger, Reference Risinger2006: 63). When the animal mutilations continued after he was imprisoned, this was attributed to Edalji directing a gang from prison (Reference Risinger2006: 73–74, 79). The Home Secretary eventually pardoned Edalji but did not provide him with compensation because of doubts about proven innocence and concerns that Edalji had brought suspicion upon himself (Reference Risinger2006: 9). Arthur Conan Doyle believed that the failure to compensate Edalji was a result of racial prejudice and likened his case to the French miscarriage of justice of Alfred Dreyfus that will be examined in Chapter 7.2 (Basu, Reference Basu2021: 170–190).

Conan Doyle also championed the case of Oscar Slater, a Jewish German immigrant. He was convicted in 1909 of a brutal murder of an eighty-three-year-old woman by a 9-6 jury verdict in Scotland on the basis of weak identification evidence. Slater’s lawyer told the jury that Slater had “to fight a most unfair fight against public prejudice, roused with a fury I do not remember to have seen in any other kind of case” (Toughill, Reference Toughill1993: 225). His death sentence was commuted because of concerns about whether he was guilty including alternative suspects within the victim’s family. In 1926, a Court of Appeal was created for Scotland on much the same terms as the English Criminal Court of Appeal. In response to a 1927 book by a journalist that proclaimed Slater’s innocence, the legislation was amended to allow him to bring a retroactive appeal to the new Court of Appeal in Edinburgh. In a pattern that will also be seen in relation to the English Criminal Court of Appeal, the new Scottish court deferred to the jury and found that its guilty verdict was reasonable and supported by the evidence. It did, however, find the trial judge committed a legal error by suggesting that Slater was not entitled to the presumption of innocence on the basis of bad character evidence (Kilday, Reference Kilday, Nash and 619Kilday2020: 156–157; Slater v. H.M. Advocate 1928 JC 94). In Scotland, as in England, the history of appellate courts is tied to the history of wrongful convictions.

5.2.2 The Criminal Appeal Act, 1907

Rosemary Pattenden (Reference Pattenden1996: 31) has written that in the wake of the high-profile Beck and Edalji cases “[r]etrial by newspaper had become so prevalent that public confidence in the courts was being undermined.” The Lord Chancellor when introducing the bill creating the criminal Court of Appeal explained that it was better to have an appeal “to Her Majesty’s judges” than “an appeal to the Press” (Nobles and Schiff, Reference Nobles and Schiff2000: 40). The bill was eventually passed after an exhausting debate in the Commons and when a Conservative majority in the unelected House of Lords decided not to provoke a constitutional crisis by defeating it. Despite a political sponsor of the bill stressing that the bill would favour the judges over the press, Pattenden concluded that given judicial opposition, “the 1907 Act represented a massive defeat for the judges…” (Pattenden, Reference Pattenden1996: 32–33).

Sections 3 and 4 of the Criminal Appeal Act, 1907 allowed the Court of Criminal Appeal to quash convictions on the basis that they were unreasonable and not supported by the evidence or on the basis that there was a miscarriage of justice on any grounds. Appeals could also be overturned on the basis of error of law made by trial judges subject to the proviso that the conviction could be affirmed if there was no substantial miscarriage of justice. The remedy of a new trial was not provided until 1964 with subsequent reforms replacing the original grounds of appeal with references to the conviction being unsafe or unsatisfactory and since 1995 simply unsafe. The original grounds for appeal are still found in Canada, New Zealand and in a number of Australian states (Sangha, Roach and Moles, Reference Sangha, Roach and Moles2010).

5.2.3 The Performance of the Court of Appeal

Judges resisted the creation of a new appeal court largely on the basis of concerns about finality, efficiency and deference to the jury: themes that continue to influence the Court of Appeal in England (Nobles and Schiff, Reference Nobles and Schiff2000).

In a 1908 case, the Court of Appeal stressed that its task was not a retrial and that given that “there was evidence on both sides … it is impossible to say that the verdict is one which the jury could not have properly arrived at” (R. v. Williamson (1908) 1 Cr. App. R. 3). This decision was subject to criticism from commentators at the time (Pattenden, Reference Pattenden1996: 143). In 1909, the Court of Appeal affirmed that “the jury are the judges of fact. The Act was never meant to substitute another form of trial for trial by jury” (R. v. Simpson (1909) 2 Cr. App. R. 128, 103). The Court of Appeal stressed that it would not overturn convictions because they were against the weight of the evidence. In other words, appellate courts had to respect the way the jury weighed the evidence after having heard the witnesses and determining their credibility (Nobles and Schiff, Reference Nobles and Schiff2000: 56–57).

The Criminal Appeal Act, 1907, for the first time allowed appeals as of right on questions of law and with leave from the trial judge or the Court of Appeal of mixed questions of law and fact and sentence (Pattenden, Reference Pattenden1996: 129). The Court of Appeal was also given broad powers both to hear new evidence and to commission others to receive new evidence if the Court determined “it necessary or expedient in the interest of justice.” Indeed, these powers seem to invite investigations of complex, new and disputed factual matters by providing:

The Attorney General at the time that the bill was introduced, Sir John Walton, stressed that the new court had “ample power to get at the truth in the way they thought proper” and even imagined that the Court would be able “to elicit vital facts and points” and hear evidence. As Pattenden (Reference Pattenden1996: 131) notes, however, “the judges had other ideas.” The Court of Appeal did not use its statutory powers to appoint investigators and experts who could assist them in discovering the truth. It resisted inquisitorial innovations at a time when there still remained a faith that English justice was far superior to that provided in continental Europe. The result often placed victims of miscarriages of justice in an impossible catch-22 position. Evidence that could have been obtained at trial would not generally be considered without a good excuse because of the perceived dangers of the accused splitting a case. On the other hand, evidence that became available after trial, including witness recantations and confessions by others, would generally only be considered on a discretionary referral from the Home Secretary. The safety value of the elected executive ordering a new appeal derived from the royal prerogative of mercy still played an important role. As will be seen, continued dissatisfaction with the performance of the elected executive and retrial by media would lead to the creation of the CCRC as another remedial institution before the end of the twentieth century.

In 1908, the new Court warned that the power to admit new evidence on appeal “was not meant to be exercised for supplementing or supporting the case made at the trial where the witnesses would have been called” (R. v. Mortimer (1908) 1 Cr. App. R. 20, 22). It was concerned about the accused splitting their case by holding back evidence. It would only rarely admit new evidence, “otherwise in every case this court would be in effect asked to carry out a new trial” (R. v. Jones (1909) 1 Cr. App. R. 27, 32). The Court of Appeal generally deferred the consideration of retractions by witnesses and even confessions by third parties to the Home Secretary, despite controversy over the elected executive’s role being one of the factors that led to the creation of the Court of Appeal (Pattenden, Reference Pattenden1996: 132–133). In one early case, however, it overturned a guilty plea when faced with evidence that the accused was already in jail at a time he confessed to and pled guilty to the crime of sacrilege and was sentenced to twelve months of hard labour (R. v. Verney (1909) 2 Cr.App. R. 107). False guilty pleas are not a new form of injustice.

The Court of Appeal refused to hear new evidence in the 1952 murder appeals of Edward Devlin and Alfred Burns (Nobles and Schiff, Reference Nobles and Schiff2000: 63). The new evidence was strong and included alibi evidence, an admission of perjury by a witness and a confession by a third party. Before he was hanged, Devlin wrote: “Don’t worry Mam. We know we’re innocent. If we die, we die as martyrs not murderers” (Skelly, Reference Skelly2012: 45). The case was also linked with a detective who was involved in the wrongful convictions of George Kelly and Charles Connolly. Connolly pled guilty to robbery to avoid the death penalty. This wrongful conviction was eventually corrected on a reference from the CCRC. (Kelly v. R [2003] EWCA Crim 2957).

In the wake of several wrongful convictions in the 1950s, Lord Donovan warned in 1965 that “in the case of an innocent person who has been wrongly identified and wrongfully convicted,” the Court of Appeal would uphold a conviction simply if an identification was “credible” (Pattenden, Reference Pattenden1996: 144). The Court of Criminal Appeal deferred to the jury even in cases where it suggested that the jury had been exposed to inadmissible evidence (R. v. Thompson [1962] 1 All ER 65). Desires were expressed in Parliament that the Court of Appeal should be less deferential to jury verdicts and more willing to accept fresh evidence (Nobles and Schiff, Reference Nobles and Schiff2000: 70–71). The Court of Appeal responded in a 1968 case involving weak identification evidence and indicated that a “lurking doubt” might make a conviction unsafe (R. v. Cooper 1969 53 Cr. App. R 82). That said, the Court of Appeal has used this ground of appeal sparingly including in the case where the new ground was announced when it affirmed a murder conviction based on the uncorroborated evidence of an accomplice and upheld the conviction on two more referrals by the Home Secretary (Pattenden, Reference Pattenden1996: 386; Roberts, Reference Roberts2017). In 2012, it rejected the idea that convictions should be overturned “on the basis of some collective, subjective hunch that the conviction is or may be unsafe” (R. v. Pope [2012] EWCA Crim. 2241 [14]). The Home Secretary eventually remitted the sentence and only decided not to grant a pardon because “it would have looked like too much of a public snub” to the Court of Appeal. This underlines that the Court of Appeal did not render the elected executive superfluous in cases of suspected miscarriages of justice.

5.2.4 Changes to the Court of Appeal

The reluctance of the Court of Appeal to allow appeals from convictions was also influenced by its lack of power to order a new trial, which was introduced in 1964. In 1966, the grounds of appeal were changed to unsafe or unsatisfactory, wording that was proposed in the original 1907 legislation but dismissed as too subjective. Michael Naughton (Reference Naughton2013: 148) argued that these changes were “catastrophic” for the factually innocent. The original grounds based on unreasonable convictions not supported by the evidence or miscarriages of justice still operate in some Australian states, Canada and New Zealand. Under these grounds, courts have been reluctant to make declarations of factual innocence even when wrongfully convicted persons request them (R. v. Mullins-Johnson 2007 ONCA 720). The harmless error proviso was also changed in 1966, so that appeals from legal errors would only be dismissed if the Court of Appeal was convinced that there was no miscarriage of justice as opposed to “substantial miscarriage of justice” (Nobles and Schiff, Reference Nobles and Schiff2000: 66). The 1966 changes also tried to liberalize the law surrounding the admission of fresh evidence by providing that even if the evidence was available at the original trial, the Court of Appeal should consider whether there was a reasonable explanation why it had not been introduced (Roberts, Reference Roberts2017: 309).

The courts have developed a complex jurisprudence on the admission and effects of new evidence on appeal that leaves much discretion. In 1999, the Court of Appeal stated that the power to admit new evidence “exists to safeguard defendants against the risk and consequences of wrongful conviction” (R. v. CCRC ex parte Pearson 1999] EWHC Admin. 452). At the same time, there are more restrictive elements in the jurisprudence. In a death penalty case from Trinidad and Tobago, the Privy Council refused to admit new evidence in a 3:2 decision concluding: “It cannot be right to allow an appeal, without more, simply on the basis that the State’s main witness has later been shown to have told an outright lie” (Dial and Dotton v. The State [2005] UKPC 4). The Court of Appeal has warned against the new admission of expert evidence because of concerns that the trial will become only a “dry run” (R. v. Kai-Whitewind [2005] EWCA Crim 1092). In the first 300 cases heard by the Court of Appeal in 2016, the Court of Appeal only allowed the fresh evidence in 19 per cent of the forty-two requests and only allowed one appeal on that basis (Roberts, Reference Roberts2017: 319–320).

In 1995, the grounds of appeal were reduced to safety from the prior references to safety and satisfactory. The Court of Appeal was required to find that a conviction “is unsafe” at the urging of the senior judiciary who were concerned that the Runciman Commission’s recommendations that a conviction be quashed if it “is or may be unsafe” would change the Court’s practice and be too generous to the accused (Law Commission, 2025: 8.24–8.29). The power to admit new evidence was made somewhat less restrictive in response to high-profile wrongful convictions in terrorism cases. The law before the 1995 legislative changes required that new evidence should be “likely to be credible,” and now all that was required was that the new evidence be “capable of belief.” This followed recommendations from the Runciman Commission (Reference Runciman1993: chapter 10, para 3) that the Court of Appeal “should be more willing to consider arguments that indicate that a jury may have made a mistake” and that it “be more prepared, where appropriate, to admit evidence that might favour the defendant’s case even if it was, or could have been, available at the trial.” The legislative changes were minor (Belloni and Hodgson, Reference Belloni and Hodgson2000: 177) and did not change the actual practice of the Court of Appeal.

Belloni and Hodgson (Reference Belloni and Hodgson2000: 175) found that the Court of Appeal allowed only 13 per cent of appeals on average between 1988 and 1997, with most leaves to appeal being denied by a single judge. The international right to appeal will be examined in Chapter 11.4, but the English practice of not allowing appeals as of right has been held to be consistent with that right (Law Commission, 2025: 4.16). The Law Commission has recommended that time to seek leave to appeal should be extended from one to two months but has not recommended that appeals be allowed as of right, as they are in many civilian jurisdictions examined in Chapter 4 (2025: 6.33). It has also not recommended fundamental change to the test for admitting new evidence.

In Lucy Letby’s appeal, the Court of Appeal rejected her challenge to a medical doctor, Dr. Dewi Evans, called by the prosecutor, who volunteered to assist the police in the case. It stressed the doctor’s experience “with decades of clinical hands-on experience with neonates” and that the evaluation of his evidence was a jury matter based on circumstantial and cumulative evidence (R. v. Letby [2023] EWCA Crim 748 [115] [118]). It also rejected new evidence that Letby wanted to call by Dr. Shoo Lee, one of the authors of a research paper used at trial, on the basis that her lawyers could have called this foreign expert during her trial and that “the interests of justice require a defendant’s whole case to be put forward at trial unless there is good reason why that could not be done” (R. v. Letby [2023] EWCA Crim 748 [186]). This affirms the Court of Appeal’s rather restrictive test for admitting new evidence. It also begs the question of whether the Court of Appeal will consider new expert evidence offered by Dr. Shoo Lee and others should the CCRC refer Letby’s conviction back to the Court of Appeal.

The Court of Appeal also stressed that the jury that convicted Letby after a ten-month trial of seven counts of murder and seven counts of attempted murder of newborns in her care would have considered all the “circumstantial evidence … in its totality…” (R. v. Letby [2023] EWCA Crim 748 [183] [187]). This defers to the jury and the finality of its decisions, but discounts the danger that they might give excessive weight to notes that Letby wrote to herself in which she appeared to blame herself for some of the deaths. This aspect of the case will be discussed in Chapter 8.5.

Guilty pleas are encouraged and used to resolve the majority of criminal cases in England. Nobles and Schiff argue that the Court of Appeal must defer to the guilty plea process, given that it only hears about 200 appeals a year compared to over 60,000 guilty pleas (Nobles and Schiff, Reference Nobles and Schiff2020: 539). They recognize that the Court of Appeal has corrected false guilty pleas in cases with clear evidence that another person committed the crime or that a crime did not occur (Reference Nobles and Schiff2020: 527). In most other cases, the Court of Appeal maintains “there is nothing unsafe about a conviction based on the defendant’s own voluntary confession in open court … the trial process is not a tactical game. A defendant knows the true facts; he ought not to admit to facts that are not true” (R. v. Asiedu [2015] 2 Cr. App. R 8 at [19] [32]). Nobles and Schiff conclude that while the Court of Appeal theoretically can remedy false guilty pleas, they are practically unable to do so because of concerns about the efficiency and finality of convictions (Nobles and Schiff, Reference Nobles and Schiff2020: 518).

In 2022, the Court of Appeal overturned two guilty pleas to fifteen counts of manslaughter by arson, entered in 1981 by a young man with cognitive challenges, who was still detained at the time of the new appeal. It did so, however, by ruling that after an otherwise valid guilty plea, as opposed to a conviction by the jury, it was necessary for the accused to establish that they did not commit the offence. Given the accused’s physical disability, it would have been impossible for him to have set the fires in the two overturned counts (Tredget v. R [2022] EWCA Crim. 108 [171–173]). The CCRC had, however, referred all the guilty pleas on the basis of new evidence about the causes of the fire, Tredget’s below normal intelligence and his suggestibility as well as the way that police had obtained his confessions, which he had soon thereafter recanted. The Court of Appeal upheld thirteen of the manslaughter convictions stating: “the critical question in this context is whether it has been established that the appellant did not commit these offences. A possible, or preferred, accidental cause does not lead to that conclusion” (Reference Nobles and Schiff2022: [186]). This high standard effectively requires proven innocence. This is a harsh approach that discounts increased knowledge about the prevalence of both false confessions and false guilty pleas. It will be suggested later in this chapter that this approach, if applied by the courts, might have resulted in sustaining many of the guilty pleas entered in the Post Office cases.

5.4.1 The Origins of the CCRC

The origins of the CCRC lie in a series of wrongful convictions obtained in the 1970s in cases involving terrorism in England committed by the Irish Republican Army (IRA). Between 1989 and 1992, the Home Secretary – the member of the elected executive responsible for ordering new appeals based on new evidence after appeals had been exhausted – referred twenty-eight cases involving forty-nine people to the Court of Appeal, including those of the Guildford Four, Birmingham Six and Maguire Seven, all terrorism cases with Irish accused. These “Irish” cases will be examined in greater depth in Chapter 7.2. In many of these cases, the Home Secretary only ordered a new appeal after a number of prior unsuccessful applications and after extensive and critical media coverage.

The Home Secretary was advised by 12 case workers and 2.5 senior officials who were not lawyers. They took a reactive approach to requests for new appeals (Hoyle and Sato, Reference Hoyle and Sato2019: 4). In 1982, a Parliamentary Committee commented that “in practice” the Home Secretary’s decision “may depend upon the amount of pressure that is brought to bear on the Home Secretary by people of influence” such as were recruited to campaign for referrals in the above-mentioned terrorism cases. This led to concerns that the success of the petition to the Home Secretary depended more on the popularity of the cause than its merits (Mansfield and Taylor, Reference Mansfield, Taylor, Walker and Starmer1993: 166–167). Journalists such as Bob Woffinden and Ludovic Kennedy created a sense that the justice system was both on trial and in crisis, despite the absence of “any systemic analysis of public opinion” (Nobles and Schiff, Reference Nobles and Schiff2000: 122). Parallels can be drawn with the period before the 1907 creation of the Court of Appeal, where many argued that legal reform was a lesser evil to continued public criticism of the justice system (Reference Nobles and Schiff2000).

The advocacy group Justice concluded that the Home Secretary “as part of the Executive” was anxious “not to be seen to be interfering with the work of the courts” and so ignored errors, such as poor police investigations, inadequate defence lawyering, “underhand tactics by the prosecution” and “poor summing-up by the judge” (Justice, 1989: 3). It recommended the creation of a new independent review body with investigative powers including the power to hold hearings and appoint counsel for applicants. It recognized that such a review body “will not be infallible” but that it should be more independent, impartial and less reactive than the Home Secretary (1989: 72). The C3 Department did not conduct its own investigations, and if it asked the police to investigate, the police report remained confidential (Belloni and Hodgson, Reference Belloni and Hodgson2000: 182). It referred only five cases a year between 1972 and 1992, including the “Irish” wrongful conviction cases and those involving the West Midlands Crime Squad (Reference Belloni and Hodgson2000: 183).

It would be a mistake, however, to only blame the Home Secretary for delays in correcting the “Irish” wrongful convictions. The Court of Appeal dismissed the first appeal by the Guildford Four in 1977, even in the face of confessions by others that they had committed the bombings. Attempts by the Birmingham Six to prove through civil litigation that they had been beaten in custody were rejected as an abuse of process. The case is best known for Lord Denning’s infamous remarks that if the six men were successful, it would produce “such an appalling vista that every sensible person in the land would say: It cannot be right that these actions should go any further” (McIkenny v. Chief Constable [1980] 1 Q.B. 283 at 323 aff’d [1982] A.C. 529 (H.L.)).

In 1993, the Royal Commission on Criminal Justice, chaired by Viscount Runciman and appointed on the day that the Birmingham Six were finally released from prison, recommended the creation of a new authority to replace the Home Secretary’s role in ordering second or subsequent appeals. It justified the creation of a new body both on separation of powers and functional grounds related to the Home Secretary’s “reactive” approach to applications and the need for a proactive body that will have “access to specialist advisors, such as scientific advisors, as necessary.” Quoting Justice, the Commission suggested that a new body would operate on “inquisitorial lines” and attempt to get at “the truth” of the matter (Runciman, Reference Runciman1993: 10.6). It rejected the idea of a body that would itself correct miscarriages of justice. At the same time, it recognized that the Court of Appeal had been reluctant to correct wrongful convictions and should be more willing to overturn convictions and admit new evidence (Reference Runciman1993: 12.25). It also recommended reforms of forensic sciences, which were implicated in many of the terrorism wrongful convictions. Although there was widespread praise for these recommendations, many academics expressed unease with the Runciman Commission’s other recommendations that stressed the need to convict the factually guilty as much as to acquit the factually innocent.

5.4.2 The Criminal Appeals Act, 1995

The Criminal Appeal Act, 1995, provided for the creation of the CCRC with no fewer than eleven members appointed on the recommendation of the Prime Minister, a third of whom would be lawyers with ten years’ experience and two-thirds with knowledge or experience with criminal justice. This established possible tensions both within the commission between lawyers, other experts and lay people. The late Professor Peter Duff, who served as a commissioner on the similarly constituted Scottish commission from 1999 to 2007, has suggested that the “lay” as opposed to the “legal” members of the Commission were more likely to be concerned about both factual innocence and factual guilt than the legal members. At the same time, he stated that he did not have a case where he was absolutely certain of the applicant’s factual innocence (Duff, Reference Duff2009: 721).

The CCRC has been called a “radical reform” designed “to provide a fresh approach to the investigation of miscarriages of justice” (Hodgson, Reference Hodgson2020: 280, 285). It was given statutory powers to obtain new expert reports and appoint investigating officers from the police to obtain new evidence. The CCRC originally only had statutory powers to obtain relevant documents from public bodies though it could exercise these powers to obtain information even if police and prosecutors claimed legal privileges designed to protect the identity of police informants or communications with legal advisors. In 2016, the CCRC’s powers were expanded to allow it to obtain relevant documents from private bodies such as private providers of forensic sciences (Criminal Cases Review Commission (Information) Act, 2016 c.17). At the same time, the CCRC was not designed to be a transparent organization. Sections 23 and 24 severely limited its disclosure of information, but there was a requirement for annual reports.

The Criminal Appeals Act, 1995 went beyond the Runciman Commission and included applications in relation to sentences as well as convictions. Justice (Reference Runciman1993: 28) disagreed with the inclusion of sentencing “on grounds of principle and practice,” arguing that sentencing should be left to the courts. It also expressed concerns that the caseload of the CCRC could be increased by applications about sentencing. Justice also opposed the idea of relying on police to conduct reinvestigations and argued “that without an in-house team of investigators, paid out of its own budget,” the independence of the CCRC “will always be open to question” (Reference Runciman1993: 35). It also opposed the referral criteria for the CCRC, which was a predictive test of whether there was a “reasonable possibility” that the Court of Appeal would overturn a conviction or sentence and proposed that the test should be “whether there is an arguable case that there has been a wrongful conviction” (Reference Runciman1993: 36).

Although the creation of the CCRC was much more warmly received than many of the other Runciman Commission recommendations, some commentators were sceptical. One concluded “apart from being independent of the executive, the proposed body looks remarkably like C3,” the former office that advised the Home Secretary (Thornton, Reference Thornton1993: 929). Rosemary Pattenden (Reference Pattenden1996: 410) was less blunt when she concluded that the Home Secretary process suffered from “four major defects.” Only two of these, the lack of independence and the personnel who assisted the Home Secretary not having suitable qualifications, were addressed by the CCRC. She warned, however, that the other two defects, reliance on police investigations and not being “open to public scrutiny,” had not been addressed by the CCRC. The referral grounds based on predictions whether there was a reasonable possibility that the Court of Appeal would admit new evidence and overturn a conviction or sentence soon became a lightning rod for criticism of the CCRC (Belloni and Hodgson, Reference Belloni and Hodgson2000: 186–186; Naughton, Reference Naughton2010). Nevertheless, the CCRC has survived as a new criminal justice institution and one that has migrated to other lands.

5.5.1 The CCRC and the Rationing of Scarce Investigative Resources

The CCRC started its operation in April 1997. It inherited 228 cases from the Home Secretary (Belloni and Hodgson, Reference Belloni and Hodgson2000: 190). It received about 1,000 applications a year in its first years (Hoyle and Sato, Reference Hoyle and Sato2019: 4). The number of applications increased by 70 per cent between 2009/2010 and 2012/2013 after it introduced an easy-read application and conducted more outreach to the prisons. There is a trade-off between the ability of the CCRC to be more accessible and the reality that only a small number of applications made to it are successful. Nobles and Schiff (Reference Nobles and Schiff2001) suggest that the CCRC initially took an extremely thorough approach to the applications it inherited from C3, but as its backlog increased, it adopted more expeditious procedures. The CCRCs are no more immune from efficiency demands than the courts.

About 40 per cent of applications are made by applicants who have made no appeals. Most of these are rejected, and few of the rejected applications go on to seek leave to appeal. Over 40 per cent of the no appeal applicants pled guilty compared to only 8 per cent of cases with appeals. No appeal applicants were more likely to be women and less likely to have lawyers (Hodgson, Horne and Soubise, Reference Hodgson, Horne and Soubise2018: 17).

Decisions to refer are made by three commissioners, but decisions to reject an application are made by only one commissioner. In their study of the CCRC, Professors Carolyn Hoyle and Mai Sato found considerable variability in the decisions made by single commissioners (Reference Hoyle and Sato2019: 93–96). They also found that some decisions seem to be influenced by strategic considerations about whether the decision would be judged so unreasonable as to be reversed on judicial review (Reference Hoyle and Sato2019: 59, 89). This is disturbing because they found that the CCRC has only lost one judicial review under standards of administrative law that defer to expert administrative agencies (Reference Hoyle and Sato2019: 62). They quoted one commissioner who examined no appeal cases which he or she characterized as “bargain basement work … you get into a certain way of thinking about cases, and you get into being cynical…”(Reference Hoyle and Sato2019: 89).

Laurie Elks (Reference Elks2008: 307), a former Commissioner, has warned, “if the Commission were to set out to review all the prosecution material in every case, its output would grind to a halt. There is no certainty, therefore, that the Commission’s investigations will pick up non-disclosure where it has taken place.” Professors Hoyle and Sato conclude that the CCRC’s powers to obtain more information are exercised in a “reasonable” and “proportionate” manner that is attentive to competing privacy concerns but also the possibility of complaint, supervision and judicial review (Hoyle and Sato, Reference Hoyle and Sato2019: 56). At the same time, they warn that something could be missed by avoiding “empirical” as opposed to “desktop” examinations. They recount two cases where visits by case review managers revealed critical information and a third where an interview with the applicant revealed new information (Reference Hoyle and Sato2019: 271).

Much of the CCRC’s staff are legally trained and did not necessarily have scientific and other skills to deal with convictions based on forensic science and CCTV identification (Reference Hoyle and Sato2019: 84). After accumulating a backlog of cases, the CCRC developed key performance indicators in 2004 that stressed efficiency (Reference Hoyle and Sato2019: 259–261). The CCRC engages in the same rationing of justice and limited resources as the Court of Appeal (Nobles and Schiff, Reference Nobles and Schiff2000) albeit in a less transparent manner because its enabling statute, unlike some more recently created CCRCs in New Zealand and Canada, prohibits it from releasing its statements of reasons in cases. Hoyle and Sato (Reference Hoyle and Sato2019: 263) found that the CCRC devoted considerable resources to preparing detailed statements of reasons, which in their sample were on average thirty-three pages long.

Not enough research has been done on the reactions of the more than 97 per cent of applicants who have their applications rejected by the commission. Its 2023/2024 annual report found that almost 65 per cent reported the review process to be “good” or “ok,” while about 35 per cent classified it as “bad” (Criminal Cases Review Commission (CCRC), 2024: 39). Hodgson (Reference Hodgson2020: chapter 10) has raised concerns that the CCRC is not open enough to participation by applicants or their lawyers and draws parallels with French inquisitorial trial procedures where the defence is also marginalized. Legal aid cuts mean that only 7 per cent of applicants have lawyers (CCRC, 2024: 39). Concerns about efficiency and not appearing as an advocate for applicants, perhaps more than any commitment to inquisitorial procedures, may explain the CCRC’s distant relationship with applicants. Legal process analysis should be applied to CCRCs. It reveals that these new institutions are small administrative agencies subject to the vagaries of the government’s budgeting and appointment decisions (LaForme and Westmoreland-Traore, Reference LaForme and Westmoreland-Traore2021; Leonetti, Reference Leonetti2023a).

5.5.2 The CCRC’s Mistakes: Victor Nealon, Andrew Malkinson, Oliver Campbell and Peter Sullivan

The limits of the CCRC, as well as its ability to correct its mistakes on a subsequent application, have been dramatically revealed in four recent cases where it rejected initial applications that eventually resulted in successful referrals. A more inquisitive and better funded CCRC might have made the successful referrals earlier, thus saving the four men many years of imprisonment and also sparing the CCRC much negative media attention. That said, the CCRC ultimately made referrals in all four cases that led to the wrongful convictions being quashed.

Elks reported only two DNA exonerations in the CCRC’s first ten years of operation (Elks, Reference Elks2008: 83). One explanation may be that the English police used DNA to exclude the innocent. Alas, the Victor Nealon and Andrew Malkinson cases raise an alternative and more disturbing explanation for the comparative absence of DNA exonerations: the CCRC may not always have ordered DNA testing even in the minority of cases where DNA evidence was present. In any event, the CCRC missed at least two DNA exonerations. In both cases, the CCRC and the courts accepted what turned out to be mistaken eyewitness identifications. A third DNA exoneration, that of Peter Sullivan, is less clear-cut because it may have depended on advances in DNA testing since the CCRC denied his first application in 2008. The Peter Sullivan and Oliver Campbell cases demonstrate how both the CCRC and the Court of Appeal failed to recognize many of the hallmarks of false confessions given by those with cognitive challenges and in Sullivan’s case unreliable bitemark analysis. The fact that three of the four cases involved sexual assaults also reveals dangers of wrongful convictions in such cases, a topic that will be revisited in Chapter 8.8.

5.5.3 The Difficulty of Overturning the CCRC’s Decisions on Judicial Review

It is inevitable that a CCRC will make mistakes. Alas, a legal process approach reveals that it will be very difficult for a decision not to refer a case back to the courts to be overturned on a successful judicial review. Hoyle and Sato (Reference Hoyle and Sato2019: 89) report that only one judicial review of the CCRC’s decisions has been successful, and they quote one person within the CCRC making strategic use of deferential standards of judicial review under Wednesbury (1948) 1 K.B. 223. They conclude that this statement was concerning because it “suggests that quality assurance of commissioner’s screening practices is primarily intended to defend the Commission against judicial review” (Reference Hoyle and Sato2019: 90). I agree with their concerns but I am less certain that their ideal of “deliberative accountability” (Reference Hoyle and Sato2019) is realistic in a small agency with an inadequate budget, performance indicators based on efficiency and a backlog of applications to process. As one of their other interviewees observed “I think it’s public bodies, by and large, are not set up to be bold and imaginative. They’re set up to follow the rules” (Reference Hoyle and Sato2019: 67). In upholding CCRC decisions on judicial review, courts have only gently chided the Commission for not sharing more information with disappointed applicants (Davis v. Criminal Cases Review Commission & Anor [2009] EWHC 2926).

The CCRCs are important new criminal justice institutions but also small administrative agencies that act as gatekeepers to judicial reconsiderations of wrongful convictions. Even in cases where the CCRC concludes that there is a reasonable possibility that a conviction or sentence will be overturned by the Court of Appeal, it has a discretion not to make a reference back to the courts. The lack of public disclosure of the decisions of the English CCRC is also concerning. Many administrative agencies that make decisions with less momentous consequences than the CCRC do disclose their reasons for decisions.

5.5.4 The CCRC’s Relationship with the Government

One of the biggest selling points of CCRCs is that they are “independent” from government. To be sure, commissioners are less dependent on government than a member of the elected executive such as the Home Secretary, who makes decisions whether to order second or subsequent appeals. That said, CCRCs are dependent on governments for their budgets, the appointment of commissioners and their reappointment. As early as 2008, there were signs of tension between the CCRC and its sponsoring department with former commissioner Laurie Elks (Reference Elks2008: 335), suggesting that the two entities had a “difficult relationship.”

One revealing but ultimately unsuccessful judicial review application claimed that the CCRC was not independent from government. The challenge came in the wake of a review by the Department of Justice designed to increase the CCRC’s efficiency and limit its costs and the decision not to reappoint a Commissioner who was known to oppose the changes the CCRC had introduced in light of the government’s review. The reviewing courts noted that the then chair of the CCRC, Helen Pitcher, agreed with the changes, which included having commissioners work on a per diem as opposed to having a salary and pension. The new daily rate for commissioners would be £358, less than the £502 made by part-time judges. The per diem commissioners would also serve shorter three-year periods before coming up for reappointment (Queen on application of Warner v. Secretary of State [2020] EWHC 1894 (Admin) [54]). The issue of principle here is that the CCRC has powers equivalent to those of the judiciary and should not be treated as junior partners or day help (LaForme and Westmoreland-Traore, Reference LaForme and Westmoreland-Traore2021).

The administrative court recognized that, both “operationally” and “constitutionally,” the CCRC must be, and be seen to be, independent from government “if the public is to have confidence in its decisions” (Queen on application of Warner v. Secretary of State [2020] EWHC 1894 (Admin) [18]). It is clear, however, that public confidence in the CCRC has plummeted since this 2020 decision. Drawing on information obtained from a critical television program, the Court noted that government briefing notes that indicated an intention not to renew a Commissioner were “troubling” and amounted to “political interference” (2021: para 64). Nevertheless, the Court rejected the application for judicial review on the basis that the disappointed applicant had not established a reasonable apprehension that the CCRC was biased. This court’s decision reveals that it is much easier to say that a CCRC is independent from government than to ensure such independence. It also reaffirms that the courts are not likely to save the CCRC from a government intent on reducing its budget or appointing its own people to serve as commissioners. It will be seen later that the Canadian government has unfortunately rejected reform proposals designed to make a Canadian CCRC less dependent on government funding and appointment decisions and to tie the payment of commissioners to that received by independent judges. (LaForme and Westmoreland-Traore, Reference LaForme and Westmoreland-Traore2021).

5.5.5 The CCRC’s Referrals of Cases to and Its Dialogue with the Courts

A CCRC has to be concerned not only with its relations with government on funding and appointment issues but also with its dependence on the courts to reverse convictions.

As early as 1999, the Court of Appeal had affirmed that the reasonable possibility of reversal test for a referral required the CCRC “to try to predict the response of the Court of Appeal if the case were referred and application to adduce the evidence were made. It could only make that prediction by paying attention to what the Court of Appeal had said and done in similar cases on earlier occasions” (R. v. CCRC ex parte Pearson [1999] EWCA (Admin) 452: [56]). This approach places a premium on legal as opposed to investigative or forensic expertise.

Much ink has been spilt over whether the CCRC is a captive of the willingness of the Court of Appeal to admit new evidence and overtime convictions. Hoyle and Sato (Reference Hoyle and Sato2019: 106–107) observed that commission staff were quite deferential to the Court of Appeal’s jurisprudence, including its unwillingness to overturn guilty pleas. They also found that the CCRC was sensitive to the Court of Appeal’s reluctance to admit new evidence from “a bigger and better expert” on appeal (Reference Hoyle and Sato2019). Hodgson (Reference Hodgson2020: 306) revealed debate within the CCRC about its referral test with one case manager stating at a meeting that “it would be nice to send it back, but I find it difficult to think that the Court of Appeal will find it enough” only to be reminded by a colleague that test for referral was only a “real possibility” that the conviction would be overturned. She also quoted one case manager that challenging the Court of Appeal is a “truly hopeless and pointless task” and another that bemoaned that “the commission is scared of the Court of Appeal” and should “not be so hung up on what the Court of Appeal will think” (Reference Hodgson2020: 307, 309).

The CCRC was frequently criticized for being too dependent on the Court of Appeal (Naughton, Reference Naughton2010, Reference Naughton2013; Nobles and Schiff, Reference Nobles and Schiff2001), but as Michael Zander (Reference Zander and Naughton2010) has argued, many of these criticisms should probably have been directed at the Court of Appeal. The Westminster Commission (2021) has recommended that the referral test be made less predictive of what the Court of Appeal will do if the case is referred to them and that they have the discretion to refer cases when they determine it is in the interests of justice to do so. The Law Commission, 2025 has provisionally recommended that the CCRC’s referral test be changed to whether a conviction “may be unsafe” to encourage the CCRC to be less concerned about the Court of Appeal’s reactions to its referrals. It noted that with the exception of the Crown Prosecution Service, all consultees had concluded that the present predictive test was hindering the correction of miscarriages of justice (2025: 11.87, 11.127, 11.166).

Laurie Elks, a former commissioner, has recognized the danger that the CCRC will “use its gatekeeper role to bar the progress of applications based upon evidence which it predicts the Court of Appeal will view negatively” (Elks, Reference Elks2008: 70). At the same time, he rejected criticism that the CCRC had been too timid. In support, he cited many cases in which the CCRC made referrals on the basis of new evidence challenging shaken baby syndrome (SBS), facial mapping, auditory recognition, machine controls, explosives, firearms, firearm residue, bloodspatter, forensic linguistics, attention deficit disorder, accident reconstruction and cell phone triangulation. From a legal process perspective, an expert commission with powers to commission new evidence should be ahead of the courts when it comes to detecting frailties and advances in forensic science and machine-based evidence.

There is some evidence of the CCRC referring less cases in response to signals from the Court of Appeal. When the vast majority of joint enterprise convictions the CCRC referred were upheld, the CCRC made far fewer references because of concerns about its reputation with the Court of Appeal (Hewitt, Reference Hewitt2023; Law Commission, 2025: 11.65). It also stopped referring SBS convictions after the Court of Appeal in 2012 upheld a conviction simply based on the controversial triad of symptoms (2025: 17.32).

There is both a formal and informal dialogue between the Court of Appeal and the CCRC. Lord Justice Judge wrote a letter to Graham Zellick, then chair of the CCRC, raising concerns that the CCRC was not applying the correct test for the admission of new evidence (Elks, Reference Elks2008: 68–69). The Law Commission has suggested that such a communication infringed on the CCRC’s independence. For example, it is clearly improper for a Minister as part of the executive to contact a judge about how they interpreted a statute (Law Commission, 2025: 11.311). This communication lends some support to those who are concerned that the CCRC is too dependent on the Court of Appeal’s approach both to the admission of new evidence and overturning convictions. The independence of the CCRC is also related to its status. The reduction in pay and the part-time status of the commissioners suggest that the CCRC will remain the junior partner in its dialogue with the Court of Appeal.

A Canadian report by two retired judges sought to respond to these concerns by recommending that the salaries of commissioners be tied in some way to those of judges and that the Court of Appeal hear any evidence that was a basis for a referral from a future Canadian CCRC while retaining its right to decide what, if any weight, to the new evidence (LaForme and Westmoreland-Traore, Reference LaForme and Westmoreland-Traore2021). This followed the frequent practice of appellate courts in hearing new evidence before deciding whether to admit it. The Canadian legislature, however, rejected these recommendations in its 2024 legislation creating Canada’s new CCRC (S.C. 2024 c.33).

There should be creative tension between a CCRC and the Court of Appeal, given their different mandates, concerns and composition. The Court of Appeal will be concerned about maintaining a manageable workload and deference to juries and other triers of fact. The CCRC should be less concerned with such matters and more concerned about cases that may result in miscarriages of justice. It should also take an interest in the frailties of evidence that have contributed to wrongful convictions. The CCRC is supposed to be a lay body that only requires a third of commissioners to be lawyers. Laurie Elks (Reference Elks2008: 109) has commented that the CCRC as “an essentially lay body … is liable to be distrusted or even resented by the Court of Appeal when it has the temerity to suggest that learned counsel has slipped up on the job.” He concedes that it is “arguable” that the CCRC has “been to some degree intimidated by the hostile approach” (Reference Elks2008: 109) of the Court of Appeal on claims of ineffective assistance of counsel (R. v. Day 2003 EWCA 1060) even though many applications to the CCRC allege legal incompetence. This reveals that the CCRC is in a difficult position between disappointed applicants and an often cautious and sometimes defensive Court of Appeal.

The CCRC as a new criminal justice institution was vulnerable to criticism from the Court of Appeal, but the English CCRC was able to withstand such criticism. As discussed earlier, the Court of Appeal strongly criticized the CCRC in 2003 for referring Ruth Ellis’ 1955 murder conviction and even invited Parliament to legislate an end to review of historical cases (R. v. Ellis 2003 EWCA Crim 3556 [90]). Parliament did not act. As will be seen, the Scottish CCRC has been subject to more heavy-handed legislative replies reminding it of the importance of finality of convictions (Roach, Reference Roach2019a: 972). The Law Commission (2025: 11.330) has observed that the Court of Appeal’s criticisms of the CCRC have undermined its independence.

Louise Hewitt (Reference Hewitt2023: 236) has argued that the CCRC’s decision to follow the Court of Appeal and refer less joint enterprise cases may have the effect of denying access to justice to Black people who are believed to be over-represented in such cases. As will be seen later, both New Zealand and Canadian discussions of the CCRC have focused on the demographic make-up of applicants compared to populations most at risk for miscarriages of justice. That said, the CCRC, to its credit, is now collecting and publishing data on the demographics of its applicants, with its 2023 annual report indicating that almost 25 per cent of applications are made by members of an “ethnic minority group,” 20 per cent from persons with a disability and 8 per cent from females (CCRC, 2024: 7).

The Court of Appeal has the statutory power to ask the CCRC to conduct investigations. As discussed earlier, the Court of Appeal has had inquisitorial investigative powers since its creation in 1907 but has rarely exercised them. Hoyle and Sato describe how the CCRC prioritizes requests from the Court of Appeal with most of the requests relating to alleged jury improprieties. Investigations can, however, include witness retractions and even alternative suspects (Hoyle and Sato, Reference Hoyle and Sato2019: 249–250). Such investigations suggest that both the CCRC and the Court of Appeal can affect the workload of the other. They also suggest that the CCRC can add inquisitorial and investigative elements to the adversarial appeal process should the Court of Appeal so desire.

The CCRC has not initiated a dialogue with the elected executive by referring cases for an executive pardon. Such an approach could, for better or worse, re-politicize some discourse about wrongful conviction, which has become more legalistic under the CCRC’s predictive test for referral and the appointment of a majority of commissioners who are lawyers. Both Hoyle and Sato (Reference Hoyle and Sato2019: 313) and Hodgson (Reference Hodgson2020: 297) have urged the CCRC to exercise its powers to make recommendations for the royal prerogative mercy in cases where the courts have been resistant to correcting miscarriages of justice. Naughton (Reference Naughton2010) also argued that this approach should be used to vindicate factual innocence. One problem is that the language of pardon is at odds with that of exoneration and innocence. At the same time, a legal process approach should remain open to the possibility that executive clemency can, in some circumstances, be more generous than judicial relief.

5.5.6 The 2015 Parliamentary Committee and Westminster Commission Reports on the CCRC

A 2015 Parliamentary Committee report found that between 2009/2010 and 2014/2015, funding to the CCRC fell from £6.5 million to £5.3 million, which amounts to a 30 per cent cut when adjusted for inflation (House of Commons Justice Committee, 2015: para 31) “The Chair of the Commission has confirmed that for every £10 that his predecessor spent on a case ten years ago, he now has just £4, which he has described as ‘the biggest cut that has taken place anywhere in the criminal justice system’” (House of Commons Justice Committee, 2015: para 31; Hoyle and Sato, Reference Hoyle and Sato2019: 18). The CCRC is vulnerable to underfunding because it has no powerful constituency supporting it.

The Parliamentary Committee recommended that the CCRC should receive an additional £1 million of annual funding “as a matter of urgency.” It also recommended that the CCRC should be “less cautious” in making referrals and engage more with applicants and that the Law Commission should review the grounds for allowing appeals from conviction in the Court of Appeal. This latter recommendation recognized the symbiotic relationship between the two institutions. The Parliamentary Committee affirmed the importance of the CCRC by stating that “the level of successful referrals from the CCRC shows that it remains as necessary a body now as when it was set up … even its strongest critics have said that they simply want it to improve. The existence of the CCRC is not enough in and of itself; it must be given the resources and powers it requires to perform its job effectively” (House of Commons Justice Committee, 2015: para 55).

The 2021 All-Party Parliamentary Group report, also known as the Westminster Commission on Miscarriages of Justice, concluded that the underfunding of the CCRC has continued. It found “adjusting for inflation, the CCRC’s funding level in 2014/15 (£5.25m) was 43% lower than it had been in 2003/04 (£9.264m). Even though funding has slightly increased more recently, to £5.936m in 2019/20, it is still significantly lower than in previous years” (Westminster Commission, 2021: 27). The result of this underfunding was that the average caseload of a case review manager was 27 in 2017, whereas it had only been 12.5 cases in 2010/2011 (2021: 28). It warned that “without increased resources the CCRC cannot examine all relevant documents, carry out enough face-to-face enquiries and take advice from external forensic experts” (2021: 6).

Partly, as a result of these two reviews and well-publicized initial failures by the CCRC in the cases of Victor Nealon and Andrew Malkinson, the CCRC’s budget has increased from £7.8 million in 2022–2023 to £9 million in 2023–2024 and a projected £9.3 million in 2024–2025 (CCRC, 2024: 46). At the same time, concerns remain that the budget is still declining in real terms and that the part-time and remote work of the commissioners amounted to only 2.26 full-time commissioners in 2023–2024 compared to 8.8 a decade earlier (Justice Committee, 2025: paras 70, 74).

Like the 2015 Parliamentary Committee, the Westminster Commission expressed concerns that the CCRC was too deferential to the Court of Appeal and recommended that its real possibility of a conviction being quashed test be replaced by a new test based on conclusions that the conviction may be unsafe or that it is in the interests of justice to make a referral. It hoped that such a change, “would encourage a different and more independent mindset” (Westminster Commission, 2021: 4–7). It also heard evidence that it was getting more difficult to have the Court of Appeal reverse a conviction and recommended that the Law Commission examine the adequacy of the Court of Appeal’s grounds for allowing appeals (2021: 40–43), a project that is now underway (Law Commission, 2025).

The Commission encouraged the CCRC to be more open and transparent with applicants. Jacqueline Hodgson has concluded that the CCRC’s “understanding of itself as an independent and inquisitorial body” has led it to be too distant from applicants who know the case (Hodgson, Reference Hodgson2020: 324). Cuts to legal aid mean that only about 7 per cent of all applicants have lawyers compared to previous figures as high as 30 per cent of all applicants. The ability of applicants, some of whom may have language and cognitive difficulties, to effectively respond to the CCRC’s provisional and often detailed reasons is very much in doubt. Hodgson (Reference Hodgson2020: 322) quotes one CCRC case manager who recognized that, given that “[o]ver 50% of those in prison are seriously challenged in literacy terms,” it is “ridiculous to expect an applicant to able to respond to what we write to them.” She also pointed out that under the 2003 legislative changes, applicants are now generally restricted to the grounds of appeal included in the CCRC’s reasons for making a referral. This makes it even more important to make “room for an adversarial lawyer” in the CCRC’s “inquisitorial review” (Reference Hodgson2020: 324). As suggested in Chapter 4, both adversarial and inquisitorial procedures have weaknesses, and an ideal system would combine elements of both.

5.5.7 The CCRC and Systemic Issues

Although the Runciman Commission anticipated that the CCRC would be well positioned to play a role in systemic reform to reduce miscarriages of justice, it has rarely engaged in such work. One reason could be a lack of resources. Another is that CCRCs, in order to maintain the credibility with courts and governments, make efforts not to be seen as an advocate for the wrongfully convicted (Duff, Reference Duff2009; Roach, Reference Roach2010b).

There are some exceptions to the general pattern of non-engagement with systemic issues. In response to a pattern of cases that was discovered for people who were convicted of immigration offences, the CCRC took active steps to educate the immigration bar about defences of being refugee (Hoyle and Sato, Reference Hoyle and Sato2019: 217–225). Even if it does not see itself as a law reformer, a CCRC has an institutional interest in making recommendations to prevent repetitive applications. At the same time, the CCRC received some pushback from the Court of Appeal for automatically referring these cases when no appeal had been taken (YY v. The Queen [2016] EWCA Crim 18 [43]) and stopped automatically referring them after that criticism (Hodgson, Reference Hodgson2020: 287). The CCRC also took some steps with the Post Office cases that will be examined later in this chapter. The Law Commission (2025: 17.72) has provisionally recommended that the CCRC should play a greater role in the investigation of systemic miscarriages of justice. It also recommended that, given that the CCRC’s own conduct may be implicated in some of these systemic miscarriages of justice, there may be a need for judicial public inquiries (2025: 17.107).

5.6.1 Scotland

The Scottish Criminal Cases Review Commission (SCCRC) was created in 1999 with a similar structure to the CCRC. A third of its members are required to be experienced lawyers, and two-thirds must have experience with criminal justice. It refers cases on the basis that a miscarriage of justice may have occurred, and it is in the interests of justice to do so. This raises some concern that the Commission may deny referrals to appellate courts even when there are concerns about a miscarriage of justice. The Scottish commission did this in one case where it concluded that, while there may be a miscarriage of justice because of the legal insufficiency of the charge, it was not in the interests of justice to refer the case because the accused was factually guilty (Cochrane v. Her Majesty’s Advocate 2006 HCJAC 27 [15]). (See also Gordon v. SCCCR 2017 UKSC 20). This decision not to refer was upheld on judicial review following patterns observed earlier of CCRC’s benefitting from deferential standards of judicial review. Both Duff (Reference Duff2009: 706–709) and Leverick, Campbell and Callander (Reference Leverick, Campbell and Callander2017: 71–72) defend this decision on the basis that a post-conviction review system should be concerned that public confidence could be eroded if a referral was made on a “technicality.” This ignores that the English CCRC has referred cases on the basis of legal errors and the vagueness and subjectivity of the concept of technicality. At the same time, the English CCRC has been heavily criticized for such “technical” referrals by those such as Naughton (Reference Naughton2013), who subscribe to a proven factual innocence model. Concerns about factual innocence or guilt remain a popular way to ration justice and one embraced with some organisations that engage with CCRC’s.

A Canadian report examined the Scottish referral standard and recommended that the government adopt its referral standard that a miscarriage of justice may have occurred but not its discretionary requirement that the commission also conclude that the referral is in the interests of justice. The report written by Harry LaForme, Canada’s first Indigenous appellate judge, and Juanita Westmoreland-Traoré, Canada’s first Black law dean, argued that those failed by the justice system were not likely to be convinced by a conclusion from a commission that it was not in the interests of justice to refer a case. They also expressed concerns based on the Scottish experience that the commission could deny referrals based on the past history, bad character or perceived factual guilt of applicants (LaForme and Westmoreland-Traore, Reference LaForme and Westmoreland-Traore2021). As will be seen, the Canadian government rejected this advice, and a new Canadian commission has identical grounds for referral as the Scottish commission. In addition, a New Zealand commission created in 2019 uses the “interests of justice” as the exclusive grounds for referral. “Interests of justice” maximizes the flexibility of the commission and avoids some of the predictive nature of the English referral test. At the same time, it remains vague and opaque and maximizes the discretion of the commissioners.

Like the CCRC, the SCCRC does not see itself as an advocate for applicants and plays no role after it refers a case to the Court of Appeal (Duff, Reference Duff2009). It has a lower “success rate” in terms of appeal court quashing convictions. As of March 31, 2024, it had referred ninety-five convictions to the Court of Appeal with forty-eight of the convictions being overturned and forty-three of the convictions being upheld (SCCRC, Case Statistics).

The SCCRC’s most controversial referral was that of Al Megrahi convicted in the Lockerbie terrorism case, an appeal that Al Megrahi eventually abandoned in return for a deal that allowed him to return to Libya. A year after this referral, legislation was enacted reminding the SCCRC to have “regard to the need for finality and certainty in the determination of criminal proceedings” and giving the appellate courts a power to refuse to hear a reference from the SCCRC on similar grounds (Criminal Procedure (Scotland) Act, 2010). This indicates that CCRCs are vulnerable to legislative pushback. The Court of Appeal used this power to refuse to hear at least one referral from the SCCRC with respect to alleged jury impropriety in conducting independent research on an accused convicted of indecent assault (Carberry v. HM Advocate 2013 SCCR 587). At the same time, the legislative restrictions were repealed in 2016, and the SCCRC subsequently made a second reference with respect to Al Megrahi’s conviction in relation to the Lockerbie bombing. In both referrals, the SCCRC provided a gist of its reasons despite statutory restrictions on the release of its full reasons. The appeal court, as in the first appeal, upheld the conviction deferring to the trial court’s factual findings. The SCCRC, like the English CCRC, is a new but resilient institution that is not always in sync with the courts.

The SCCRC’s more aggressive approach in referring convictions in cases where the Court of Appeal subsequently upholds the conviction is related to both how the Commission and the Court of Appeal approach their task. This follows the basic legal process insight that it is a mistake to focus on CCRCs and courts to the exclusion of the other. The Scottish courts, no less than the English Court of Appeal, have been concerned about the finality of convictions. In one early and important case, the court rejected overturning convictions on the basis of a lurking doubt about guilt (Harper v. HM Advocate [2005] HCJAC 23). In a recent reference, the Court of Appeal upheld rape convictions of three men on the basis that while there was new research on whether the complainant’s injuries demonstrated a lack of consent, the new material was not sufficient to demonstrate a miscarriage of justice (Meighan et al v. Her Majesty’s Advocate [2021] HCJAC 68 [85]). The Court of Appeal similarly rejected new evidence offered on a referral to support a woman’s claim of diminished responsibility when killing her spouse (Graham v. Her Majesty’s Advocate [2018] HCJAC 57). Like the English courts, the Scottish courts value finality over new expert opinions.

Although the Scottish commission has been more aggressive than the English one, the late Professor Peter Duff (Reference Duff2009: 722–724), a former member of the SCCRC, urged it to be even more aggressive and to use its powers to cause the court to reconsider its jurisprudence. He also noted that in cases where the commission made no reference to the courts, the courts deferred to the commission even to the point of saying that it was up to the commission to decide under its vague “interests of justice” test (Duff, Reference Duff2009: 722–724). Lisa Griffin (Reference Griffin2013) concluded that the SCCRC was prepared to make its referrals on its own view about fresh evidence and the effects of disclosure violations, but not on the basis of lurking doubt about guilt once the Scottish courts had rejected this ground of appeal.

The SCCRC made a particularly interesting referral in a case where the only identification of the accused was an in-dock identification in court. As discussed in Chapter 3.2, some reform legislation has required that easy in-dock identifications be placed in the context of the history of any prior attempts at having the witness identify the accused. Alas, the Scottish courts upheld the conviction (Gage v. HM Advocate [2012] HCJAC 14). The SCCRC has also made unsuccessful referrals asking the Court of Appeal to consider evidence that the appellant was highly suggestible (Beattie v. HM Advocate [2009] HCJAC 22; Murray v. HM Advocate [2009] HCJAC 58); that the complainant in a sexual case should have been questioned about her sexual history (Thomson no 2 v. HM Advocate [2010] HCJAC 11) and with respect to ineffective defence representation (Swankie v. HM Advocate [2008] HCJAC 59). The SCCRC has not made recommendations on wrongful conviction reform, but it has been somewhat more active in commissioning research on its own processes than the CCRC (Roach, Reference Roach2019a: 973). The SCCRC has not been subject to the same criticism as the CCRC for being dependent on the Court of Appeal (Callander and Leverick, Reference Callander and Leverick2019).

5.6.2 Norway

The Norwegian Criminal Cases Review Commission has been operating since 2004. Although it has a structure similar to the English and Scottish commissions in combining legally trained and other experts in criminal justice, its members have traditionally included mental health professionals. This helps explain why about 40 per cent of the Commission’s referrals have involved new mental health evidence. It also raises questions about how the conduct of various commissions is shaped by the expertise of the commissioners and the staff working for the commission. The idea that the background of judges matters has been accepted since the legal realists, but it has featured less with respect to executive agencies, which have been based on undifferentiated concepts of and deference to expertise. Although commissions can hire outside experts, some degree of expertise may be required to perceive the need for additional expertise.

In 2021, thirty-one of thirty-four convictions reopened were as a result of new evidence about the applicant’s mental health. This pattern reflects Norway’s penal culture both with respect to recognizing any psychotic state as establishing the mental disorder defence and allowing “a mild mental disability” to mitigate sentence (NCCRC Annual Report, 2021). It underlined the need to examine how similar institutions grow in the soil of different penal cultures. Penal cultures are not, however, static. As suggested in Chapter 4.3.2, the pattern of referrals in Norway may change with the 2020 restrictions on the Norwegian mental disorder defence.

In 2021, the Norwegian Commission received 263 applications, its highest amount ever received. It ordered that cases be reopened for about 18 per cent of all applications it receives, a significantly higher rate than the other commissions (NCCRC Annual Report 2021). The North Carolina commission, which, as discussed in Chapter 6.3.2.1, is limited to claims of factual innocence, has the lowest referral rate of all the commissions, but the Norwegian rate is much higher than the other commissions (Leverick, Campbell and Callander, Reference Leverick, Campbell and Callander2017). The Norwegian Commission appears to have sufficient funds to appoint counsel when required to assist both applicants and victims. In this way, it can combine features of both inquisitorial investigation and adversarial challenge.

5.6.3 New Zealand

In 2005, a New Zealand judge and former prosecutor, Sir Thomas Thorp published a research report, which raised concerns about continuing to allow the elected executive to decide whether a second or subsequent appeal was required because of concerns about miscarriages of justice. He stressed the disadvantages of the “reactive” procedure for Indigenous Maori people, who at the time constituted about 60 per cent of the prison population most at risk for wrongful convictions. Only 2.25 Ministry staff were devoted to handling ten to fifteen applications each year. Thorp observed that only 9 per cent of applicants between 1995 and 2002 were Maori, and only 2 per cent were Pacific Islanders (Thorp, Reference Thorp2005: 53, 93–94). He stressed that Maori may have particular reasons not to have confidence in the justice system (Reference Thorp2005: 57). Simon Mount (Reference Mount2009: 466) also advocated for replacing petitions to the elected executive with applications to a publicly funded commission. He also found few applicants to the Minister of Justice were Indigenous (Reference Mount2009: 468–469).

Malcolm Birdling (Reference Birdling2012: 75–77) confirmed the reactive nature of the New Zealand process. He documented one case where the advisors to the Minister believed that a case should not be referred to the Court of Appeal, but an outside barrister did. The case was referred, and the conviction was quashed. In another case, the Minister’s advisors recommended denying the application without telling the Minister that an outside barrister who reviewed the application initially concluded that the conviction should be referred back to the court. He also found that the test applied by the Minister of Justice generally required new evidence and was based on a prediction of whether the Court of Appeal would overturn the conviction.

In 2019, the coalition government of Jacinda Arden enacted the New Zealand Criminal Cases Review Commission Act no 66 of 2019. Only the conservative National Party opposed the bill (Hoyle, Reference Hoyle2020: 211). The law followed the English model by requiring one-third of commissioners to be experienced lawyers and two-thirds to have expertise relevant to the Commission’s work. What was different was that under s.9(2) of the act at least one “member must have knowledge or understanding of te ao Māori and tikanga Māori.” This was a recognition of Thorp’s (Reference Thorp2005) important point that Maori were grossly over-represented in prison and had good reason to be as distrustful of an institution dedicated to redressing miscarriages of justice. Two of the first New Zealand Commissioners were Maori, and another was a Pacific Islander. This may change as the new National Party government decided not to reappoint some of the original commissioners and replaced the Chief Commissioners with a retired judge, despite concerns that this presented a conflict of interest, given that the commission reviews the work of the judiciary (White, Reference White2024). The frequent description of commissions as “independent” obscures the important role of governments in making funding and appointment decisions. Commissions may professionalize and bureaucratize wrongful conviction discourse, but they are not immune from politicalization.

The New Zealand Commission has broad powers to refer convictions and sentences for a new appeal if they determine it is in the interests of justice. As suggested earlier, “interests of justice” is a vague and opaque term. The New Zealand commission, unlike the English and Scottish ones, has statutory requirements to disclose its decisions, but such disclosures so far only have revealed the statutory ground with most rejected cases simply being listed as “not in the interests of justice” with others being listed as “incomplete application” and “appeal process available” (NZ CCRC, Completed Reviews). There is statutory guidance that provides that the CCRC should consider whether there has been an appeal; whether there was a new argument or evidence; the prospects of the Court of Appeal allowing the appeal and any other matter that the commission considers relevant (Reference Worth2024: s.17). Carolyn Hoyle (Reference Hoyle2020: 216) has suggested that this may force the New Zealand commission to be almost as predictive as the English commission, though the New Zealand test leaves more flexibility and discretion to the Commission.

Section 12 of the New Zealand Act allows for the Commission to conduct systemic inquiries, and the Commission has announced its first such inquiry into the frailties of eyewitness identification, including cross-racial identifications. Hoyle (Reference Hoyle2020: 219–220) suggests that the English commission has done some similar work, especially with respect to repetitive applications it received where the defence of being a refugee was not raised as a defence. One difference, however, is that this work seemed aimed at reducing the number of applications the English commission received, whereas New Zealand’s inquiry has the potential both to increase the number of applications and to lead to law reform.

Although the commission secured four years of stable funding at $4 million each year, it received more applications, about nine a month, than anticipated. In its first year, it received 220 applications, which were more than the 172 applications made to the elected executive between 1995 and 2020. This created a backlog and revealed the difficulties of ensuring that CCRCs have an adequate annual budget. The first Chief Commissioner, Colin Carruthers warned that “the organization will fail unless it’s got a proper budget” (White, Reference White and Robins2023b: 207). A Canadian report found that one benefit of leaving the task to the elected Minister of Justice was that the group advising the Minister could simply draw on the Department of Justice’s large budget (LaForme and Westmoreland-Traore, Reference LaForme and Westmoreland-Traore2021).

As in England where courts have relied on the “safety net” of the CCRC as a justification for rejecting post-conviction disclosure rights (Nunn, 2014 UKSC 7 [39]), the New Zealand courts have already used the existence of a CCRC as a reason to restrict some rights of the accused. The New Zealand Supreme Court has twice cited the availability of the commission as reasons not to recall judgments or order second appeals (Jolley v. R. [2022] 1 NZLR 595; Uhrle v. R. [2020] NZSC 62). This raises the disturbing possibility that the existence of a CCRC could cause the courts and perhaps the press to become less rather than more concerned with miscarriages of justice.

The New Zealand commission has been gifted and uses a Maori name, Te Kāhui Tātari Ture. It has conducted outreach to Maori populations. As of the end of 2024, about 40 per cent of the 498 applications received by the Commission (NZ CCRC, Application Statistics) have been Maori, demonstrating a marked improvement in access to justice than the 9 per cent documented by Thorp (Reference Thorp2005) or the slightly increased percentages later documented by Mount (Reference Mount2009). At the same time, the percentage of Maori applicants has still not matched over-representation in prison.

The commission’s second referral involved a Maori man, Mikaere Oketopa (formerly Michael October). He was convicted by an all-white all-male jury with two other Maori men of raping and murdering a white woman in 1994 even though the other two men said they did not know him. New evidence about a possible false confession was obtained by the commission from a leading expert on false confession, Gisli Gudjonsson. The Commission also stated that the police had failed to interview two alibi witnesses and may have been affected by tunnel vision (New Zealand Herald, 2023). The applicant Oketopa was not a popular person having taken prison guards hostage to protest his wrongful conviction. He was also convicted of domestic assault while paroled.

Another early referral was based on concerns about mistaken eyewitness identification and concerns that the police who conducted the identification line-up had informed the complainant in a sexual offence case about the identity of the suspect. The conviction was later quashed with the consent of the prosecution. The court confirmed the commission’s concerns and found deliberate police misconduct (X v. R 2024 NZHC 467 [18]).

The New Zealand commission has so far been able to achieve its intended goals of increasing the number of applications from Maori people, who are dramatically over-represented in prisons. It has already struggled with budgetary and delay issues because of a higher than expected applications, and a change of government has led to controversy over the government’s refusal to reappoint some commissioners and its appointment of a retired judge to head the Commission (White, Reference White2024). The English practice has not been to appoint former judges because of perceived conflicts of interests, and the Canadian report made similar observations (LaForme and Westmoreland-Traore, Reference LaForme and Westmoreland-Traore2021: 35, 67). Concerns have been expressed about delays and low rates of referrals by the New Zealand commission (White, Reference White2025).

5.6.4 Canada

From 1989 to 2008, seven public inquiries in Canada recommended that the role of the elected executive, the federal Minister of Justice, in ordering new appeals or trials, be replaced with a CCRC-like body. The creation of a commission was not a political priority until the issue was linked to the over-representation of Indigenous people in prison (LaForme and Westmoreland-Traore, Reference LaForme and Westmoreland-Traore2021; Pate, Reference Pate2022). There has been some use of executive pardons in Canada, but it has been criticized for its politicalized and closed process and for not being clearly focused on wrongful convictions (Trotter, Reference Trotter2001).

The provincial inquiry into David Milgaard’s wrongful 1969 murder conviction was unable to determine the exact basis for an initial decision by the Minister of Justice in 1991 to deny Milgaard a second appeal but a reversal of that decision a few months later after Milgaard’s mother Joyce pleaded with the Prime Minister in a highly publicized encounter in a hotel lobby (MacCallum, Reference MacCallum2008). This inquiry echoed those who defended the creation of a Court of Criminal Appeal in 1907 as a way to remove trial by media. It made harsh conclusions that Milgaard’s public and ultimately successful campaign for a second appeal risked bringing the administration of justice into disrepute (Roach, Reference Roach2017a).

In 2002, a Liberal government refused to create a CCRC but formalized the process for the Minister of Justice granting new appeals or new trials. It set a high standard for such referrals, namely, that a miscarriage of justice “likely occurred.” This resulted in only twenty-five referrals from 2002 to 2024, almost all of which resulted in the convictions being reversed either by an appeal court or by decisions by prosecutors not to prosecute a new trial. The Minister only received on average five complete applications each year, but this increased after 2018 to about eighteen complete applications a year (LaForme and Westmoreland-Traore, Reference LaForme and Westmoreland-Traore2021: 59). The process was reactive, and most applications were rejected before the formal investigative phase. If an application was accepted, the investigation was often very slow and could take five to six years. In response to this delay, Canadian courts invented a new process of granting bail to applicants while they waited in the investigative stage (Reference LaForme and Westmoreland-Traore2021: 145–149). This mitigated delay but also forced victims of miscarriages of justice to live under restrictive and sometimes humiliating bail conditions (Roach, Reference Roach2025a: 179–181).

Public consultations were held in 2021 that led to a law enacted in 2024 that created a Miscarriage of Justice Review Commission to replace the Minister of Justice’s role in ordering new trials or second or subsequent appeals. This continued a Canadian innovation of allowing the executive to select between an appeal and a new trial, whereas in other countries, an appeal is the only option. The Canadian new trial option, especially when used in historical cases, generally results in no subsequent prosecution and comes close to allowing an executive body to overturn a conviction.

The name of the new Canadian commission reflects concerns expressed by exonerees in the public consultation about the CCRC name, in part inspired by increased criticisms of the English CCRC. The exonerees stressed that they are not “criminal cases” but people with families and that they want their cases investigated and not “reviewed.” Also, in response to concerns from exonerees, the Commission will have broad powers to provide a variety of legal and social supports to applicants (LaForme and Westmoreland-Traore, Reference LaForme and Westmoreland-Traore2021: 7).

The standard for referral for new appeals or trials was lowered to a miscarriage of justice “may” have occurred as opposed to 2002 standard of “likely” occurred. The Conservative party voted against the bill because of concerns that this would open the floodgates and create a parallel justice system even though the Canadian referral standard is identical to that used by the SCCRC. As discussed earlier, the Canadian government rejected arguments by Justice LaForme and Westmoreland-Traore that the “interests of justice” were too vague.

The new Canadian law also specifically rejected the American proven innocence model by providing that the Commission may grant a new appeal or a new trial “even if the evidence does not establish the innocence of the applicant” (Criminal Code of Canada s.696.1(7) as amended by S.C. 2024 c.33). This embraces a focus on miscarriages of justice that includes but is not limited to factual innocence. It underlines the exceptional nature of the American proven innocence model that is used by the first CCRC-like body, the North Carolina Innocence Inquiry Commission, which was created by the North Carolina legislature in 2006 and will be discussed in Chapter 6.3.2.1.

The Canadian legislation requires that the government consider gender and diversity in appointing commissioners but, contrary to the recommendations made by Justice LaForme and Westmoreland-Traore, does not require the appointment of at least one Indigenous and one Black commissioner to reflect over-representation of such groups among the prison population most at risk. It also does not respond to their concerns that stakeholders should have a role in appointing commissioners; that the commission should be treated like the judiciary to guard against underfunding; that it should have the same powers as the English commissioner to obtain access to privileged information and that Appeal Courts should be required to admit and consider the new evidence that was used by the Commission while also being able to decide what, if any, weight to attach to it. Their recommendations to broaden the grounds for Courts of Appeal to overturn convictions and to provide for a quicker system of compensation not tied to factual innocence were also not pursued by the Canadian government (Roach, Reference Roach2025a: 340–351). This meant that an ambitious plan that had been praised for its potential to create the most effective CCRC yet (Leonetti, Reference Leonetti2023a) was not pursued even though it was widely recognized that there would not be any more wrongful convictions reforms at the federal level for the foreseeable future. The new Canadian commission, unlike the English CCRC, can only refer convictions and not sentences back to the courts and does not have the statutory power to refer cases to the executive to consider issuing a pardon. Its relation with the government will be complicated, and its relations with the courts will be complicated because each of Canada’s ten provinces has its own Court of Appeal.

Amendments were made after the bill was introduced to confirm that it could consider applications from those who did not appeal and that it could make recommendations about systemic issues that may lead to miscarriages of justice. The new Canadian commission, unlike the English, Scottish and New Zealand commissions, does not have jurisdiction to hear cases based on sentencing. This rejected reform proposals that stressed that Indigenous people may, in particular, suffer from being sentenced or denied parole on the basis of inadequate facts about their circumstances (LaForme and Westmoreland-Traore, Reference LaForme and Westmoreland-Traore2021; Pate, Reference Pate2022; Roach Reference Roach, Maniksis and Watson2025b). The omission of sentencing may represent a more punitive North American legal culture with the Liberal government fearing that including sentencing could lead to it being depicted as soft on crime in the 2025 election.

5.6.5 The Counter-Example of Australia

Australia has so far rejected calls for the creation of a CCRC to replace the power of the elected executive to order new appeals. Between 2007 and 2014, New South Wales had a commission with powers to refer convictions to the court, but it was limited to DNA cases and was repealed in 2014 after failing to refer any convictions back to the court (Hamer, Reference Hamer2014).

After rejecting a bill in 2010 to create a CCRC with its Attorney General, arguing that “South Australia is not Texas” and is “not awash in wrongful convictions” (Hamer, Reference Hamer2014: 278), South Australia in 2013 created a new right of the accused to bring a second or subsequent appeal on the basis of fresh and compelling evidence that reveals a substantial miscarriage of justice. This was “a minimalist reform” (Sangha and Moles, Reference Sangha and Moles2015) that, like the revision procedures in Sweden examined in Chapter 4.3, placed the onus on the convicted person to find new evidence. As will be discussed in Chapter 11.4, one impetus for this legislation was a suggestion that Australia was not complying with the right to appeal in Article 14(5) of the ICCPR, given that the High Court refuses to hear new evidence and Australian courts have held that those who petitioned the elected executive for a second or subsequent appeal have no legal rights in what is essentially a petition for mercy (Sangha and Moles, Reference Sangha and Moles2015). Retired High Court Justice Michael Kirby has argued that all Australian jurisdictions must either create a right to the second or subsequent appeal or a criminal cases review commission in order to comply with the ICCPR (Kirby, Reference Kirby2019).

The Australian approach is a neoliberal reform that avoids having to create a state-funded CCRC and having it serve as a gatekeeper to the courts. The Australian courts have interpreted the new right very restrictively in part because it is patterned after provisions enacted in England in 2003 and followed in many Australian states that allow acquittals to be reopened but only if there is fresh and compelling new evidence that could not have been discovered with reasonable diligence at trial. Any new evidence must be reliable, substantial and highly probative (Van Beelan v. The Queen [2017] HCA 565 at 577). It must also be in the interests of justice to admit the new evidence. The new evidence must also be “fresh,” which in the South Australian legislation is defined to require not only that it was not adduced at trial but that it could not have been adduced “even with the exercise of reasonable diligence.” The statutory requirements that the new evidence be fresh and compelling are mandatory unlike tests for admitting evidence on the first appeal where courts can waive requirements in the interests of justice (Sangha and Moles, Reference Sangha and Moles2015: 183). Finally, the original verdict will only be disturbed if the result would be “substantial” miscarriage of justice (Sangha and Moles, Reference Sangha and Moles2015: 177). As Australian law professors Bibi Sangha and Roberts Moles argue, such restrictions are understandable with respect to incursions on the right against double jeopardy that is affected by reopening acquittals, but they can defeat meritorious claims of wrongful convictions (Reference Sangha and Moles2015: 200–202). The Australian approach seems based on equal treatment of compelling new evidence of guilt or innocence. This is contrary to the asymmetrical treatment of the accused and prosecutors under the presumption of innocence as discussed in Chapter 2.5.

The results of the accused’s new right to second and subsequent appeals so far have been modest. Since 2013, there have only been eight applications in South Australia, and the majority have failed. Since 2015, there has only been one failed application in Tasmania, and since 2019, only two of nine applications have been successful in Victoria (Ruyters and Bartle, Reference Ruyters and Bartle2024: 291). In a 3:2 decision, the Australian High Court refused to allow the country’s longest-imprisoned Indigenous person, Derek Bromley, an appeal on the basis that new evidence, showing that the prosecution’s key witness was psychotic, was not fresh and compelling enough and that the jury had considered the witness’s mental health before convicting Bromley in 1984 of murder (Roach, Reference Roach2025a: 338–340). Two of the five judges who heard the case dissented. They stressed that given the new evidence, there was “a significant possibility that an innocent person has been convicted” (Bromley v. King [2023] HCA 43 [216]).

In addition, Australian states that have most recently added a right to a second and subsequent appeal have given preference to proven innocence. For example, Western Australia provided in 2022 legislation: “The Court of Appeal must allow an appeal based on new and compelling evidence if it is satisfied on the balance of probabilities that, in light of all the evidence, the evidence establishes that the offender is innocent” (Criminal Appeals Act, 2004 s.35H(4)). The legislation still allows convictions to be overturned on the basis of “substantial miscarriages of justice” (Roach, Reference Roach2025a: s.35H (2–3), but this new provision prefers innocence over other miscarriages of justice. Queensland in its 2023 legislation also instructs appeal courts to allow appeals from conviction on a second or subsequent appeal if “on the balance of probabilities, it is of the opinion that, in considering all of the evidence, the appellant was not guilty of the offence of which the appellant was convicted” (Queensland Criminal Code s.671AE(3)). This avoids the word innocence but seems to require proof of innocence on a balance of probabilities. These new laws represent a migration of the American proven innocence model to Australia, albeit on a slightly lower standard than the clear and convincing evidence standard that, as discussed in Chapter 6.5, is frequently required by American courts. The picture that emerges, however, is one where proven innocence standards can eclipse more generous miscarriage of justice standards as a means of rationing justice and where new rights to second and subsequent appeals can be a substitute reform for the creation of a permanent and publicly funded CCRC with investigative capabilities.

The traditional asymmetry of the criminal law and its preference for allowing wrongful acquittals over wrongful convictions suggest that thought should be given to reducing a second and subsequent appeal against conviction to only require “new” as opposed to “fresh and compelling evidence” that reveals “a miscarriage of justice” as opposed to a “substantial miscarriage of justice” (Sangha and Moles, Reference Sangha and Moles2015: 206). Even such a reform, however, would still require the convicted person to find the necessary new evidence without the public funding or powers of a CCRC.

5.6.6 Summary

Like the Court of Appeal, there has been a heavy export market in the CCRC. At the same time, there is a need for countries both to adopt CCRCs to their local conditions and be sensitive to the weaknesses of the English CCRC. The Scottish commission demonstrates that commissions can be more aggressive in making referrals that challenge the Court of Appeal’s jurisprudence on admitting new evidence and overturning convictions. The Scottish commission, like the New Zealand and Canadian commissions, has a referral test that appears to be less predictive of whether the Court of Appeal may overturn the conviction. The Norwegian Commission demonstrates how a well-resourced commission can more frequently appoint lawyers to assist applicants and how specific expertise on the Commission, in that case, in mental health, can produce many referrals in those areas. This raises questions of where Commissions should allocate their limited resources in creating in-house expertise. The ability to critically examine guilt produced by machines and statistics may be a neglected area of expertise.

The English CCRC has not used its statutory powers to circumvent the courts by recommending to the elected executive that a pardon be granted (Hoyle and Sato, Reference Hoyle and Sato2019). The Australian approach has avoided creating a new state-funded CCRC and has placed the burden on convicted persons to find fresh and compelling new evidence. It is patterned after legislation that places the same demanding standards on the state when it wishes to reopen acquittals in derogation of the accused’s protections against double jeopardy. This seems to be based on an equivalence of the importance of new evidence about factual guilt or innocence even though traditional liberal principles are supposed to prefer wrongful acquittals to wrongful convictions.

The New Zealand legislation recognizes the need for expertise about Indigenous people over-represented among the prison population most at risk for miscarriages of justice. It has had some success in attracting more applications from Maori people than the previous system that relied on the elected executive reacting to applications for the Crown’s mercy. Recommendations for the need for Indigenous and Black commissioners (LaForme and Westmoreland Traore, Reference LaForme and Westmoreland-Traore2021) were not followed in the Canadian legislation, as were attempts to insulate its new CCRC from government-dominated budgeting and appointment processes. From a legal process perspective, all commissions remain vulnerable to underfunding and the appointments made by the government of the day. Indeed, the new conservative government in New Zealand has attracted criticism for its refusal to reappoint existing commissioners and the appointment of a retired judge to be a new Chair of its commission (White, Reference White2024).

5.7.1 Factual Innocence and the Failed 2006 New Labour Consultation

In 2006, the UK government conducted a consultation on whether convictions should no longer be quashed on appeal on the basis of procedural or legal error but only on the basis of factual innocence. Prime Minister Tony Blair himself foreshadowed his government’s concern when introducing a scheme to “rebalance” the justice system in 2002, he argued, “Its perhaps the biggest miscarriage of justice in today’s system when the guilty walks away unpunished” (Blair, Reference Blair2002). His appeal to factual innocence was also matched by an appeal to factual guilt foreshadowing 2003 legislation that would allow acquittals to be reopened in the face of new and compelling evidence of factual guilt (Criminal Justice Act, 2003 c.44).

The government’s 2006 Quashing Conviction consultation noted with some horror that

The Blair government with its popular tough on crime approach wanted to deny appeals to those who were not innocent.

Laurie Elks (Reference Elks2008: 47–48), who was a member of the CCRC in 2006, subsequently wrote that the adoption of the Blair government’s proposal “would have required the re-writing of many school and university textbooks on British constitutional law and history” (Reference Elks2008: 47–48). He was not alone in his opposition. Most groups representing the bar, as well as civil society organizations such as Liberty and Justice, vehemently opposed the 2006 proposal. The CCRC in a very rare act of engagement in policy debates opposed the proposal, arguing “if the legislation were to enshrine the principle that any irregularity could be overlooked – provided that the convicted persons are deemed plainly guilty – this would inevitably cause collateral damage to the criminal justice process by which innocent defendants would be more likely to be convicted” (as quoted in Hoyle and Sato, Reference Hoyle and Sato2019: 179; Roach, Reference Roach2010b: 115–116).

The government’s approach had some perhaps unlikely support. The Runciman Commission, which had recommended the creation of the CCRC, had equally stressed the punishment of the guilty along with the acquittal of the innocent as goals of the justice system. Other bodies recommended that in order to respect the rights of victims, acquittals should be re-opened if there was new and compelling evidence. Michael Naughton (Reference Naughton2013: 176), the head of a network of innocence projects in the UK, raised concerns that the CCRC was failing the innocent and winning points for victories on legal technicalities including a 2004 referral that led to the reversal of an order for the destruction of Dino the dog.

Hannah Quirk (Reference Quirk2007: 768) warned that while claims of innocence may have a use in campaigning, if implemented into law, they would create “two tiers of successful appellants…: the innocent and those who ‘escaped on technicalities.’” She also argued that changing the test to innocence would undermine legislative reforms designed to prevent police and prosecutorial conduct that had been associated with famous wrongful convictions such as the Birmingham Six. She also drew on her experience as a former CCRC case reviewer that “clear-cut cases of innocence” are “rarely seen” and that a move to an innocence model “would be a regressive step” (Reference Quirk2007: 776).

In response to all the criticism it received, the Blair government withdrew a proposal in a bill that would have allowed the Court of Appeal to uphold a conviction if it was satisfied that the appellant was guilty (Law Commission, 2025: 8:57). In 2007, the government announced that it would retain the safety test for quashing convictions.

5.7.2 Compensation: From Miscarriages of Justice to Proven Innocence

In 1988, England enacted legislation that provided for compensation if the applicant could prove beyond a reasonable doubt that they suffered a miscarriage of justice. Between 1999 and 2008, over £83 million was paid in compensation (Law Commission, 2025: 16.18). In 2008, a Labour government amended this law to provide statutory caps on damages, so that those wrongly imprisoned for less than ten years could receive no more than £500,000 and those imprisoned longer could receive no more than a £1 million (Quirk, Reference Quirk, Jasinski and Kremens2023). Consistent with Labour’s rebalancing of the criminal justice system, this was defended as producing parity between compensation available to victims of crime and the wrongfully convicted. In the world of politics, there is parity, and sometimes competition, between these two sets of victims. This is supported by public opinion polling suggesting that many people view wrongful acquittals and wrongful convictions as equal harms (Garrett and Mitchell, Reference Garrett and Mitchell2023). On such a view, it does not matter that much if the state will generally bear much more direct responsibility for a wrongful conviction than for a failure to prevent crime. In any event, the 2008 cap has not been revised as of early 2025, causing Andrew Malkinson to argue that the £1 million cap should be doubled to keep up with inflation (The Guardian, 2025b). From 2008 to 2014, only £6. 3 million was paid in compensation (Law Commission, 2025: 16.18).

As discussed in Chapter 2.3.3, the 1988 English legislation was interpreted restrictively by the courts to exclude miscarriages of justice, such as illegal extraditions and unsafe convictions (Mullen 2004 UKHL 18; Adams [2011] UKSC 18). The courts also upheld the ability of the state to deduct “room and board” costs for the time in which a person had been wrongfully imprisoned (R. (on the application of O’Brien) v. Independent Assessor [2007] UKHL 10). This was only changed after Andrew Malkinson complained about its injustice shortly after his release from prison (Law Commission, 2025: 16.2). The change was not applied retroactively, and the government has refused to pay Paul Blackburn more than £100,000 that was deducted when he was compensated after serving twenty-five years after he was wrongfully convicted for murder at fifteen years of age and on the basis of evidence fabricated by the police (British Broadcasting Company, 2024). Compensation decisions can impose additional harms on the wrongfully convicted.

Despite many existing judicial and legislative restrictions on compensation, the Cameron Conservative government amended the compensation legislation again in 2014 to require claimants to establish “beyond a reasonable doubt” that they “did not commit the offence” (Criminal Justice Act, 1988 s.133 (1ZA)), as amended by Anti-Social Behaviour, Crime and Policing Act, 2014 c. 4 ss.61(3, 153). The original version of the bill required proof beyond a reasonable doubt that the claimant was innocent. This was criticized by many legal elites for inverting the presumption of innocence. As Hannah Quirk (Reference Quirk, Jasinski and Kremens2023: 14) noted, however, the final version “is largely a distinction without a difference.” Jacqueline Hodgson (Reference Hodgson2020: 294–295) pointed out that there is a “surreal mismatch” now between the standards needed for compensation and those used by courts when overturning unsafe convictions. The Law Commission (2025: 16.35) has noted that this is the only point where accused persons “need to prove their innocence, let alone to do so to the high standard of beyond reasonable doubt.”

If the new innocence requirement was introduced to conserve funds, it certainly worked. In 2017 and 2018, no compensation was awarded under the amended English legislation. Between 2019 and 2021, only £1.5 million was paid out compared to £41 million being paid out between 1999 and 2001, a period before Parliament had imposed both caps on damages and innocence requirements (Roach, Reference Roach2024a). Under the 2014 law, compensation was denied to Victor Nealon after his sexual assault conviction was quashed on the basis of a DNA exclusion. Nealon failed in his attempt to challenge the new law as infringing the presumption of innocence despite concerns that the denial of compensation on innocence grounds inflicts both financial and moral harms of the wrongfully convicted (Hoyle and Tilt, Reference Hoyle and Tilt2020; Nealon v. United Kingdom, 2024 ECHR 514).

The Law Commission (2025: 16: 45) has recognized the injustice and frequent impossibility of requiring proof of innocence beyond a reasonable doubt. Its rather timid provisional response, however, was to suggest that applicants still have to prove factual innocence but on a balance of probabilities even though their convictions will be overturned on the different basis that they were unsafe. The elected executive would continue to make decisions about compensation, whereas under many American compensation laws, it is administrative agencies or courts that determine whether the convicted person has established their innocence and other qualifications for receiving compensation. More fundamentally, it is disappointing that the Law Commission did not recommend a change to the American-style factual innocence requirement despite documenting drastic reductions in compensation paid since its 2014 introduction.

The Law Commission was on sounder ground in finding that there is a need for non-financial supports, especially given that the wrongfully convicted receive less post-release rehabilitative supports than the rightfully convicted (2025: 16: 144). It also noted that the wrongfully convicted are not necessarily accorded the highest priority for receiving housing assistance (2025: 16: 132). Hopefully, this will result in reforms. The limits of financial compensation as a means to respond to the many irreparable harms of miscarriages of justice will, however, be examined in Chapter 11.6.

Both the failed 2006 and successful 2014 moves towards proven innocence underlined that, while that model is most firmly entrenched in the United States, it is not limited to America. It is also a reminder that the wrongfully convicted are a politically vulnerable constituency that can be subject to changes that will less generously allocate post-conviction review and funds to compensate for miscarriages of justice.

5.8.1 The INUK 2004–2014

The INUK was formed in September 2004 by Innocence Projects at the Universities of Bristol and Leeds. It was influenced by American innocence projects. Michael Naughton and Carole McCartney (Reference Naughton and McCartney2004: 152) raised concerns that the CCRC “can be said to have diverted the public gaze away from the problem. The kind of moral outrage and public crises of confidence once associated with miscarriages of justice has been replaced with an almost perverse need for miscarriages of justice as we now have an official organisation to deal with them!” This raised legitimate concerns that wrongful conviction discourse would become professionalized and bureaucratized with the advent of the CCRC.

Before it closed operations in July 2014, INUK at its peak included twenty-six university innocence projects and one in a corporate law firm. The guiding force behind INUK was University of Bristol sociologist Michael Naughton, who is also a prolific wrongful conviction scholar. He championed a “lay” approach that focused on factual innocence and guilt (Naughton, Reference Naughton2013). In 2006, he argued against lumping “together victims of miscarriage of justice and, potentially, innocent victims of wrongful convictions.” He was concerned that both groups could be perceived “as ‘getting off on technicalities,’ as, indeed, all criminal appeals are highly technical affairs governed by strict rules and procedures.” He expressed concerns that the establishment of the CCRC had “blurred into one amorphous issue general concerns about human rights, civil liberties and social justice under a banner of miscarriages of justice” (Naughton, Reference Naughton2006).

Like many American innocence projects, INUK limited its cases to factual innocence and did not consider claims of “technical miscarriages of justice,” including sentencing, diminished responsibility or guilt to a less serious crime (Naughton and Tan, Reference Naughton and Tan2010: 61). Professor Naughton trained those working in the projects to investigate factual innocence, and “it goes without saying that if evidence turns up in the course if an innocence project investigation that a ‘client’ is not innocent, the case is terminated…” (Naughton, Reference Naughton2010: 31). At the end of 2011, INUK called for the CCRC’s real possibility test to be replaced by one that would require a reference in any cases where the CCRC believes that the applicant “is or may be innocent.” That would have likely reduced the number of referrals to the Court of Appeal, but INUK was not opposed to ending references in both “non-innocence” and sentencing cases. Some subsequent research into the CCRC lends some support to some of INUK’s concerns. For example, Hodgson (Reference Hodgson2020: 291) quoted an interview with a CCRC case manager who observed the decline in “genuine” miscarriages and the rise of “many more technical miscarriages now where somebody, somewhere hasn’t followed the rules, [but] there is not actually much doubt that the appellant was guilty as hell.”

The INUK also believed that the CCRC was too dominated by lawyers and called for the appointment of “former forensic scientists, investigative journalists and academics as Commissioners and Case Review Managers, which could help to promote a more factual investigative culture” (INUK Inquiry, 2014: 11). The INUK was not alone in growing criticisms of the CCRC. Former commissioner and investigative journalist David Jessel stated that “my former colleagues at the CCRC are lawyers all, even though the 1995 Act said only that a third of them need be.” He was concerned that the CCRC, under the sway of lawyers, could become “the handmaiden of the Court of Appeal.” Drawing on his experience as an investigative journalist, Jessel concluded:

Jessel warned that the majority of the CCRC’s cases “aren’t investigated at all. They are reviewed, yes, but a review may often take the form of reading an application form and deciding – there and then – that there’s nothing in the case.” He conceded that “this ruthless triage, or filter, may be inevitable when you have 1000 cases a year” but warned “it’s also a racing certainty that by not investigating so many cases, genuine miscarriages of justice slip through” (2012: 11). Jessel’s concerns, like Naughton’s, have been vindicated with the high-profile mistakes made in the Nealon, Malkinson, Campbell and Sullivan cases examined earlier. To some extent, the CCRC’s mistakes were inevitable, given its limited investigative resources (Nobles and Schiff, Reference Nobles and Schiff2001).

Naughton criticized both the Court of Appeal and the CCRC for not being “concerned with whether appellants are factually innocent or factually guilty” and argued that the original vision of both institutions was to provide remedies for the factually innocent (Naughton, Reference Naughton2013: 152, 168). Although victims of unfair processes could have their convictions quashed “from a lay perspective, successful appellants are only victims … if they are considered to be factually innocent” (Reference Naughton2013: 20).

The INUK released a dossier of forty-four cases in which it believed applicants to the CCRC to be innocent, but the CCRC did not make a referral of the case to the Court of Appeal. The breakdown of relations between the CCRC and INUK was no doubt related to the personalities involved, but it also spoke to a more pervasive pattern of the CCRC “distancing itself from the interests of the applicant and the applicant’s lawyers” (Hodgson, Reference Hodgson2020: 311) in an attempt to seem independent from them. Naughton criticized the CCRC for “not undertaking thorough investigations to determine whether claims of innocence are true” and for not referring such cases to the executive for a free pardon (INUK, 2013a: 21). At the same time, Emily Bolton and Glyn Maddocks of the Centre of Criminal Appeals stated that it would have “grave concerns” that an innocence standard would make it more difficult for innocent people to have their convictions overturned. They criticized INUK for effectively saying, “people are having troubling hitting the bullseye with the CCRC so lets draw it smaller.” Innocence, “while appealing from a campaigning perspective, leads to a nightmarish standard for a wrongfully convicted person to have to meet, and a gift for those prosecutors (or courts) with an interest in defending convictions” (Reference Jiang2013a: 48–49).

A 2013 INUK report included a statement by Paddy Joe Hill, one of the wrongfully convicted Birmingham Six, that: “For people in prison, miscarriages of justice are people who are innocent and who have been wrongly convicted. As things stand, the CCRC is referring cases of guilty people on technicalities. This is not its intended function” (INUK, 2013b: 26). Hill also questioned the CCRC’s independence: “But how is it supposed to be independent? The CCRC is financed by government. Its senior members are appointed by government, and a number of them used to work for the Crown Prosecution Service. The CCRC is not as independent as it claims to be. It helps to maintain the status quo of the judicial system” (Reference Jiang2013b: 26). In other words, Hill was raising the possibility that reforms such as the CCRC prompted by the own injustice he suffered were providing only legitimating illusions of independence, reform and commitment to the truth.

The INUK’s criticisms of the CCRC affirm Nobles and Schiff’s (Reference Nobles and Schiff2000: 260) observations that miscarriages of justice may “generate hostility and impatience towards bodies with official responsibility for remedying mistakes” even though such criticisms may be better directed at first-order institutions such as police and forensic service providers. They may also have made the CCRC defensive and less welcoming of alliances with innocence organizations even though such alliances have the potential to improve the appointments and budgets that the CCRC received from the government. A Canadian report has recommended that a CCRC should have an advisory committee that would represent innocence organizations and other civil society groups. It proposed that such an advisory committee could play an important role as an advocate for the Commission with the government; that it could play a role in appointments to the Commission and that it could be a public advocate for reforms that could prevent wrongful convictions (LaForme and Westmoreland, Reference LaForme and Westmoreland-Traore2021: 75). For its part, the English CCRC had some non-executive directors, but they were generally specialists in budgetary matters, not wrongful convictions.

In 2014, INUK announced it was disbanding in part because it was running out of cases of provable factual innocence. This accords with statements by commissioners on both the English commission (Kyle, Reference Kyle2004) and the Scottish commission (Duff, Reference Duff2009) that they also saw few cases of provable factual innocence (Quirk, Reference Quirk2007). The missed cases of Victor Nealon, Andrew Malkinson, Oliver Campbell and Peter Sullivan, however, suggest that at least some of INUK’s criticisms were valid and that it still had work to do including in being a watchdog of the CCRC. In comparative law terms, INUK can ultimately be seen as a failed attempt to transplant American innocence projects to England’s very different legal and penal culture.

5.8.2 The Future of English Innocence Organizations and Movements

It would be a mistake to conclude that the collapse of INUK means that there is no room for innocence organisations outside of the United States. Lynne Weathered and Stephanie Roberts (Reference Weathered and Roberts2009: 9), while sympathetic to the innocence project’s focus “on lay notions of factual innocence,” have attempted to bridge some of the stark gulf that Naughton and INUK drew between factual and legal innocence by noting that the conviction of a proven factually innocent would be found unsafe by English courts. They characterized innocence projects, as Nobles and Schiff (Reference Nobles and Schiff2000) had characterized the media, as “linkage institutions” between legal and lay systems (Reference Nobles and Schiff2000: 10). Innocence projects in their view could be bilingual in the sense that they would talk to the media in terms of factual innocence but make applications to the CCRC and courts in terms of legal innocence. Innocence organisations can also lend legitimacy to the CCRC, especially with respect to applicants who may be frustrated by delay. Criticisms from innocence organisations can also help delegitimatize CCRCs. They can be criticized by governments for not being efficient with limited budgets, by courts for making too many referrals and by innocence organisations for not making enough or the right referrals.

There is wide agreement that there is still a legitimate role for innocence organisations (Greenwood, Reference Greenwood2021; Hewitt and McGourlay, Reference Hewitt and Claire2021). Cutbacks in legal aid and underfunding of the CCRC make innocence projects important. Hewitt and McGourlay reported in 2020 that there were twelve innocence projects operating in the UK. The innocence projects that they head in London and Manchester are part of the Innocence Network and, as such, require factual innocence as criteria for the cases they accept (Reference Greenwood2021: 227, 238).

In my view, the Innocence Network should rethink the requirement of factual innocence. To be sure, factual innocence can be justified as criteria adopted by charitable institutions for rationing their scarce resources. At the same time, the Innocence Network should be aware of the possibility that, while their “tragic choice” rationing may be perfectly acceptable in the United States, it may be giving innocence projects outside the United States a next to impossible task. It should also recognize that while proof of factual innocence is often necessary in the United States, it is less necessary in other legal systems and may have regressive effects.

Holly Greenwood (Reference Greenwood2021: 468) also notes that American innocence projects, unlike English ones, are not run by full-time academics and that they are “more akin to specialized law firms which receive significant philanthropic funding” (Reference Greenwood2021: 477). Americans should not assume that the charitable funding that is available and encouraged by their low taxation rates is necessarily as readily available in other countries. In addition, American innocence organisations “benefit” from much higher prison populations, thus increasing the pool of possible candidates. Each innocence movement can take inspiration from the US model, but it should respond to local conditions and not necessarily have the same goals or expect to achieve the same results as American innocence projects. Innocence projects can also play a watchdog role with respect to CCRCs. To be sure, INUK’s aggressive criticisms of the CCRC were controversial, but to some extent, they have been validated by the high-profile mistakes made by the CCRC in the Victor Nealon, Andrew Malkinson, Oliver Campbell and Peter Sullivan cases.

5.9.1 Systemic Problems, Mass Exonerations and False Guilty Pleas

The Post Office started using the Horizon computer system in 1999 and began prosecuting sub-postmasters that year for shortfalls in remitted cash identified by the new computer system. Horizon was very complex. It was the largest non-military computer system developed in Europe. Despite many complaints about the system, the Post Office denied problems. Moreover, it continued to use the system and prosecute sub-postmasters.

Computer Weekly, a specialized publication, was the first to write about the scandal in 2009. It featured the plight of seven sub-postmasters and quoted one of them, Jo Hamilton, as saying: “I didn’t understand what was happening, and I’m so rubbish with IT that at the time, I thought it was somehow my fault.” (Thomson, Reference Thomson2009) Hamilton was charged with theft but pled guilty to false accounting. She avoided jail when her customers loaned her £9,000 to repay the Post Office.

The Post Office defended Horizon as “robust” and noted that none of the postmasters “have firm evidence” that the computer system was to blame. The story was unable to reveal any flaws in the system, but quoted IT experts about the need for additional investigation (Thomson, Reference Thomson2009). Many of the flaws in the system only came to light as a result of group or class action civil litigation that resulted in findings by a tech-savvy judge who ruled against the Post Office in 2018 and 2019. After the Post Office’s unsuccessful attempts to appeal, the litigation was settled for £58 million, but after the fees of the private funders and lawyers, the 555 sub-postmasters only received £11 million.

The scandal underlines the pervasive danger of false guilty pleas and how even small business people like the postmasters lack the resources to challenge allegations of guilt made by large organizations and often based on the work of complex machines or science. Although the largest, the Post Office cases were not the first mass exonerations in England. Previous cases involved people who pled guilty to impaired driving before systemic problems were revealed in systems used to test their blood. In the late 1980s, almost 1,500 cases involved problems in the Manchester police lab. Only fifty-two people sought judicial review of their guilty pleas, and sixty to seventy received compensation. At least two people attempted suicide, and one lost his business (Law Commission, 2025: Appendix 3: 3.43ff). Another case of lab error affected over 10,000 cases involving forty-two police services. Again, it only resulted in a few guilty pleas being declared unsafe (2025: 3.48, R. v. Senior 2018 EWCA Crim 837). The attention that the sub-postmasters eventually received made them lucky compared to these other groups.

In the largest Post Office criminal appeal, the Court of Appeal decided that the Post Office’s prosecutions were an abuse of process. It recognized that those who made false guilty pleas were “in the very difficult position of being charged with offences of dishonesty committed in breach of their employer’s trust. They are likely to have been advised that imprisonment is very often imposed for such offences, and that the mitigation which would be available to them if they pleaded guilty could therefore be of particular importance” (Hamilton and Others v. The Post Office, 2021 EWCA Crim 577 at para 125). In other words, false guilty pleas are often the result of prosecutor’s offers that no reasonable accused can refuse.

A year later, however, in a non-post office historical case involving multiple counts of manslaughter, the Court of Appeal decided that in order to give meaning to a valid guilty plea, the appellant must establish that he did not commit the offence (Tredget v. R [2022] EWCA Crim. 108 [171–173]). This high standard would have been an impossible-to-meet standard in the Post Office cases, where the full facts of the scandal are the subject of an ongoing public inquiry. The Court of Appeal and subsequently Parliament were much more sympathetic to the position of the sub-postmasters as small business people than Peter Tredget, who as a disturbed teenager pled guilty to a series of arsons and manslaughter. In Chapter 6.7, the American experience with mass exonerations will be examined. Such cases reveal much about the standard operating procedures of the justice system and the systemic flaws of criminal justice institutions, such as the police, prosecutors and forensic labs. They also reveal the limits of more individualistic case-by-case corrections of wrongful convictions.

5.9.2 The Failures of the CCRC and the Court of Appeal to Correct Injustices in a Case-by-Case Manner

The CCRC became aware of problems in the Post Office cases in 2013 but only engaged in more proactive outreach to potential applicants after the conclusion of the successful civil litigation in 2019. The CCRC did not exercise its power under ss.19 and 20 of the Criminal Appeals Act, 1995, to appoint an investigating officer (Flinders, Reference Flinders2019; Law Commission, 2025: 17.36, 17.69). It mailed letters to over 350 people but only received fifty-one responses and twenty-one of them indicated that they wanted no contact with the CCRC. By the time exonerating legislation was enacted in 2024, the CCRC had completed 135 cases and had decided not to refer 62 cases because of concerns about guilt independent from Horizon. Its referrals had resulted in sixty-three convictions being quashed but five convictions upheld (CCRC, 2024: 33). Case-by-case resolution of the miscarriages of justice was slow and not certain. Many people who pled guilty appeared not to want any additional involvement with a justice system that had already failed them.

In December 2023, the government’s advisory committee on the scandal recommended legislation. It argued that the “balance clearly favours overturning convictions of the innocent as against the risk that some guilty might go free (they have been punished already)” (Horizon Compensation Advisory Board, 2023). It also pointed out that many of the postmasters were not willing to apply to the CCRC or appeal “given their understandable deep distrust of authority” and research that revealed that over 60 per cent of those affected had post-traumatic and depressive symptoms. The Post Office had also informed the Advisory Board that in over 300 cases it might not be possible for the courts to determine that the convictions were unsafe. The Advisory Board raised fundamental concerns about the ability of both the CCRC and the Court of Appeal to resolve the scandal when it stated:

The whole affair raised the larger question “whether the legal system (at each of the stages of investigation, prosecution, sentencing and review of appeals) is capable of differentiating between the innocent and the guilty – and the Horizon story raises serious concerns in this regard” (2023). This could be seen as another recognition that the criminal justice system, especially when confronted with complex systems, could not honour its rhetorical commitment to convicting only the guilty and acquitting the innocent.

5.9.3 The Importance of Mr. Bates v. The Post Office to the Remedial Legislation

Despite all of their shortcomings, the CCRC and the Court of Appeal might have continued to resolve the Post Office cases had not over 13 million viewers spent part of the 2023–2024 Christmas and New Year’s holiday watching an entertaining and BAFTA award-winning television drama “Mr. Bates v. the Post Office.” Toby Jones played Alan Bates, and Monica Dolan played Jo Hamilton. The impact of the television series reveals, again consistent with Nobles and Schiff’s approach, the continued role of the media in creating a crisis that eventually forced Parliament to enact. At the same time, history did not simply repeat itself. The media was not the sustained investigative reporting that led to terrorism wrongful convictions eventually being corrected in the 1990s, but a cozy drama that built on a growing frustration with the government after the policy failures with respect to the Covid epidemic and Brexit.

Five days after “Mr. Bates v. the Post Office” was aired, Rishi Sunak’s Conservative government indicated it would enact emergency legislation “to make sure subpostmasters convicted will be quickly exonerated” (Flinders, Reference Flinders2024). The exoneration legislation received pushback from the senior judiciary with Chief Justice Carr telling a Parliamentary Committee on January 16, 2024, it was “not factually correct” that the courts could not cope with the scandal and “any suggestion that the judiciary has given any proposed legislation the green light is simply not true” (Siddique, Reference Siddique2024). Isobel Plumstead, a retired former chair of Circuit Court judges, was blunter when she stated:

A representative of the Law Society added: “It breaches a fundamental principle which is effectively the government legislating against decisions, against the independence of the courts. These are exceptional circumstances, it is an extraordinary measure. It must not, must never be seen as a precedent” (The Guardian, 2024d).

In introducing the bill in Parliament, Kemi Badenoch, then Conservative Minister responsible for Business and Trade, said it was designed to avoid “a case-by-case, file-by-file assessment of each prosecution” in order to achieve efficiency. Some in the legal community continued to express constitutional concerns about the use of legislation to nullify convictions in the courts. David Dennis recognized the risks of “lumping the genuinely innocent majority with a very small potentially guilty minority” but supported the bill nevertheless. In discussing the bill on second reading in the House of Lords, Lord Falconer stated that the numbers affected range from 500 to “maybe 900” and stated: “the destruction of the lives and livelihoods of the sub-postmasters and their families is unspeakable.” As members of the then Labour opposition noted, the legislation would not have been enacted if not for the ITV drama Mr. Bates v. the Post Office. This raises difficult questions about whether Parliament would come to the rescue of a less sympathetic group. There certainly was no legislation to exonerate migrants who were wrongfully convicted and were subject to mass and successful referrals back by the CCRC back to the courts.

Special legislation was enacted not only to nullify the convictions but to compensate the sub-postmasters. The Law Commission (2025: 16.76, 16.80) has correctly noted that this special legislation is far more generous than the 2014 compensation law that effectively required proof of innocence, something that might be impossible given the complexity of the computer system and the Post Office’s own destruction of relevant materials. In September 2023, the government offered postmasters who had their convictions overturned £600,000 in compensation (Flinders, Reference Flinders2023). This was over the £500,000 cap under the regular compensation for those who had been wrongfully imprisoned for under ten years. The Post Office has estimated that its total compensation bill may be in excess of £1 billion, especially after claims increased after the airing of Mr Bates v. The Post Office (The Guardian, 2024e).

The postmasters were sympathetic and certainly harmed, but most served no time in prison. The special compensation legislation harkened back to the use of ex gratia payments before compensation legislation was first enacted in 1988. On the one hand, there was flexibility, but on the other hand, there was a potential for discrimination. Sympathetic and well-publicized business people like the postmasters received generous compensation, while Victor Nealon, who had wrongfully served seventeen years in prison for sexual assault, received nothing. As with the exoneration legislation, Parliament has been prepared to enact exceptional legislation because of the inadequacy of the existing legislation without dealing with the inadequate status quo that will apply to the vast majority of victims of miscarriages of justice.

5.9.4 Summary

The Horizon Compensation Advisory Board’s December 15, 2023, letter recommending wholesale exoneration legislation is a damning indictment of the ability of the legal system to correct miscarriages of justices. It also makes clear that both the CCRC and the Court of Appeal would have upheld a non-trivial number of the convictions overturned by the legislation. This is especially true if the Court of Appeal was not prepared to rule the prosecutions to be an abuse of process. If, as the Court of Appeal suggested in a case involving a deceased postmaster “the usual principles must apply equally in all cases” (White v. Post Office [2022] EWCA Crim 435 [15]), then each postmaster would have had to establish not that their conviction was unsafe but that they did not commit the offence (Tredget v. R [2022] EWCA Crim. 108 [171–173]). For perhaps most, this would have been an impossible burden.

My concern is not so much that some postmasters may have received lenient treatment or even that some might have been factually guilty, though that may have occurred. Rather, it is a concern rooted in equality (Dworkin, Reference Dworkin1985: ch 3) that others, perhaps more disadvantaged than the postmasters and who are not the subject of sympathetic media, might receive unequal treatment under the Tredget precedent and compensation legislation that similarly requires proof of innocence.

There are concerns about the fate of miscarriages of justice in what might be characterized as an increasingly politicized and perhaps even post-truth world. As machines and complex science are increasingly used in prosecutions and legal aid is cut, the Post Office Scandal, like similar and even larger mass exonerations in the United States that will be discussed in Chapter 6.7, raises existential questions about whether all accused really can ever have the fair trials that are promised to them. The exonerations and compensation provided to the postmasters may have been a form of rough justice, but it was only justice for some.