4.2.1 The Inquisitorial Hope

Writing in 1932, Edwin Borchard praised European countries for their willingness to correct wrongful convictions and to compensate victims. He wrote: “In Europe, the opportunity to assert and prove innocence is practically always available before the courts, on a petition for revision of the sentence” (Borchard, Reference Borchard1932: 405). His praise of European systems reflected their willingness to codify possible causes of wrongful convictions and to allow appeals on factual grounds. Like many American commentators after him, Borchard bemoaned American appeals focused on legal errors as one that both allowed “unjust verdicts” and new trials to be ordered “on technicalities where no substantial injustice can be shown” (Reference Borchard1932: xxi). Borchard clearly saw European systems as more concerned about truth and innocence than Anglo-American adversarial systems.

In 1957, Jerome and Barbara Frank related wrongful convictions to the adversarial idea that the criminal trial was “a sort of game or sporting event” (Frank and Frank, Reference Frank and Frank1957: 34). The defence was often mismatched in such a game, and it was possible for a person to be wrongfully convicted “for the crime of mistakenly selecting a poor lawyer” (Reference Frank and Frank1957: 35). They praised Scandinavia for its public legal aid just as Borchard had praised the Nordic countries for their generous compensation systems.

Because defence lawyers were often outmatched by prosecutors, they often resorted to “bargaining” with the prosecutor even though their clients might be innocent (Reference Frank and Frank1957: 114). Jerome Frank also drew on his own experience as an appellate court judge to reject the idea that wrongful convictions would be corrected on appeal as the “Upper Court Myth.” He explained that appeal court judges “are obliged to accept the jury’s verdict” and “take as true the testimony of the government witnesses, even if that court suspects they lied or were mistaken” (Reference Frank and Frank1957: 33). This was the opposite of the inquisitorial revision proceedings, which allowed appeal courts to revisit factual findings.

In 1996, Daniel Givelber: 1371 observed that while the American system granted the accused many rights, it failed to impose obligations on the police or prosecutors “to conduct a thorough investigation, to maintain comprehensive records” or to screen charges. This created the possibility of convictions that were “meaningless” and in error. He also elaborated how protestations of innocence were often ignored (Givelber, Reference Givelber2000) as the appeal and post-conviction system focused on legal error. Like other critics of the American adversarial system, he demonstrated impatience with the willingness of American courts to correct even minor legal errors while ignoring factual errors.

In 1999, Colorado law professor William Pizzi wrote Trials without Truth in which he argued that plea bargaining in the United States avoided trials at all costs with the truth being “compromised right and left.” Judges were reluctant to explore the factual accuracy of a guilty plea lest a deal carefully worked out between the lawyers fall apart (Pizzi, Reference Pizzi1999: 73). Although acknowledging the diversity of inquisitorial systems, a common feature was the existence of an investigative file that did not “belong to one side” (Reference Pizzi1999: 113) and was disclosed to both sides without adversarial “games” about disclosure. Professor Pizzi had a point that a failure to disclose relevant evidence to the accused was a frequent immediate cause of wrongful convictions and that such disclosure violations continue despite much adversarial litigation over them. He was as much concerned with the exceptionalism of the American criminal justice system as with any “inquisitorial advantage” in discovering the truth. In 2021, he wrote another book relating mass imprisonment to the complexity of American trials that, combined with high mandatory sentences, drove the vast majority of accused to accept a plea bargain (Pizzi, Reference Pizzi2021).

Pizzi also criticized the lack of judicial leadership in the United States in questioning witnesses and in giving guidance to juries. Here again, the American justice system can be seen as hyper-adversarial because judges in England and Canada often provide extensive guidance to juries on the facts, as well as on the law (Friedland and Roach, Reference Friedland and Roach1997). Pizzi ignored the possibility, however, that “judicial leadership” or ownership of the investigation could make judges susceptible to the natural human process of confirmation bias (Roach, Reference Roach2010a). He also contrasted the inquisitorial commitment to give full reasons for decisions to the absence of reasons for the jury’s verdict to convict or acquit (Pizzi, Reference Pizzi1999: 95).

In 2008, George Thomas III wrote a book that picked up on many of the themes articulated by Pizzi but focused more directly on the risk of convicting the innocent. Like Pizzi, Thomas was concerned that in the American criminal justice system “if the defendant is ‘procedurally’ guilty, his substantive innocence is besides the point” (Thomas, Reference Thomas2008: 4). He argued: “At no stage in the American process does a judge engage in any meaningful screening of the case against the accused. Why not? The only explanation is the contest culture, where the parties fight it out before an impartial judge or referee” (Reference Thomas2008: 171). Thomas looked to France as an example where there were better pre-trial screens to determine if the accused was guilty and where professional and non-elected prosecutors and judges dominated the trial process (Reference Thomas2008: 174). Thomas argued that in France, the focus was on the discovery of the truth as opposed to the adversarial application of procedural rights and the rules of evidence.

At the same time, Professor Thomas’s (Reference Thomas2008) book ignored the Outreau affair in France, where an appeals court in 2005 overturned child abuse convictions. Subsequent inquiries in France related the miscarriages of justice to the role of an investigative judge who relied on testimony from children and psychological experts without sufficient adversarial challenge (Hodgson, Reference Hodgson2020). This, of course, demonstrates the “grass is greener” nature of the debate. Those within the American adversarial system look to inquisitorial reforms as a means to better prevent and remedy wrongful convictions often without full knowledge of the problems of inquisitorial systems. As will be seen, those within inquisitorial systems were also susceptible to the “grass is greener” effect.

Thomas, like both Pizzi and Givelber, was disappointed that the extensive resources devoted to protecting due process rights in the United States had not prevented wrongful convictions. Instead of relying on ex-post challenges through the exclusionary rule, he proposed that a judicial official should supervise police investigations including identifications, the taking of statements and the disclosure of evidence to the accused. Thomas ignored that some adversarial systems, including the English system for interrogations discussed in Chapter 3.4, placed more emphasis on legislative and executive regulation of police conduct than the American system. He also ignored that in inquisitorial systems, the supervision of police investigations by judges or quasi-judicial prosecutors did not always result in full and fair investigations. The scepticism about due process reflected a long-standing American focus on factual innocence (and its converse, factual guilt; Packer, Reference Packer1968) and dismay that the resources that the liberal Warren Court devoted to due process had not prevented wrongful convictions (Findley, Reference Findley2009a).

Like Pizzi, Thomas viewed plea bargaining as a threat to the innocent. He argued that judicial supervision could avoid “the coercion and unequal bargaining power that infects the American plea-bargaining system.” He elaborated that “not being a party, the magistrate would have no incentive to bargain. He would set a fair penalty, and if the defendant wanted to accept, he could. If the defendant does not accept the proposed penalty, the case goes to trial” (Thomas, Reference Thomas2008: 207). As will be seen, this proposal resembles the use of penal orders in Europe, where concerns have been raised that even judicial officials confirm guilt under efficiency pressures.

In 2008, Mirjan Damaska (143) wrote that the adversarial reliance on juries to determine evidence questions was challenged by the increasing “scientization of factual inquiry.” Rather than relying on adversaries, including competing experts, or amateur judges or juries, “a better way of conveying scientific data” (Damaska: 146) would be to allow a specialized investigative judge to study such complex matters. Stephen Thaman (Reference Thaman2010) similarly argued that inquisitorial systems which rely on professional fact finders who have to write reasons have an advantage over adversarial systems in dealing with increasingly complex scientific evidence.

At the same time, miscarriages of justice will occur in those cases where the safeguards of any system are at their weakest. The European Registry of Exonerations includes 26 of 126 cases where false or misleading forensic evidence played a role in a wrongful conviction. Incorrect use of statistical evidence led to the wrongful conviction of a Dutch nurse Lucia de Berk in 2003 for killing child patients in the hospital. Faulty pathology evidence about cause of death led to a 2010 wrongful murder conviction of Manfred Genditzi in Germany that was only corrected with new expert evidence in 2023. Pietro Paolo Melis was wrongfully convicted of kidnapping in Italy in 1999 on the basis of flawed voice analysis with new evidence leading to his acquittal in 2016 after 18.5 years imprisonment (European Registry). Such cases do not refute the main points made by Professors Damaska and Thaman that inquisitorial systems have the potential to better use expertise in fact-finding than adversarial systems, but they do temper them.

Michael and Lesley Risinger (Reference Risinger and Risinger2011) have proposed that accused be allowed to plead factually innocent as an alternative to not guilty. This new plea of innocence would trigger an inquisitorial-style investigation by a judicial official. In such cases, there would also be more fact-based appeals similar to revision proceedings used in inquisitorial systems. Elsewhere, Michael Risinger (Reference Risinger2004: 1314–1321) has proposed a move from a narrow appellate focus on sufficiency of evidence, which assumes that the jury is correct in drawing inferences to one based on safety. This again demonstrates how many American critics of adversarial systems have been influenced by the particularities of the American system given that England has long used the safety standard for reviewing appeals.

Like the Risingers, American law professor Tim Bakken has argued that the criminal justice system would be better at investigating the facts of the case if the accused had the alternative of pleading “innocent” as well as “not guilty.” The new plea of innocence would covert the judicial process “from an adversarial contest to an inquisitorial inquiry” (Bakken, Reference Bakken2008: 561–562). Judicial supervision of the investigation “will help to lessen the combative ethos of prosecutors and defense lawyers, who may have different views about how to conduct an innocence investigation” (Bakken, Reference Bakken2022: 162). “The government should be responsible for searching for exonerating facts rather than being satisfied with due process” (Reference Bakken2022: 3). He also expressed his dissatisfaction at the Alford plea that allows a court to accept a guilty plea from accused who maintain their innocence. Most adversarial systems, however, would also reject such pleas, which in the Alford case were made to avoid the death penalty. Bakken argues that countries such as the Netherlands, Italy and Germany have many less remedied wrongful convictions than the United States (Reference Bakken2022: 77). This is true but begs questions about whether the American system is exceptional both in its use of the criminal sanction and the legal and civil society resources it devotes to correcting and compensating wrongful convictions.

Professor Bakken was sceptical about due process and argued that those who plead innocence would have to waive their right to silence and the confidentiality of lawyer-client privilege (Bakken, Reference Bakken2008: 563). The North Carolina Innocence Inquiry Commission similarly requires those making claims of factual innocence to waive all procedural rights including attorney client privilege. Like other critics of adversarial systems, Bakken believes that inquisitorial systems would be better for determining factual guilt or factual innocence. He was willing to trade the American due process focus on legal guilt (Packer, Reference Packer1968) for an inquisitorial focus on factual truth without contemplating that the inquisitorial systems in continental Europe at least are also concerned about due process. The grass looks greener on the other side of the adversarial/inquisitorial divide because commentators cannot clearly or fully see what is on the other side. American commentators in particular assume that their system is the adversarial norm.

In a recent book, Ralph Grunewald compared the German criminal system favourably to the American system. He stresses the danger of an adversarial system in which prosecutors and defence lawyers engaged in a “regulated story telling contest” (Grunewald, Reference Grunewald2023: 93) with the jury selecting between the two stories by filling in the blanks in the stories. The jury produces a verdict that, unlike written judgments, is “de facto unreviewable” and does not attempt to ensure “narrative accuracy” (Reference Grunewald2023: 123). By contrast, professional judges in Germany are “driven by an impartial and independent investigation into the truth of a case. Factual truth is a value of constitutional magnitude. Even the concept of finality is interpreted in light of factual truth and finds its limits when a new, more accurate historical truth can be proven” (Reference Grunewald2023: 9). He concludes the German process “has a high degree of transparency and allows for review of fact and law. Judges must write out their verdicts and must explain which fact is used for which element of the crime” (Reference Grunewald2023: 150). His accounts reflected the high reliance on juries in the United States. In contrast, jury trials in Canada are quite rare and courts hold the required written reasons of judges to a higher degree of appellate review than the unexplained verdicts of juries (R. v. Lohrer, [2004] 3 S.C.R. 732). The United States may be the dominant adversary system, but it is not the only one in the world.

To his credit, Grunewald examined a famous German wrongful conviction where a wife, son-in-law and two daughters were wrongfully convicted of murdering the family’s unpopular patriarch, Rudolph Rupp, even though his body had not been discovered at the time. He noted that this wrongful conviction was based on a 1,650-page dossier and 240-page judgement (Reference Grunewald2023: 190), a far cry from the absence of reasons delivered by lay jurors in many adversarial systems. The trial lasted twenty-four days over five months and was heard by three professional and two lay judges (Reference Grunewald2023: 182). Despite this evidence of the acceptance of untrue narratives in the German system, Grunewald still maintains that the American system has a “degree of narrative latitude” that promotes “untruth storytelling” and makes wrongful convictions more likely than in Germany.

Grunewald’s analysis is provocative but can be questioned. America’s high rate of wrongful convictions should not simply be blamed on adversarial systems but on its use of mass imprisonment. The investment of resources as represented by the trial dossier is impressive, but such a sustained, singular and official narrative may result in confirmation bias and tunnel vision (Roach, Reference Roach2010a). As in the United States, an appeal in the German wrongful conviction was dismissed with the appellate court holding that any errors made in the trial were harmless (Grunewald, Reference Grunewald2023: 187). The reopening of the case might not have occurred had the car with the deceased’s body not been discovered (Reference Grunewald2023: 188). The four accused were acquitted only after a four-month new trial that did not accept that the death was accidental because the judges at the new trial “sided with the initial story as it was imagined by the officers who started the investigation.” As Grunewald concedes, this may have reflected “confirmation bias” (Reference Grunewald2023: 190–192). The arrogance of a former German judge who responded to journalistic concerns about wrongful convictions by arguing that “the likelihood that journalists of all people know what the truth and a false conviction are is similar to that of a social worker knowing the best method of cancer therapy in an individual case” is also troubling (Reference Grunewald2023: 161). As will be seen in Section 4.2.2, inquisitorial critics of their own systems have recognized similar weaknesses in the dossier-based approach. They often look to the jury’s undoubtedly freedom in reaching a verdict as a sign of a commitment to a more democratic approach to justice.

The grass is greener inquisitorial hope is not solely the preserve of American academics. The Royal Commission on Criminal Justice chaired by Viscount Runciman acknowledged that its recommendations to counter miscarriages of justice were designed “to move the system in an inquisitorial direction” (Runciman, Reference Runciman1993: 3). It recommended that trial judges should be prepared to exercise their powers to call witnesses (Reference Runciman1993: 123). This responded to concerns that defence lawyers had been unable effectively to challenge the state’s forensic evidence in some well-publicized wrongful convictions including the Birmingham Six and the Guildford Four. It also recommended the creation of what was to become the Criminal Cases Review Commission (CCRC) to play a “proactive” role in investigating alleged miscarriages of justice as opposed to the “reactive” role of the Home Secretary (Reference Runciman1993: 182).

A 2008 Canadian commission of inquiry similarly recommended the creation of a CCRC in order to bring “a non-adversarial and inquisitorial approach to the process of conviction review” (MacCallum, Reference MacCallum2008: vol. 1, 390). The commission praised the English CCRC and assumed that it had the resources to investigate all meritorious cases. It ignored evidence that applicants to the CCRC who were represented by lawyers were more likely to receive remedies than the majority who did not. There is a danger that much enthusiasm for CCRCs in adversarial systems is based on somewhat naïve views that they will have the resources to investigate every case and find the truth.

4.2.2 The Adversarial Hope

Many working within inquisitorial systems have looked to adversarial systems as a means to better prevent wrongful convictions. There are, however, some interesting exceptions. In an 1895 book, French lawyer Gaston Pean praised a recent trend in countries with inquisitorial systems to allow the correction of factual errors at trial through revision procedures tied to codified grounds for reopening convictions such as subsequent contradictory civil judgments or criminal convictions of justice participants. Writing before the creation of the English Court of Criminal Appeal, he argued that, in contrast, English law provided no remedies for errors of fact (Pean, Reference Pean1895: 291–292).

The German lawyer and academic Max Hirschberg expressed a preference for the adversarial system. He wrote that he had been “counsel in hundreds of cases in Germany, where the judge could devote only as many hours on one case, as American court devotes days or even weeks. The writer was greatly surprised, nay delighted, to find” an emphasis on cross-examination and stricter laws of evidence as a means to determine the truth (Hirschberg, Reference Hirschberg1940: 43–44). Hirschberg embraced Wigmore’s views about the importance of cross-examination to the discovery of the truth. He also stressed the need for expert witnesses to be cross-examined, stating: “the greater the self-confidence of the expert witness, the greater his arrogance … A real student of science is too well aware of the fallibility of scientific knowledge to presume infallibility, while a charlatan tries to force his infallibility on his public” (Reference Hirschberg1940: 34). As in many defences of the adversarial system, his comments assumed an equality of arms between the adversaries that is often lacking.

Many countries in South America in the 1980s and 1990s moved away from inquisitorial procedures including the use of investigative judges towards procedures that stressed more participation by the parties with a greater focus on oral hearings and the rights of the accused (Brooks, Reference Brooks2012: 1119). As Maximo Langer (Reference Langer2007) has documented, the movement was made for a number of reasons including to augment transitions to democracy, increase efficiency and decrease corruption. These transitions were not done with the aim of preventing wrongful convictions.

A Chilean law professor stressed the colonial and authoritarian roots of that country’s inquisitorial system. Evidence was gathered without defence input in a “secret investigation” in “a secret pretrial phase” (Cousino, Reference Cousino1998; 325). He argued that the bureaucratic nature of inquisitorial systems and its tradition of mandatory prosecution deny “the controversial meaning of criminal enforcement” (Reference Cousino1998: 331). Similarly, an Argentinian law professor, Julio Maier, criticized the role of the investigative judge as an impossible combination of prosecutorial and adjudicative functions. He stressed the need to move away from mandatory pre-trial detention and mandatory prosecution and towards more open trials as Argentina transitioned to democracy (Langer, Reference Langer2007: 637–640).

At the same time, the implementation of the adversarial hope in South America is often dependent on the development of effective defence lawyering, something that has proven difficult. After examining the failures of defence lawyers and rising prison populations in Chile as it adopted more adversarial measures, one commentator has concluded that “adversarialism without meaningful defense lawyering can be as inhumane and lead to as many injustices as the now rejected inquisitorial systems” (King, Reference King2018: 893). This is a relatively rare recognition in the theoretical literature comparing adversarial and inquisitorial systems that each system has its own weaknesses. Similarly, Mauricio Duce (Reference Duce2015) has noted that while Chilean law reform was designed to increase the importance of the oral trial, the time actually devoted to contested oral trials is generally inadequate to ensure accuracy. He also notes that many wrongful convictions happen through the use of summary procedures.

Martin Killias (Reference Killias and Killias2008: 141) has observed that in Switzerland, defence lawyers “do not play a very active role in the factual side of the case.” This was illustrated in one case where a Turkish immigrant was wrongfully convicted of possessing 1 kg of heroin. Lab results indicating that the substance had less than 1 per cent heroin were not disclosed until after the accused had served most of his six-and-a-half-year sentence. He also noted that in 230 successful revision cases between 1994 and 2004 on behalf of the accused, 159 cases involved penal orders in minor cases that the accused did not initially contest. Indeed, 129 of the overturned convictions involved traffic offences and in only 42 cases was the sentence longer than six months (Reference Killias and Killias2008: 150–151). Professor Killias concluded that his study demonstrates “how vulnerable decisions are whenever defendants or other key figures are not adequately heard … when justice becomes ‘too summary’” (Reference Killias and Killias2008: 151). Based on the same study, Gillieron (Reference Gillieron2012: 243) argued that “wrongful conviction by penal order mainly concerned factually innocent defendants” and were often based on the “confusion of names” and inadequate investigations. Similar findings were made in a subsequent study in Chile (Duce, Reference Duce2015). Writing from an inquisitorial tradition, all of these authors stress that their findings confirm the need for more adversarial procedures. At the same time, many observers from adversarial systems might marvel at the fact that wrongful convictions for such minor crimes were even corrected at all. In both the United States and the UK, it often takes a major scandal such as the Post Office scandal to correct wrongful convictions for minor crimes.

Chrisje Brants (Reference Brants2012: 1079), a leading Dutch scholar of wrongful convictions, has argued that “from a scientific point of view, the presentation of and attempt to falsify two versions of the events” is better at discovering truth than “verification of the prosecutor’s version by the judge” in inquisitorial systems. She notes that “compared to adversary procedures, a Dutch trial is short, document-based and with limited debate in court” (Brants, Reference Brants, Huff and Killias2013: 168). The accused’s lawyer cannot call expert witnesses or any witnesses at all (Reference Brants, Huff and Killias2013: 1083) and only receives the dossier ten days before trial (Brants, Reference Brants, Huff and Killias2013: 169). Judges may be subject to confirmation bias based on the prosecutor’s file and the opinion of experts called by the judge (Brants, Reference Brants, Huff and Killias2008: 172). Such confirmation bias is also fostered by traditions of short judgments including on appeal (Reference Brants, Huff and Killias2008: 174). Professor Brants found all of these features to be a factor in four high-profile wrongful convictions in the Netherlands (Brants, Reference Brants, Huff and Killias2013: 178). In three of these cases, false confessions were obtained and “figured prominently in the court’s reasoning” even though they were quickly retracted by the accused (Reference Brants, Huff and Killias2013: 180).

Brants (Reference Brants, Huff and Killias2008: 168–169) has expressed concerns that inquisitorial procedures assume that police, prosecutors and judges are all engaged in “nonpartisan gathering of evidence” and adhere to “professional ethics.” But such assumptions may no longer be valid in an era when trust in state actors and “professional elites” is declining. She has warned that “professionalism has its darker side in peer solidarity and routine” with Dutch judges valuing collegiality so that they rarely dissent (Brants, Reference Brants, Huff and Killias2008: 178; Brants, Reference Brants, Huff and Killias2013: 182). Brants also noted that a commission created to investigate possible wrongful convictions in the Netherlands was placed within the prosecutor’s office as opposed to being the sort of independent commission used in a number of countries with adversarial systems (Brants, Reference Brants2012: 1106–1108).

Somewhat similar to Brants, German law professor Carsten Momsen criticized the lack of the “checks and balances” of a strong adversarial defence and the unwillingness of German prosecutors and judges to admit error (Momsen, Reference Momsen2023: 243). Although convictions can be reopened on the basis of new facts, this imposed a high burden on often impoverished convicted persons. In addition, such challenges were often resisted by what he calls the “inquisitorial authoritarian” (Reference Momsen2023: 241) culture of German criminal justice. Momsen looks with admiration to innocence organizations and conviction integrity units that exist in the United States and some other countries that use an adversarial system. His analysis is especially helpful in going beyond an exclusive focus on legal procedures and raising concerns about what he sees as a lack of comparative awareness of wrongful convictions in Europe as opposed to North America. At the same time, the increased American awareness of wrongful convictions may in part be related to the greater use of the criminal sanction in the United States than in Europe. In 2024, the United States imprisoned 531 people per 100,000 for a total prison population of 1.7 million compared to Germany, which imprisoned 67 per 100,000 for a total prison population of about 56,000 people (World Population Review).

4.2.3 Summary

The “grass is greener” debate both idealizes and dichotomizes two systems that have much in common (Hodgson, Reference Hodgson2020). This is a special danger in the wrongful conviction context. Even the highest estimates of the prevalence of wrongful convictions suggest that they only occur in a small minority of cases. Even if “pure” adversarial or inquisitorial systems existed, each would be liable to breakdowns that could result in wrongful convictions. Adversarial systems will always be vulnerable to an inequality of arms between the accused and the state and to ineffective or non-existent assistance of defence counsel. Inquisitorial systems will always be vulnerable to the dangers of confirmation bias or tunnel vision. The idea of the court selecting its own experts may seem appealing but could result in the court selecting the sort of expert witnesses that have been responsible for multiple wrongful convictions. Grass is greener debates can, however, be illuminating in revealing some adversarial and inquisitorial features that may be susceptible to error and some reforms influenced by the other system to counter such weaknesses. But first, it is necessary to dispel the somewhat simplistic binaries created by the “grass is greener” debate by examining some differences between inquisitorial systems. Within the broad range of countries that use inquisitorial procedures, some may be better at preventing and remedying wrongful convictions than others.

4.3.1 Sweden

As in many inquisitorial systems, Sweden provides a codification of the reasons for reopening convictions. Chapter 58, Section 2, of the Code of Judicial Procedure places the onus on those seeking to overturn convictions to establish their eligibility under one of five codified grounds, the most general one being that there is “a circumstance or item of evidence that was not presented previously is invoked and its presentation probably would have led to the defendant’s acquittal or that the offence would have been linked to a sanction provision milder than that applied…” (emphasis added). In both adversarial and inquisitorial traditions, the discovery of new evidence is often an essential step in correcting a wrongful conviction. Minna Grans (Reference Grans and Robins2023: 89–90) has noted that “the threshold of ‘probably’” in the above provision “is extremely high” and relates this to a concern about maintaining the finality of convictions.

Sara Hellqvist (Reference Hellqvist2021: 328) found that only 6 of 212 applications made in Sweden in 2015 to reopen convictions were successful. Three of the successful applications were made by individuals, but three were made by prosecutors and were based on legal errors (Reference Hellqvist2021: 330). Most of the applications were made without legal advice and simply asserted the applicant’s innocence with 37 of the 212 applications being incomprehensible (Reference Hellqvist2021: 334). Only 28 per cent of the applications included any reference to new evidence or new circumstances. Although inquisitorial systems may often seem to be more advanced than common law adversarial countries by codifying specific grounds for reopening convictions and thus officially recognizing the possibility of error, this research underlines the differences between having a legal right to reopen a conviction and actually being able to realize such rights.

Many of the wrongful convictions that have been corrected in Sweden were initially uncovered by investigative journalists (Martinsson, Reference Martinsson2020). In one case, a wrongfully convicted man had to present three applications for revision to the Supreme Court and serve thirteen years in prison before the Court reluctantly concluded that while each application itself was insufficient, their cumulative effect was sufficient to require a new trial, which resulted in an acquittal. What seemed to push the case towards successful revision was when one of the witnesses contradicted his testimony at trial on a true-crime podcast (Reference Hellqvist2021: 96; European Registry: Kaj Linna).

Between 2000 and 2020, only forty-five applications in Sweden to reopen convictions were granted with all of the successful applications either resulting in acquittals at new trials or no prosecutions (Grans, Reference Grans and Robins2023: 94–95). This reveals that applications are only granted in cases of clear miscarriages of justice. The forty-five successful applications were mostly for serious offences including fifteen murder cases and nine sexual offences cases. Grans concludes that these cases “may be the tip of the iceberg, even if it is impossible to make any definitive statements about the total number of miscarriages of justice” (Grans, Reference Grans and Robins2023: 96). Sweden allows convictions to be reopened, but convicted individuals are generally left to their own devices in finding new evidence that if considered at trial would probably have led to their acquittal.

4.3.2 Norway

Section 391 of the Norwegian Code of Criminal Procedure is similar to the above Swedish procedure with respect to the grounds for reopening a conviction. A conviction can also be reopened on the basis of a new decision by international courts including with respect to procedural error. This embraces a concern about due process and a broader concern about miscarriages of justice and compliance with European Court of Human Rights decisions that is often discounted by American champions of inquisitorial reforms who imagine Europe as a place more concerned with truth and less concerned with rights than the United States.

The critical new evidence provision is similar to the Swedish standard because it sets a high standard by requiring that “a new circumstance” or “new evidence” to be presented “which seems likely to lead to an acquittal or summary dismissal of the case or to the application of a more lenient penal provision or a substantially more lenient sanction” (Stridbeck and Magnussen, Reference Stridbeck and Magnussen2012). Norway, like Sweden, demands that new evidence sufficient to reopen a conviction probably would change the result of the first trial.

But the law on the books does not always represent the law in action. In response to a high-profile murder wrongful conviction in the Liland case and a subsequent report on the case, Norway created a Criminal Cases Review Commission that came into operation in 2004. It has powers to order the production of evidence and commission new expert evidence. It also has access to national data banks containing fingerprints and DNA. It can appoint lawyers for both applicants and affected crime victims. It receives about 170 applications each year and has reopened about 18 per cent of these applications with the convictions being overturned in about 80 per cent of all cases referred (Norway Criminal Cases Review Commission Annual Report, 2021: 1; Stridbeck and Brennan, Reference Stridbeck, Brennan and Robins2023: 75). This has resulted in a far higher number of overturned convictions than in Sweden where only forty-five convictions were successfully overturned between 2000 and 2020 (Grans, Reference Grans and Robins2023: 94–95). Thus, similar grounds for reopening convictions in the Swedish and Norwegian codes result in many more convictions being successfully reopened in Norway. The creation of a well and publicly funded commission has increased access to justice for the wrongfully convicted in Norway as compared to neighbouring Sweden.

At the same time, Norway is not immune from long waits for exoneration. Viggo Kristianson was convicted in 2001 of murder and rape of two young girls mainly on the basis of the testimony of an alleged accomplice. Kristianson unsuccessfully attempted six times to reverse his conviction through the courts. It was only after the Norwegian CCRC referred his case in 2021 that his conviction was overturned, and he was acquitted on the basis of DNA evidence and new evidence related to alibi (European Registry: Viggo Kristianson).

It would be a mistake to conclude that the cases revealed by the Norwegian commission are similar to cases revealed by other commissions. They are shaped by Norway’s distinct but perhaps changing legal culture. About 40 per cent of cases reopened in Norway by the commission have involved new psychiatric evidence often related to applicants who are factually guilty but may use the new evidence to claim a mental disorder defence (Stridbeck and Brennan, Reference Stridbeck, Brennan and Robins2023: 73, 79; Stridbeck and Magnussen, Reference Stridbeck and Magnussen2012: 1382–1383). Some concerns have been raised whether this represents the best allocation of limited investigative resources (Grondhal and Stridbeck, Reference Grondahl and Stridbeck2016). In any event, this pattern of cases may change given Norway’s 2020 restrictions on the mental disorder defence in response to concerns that mass murderer Anders Brevik might have benefitted from the previous generous defence.

4.3.3 Summary: Differences between Sweden and Norway

The differences between the Swedish and Norwegian systems reveal some practical access to justice barriers to fact-based revision procedures which appear to be an inquisitorial advantage over the law-based appeal and post-conviction procedures found in adversarial systems. Even though many inquisitorial systems are prepared to hold new trials in the face of new evidence, it is practically difficult for convicted people to find new evidence.

Thaman, Reference Thaman2019: 944 has observed that the “classic form of appeal” in inquisitorial systems “in which the appeals courts rehears the first instance evidence and takes new evidence … is rapidly disappearing” with an increasing focus on legal error and the transcript of the first trial. In this way, inquisitorial advantages with respect to fact-based appeals are diminished by increasing legal and practical burdens on the convicted person to demonstrate error at the first trial.

This comparison between Sweden and Norway has stressed what is particular about Norway’s criminal case review commission. It has taken an institutional legal process approach, which stresses the important role of Norway’s state funded commission in reopening convictions compared to reliance on private resources and the media in Sweden to find new evidence that will probably result in overturning a conviction.

4.3.4 China

China’s 1996 Criminal Procedure Law (CPL) was praised as a move in an adversarial direction that restricted administrative and investigative detention and limited the pre-trial role of investigating judges. Mike McConville’s (Reference McConville2011: ch 6) empirical research, however, suggests that there remained considerable “pre-trial activism” by Chinese judges who also heard the case.

In 2018, Article 15 of China’s Criminal Procedure Code was amended to provide that when a criminal suspect or defendant voluntarily and truthfully confesses to his or her crime, admits to the facts of the crime that he or she is charged with, and is willing to accept punishment, the criminal suspects or defendant “may be granted leniency in accordance with the law.” There is a focus on efficiency under this new system. Pleas are supposed to be accepted by a single judge within fifteen days after the prosecutor applies for such a procedure (Lu, Lin and Li, Reference Lu, Lin, Li, Li, Yuan and Zhang2024: 19). Guilty pleas are currently restricted to cases involving three years’ imprisonment or less (Reference Lu, Lin, Li, Li, Yuan and Zhang2024: 28).

Professor Liling Yue (Reference Yue2021: 26–27) of the China University of Law and Political Science notes that in the first six months of 2019, guilty pleas accounted for 80 per cent of cases approaching the high levels of guilty pleas in the United States. At the same time, she suggests that the Chinese system is similar to the German because the accused must admit guilt to all offences and not engage in charge bargaining. She also notes that while defence lawyers can advise their clients with respect to guilty pleas, they cannot examine the evidence or case files or even attend the critical negotiations between the accused (generally held in pre-trial detention) and the prosecutor. It is not difficult to see false guilty pleas by people who are innocent or have valid defences emerging from such a coercive and one-sided plea bargaining process. The danger that plea bargaining has been transplanted in China from adversarial systems, minus defence lawyers, will be examined in greater depth in Chapter 9.6.

Lu, Lin and Li (Reference Lu, Lin, Li, Li, Yuan and Zhang2024: 13–14) have more recently found that plea bargaining rates in China have increased to 90 per cent with courts approving the prosecutor’s suggested sentence in over 98 per cent of cases. They concluded that the 2018 amendments have converted prosecutors into de facto judges. Moreover, “plea leniency exhibits no elements of a ‘bargain’ because it is based on a ‘one-sided consensus’ imposed solely by ‘procuratorates’ or prosecutors who generally does not allow charge bargaining” (Reference Lu, Lin, Li, Li, Yuan and Zhang2024: 20).

Professor Yu Grace Mou of the School of Oriental and African Studies in London notes that 86% of police officers, 73% of judges and 68% of prosecutors surveyed in China believe that lower standards of proof can be used when the accused pleads guilty and that “the idea that the abbreviated procedures do not require strong evidence has been firmly entrenched in practice” (Mou, Reference Mou2020: 142). She observed the following “plea bargaining” session in the course of her observational studies:

A senior prosecutor told Mou that one in three prosecution cases would be dropped if guilty pleas were not available, and this was confirmed in her own study of sixty-four cases in which 25 per cent of the cases involved “hopelessly weak” cases but all resulted in guilty pleas (Reference Mou2020: 156). If China has any inquisitorial advantage with respect to false guilty pleas, it is only that the 2018 procedure is limited to cases involving less than three years’ imprisonment.

4.3.5 Taiwan

Taiwan amended its Code of Criminal Procedure to introduce plea bargaining in 2004, fourteen years earlier than China. As in China, plea bargaining is restricted to less serious cases that in Taiwan involve two years’ imprisonment or less. Article 451-1 of Taiwan’s CPL also empowers judges to reject plea bargains on a variety of grounds including lack of voluntariness and factual inaccuracy. Chin and Lin (Reference Lin, Li, Yuan and Zhang2024: 139) have argued that Taiwan’s inquisitorial tradition “grants judges exclusive control over case narratives” and helps explain cultural resistance to plea bargains in which “many still adhere to the intuitive pursuit of objective truth. Greater consensus among legal practitioners will be necessary for the plea bargain system to become more well-known and accepted by the general public” (Reference Lin, Li, Yuan and Zhang2024: 149).

Mao-hong Lin (Reference Lin, Li, Yuan and Zhang2024: 161–162) reports that from 2017 to 2021 only 2.54 per cent of all cases in Taiwan were resolved by plea bargains (Lin, Reference Lin, Li, Yuan and Zhang2024: 133). Prosecutorial resistance to plea bargaining is based on a belief that “justice is about truth, not compromise” as well as requirements that victims be consulted and senior prosecutors approve before a plea bargain is proposed explain the low rates of plea bargaining in Taiwan as compared to China.

4.3.6 Summary: Differences between China and Taiwan

Enshen Li, Xiaoyo Yuan and Yan Zhang (Reference Li, Yuan and Zhang2024) have explained the important difference between China and Taiwan with respect to plea bargaining on the basis of Taiwan’s commitment to a judge-centred inquisitorial system that dates back to its occupation by Japan. They label this approach “judicial paternalism.” On the other hand, the success of plea bargaining in China since 2018 is related to the need for efficiency in its high volume system, quotas placed on prosecutors to resolve cases quickly and a lack of strong defence lawyers. The differences between Taiwan and China’s approach to wrongful convictions will be discussed in greater depth in Chapter 9, but this comparison of the two countries underlines important differences despite their common use of inquisitorial systems. With respect to guilty pleas, Taiwan’s system seems more traditionally inquisitorial than China’s in its continued suspicion and low use of plea bargaining resulting in guilty pleas.

4.4.1 Germany

Professional and lay judges often sit together in Germany. Judges examine the witnesses and appoint expert witnesses, but other parties also can ask questions. There are broad rights of appeals which generally can be in the form of retrials (Weigend, Reference Weigend and Bradley2007). Civilian systems have generally been more inclined to grant wider rights of appeal to both the accused and the prosecutor as of right than the common law system, and this has been related to their greater commitment to the truth (Djukic, Reference Djukic2018; Marshall, Reference Marshall2011).

Following civilian traditions, Article 359 of the German Code of Criminal Procedure codifies the basis for reopening a conviction for the accused’s benefit. It provides the following:

The reopening of proceedings concluded by a final judgment shall be admissible for the convicted person’s benefit

1. 1. if a document produced as genuine, to his detriment, at the main hearing was false or forged;

2. 2. if a witness or expert, when giving testimony or an opinion to the convicted person’s detriment, was guilty of intentional or negligent breach of the duty imposed by the oath or of intentionally making a false, unsworn statement;

3. 3. if a judge or lay judge who participated in reaching the judgment was guilty of a culpable breach of his official duties in relation to the case, unless the violation was caused by the convicted person himself;

4. 4. if a civil court judgment on which the criminal judgment is based is quashed by another final judgment;

5. 5. if new facts or evidence were produced which, independently or in connection with the evidence previously taken, tend to support the defendant’s acquittal or, upon application of a more lenient criminal provision, a lesser penalty or a fundamentally different decision on a measure of reform and prevention; and

6. 6. if the European Court of Human Rights has held that there has been a violation of the European Convention on the Protection of Human Rights and Fundamental Freedoms or of its Protocols and the judgment was based on that violation.

The traditional grounds for reopening convictions in Article 359(1)–(4) are based on proven fault of criminal justice actors or inconsistent civil or criminal verdicts. These grounds are not widely used but should not be deprecated because they allow for a convicted person to reopen a conviction by launching a successful collateral attack on the conviction. This has been attempted in adversarial systems but has been shut down on vague abuse of process grounds most infamously in the case of the Birmingham Six (Hunter v. Chief Constable of the West Midlands Police [1981] UKHL 13). The focus in many inquisitorial codes on responding to inconsistent judgments underlines their commitment to truth and consistency with the values of the law (Damaska, Reference Damaska1986). It also demonstrates a commitment to the potential civil or criminal accountability of justice system participants who are at fault for causing wrongful conviction. This accountability is often lacking in adversarial systems.

The new facts provision in Article 359(5) rejects the American focus on proven innocence because it relates to any new facts and evidence that would support either the accused’s acquittal or a more lenient offence or punishment. This is a broader concept of miscarriage of justice than is the norm in the United States and may be related to a less punitive approach to criminal justice in Germany and a greater concern with rehabilitation.

Article 359(6) also contemplates a commitment to a broader understanding of miscarriage of justice by allowing convictions to be reopened because of a European Court of Human Rights decisions. It seems to contemplate fully retroactive justice in applying new human rights precedents. In contrast, North American courts frequently will apply human rights precedents with prospective effects only because of the concerns that retroactive application will cause too much disruption to the finality of convictions (Linkletter v. Walker 381 U.S. 618 (1965); Stovall v. Denno 388 U.S. 293 (1967); R. v. Sarson [1996] 2 SCR 223).

Article 359(5) sets a slightly lower standard by only requiring new evidence that “tend to support the defendant’s acquittal” compared to the Danish and Swedish Codes examined earlier that require a probability that the new evidence would lead to an acquittal. As Ralph Gruenewald (Reference Grunewald2014: 1150) has argued, Article 359 of the German Code of Criminal Procedure demonstrates a greater commitment to truth by allowing new evidence to be introduced compared to the reluctance of the United States Supreme Court in Herrera v. Collins, 506 U.S. 390 (1993) to recognize even a freestanding claim of innocence to be proven by clear and convincing evidence or possibly even a higher standard.

Although wrongful convictions have not captured the public imagination, there is a long tradition of research into them in Germany. As discussed in Chapter 2, Eric Sello and Max Hirschberg, both wrote treatises in German on wrongful convictions or what Hirschberg called the pathology of criminal justice before Edwin Borchard published a similar book in the United States cataloguing wrongful convictions (Borchard, Reference Borchard1932). Professor Peters published research into wrongful convictions in the early 1970s before the explosion of writing about miscarriages of justices in adversarial systems starting in the 1990s. Peters’ research confirmed that many of the same immediate causes of wrongful convictions found in adversarial systems also occur in Germany. Other studies of retrials between 1951 and 1964 in Western Germany found false witness statements in 22.7% of successful retrials, false confessions in 7%, and expert error in 6.6% of cases (Bliesener et al., Reference Bliesener2023: 148). A 2019 study of thirty-one successful retrials found false witness statements to be the leading cause followed by faulty expert evidence and false confessions (Reference Bliesener2023: 149).

A study of 512 retrials in Germany between 2013 and 2015 found a 40 per cent success rate with a quarter of the successful cases involving mistaken eyewitness identifications and a quarter involving false witness statements. Over 55% of these cases involve property crimes and 10% involved traffic offences with violent crimes only constituting about 20% and homicide cases less than 6% of the cases where convictions were overturned and retrials held (Reference Grunewald2023: 152). Retrials supported by the prosecutor’s office had almost a 92% success rate compared to a 40% success rate for retrials initiated by the defence (Reference Grunewald2023: 153). This suggests that the prosecutor as a judicial official still carries weight in Germany. A large percentage of successful cases (42%) involved errors in calculating sentences with only about 38% of successful cases involving factual errors. Even these factual errors often involve state errors in ascertaining the correct identity of the accused or whether they had a required license (Reference Grunewald2023: 156–157). Similar findings have been found in successful retrials in Chile (Duce, Reference Duce2015) and Switzerland (Gillieron, Reference Gillieron2012). This reveals concerns about factual accuracy in minor cases often lacking in adversarial systems. It may also reflect penal culture where limited resources devoted to correcting wrongful convictions can be allocated to less serious cases.

Another German study identified only twenty-nine wrongful convictions that were corrected in Germany between 1990 and 2016. It was limited to those who had served some part of prison sentences and were subsequently acquitted at a retrial. The divergence between the numbers in this study and the much higher numbers found in studies of all retrials reveals a factor often ignored in the study of wrongful convictions: the use of long prison sentences likely increases the number of remedied wrongful convictions by increasing both the incentive and practical time for the wrongfully convicted to seek remedies. The twenty-nine wrongful convictions where the accused had served prison time involved twelve cases of sexual offences, eleven cases of other forms of severe violence including eight homicide cases but also included five property offences (Leuschner, Rettenberg and Dessecker, Reference Leuschner, Rettenberg and Dessecker2020: 697).

In eight of the twenty-nine cases, the wrongfully convicted person committed the criminal act but was subsequently held not to be responsible because of mental disorder (Reference Leuschner, Rettenberg and Dessecker2020: 701). As in Norway, this demonstrates a broader understanding of miscarriage of justice than is not limited to factual innocence. This reflects a less punitive penal culture than in the United States. In six of the eight cases, the court did not ask for an expert opinion on mental disorder at the original trial (Reference Leuschner, Rettenberg and Dessecker2020). This underlines that inquisitorial systems are not foolproof if responsible officials fail to conduct relevant investigations (Reference Leuschner, Rettenberg and Dessecker2020: 698). Those eventually held not to be criminally responsible because of mental disorder had served on average forty months in prison before their conviction was overturned (Reference Leuschner, Rettenberg and Dessecker2020: 701). As discussed earlier, the Norwegian CCRC has discovered even more cases where the accused was convicted but was subsequently on new evidence found not to be criminally responsible because of mental disorder. This suggests that the German revision procedure may suffer from access to justice problems as discussed earlier with respect to Sweden.

As will be discussed in Chapter 8.8 on the role of gender, wrongful convictions in sexual assault cases raise many complex and sensitive issues. The German study found twelve cases of false accusations against men in sexual cases with the female complainants subsequently admitting that they had lied. In seven other cases, the case was reopened because of the complainant’s subsequent behaviour (Reference Leuschner, Rettenberg and Dessecker2020: 699). One case involved Ralph Witte, who was imprisoned for five years before expert evidence was used to overturn his sexual assault conviction on the basis that the young complainant had a personality disorder. Another case involved Horst Arnold also imprisoned for five years before his conviction was overturned on the basis of subsequent lies and false accusations made by the complainant, a work colleague (Leuschner and Rettenberg, Reference Leuschner, Rettenberg and Robins2023: 37). Sexual assault wrongful convictions involving young complainants also frequently involved expert evidence at trial that the complainant was credible that was subsequently found to be faulty (Reference Leuschner, Rettenberg and Robins2023: 700). Adversarial systems may have been less willing to allow experts vouch for the credibility of young complainants but may also have deferred to the original conclusion at trial that the complainant was credible.

Data from the European Registry of Exonerations reveals thirty-eight wrongful convictions from Germany as of the end of 2024. Some of the cases seem to resemble false guilty pleas. For example, a couple, Michael and Karolina Lubitz admitted to sexual abuse of their foster child on the apparent understanding that they would receive a one-year sentence that would not be served in prison and allow one of them to keep a civil service job. Because sentences of less than a year are generally suspended in Germany, the couple served no time in prison before their wrongful convictions were corrected (European Registry, Michael and Karolina Lubitz). This is a reminder that wrongful convictions are influenced by the broader penal culture.

Of the thirty-eight cases, twelve involved false confessions including to serious crimes such as murder and rape. Six of the thirty-eight cases involved flawed forensic evidence. One of these cases involved an expert who identified Donald Stellwag as the suspect in a video of a bank robbery. Stellwag was only exonerated after six years in prison when someone else confessed to the robbery. It remains an open and important question whether well-resourced inquisitorial systems can outperform adversarial systems with respect to complex forensic and machine-based evidence including video identifications. Well-resourced litigants in adversarial systems may be able successfully to challenge such evidence but inquisitorial prosecutors and judges who are well-trained about the frailties of such evidence could have wider benefits. In any event, errors in identification of video evidence are a new and likely significant immediate cause of wrongful convictions in both adversarial and inquisitorial systems (Edmond and Vuille, Reference Edmond and Vuille2014).

Carsten Momsen (Reference Momsen2023: 242) has recounted how juries in Germany were eliminated in 1924 in part because of a 1919 wrongful murder conviction and because of concerns that the jury’s verdict was not fully documented. At the same time, he argued the judge’s verdict in Germany today is also not fully documented. This is especially the case with respect to video evidence that judges frequently rely upon (Reference Momsen2023: 247). His criticism regarding the lack of documentation of factual reasoning in German cases suggests that the advantages of reasoned judgments might be overstated, at least in cases of wrongful convictions.

4.4.2 Italy

Since 1988, Italy has added adversarial elements to its criminal justice system. Reliance on investigative judges who may be particularly susceptible to confirmation bias has been reduced. Serious offences are tried before juries of two judges and six lay persons. The defence has rights to insist that evidence be called and that it be subject to cross-examination. At the same time, Italy has retained broad rights of appeals in the form of retrials typical of many inquisitorial systems.

Article 630 of the Italian Code of Criminal Procedure, as is the norm in civilian systems, sets out grounds for fact-based revisions as follows.

The revision request may be made either by the convicted person, their next of kin, or the prosecutor. Articles 630(a–b) and (d) follow civilian traditions in basing revision on inconsistent judgments and false documents or statements. Article 630(c) provides for revision on the basis of new evidence that “proves that convicted person” should have had charges dismissed. This is a higher standard than a similar provision used in Germany and comes quite close to a restrictive proven innocence standard. In practice, the request for revision is made to an appellate court of professional judges that will weigh the novum or new evidence “in terms of reliability, persuasiveness and congruence against the evidence already assessed in the previous judgment.” Between 1992 and 1999, the Italian courts were conflicted about what constitutes new evidence under Article 630(c), but they now appear to have returned to a generous definition that includes any evidence not previously considered by the court even if it existed or was even collected at the time of the original trial (Luparia, Cagossi and Pittiruti, Reference Luparia, Cagossi, Pittriruti and Robins2023: 110, 113). At the same time, Article 637(3) provides that “The Court shall not order the dismissal solely on the basis of a different evaluation of the evidence gathered in the previous trial.”

The Constitutional Court has added to this a revision request when necessary to comply with a judgment of the ECtHR. As in Germany, this reflects a commitment to a broader concept of miscarriages of justice that includes unfair trials. Rights violations are also relevant in American post-conviction habeas corpus relief but as will be seen in Chapter 6.5, applicants to be successful in some cases have to establish both a rights violation and their innocence. Many of the American “grass is greener” commentators examined earlier are mistaken in believing that inquisitorial systems prefer truth over rights. In continental Europe at least, both matter.

Italy is a leader in compensating miscarriages of justice with Article 24 of the Italian Constitution providing for compensation for judicial errors. Between 1991 and 2021, 214 cases of wrongful convictions received compensation in Italy for total payments of 76 million euros (Reference Luparia, Cagossi, Pittriruti and Robins2023: 116). To keep this in perspective, during the same period over 27,000 people received compensation in Italy for wrongful pre-trial detention (Lonati, Reference Lonati, Jasinki and Kremens2023). This again demonstrates Italy’s commitment to a broader meaning of miscarriage of justice not limited to wrongful convictions.

The European registry has eighteen wrongful convictions from Italy as of the end of 2024. They include the case of Domenico Morrone, who was wrongfully convicted of killing two boys and had four attempts to reopen his case rejected by the courts before the true perpetrators confessed. Morrone was convicted despite alibi evidence and despite being apprehended by the police twenty minutes after the shooting even though he did not match the description given by the eyewitnesses. Maurizo Bovo also failed in four attempts at revision before his murder wrongful conviction was corrected after the real perpetrator had confessed. Daniele Brilla was wrongfully convicted of drug offences and imprisoned for almost eight years before his conviction was overturned and he eventually received 4 million euros in compensation.

Italy has fairly elaborate right to counsel and self-incrimination warnings and a more or less absolute exclusionary rule. Nevertheless, false confessions, including in the Amanda Knox case, are still possible. As in the United States, there is a difference between the legal right to have defence counsel and the practical availability of defence counsel. There have been cases where false confessions have been made by suspects who are mentally ill, who are confessing to save others and who are pressured by the police (Luparia, Cagossi and Pittiruti, Reference Luparia, Cagossi, Pittriruti and Robins2023: 105).

The Italian Code of Criminal Procedure makes identification by witnesses a judicial responsibility but unfortunately regulates such procedures in an unsatisfactory manner requiring only two foils or people who resemble the suspect. Moreover, witnesses are allowed to identify the accused in court even though “the suspect is standing alone, next to their lawyer, on the defendant’s bench, or worse, in a cage” (Reference Luparia, Cagossi, Pittriruti and Robins2023: 108). There are a number of remedied wrongful convictions in Italy that involve flawed eyewitness identification. Angelo Cirro was wrongfully convicted of robbery in 2005 on the basis of mistaken eyewitness identification, even though his DNA was excluded from a cigarette found at the scene. His conviction was only overturned when the DNA was traced to another person who confessed to the robbery.

The Amanda Knox case has perhaps generated more heat than light. Knox and her companion were convicted in 2009 of murder. The trial court issued a 419-page decision but also used character evidence not generally admissible in adversarial systems. The court held confession evidence was inadmissible in the criminal trial because of the pressure placed on Amanda Knox during the interrogation. At the same time, this evidence was still known to the court that convicted Knox because her murder case was tried together with civil lawsuits by a person who claimed he was libelled by Knox’s false confession (Mirabella, Reference Mirabella2012). The ability of crime victims to participate as parties in a number of inquisitorial systems and its effects on the risk of wrongful conviction deserve additional study.

Knox’s conviction was overturned by a second appeal in 2011, which engaged in a searching factual inquiry that would not be available in adversarial systems. The appeal court reheard testimony from witnesses including from an unhoused person who said he observed Knox near the scene at the time of the murder. After rehearing this evidence, the appeal court concluded that the trial court had erred and that if the witness had seen Knox, it was on a different day than the day of the murder. The appeal court also found unreliable the testimony of another witness who belatedly went to the police to say that Knox had tried to purchase cleaning materials the day after the murder. The appeal court also determined that the trial court had erred in determining Meredith Kercher’s time of death. Appeal courts in adversarial systems may well have deferred to all of these factual findings at trial. The appeal court also reheard evidence from DNA experts and commissioned new DNA testing that revealed that a DNA match used at the first conviction may have been the result of contamination (Lenthe, Reference Lenthe2013). Amanda Knox’s successful first appeal reveals the willingness of appellate courts in many inquisitorial systems to take a harder and closer look at the facts than appellate courts in many adversarial systems.

Knox’s first successful appeal was not the end of the story. The prosecutor was able to successfully appeal the acquittal from the first appeal. The prosecutor’s ability to appeal acquittals tends to be more generous in inquisitorial systems than adversarial systems and especially the American system. This may be related both to an increased commitment to truth and interpretation of equality of arms between prosecution and the defence that does not embrace the asymmetrical standards of the presumption of innocence. The American press was extremely critical of this second appellate decision while generally not taking note that Knox’s first fact-based appeal would not have been possible in the United States (Lenthe, Reference Lenthe2013; Mirabella, Reference Mirabella2012).

Knox was convicted again by a jury at a retrial in 2014 but acquitted by an appellate court in 2015. The ultimate appellate ruling acquitting Knox also found factual errors that would be difficult for an adversarial appeal to reverse. These included findings that rejected trial findings that Knox had a motive to commit the murder, errors about the time of death, errors in DNA analysis and concerns about the illogical nature of the prosecutor’s case, as well as new evidence that Kercher’s stolen phone was found along a route taken by another person convicted of her murder.

4.4.3 France

In France, most cases are tried by judges, but the most serious cases are tried by three judges and six lay persons who deliberate together in the cour d’assisses. Appeals are generally heard by way of new trials. Investigating judges are still used for the most serious cases, and prosecutors are also considered judicial officials trained at the national judicial school. This has led to some confusion of prosecutorial and judicial roles. One juge d’instruction admitted to confusion about whether he or she was “a judge, investigator or partner of the police” (Hodgson, Reference Hodgson2020: 127). In the Outreau case, over sixty magistrats were involved but none of them were able to check the work of the other to stop the wrongful convictions of child sexual abuse (Reference Hodgson2020: 131).

Following civilian traditions, there have been attempts to predict and codify reasons in French law for reopening convictions and applying for new trials since 1808 (Etrillard and Rossant, Reference Etrillard, Rossant and Robins2023: 21–22). These original grounds included that the victim of a murder was found alive and that inconsistent judgments had been made by other courts. These grounds were not used very much because they were “too specific and thus too restrictive” (Dongois, Reference Dongois, Huff and Killias2008: 252, 254).

Under a 2014 law, new evidence is now the exclusive ground for reopening a conviction. It can be requested either by the accused or their family or by prosecutors. Article 622 of the Code of Criminal Procedure as modified by Law 2014-640 provides for the entry of new evidence that “établir l’innocence du condamné ou à faire naître un doute sur sa culpabilité.” In other words, this provision contemplates a reopening both where the new evidence establishes innocence or raises a doubt about culpability. This change was designed to make revision more accessible with a previous reference to new evidence raising a “serious doubt” change to only require a “doubt” or “un doute” (Verhesschen and Finjuat, Reference Verhesschen and Finjuat2020: 24). The National Assembly’s proposal rejected in 2014 to allow convictions to be overturned if there is the slightest doubt seems to have dropped off the table (Verhesschen and Finjuat, Reference Verhesschen and Finjuat2020: 30). As discussed in Chapter 2, fine and legalistic distinctions about what constitutes a miscarriage of justice can be a dynamic way to ration justice.

The convicted person was also given the ability under Article 626 to ask the prosecutor to conduct additional investigative measures. The application for revision is filed before a new court of revision and re-examination composed of eighteen judges of the Cour de Cassation. Five of these judges play an investigative role, and thirteen adjudicate the case with the ability to uphold or nullify a conviction or order a new trial. There is no appeal from this new court’s decision (Reference Verhesschen and Finjuat2020: 26–28). Although the 2014 reforms were designed to make it easier to reopen a conviction, there are reasons to think Nathalie Dongois’s (Reference Dongois, Huff and Killias2008: 249) pessimistic conclusion in 2008 that the rules for revisions are so strict “that it becomes virtually impossible to obtain a new trial” may still hold true. In 2015, 8 of 145 requests for revision were referred; in 2016, two convictions were nullified, but 130 requests were rejected. Four convictions were nullified in each of 2018 and 2019 with 118 and 138 requests, respectively, being rejected, generally for lack of new evidence (Verhesschen and Finjuat, Reference Verhesschen and Finjuat2020: 30). Since 1990, only 52 convictions have been overturned out of over 3,000 requests for revision (Etrillard and Rossant, Reference Etrillard, Rossant and Robins2023: 22). This is higher than the 6 of over 2,000 demands for revision under the 1989 law resulting in acquittals but not substantially higher (Dongois, Reference Dongois, Huff and Killias2008: 250). The overturned cases also include cases where, contrary to the American factual innocence model, the conviction was overturned because accused were mentally incapacitated at the time they committed the criminal act (Reference Dongois, Huff and Killias2008: 252).

The European Registry only contained ten remedied exonerations from France as of the end of 2024. Patrick Dils was convicted of murdering two young children and served fifteen years in prison. It was only after three attempts at revision that he was acquitted after a serial killer had confessed to the murders. Dils was sixteen years of age at the time of his conviction and had a mental age of eight. He made a false confession which he withdrew a month later (Hodgson, Reference Hodgson2020: 255). The number of false confessions taken from young accused and those with mental disabilities or mental illness in both France and Germany is alarming.

Other remedied French cases include the case of Rida Daalouche convicted of murder before new evidence confirmed that he was at a psychiatric hospital at the time of the crime. Marc Machin was convicted of murder and served seven years’ imprisonment before the real perpetrator confessed. At least two sexual offence convictions have been overturned when young complainants have retracted their original statements. In France, there is also the possibility of a presidential pardon used famously in the Dreyfus case discussed in Chapter 7.2 but also in 2016 with respect to Jacqueline Sauvage, a woman who killed her abusive partner in 2012. As in adversarial systems, executive clemency is an alternative to judicial remedies.

Etrillard and Rossant (Reference Etrillard, Rossant and Robins2023: 25–26) have argued that some judges, often without a sufficient background in the area of expertise, consider expert reports that they commission “to be factual truths.” They point to the persistence of the use of shaken baby syndrome (SBS) in French courts where the courts rely on a handful of experts who have bolstered their expertise by contributing to guidelines on SBS issued by public health authorities. The same experts also offer training to judges and prosecutors. Defence experts are possible but rarely commissioned and often discounted. They may also have difficulties accessing the full case file. This suggests that confirmation bias is still a problem in the relatively closed inquisitorial system.

France’s most significant miscarriage of justice scandal was the Outreau affair in which ten people were convicted in 2004 of sexual abuse of children. A key witness, the mother of some of the children, recanted her testimony during a 2005 appeal where six of the ten convicted persons were acquitted. The defence also challenged expert reports that helped to secure the convictions. The eventual overturning of the wrongful convictions led to three commissions of inquiry. Many of the recommendations were designed to increase adversarialism or the contradictoire of the system by giving the defence more access to the investigative file and more opportunities to challenge the state’s case. Proposals to abandon the role of the investigative judge were, however, rejected.

As Jacqueline Hodgson (Reference Hodgson2020: 263–264) concludes, the Outreau inquiries found that the prosecutor “dominated the investigation rather than acting as any form of check” and both the prosecutor and the juge d’instruction “employed a wholly prosecution-oriented perspective …” with the juge simply copying the prosecutor’s reasoning. The fact that both officials were classified as judicial did not prevent the wrongful convictions. Moreover, the judge refused defence requests for additional information and different experts to examine the young complainants. Defence lawyers were treated as “an unwelcome outsider” (Reference Hodgson2020: 264–265). This was all done while seven of the accused were kept in pre-trial custody for two and a half years or more (Reference Hodgson2020: 267). Professor Hodgson writes that even in the face of recommendations designed to increase the role of defence lawyers and new requirements that juge d’instruction invite comments from the parties and specify the evidence both for and against the accused, there is still no effective means “for testing out the evidence or challenging the case theory during the investigative stage” (Reference Hodgson2020: 273).

4.4.4 Japan

Japan has one the world’s lowest per capita prison populations but it has a wrongful conviction problem. The Japanese Criminal Procedure Code follows inquisitorial procedure by allowing judges to conduct searches, call expert evidence and examine witnesses in the search for truth. It prohibits convictions on the basis of uncorroborated confessions and allows judges to alter the charges laid by prosecutors to correspond with the facts. Kazuko Ito (Reference Ito2012: 1248) has stressed that judges and prosecutors frequently exchange personnel. They are in the “same bureaucratic circle” and that “judges tend to rely on the prosecutor’s argument.” Similar concerns have been expressed about the close relations between judges and prosecutors with respect to France’s Outreau affair.

The strength of the Japanese judiciary is seen by the fact that even when juries were reintroduced in Japan in 2009, they were composed of six lay people who deliberate with three judges. The jury does not have to be unanimous, but at least one professional and one lay judge must agree to a majority decision to convict. The use of juries has only led to a slight reduction of the conviction rate from 99.4% to 99.1% and some declines in the use of death sentences (Johnson and Vanoverbeke, Reference Johnson and Vanoverbeke2020: 456, 465).

The remedy for a wrongful conviction under Article 435 of the Japanese Code of Criminal Procedure is not a new appeal, as is the case in many common law countries, but a retrial. Following civilian traditions, a retrial can be based on a variety of codified factors including convictions of criminal justice personnel in relation to the original case and proof through judgments that testimony at trial was false. The new evidence provision in Article 435(vi) requires “clear evidence which should make the court render an acquittal or a dismissal” or lead to the conviction of a lesser crime. Article 453 contemplates some limited rehabilitation of those acquitted on a retrial by providing that the judgment acquitting the person “shall be made public in the Official Gazette and newspapers.”

Under Article 439 both the prosecutor and the accused can request a retrial and Article 441 allows the retrial to be requested even after the sentence was served. In theory, these provisions represent a commitment to truthful adjudication. In practice, however, even when an applicant convinces a court to order a retrial on the basis of “clear evidence,” the prosecutor can and often has appealed this ruling to appellate courts under Article 450, adding more delay. Article 442 affirms prosecutorial power by providing that a request for a retrial does not suspend a sentence unless the prosecutor consents to such a suspension. Despite the right to retrials, the Japanese system is characterized by very long waits to correct wrongful convictions.

David Johnson (Reference Johnson2015, Reference Johnson2020, Reference Johnson2022) has argued that Japan does a much worst job than the United States in remedying wrongful convictions because of a “culture of denial.” He cites research showing thirty-one remedied wrongful convictions in Japan in both the 1970s and 1980s, but only sixteen and seventeen in the 1990s and 2000s, respectively (Johnson and Vanoverbeke, Reference Johnson2020: 68). Only eleven persons in nine cases in the post–World War II era have been exonerated from Japanese death row and it took an average of thirty-two years to correct the errors (Johnson, Reference Johnson2022: S10).

False confessions played a role in 82 per cent of wrongful convictions between 1945 and 1991 (Johnson, Reference Johnson2022: S13). Japanese law allows lengthy interrogations and up to twenty-three days in police custody without counsel present and without being able to phone counsel or family (Ibusuki, Reference Ibusuki2019: 837–838). Unsurprisingly, such long periods of detention, combined with often aggressive questioning, have produced false confessions associated with wrongful convictions. The Japan Federation of Law Societies has commented that the use of “illegal and unreasonable interrogations” including threats is not “uncommon” (Human Rights Watch, 2023: 67–68).

In the Hakamada case, the Shizouka District Court concluded in 2014 that “when the high probability of the defendant’s innocence has been made clear, detaining him any longer would violate justice to an intolerable extent” (Johnson, Reference Johnson2020: 62). It then released Iwao Hakamada, who had been imprisoned since 1968 for the murder of four people. The former boxer was convicted after the court admitted only one of forty-five different confessions that the police had prepared. Hakamada had confessed on the nineteenth day of his interrogation. He soon after retracted the confession and maintained his innocence.

Shortly after Hakamada’s 1968 conviction, the most junior of the three judges resigned from the bench. The judge, Norimichi Kumamoto, had written a 350-page draft judgment that Hakamada was not guilty, but was required by the senior judges to withdraw it and write the judgment convicting Hakamada. After his resignation, the judge became homeless and suffered from addictions. It was only after this judge went public with his story in 2007 that Hakamada’s second attempt to obtain a retrial was successful (Johnson, Reference Johnson2015). Professor David Johnson, a leading commentator on Japanese criminal justice at the University of Hawaii, has commented, “it took the Tokyo High Court and the Supreme Court twelve years to reject Hakamada’s original appeals and finalize his sentence of death. Then, after Hakamada’s defense lawyers filed their first request for a retrial in 1981, it took the three levels of Japan’s judiciary 27 years to reject it” (Reference Johnson2015).

Hakamada was only acquitted at his retrial by a District Court in 2024 when he was eighty-eight years of age. He was the world’s longest-serving death row prisoner. His mental health deteriorated because of the constant threat of execution, which in Japan happens without prior notice to prisoners sentenced to death, and he was exempted from participating in his retrial because of his mental state. In acquitting him, the judge criticized the use of “inhumane interrogations meant to force a statement … by imposing mental and physical pain” and infringing his “right to remain silent, under circumstances extremely likely to elicit a false confession” (Guardian, 2024c) Prosecutors decided not to exercise their right to appeal but also perversely suggested that Hakamada was guilty; that the retrial decision was mistaken and that it was simply exercising mercy. Even such a high-profile and tragic case has not ended Japan’s “culture of denial” about wrongful convictions (Johnson, Reference Johnson2024b: 14).

Japanese criminal procedure has adopted some adversarial features including 2004 reforms that require prosecutors to disclose material to the accused. The Supreme Court in 2007 required the disclosure of police notes to the accused (Ito, Reference Ito2012: 1266–1267). Prosecutors in the Ministry of Justice have opposed reforms to eliminate their discretion not to disclose material, and problems of non-disclosure by prosecutors persist. Still, as a former Japanese judge has stated, prosecutors and police have “canons” while the defence has “air guns” (Johnson, Reference Johnson2024b: 10).

A series of high-profile wrongful convictions based on false confessions led to some law reform in 2016 that required video recording of interrogations, but only in the most serious cases that are eligible for lay participation, about 3 per cent of all cases. The committee drafting the reforms included an exoneree and a director of a movie about a wrongful conviction (Sasakura, Reference Sasakura and Takahashi2019: 32, 44). These 2016 reforms were also balanced with reforms in the same bill increasing police powers, allowing some plea bargaining and increasing the rights of crime victims. Defence lawyers still cannot be present during Japanese interrogations.

Concerns have also been raised that the use of videos of interrogations may make potentially unreliable confessions look more reliable. One example was the acceptance in 2016 of a seven hour confession in a successful murder conviction despite the accused’s arguments that the entire interrogation was not recorded and that the confession was unreliable because the accused was not fluent in Japanese and was badgered by a prosecutor who told the suspect that the victim’s family wanted him to confess and “don’t be dishonest. Don’t cheat us” (Ibusuki, Reference Ibusuki2019: 836–837). A focus of the video camera on the suspect also may make the accused seem more guilty. Japanese law professor Makoto Ibusuki warns that videos “do not necessarily tell the truth” (Reference Ibusuki2019: 846). As discussed in Chapter 3.4, recording requirements alone do not change the style of interrogations.

A 2023 report by Human Rights Watch details the problem of extensive pre-trial detention and harsh interrogations, which lead to wrongful convictions in Japan. The report points to a 99.8 per cent conviction rate in 2021 (Human Rights Watch, 2023: 79) with trials often being reduced to “ceremonies for ratifying prosecutors decisions” (Johnson, Reference Johnson2002a: 36). This is called “kensatsukan-shiho” or “prosecutor’s justice.” This is a particular danger in inquisitorial systems where prosecutors are regarded and trusted as if they are judicial officials.

Despite some moves towards more adversarial procedures and wrongful conviction reforms such as recording confessions in serious cases, the Japanese criminal justice system still employs practices such as intense interrogations that can produce wrongful convictions. It also remains dominated by prosecutors and judges who are both considered to be judicial officials. As in other inquisitorial systems, there are rights to reopen convictions, but the prosecutor’s ability to appeal orders of retrial makes the correction of wrongful convictions in Japan a very slow process.

4.4.5 Chile

Chile, like many other countries in South America, has adopted a new and more accusatory criminal procedure law as part of its transition to democracy. It created new prosecutorial and defence institutions and placed more emphasis on the rights of the accused. At the same time, these reforms were made for reasons not related to wrongful convictions and without consistent attention to the dangers of wrongful convictions in more adversarial systems. An early study conducted for the Chilean government declared the project a success in part because of evidence that cases were processed more quickly and with higher conviction rates (Vera Institute, 2004). Some have raised concerns that moves towards adversarial systems in Latin America have been overly influenced by American reforms, frequently with ethnocentric assumptions (Hodgson, Reference Hodgson2020: 33).

One of the pressing problems in South America is high rates of pre-trial detention. In Chapter 10, similarly high rates of pre-trial detention in India will be examined. The use of unwarranted pre-trial imprisonment reveals the limits of an exclusive focus on wrongful convictions and the need to take a broader approach to miscarriages of justice. Despite the adoption of adversarial codes, there are continued high rates of pre-trial detention in South America, with some evidence that public defenders do not make bail requests as often as private lawyers (Fondevila, Gustavo and Quintana-Navarette, Reference Fondevila and Quintana-Navarette2021). In Chile, however, the more adversarial reforms have been associated with reductions in pre-trial detention. In 1999, 51 per cent of Chile’s prisoners were awaiting trials, but by 2007, this percentage had been reduced to 25 per cent (Brashear-Tiede, Reference Brashear-Tiede2012; Duce et al., Reference Duce2010). This is a significant achievement. At the same time, Chile has one of the highest rates of per capita imprisonment in South America (World Population Review).

Between 2000 and 2005, Chile phased in a new criminal procedural code to replace an inquisitorial system inherited from Spain. The “old process was paper based” and involved a sumario phase of investigation and charge development that was done in secret and often involved pre-trial detention. The second plenario stage could involve submissions from prosecutors, the accused and victims (Oyanedel, Reference Oyanedel2019: 22). There were few judges, but they were “at the same time both prosecutors and adjudicators, leaving little space to challenge their own conclusions …” (Reference Oyanedel2019: 23). Proceedings were “bureaucratic” and the “the existence of a written file” was “the equivalent of the trial” (Veliz, Reference Veliz2012: 1365). As seen above, similar problems have been identified in France in the wake of the Outreau affair and in Japan.

Chile’s new Code of Criminal Procedure required investigation to be conducted by a new Public Prosecutor’s Office with charges, issues of bail and summary procedure cases being decided by a single judge in the Guarantee Court. This approach could also result in plea bargaining, guilty pleas and sentencing before a single judge in the Guarantee Court in cases involving less than five years in prison. In other cases, the final phase would be an oral trial before three judges. In both phases, the accused would be represented by a new Public Defender office. Many more judges were appointed to staff both the Guarantee and Oral Trial courts. At the same time, the effects of the move to oral and more adversarial hearings should not be overestimated. In 2006, the average duration of the oral hearing was 37.4 minutes and by 2014 it was only 16.4 minutes, thus suggesting that defence lawyers are often passive (Duce, Reference Duce2022). What could be classified as a summary procedure in some inquisitorial systems would look like a guilty plea to observers from adversarial systems.

Chile’s Code of Criminal Procedure continues the inquisitorial tradition seen in European codes of enumerating specific types of errors including contradictory judgments or convictions of witnesses or others who were involved in the original trial. As in other systems, the most important ground for reopening and overturning convictions is new evidence. Article 473(d) of the Code refers to new facts or unknown documents that is sufficient to “establecer la inocencia del condenado …” or to establish the innocence of the convicted person. This may reflect the influence of the American proven innocence model. Under Article 478, courts that find the accused innocent may also award compensation at the same time. This can increase access to justice and mitigate long delays before the wrongfully convicted receive compensation. In Chapter 10.6, it will be seen that some courts in India have taken a similar approach in recognition of access to justice problems created when the wrongfully convicted have to engage in separate litigation to obtain compensation.

Mauricio Duce, Chile’s leading scholar on wrongful convictions, has found 48 out of 601 applications to the Supreme Court between 2007 and 2016 where the court declared “that people convicted under the new accusatory system were innocent” (Duce, Reference Duce2022: 328). Most of these successful requests were made by the prosecutor and not the accused. Prosecutors in some adversarial systems also have taken steps to correct wrongful convictions (Roach, Reference Roach and Berger2017b), but the high number of retrial requests by prosecutors suggests that prosecutors in Chile have not become completely adversarial under Chile’s new system. This may be an important holdover from the inquisitorial systems, especially when a strong defence is not provided.

Many of the wrongful convictions involve cases where property crime convictions were recorded against an innocent person because the perpetrator falsely gave the name of that person when apprehended by the police. Most of the wrongful convictions were decided quickly through summary procedures in the Guarantee court. Errors were made by police and prosecutors, but they were not detected or corrected by defence counsel (Duce, Reference Duce2015). It has also been observed that “[o]ften, courts give too much credit to information provided by the police and are unduly reliant on this information” (Veliz, Reference Veliz2012: 1371). This may be true in both adversarial and inquisitorial systems.

In one Chilean case, a person was imprisoned for almost two years for a firearms offence before new evidence was introduced that he had a proper license for the weapon. Another case involved Pedro Lobos, whose robbery conviction and ten-year sentence were overturned after new evidence emerged that another person had confessed to the crime. In that case, the accused had already served more than three years in prison. Another person, Julio Robles, convicted of robbery and sentenced to five years, was successful on revision with new evidence that he was imprisoned at the time of the robbery. Tunnel vision by the police may also have been a factor in these two wrongful convictions for robbery (Duce, Reference Duce2015).

Professor Duce appropriately recognizes that the “anecdotal” information provided by remedied wrongful convictions does not identify the extent of unremedied wrongful convictions. He notes that expert evidence is used by prosecutors in about 28 per cent of the serious cases that go to trial. He warns that expert evidence in Chile suffers from the same limitations found throughout the world. He points to the wrongful conviction of Rodrigo Saaveda in 2007 for raping his daughter on the basis of medical evidence that was found faulty by the Supreme Court in 2013. Over half of trials involving sexual crimes against children used expert evidence about credibility, “whose validity is widely debated in science” (Duce, Reference Duce2015). As will be discussed in Chapter 8.8, a number of similar exonerations in child sexual abuse cases are found in the European Registry of Exonerations. In both cases, inquisitorial systems allow experts to opine on the credibility of complainants in a manner that would generally not be allowed in adversarial systems. By allowing such expert evidence, they introduced a new source of error into the proceedings. At the same time, this practice also allows new evidence to be subsequently introduced that challenges the original expert evidence. In many adversarial proceedings, similar convictions might not be amenable to challenge because of appellate deference to the jury.

Although Professor Duce was one of the architects of the new accusatorial system, he admits that the admissibility of expert evidence is rarely challenged in Chile’s new oral trials. He ultimately relates this failure to the continued impact of civilian systems of free proof evidence. This undoubtedly is a factor, but so too is the weakness of the defence in both common law and newly accusatorial systems. Concerns have been raised that adversarial systems are too lenient in allowing expert evidence of unknown reliability (Edmond et al., Reference Edmond, Cole, Cunliffe and Roberts2013).

Innocence projects in South America have been promoted through the work of an alliance Red Innocente, involving the California Innocence Project and its director Justin Brooks. They have provided training for defence lawyers, prosecutors, judges and police in South America. Brooks recognizes that innocence work differs in South America and America with the former being part of a “rule of law movement” and a “shift from inquisitorial to adversarial systems” (Brooks, Reference Brooks2012: 1118–1119). At the same time, the Chilean Public Defenders’ office, which is publicly funded, has formed its own Innocence Project, which helped inform Duce’s registry of corrected wrongful convictions. Others, however, have raised concerns that the Public Defender’s office does not have sufficient funds for post-conviction work, especially given that the requirements of revision under Article 473 “are very difficult to accomplish” (Veliz, Reference Veliz2012: 1373). Another problem in Chile is inadequate rules for the retention of biological evidence (Beltran-Roman, Reference Beltran-Roman2021). The American model of Innocence Projects supported by charitable donations may not be the optimal model for innocence work in other countries, especially those that are prepared to invest in public defenders or something more akin to conviction integrity units in prosecutor’s office.

Given the important role that aid from the United States has played in the development of more adversarial processes in South America, it will be important that such countries recognize some of the weaknesses in adversarial systems that can result in wrongful convictions, including false guilty pleas. It will also be important that they do not embrace the restrictive American model of proven innocence. At the same time, it is possible to find inspiration for a proven innocence requirement not only from the United States. Article 954(1)(d) of the 2016 Spanish Code of Criminal Procedure allows revisions in cases where “after judgment, facts or evidence become known which, if they had been provided, would have determined acquittal or a less severe sentence.” This seems not to require proven innocence. At the same time, Article 958 then provides that the sentence should be annulled if “the innocence of the convicted person is proven.” As suggested in Chapter 2, this would be uncontroversial if proven innocence was seen as one example of a larger category of miscarriages of justice or wrongful conviction where new facts would require an acquittal.

As suggested earlier, Chile appears to require proven innocence in its revision provisions. In Argentina, there is a “lack of clarity on the standard of proof required to reverse a conviction” with some support for a demanding proven innocence standard. Nevertheless, a case decided in 2015 ordering a new trial on the basis of new evidence using an “imprecise standard” that ranged “from doubt in favour of the defendant, through the reasonable probability of a different to the probability that the result would change” (Garrido, Calvo and Benitez, Reference Garrido, Calvo, Benitez and Robins2023: 163). As discussed in Chapter 2, imprecise standards of injustices that deserve remedies can be a sign of the rationing of scarce resources including concerns for preserving the finality of convictions.

South American jurisdictions may be attracted to proven innocence, especially given the crime challenges that many of those countries face and the American influence. At the same time, the aims of South American adversarial reforms in solidifying transitions to democracy could and should also support a broader focus on miscarriages of justice that include unfair trials and also give weight to the reasonable doubt principle.

4.5.1 German Guilty Pleas and Summary Procedures

In some respects, Germany is an ideal candidate to eschew plea bargaining in the name of the inquisitorial values of mandatory prosecution and the discovery of the truth. As late as 1979, the American law professor John Langbein published an admiring article called “Land without Plea Bargaining: How the Germans Do It” (Langbein, Reference Langbein1979). He defended Germany from claims that defence concessions and shortened trials played a similar role in promoting efficiency in Germany as plea bargaining in the United States (Goldstein and Marcus, Reference Goldstein and Marcus1977). Today, however, Professor Langbein warns that German plea bargains are not really voluntary and that they run a risk of false guilty pleas. In support, he cited a study where 55 per cent of defence lawyers said they had clients who falsely confessed at the plea stage to receive a sentence discount (Langbein, Reference Langbein2022: 146–147). In short, he now admits that Germany is a land with plea bargaining.

Sections 153 and 153a of Germany’s Code of Criminal Procedure provide for settlement in the pre-trial stage, even in some cases without the consent of the court. This moves away from traditionally inquisitorial procedures and seems to rely on the idea that the prosecutor can make judgments about whether trials are in the public interest. This is common in adversarial systems and may become common in inquisitorial systems where prosecutors have the extra prestige of being seen as judicial officials.

Sections 407–411 of the German Code provide for the use of “penal orders,” and this procedure is used to dispose of more cases than trials. In some ways, the use of written penal orders is consistent with the traditional inquisitorial focus on the dossier because it involves a written order, often prepared by police and prosecutor, that is ratified by the judge without an oral hearing. The result is then mailed to the accused who has a short time to decide whether to accept the penal order or challenge it through a trial. It is possible that a German version of the Post Office scandal could have been resolved through this procedure, especially given that many of the cases resulted in non-custodial sentences. German equivalents of the English sub-postmasters might not have requested a trial, given the near impossibility of individual accused challenging the faulty computer system.

Less used in Germany than penal orders are negotiated settlements under Article 257(c) of the Code. The court must agree to a guilty plea, and it can indicate upper and lower punishments in the case. Unlike penal orders, this procedure can apply to all crimes. Article 257(c) (4) provides that the “court ceases to be bound by a negotiated agreement if legally or factually relevant circumstances have been overlooked or have arisen and the court is therefore then convinced that the prospective sentencing range is no longer appropriate to the severity of the offence or the degree of guilt.” This provides some commitment to factual accuracy, but the 2018 reforms to China’s Code of Criminal Procedure also contain similar commitments. In theory, prosecutors and judges should both be concerned about factual accuracy, but they may abandon this concern in the face of a growing case load. Defence counsel can also play a role but may respond to the accused’s concerns about quickly resolving a matter and obtaining the lowest possible sentence.

The German courts have been cautious in their formal pronouncements about departures from full trials. In 2005, the German Federal Supreme Court appeared to accept the inevitability of plea bargaining for less serious cases while also prohibiting the accused from waiving their rights to appeal (which as discussed earlier, include fact-based revision appeals) as part of the bargain that would bind the court. The Court conceded that in some circumstances, efficiency concerns could trump the ideal of mandatory prosecutions in a manner reminiscent of the United States Supreme Court’s cheerful acceptance of plea bargaining when it warned: “The criminal justice system could no longer function effectively if the courts were generally forbidden from negotiating with the parties the contents of the judgment handed down” (Brown, Reference Brown2016: 94). The Court also cited concerns about the need to respect rights to “speedy justice” and called for more legislative regulation of plea bargaining. At the same time, Maximo Langer has argued that the German system “translated” plea bargaining into the idea that the accused was making a confession with some form of a trial and judicial determination of truth still occurring. My colleague Marcus Dubber (Reference Dubber1997: 598) has suggested that requirements for a confession “serves not only to deter bargains in general, but false convictions in particular.”

In 2013, the German Constitutional Court upheld Article 257(c) but on the basis that it only authorized deals about sentences and not charges and required that rights to appeal not be waived (Weigend and Turner, Reference Weigend and Turner2014: 96). The Court stressed the safeguards in the law which contemplate that “the courts continue to be bound by their duty to investigate ex officio” and that there be “exhaustive monitoring of the plea bargaining process by the public, the prosecution and the appellate courts.” In the end, negotiated settlements could in the Court’s view be reconciled with the traditional inquisitorial “aim of ensuring that criminal proceedings continue to be committed to ascertaining the substantive truth and to arriving at a just punishment that is commensurate with the crime” (German Federal Constitutional Court [BVerfG], Judgment of the Second Senate, March 19, 2013, 2 BvR 2628/10, para 65). Countries with inquisitorial traditions seem resistant to charge bargaining because it so overtly conflicts with the value inquisitorial systems place on finding the truth.

These appellate court decisions raise the questions about possible differences between formal law and actual practice. Professors Weigend and Turner (Reference Weigend and Turner2014) expressed scepticism about whether the courts will be able to satisfy the regulatory aims articulated by the Constitutional Court. At the same time, they acknowledged that the 2013 Constitutional Court decision allowed far more robust judicial regulation of plea bargaining than in the United States, where the Supreme Court has allowed pleas without full disclosure, waivers of appeals or even warnings about the need for counsel (Reference Weigend and Turner2014: 104).

The German advantage identified by Langbein in 1979 has changed in order to promote the efficiency of the court system with German courts recognizing that full trials in every case would lead to the collapse of their justice system. The frequent use of written and mailed penal orders in Germany has the potential to produce mass miscarriages of justice similar to the English Post Office cases. Nevertheless, starting points and penal culture still matter. Both German penal orders and settlement in the pre-trial stage are limited to minor cases resulting in non-custodial dispositions or custodial sentences that are suspended in favour of probation. Article 257(c) is not so limited, and this presents dangers of false guilty pleas in serious cases. At the same time, it provides for some judicial scrutiny of the accuracy of the charge and the Constitutional Court stressed the importance of these safeguards.

4.5.2 French Guilty Pleas and Summary Procedures

In 2004, sentence bargaining was introduced into French law. In 2016, this procedure was used to resolve 87,733 out of 1.4 million prosecutions (Hodgson, Reference Hodgson2020: 20). French plea bargaining is different from American court hallway negotiations. It requires that the accused be given a leisurely ten days to agree on a proposed sentence reduction. Article 495-8 of the Code of Criminal Procedure requires that the accused be represented by a defence lawyer with access to the case file. Hodgson (Reference Hodgson2020: 20) writes that “this procedure is more party centred than judge centred and therefore is seen as relating more easily to an adversarial process.” At the same time, a judge is, as in Germany, instructed to determine the factual and legal sufficiency of the charge and the sentence. Hodgson, like Weigend and Turner, expresses some scepticism about the intensity of this judicial supervision given the need for efficiency (Reference Hodgson2020: 22).

The above form of sentence plea bargaining is not available for the most serious offences such as murder, manslaughter and rape. The maximum sentence under the sentence bargaining procedure is ten years imprisonment (Langer, Reference Langer2021). As in Germany, the French courts have warned that the role of the judge is not “to rubber-stamp these negotiated agreements, but to continue to enquire fully into the facts as is the duty of the trial judge” (Reference Langer2021: 604). Hodgson also notes that “if the accused is not content to admit the offence, there is no question of persuasion or negotiation; the case will simply be listed for ordinary trial. This is an important difference from adversarial plea bargaining” (Hodgson, Reference Hodgson2020: 20).

4.5.3 Italian Guilty Pleas and Summary Procedures

Article 444 of the Italian Code of Criminal Procedure allows the prosecutor and the accused to agree to a reduced sentence by up to a third. This procedure is used in about 15 per cent of all cases (Hodgson, Reference Hodgson2020: 27). As in France, the accused must be represented by counsel. The maximum possible reduced sentence is five-year imprisonment. These limits on plea bargaining demonstrate some residual distrust of resolving cases through consent. It may result in false guilty pleas but at least not in the most serious cases.

The Italian Code also allows summary procedures. This procedure is used in about 10 per cent of all cases (Reference Hodgson2020: 27). This procedure can be used in all cases raising some of the same concerns as Article 257(c) of the German Code. At the same time, it has been argued that this procedure is not really a form of plea bargaining because a trial – albeit a summary trial – still takes place (Czerwinska, Reference Czerwinska2023: 282–283).

Steven Thaman (Reference Thaman2010: xxvii) has pointed out that moves in Italy and other inquisitorial systems “to increase the orality and immediacy of trials” have ironically “undermined the usefulness of the preliminary investigation dossier for providing a factual basis for guilt in the context of plea bargaining.” At the same time, he notes that under the Italian procedures, the accused also has the option to ask the judge to call witnesses and to be questioned in court (Reference Thaman2010: 381) procedures that may have the potential to discover the truth including false guilty pleas. Some in Italy have criticized this potential judicial review of guilty pleas as resurrecting the investigating magistrate that were rejected in the 1988 Italian reforms designed to make the system more adversarial. From the perspective of preventing wrongful convictions, however, an active judicial role is a strength.

4.5.4 Japanese Guilty Pleas and Summary Procedures

Until recently, plea bargaining in Japan was denied though concerns were raised that false confessions were made in an attempt to obtain leniency and avoid the death penalty (Johnson, Reference Johnson, Feeley and Miyazawa2002b). As discussed earlier, Japanese judges have inquisitorial powers to conduct investigations, order expert testimony, examine witnesses and even conduct searches (Turner, Reference Turner2009: 180). Once a charge is brought, the inquisitorial concern with factual accuracy allows the judge to order the prosecutor to add other counts or disapprove of the addition of counts if the new charges are not factually accurate (Reference Turner2009: 180–181). The Japanese Supreme Court held in 1995 that because of concerns about truthfulness, testimony should not be considered from a witness who had been granted immunity from prosecution as a condition of testifying (Reference Turner2009: 183). This illuminates a willingness to prevent the use of such incentivized witnesses, something as discussed in Chapter 3.3 is not done in many adversarial systems.

In 2006, summary procedures that allow cases to be disposed without the usually extensive reasons given by trial judges were introduced. They were, however, restricted to cases with a punishment under a year. A new Japanese law in 2018 allowed a form of plea bargaining but one that excludes cases involving death or bodily harm. The law does not allow an accused to receive a plea bargain by pleading in his or her own case. Only three cases have used this procedure as of 2023, generally in white-collar and organized crime where insider witnesses are often necessary. One unnamed Japanese judge told The Japan News that judges “will be sensitive to the risk of implicating innocent people and will be hesitant to take them up … statements not supported by objective evidence are no good” (Suzuki and Takada, Reference Suzuki and Takada2023). This demonstrates a traditional inquisitorial scepticism to plea bargaining as compromises of the truth, albeit one that is somewhat ironic given Japan’s wrongful convictions in murder cases on the basis of false confessions discussed earlier in this chapter.

Efficiency demands are felt by all criminal justice systems (Hodgson, Reference Hodgson2020), but Japan follows a pattern seen in other inquisitorial systems examined in this chapter by limiting the availability of guilty pleas and summary procedures to less serious cases. This can at least limit the damage done by false guilty pleas.

4.5.5 Chilean Guilty Pleas and Summary Procedures

The Chilean system also allows the disposition of criminal cases without a full trial. But there are restrictions not found in adversarial systems. Guilty pleas are generally only allowed in cases involving five years’ imprisonment or less, though abbreviated procedures in theft and robbery cases can result in up to ten years’ imprisonment. Moreover, the court can reject a guilty plea because of concerns about its factual accuracy (Langer, Reference Langer2021: 16).

Professor Duce’s (Reference Duce2015) research had shown that most remedied wrongful convictions occur in the Court of Guarantees with many involving a confusion about the accused’s name. He also noted that similar results have been found with respect to the use of summary procedures in Switzerland. Such wrongful convictions can cause harm and could even result in mass miscarriages of justice similar to England’s Post Office cases or mass drug lab and police corruption exonerations in the United States. Nevertheless, the harms caused by such false guilty pleas are limited because of statutory limits on the seriousness of offences and/or the sentence that can be given. By contrast, a number of false guilty pleas in adversarial systems involve the most serious cases. In Canada, many of the remedied false guilty pleas have involved people charged with murder and facing mandatory life imprisonment if convicted who have accepted pleas to manslaughter (Roach, Reference Roach2023a: ch 2).

Langer and Sozzo (Reference Langer, Sosso, Langer, McConville and Marsh2023) have found that some plea bargaining mechanisms were introduced in thirteen of nineteen South American countries as they moved towards more adversarial systems. They conclude that while influenced by the American example of plea bargaining, South American countries have translated them in accordance with their own traditions, which include a commitment to discovery of the truth. Like many other countries with inquisitorial systems, countries in South America tend to be suspicious of charge bargaining or resolving the most serious cases through guilty pleas. Langer and Sozzo warn, however, that a limited number of empirical studies suggest that inquisitorial safeguards on the books may not be respected in practice.

4.5.6 Summary

It is not necessary to accept the myth that inquisitorial systems remain loyal to principles of mandatory investigation, prosecution and trials to conclude that their starting points and theoretical ideals generally make them more sceptical of plea bargains than most adversarial systems (Damaska, Reference Damaska1986). Professor Hodgson has documented much convergence between inquisitorial and adversarial systems with respect to guilty pleas, but she still noted that “the concept of a plea bargain is problematic” within judge-centred inquisitorial systems with trials often being shortened as opposed to avoided altogether (Hodgson, Reference Hodgson2020: 17).

In many but not all inquisitorial systems that allow summary procedures or guilty pleas, these procedures are not allowed in the most serious cases. Germany allows pre-trial settlements for all offences, but the German Constitutional Court has only approved this procedure by stressing the role of the court in ensuring factual accuracy. Even critics of the German schemes (Langbein, Reference Langbein2022; Weigend and Turner, Reference Weigend and Turner2014) acknowledge that they do not produce the same risk to factual accuracy as adversarial and especially North American systems that rely on both charge and sentence bargaining often tied to avoiding mandatory penalties (Dubber, Reference Dubber1997; Roach, Reference Roach2023a). Judges in inquisitorial systems often have powers to inquire and even call witnesses before accepting guilty pleas. In Chapter 11.3, similar powers possessed by the International Criminal Court will be examined.

Starting points and traditions still matter. Legislative and other restrictions on the acceptance of guilty pleas and the use of summary procedures reflect the diminishing residue of the traditional concerns in inquisitorial systems about ensuring the accuracy of the court’s verdict.