Jump to ContentJump to Main Navigation
The Metamorphosis of Criminal JusticeA Comparative Account$

Jacqueline S. Hodgson

Print publication date: 2020

Print ISBN-13: 9780199981427

Published to Oxford Scholarship Online: April 2020

DOI: 10.1093/oso/9780199981427.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 07 August 2020

Miscarriages of Justice and Procedural Change

Miscarriages of Justice and Procedural Change

L’affaire d’Outreau

(p.251) 8 Miscarriages of Justice and Procedural Change
The Metamorphosis of Criminal Justice

Jacqueline S. Hodgson

Oxford University Press

Abstract and Keywords

This chapter introduces some of the common themes in miscarriages of justice across jurisdictions, before going on to focus on responses to the Outreau affair, the investigation of an alleged pedophile ring in northern France in which the weaknesses of the instruction process, ultimately, led to the collapse of the case and to three separate commissions of enquiry. The high-profile Outreau affair became a touchstone for the weaknesses of the French system, demonstrating the dangers of concentrating power in the hands of magistrats who, despite their constitutional independence and truth-seeking ideology, routinely failed to challenge or interrogate the perspectives of fellow judges or prosecutors, replicating the tunnel vision of the police investigation and undercutting the role of the defense. Reforms tempered but did not challenge the inquisitorial ideal, extending the principle of contradictoire. Ultimately, however, despite public attention and three commissions of enquiry, many proposals were never implemented.

Keywords:   miscarriage of justice, wrongful conviction, Outreau, case construction, principle of contradictoire, confirmation bias, judicial corporatism

Miscarriages of justice and the malfunctions they reveal often cause us to reassess systemic problems as well as those that are case-specific.1 They illustrate weaknesses in processes, practices and the legal and occupational cultures of criminal justice. Just as the North American disillusionment with plea bargaining and mass guilty pleas in the 1970s and 1980s led researchers to look to Europe for better ways of doing things,2 so too those in France, Britain, and elsewhere have looked to adapt their procedural models in the wake of miscarriages of justice.3 This has included reforms of the processes of investigation and prosecution, but also of the mechanisms to identify and remedy wrongful convictions.4 These adaptations do not seek to shift systems across to a different procedural model, nor are they influenced explicitly by comparative studies. Rather, they present ways to avoid some of the worst excesses or malfunctions of different processes of criminal justice. An analysis of how and why different systems break down and of the mechanisms identified to correct this provides new insights into procedural models and their limitations.

The chapters in this section reflect on the common features of miscarriages of justice across both adversarial and inquisitorial procedural traditions, noting that they have been found to malfunction in broadly similar ways in practice. (p.252) Neither the party-led adversarial model nor the centralized judicial enquiry of the inquisitorial model has been able to overcome the tendency to focus prematurely upon one suspect and to then set about constructing a case around this initial thesis. The presumed guilt of the suspect becomes the lens through which evidence is assembled and interpreted, with information that contradicts this being ignored or rejected as not credible. Whether under the direction of the police, prosecutor, or judge, the investigative phase becomes skewed toward this initial viewpoint, carrying over to the prosecution and trial. The defense role has also been undercut in both procedures, preventing the kinds of safeguards and cross-checking designed to test the reliability of evidence and to ensure equality of arms. In France, as a further safeguard, the concentration of power and authority in the hands of the juge d’instruction is understood to be tempered by a series of procedural checks and appeals adjudicated by other magistrats, but these typically confirm rather than challenge the investigation, lending it further credibility without scrutinizing the nature or reliability of the evidence or procedures. The reluctance of judges to contradict or to challenge the colleagues alongside whom they work and have trained is often referred to as a kind of judicial corporatism, a loyalty that is hard to break.

In seeking to address this systematic tendency toward case construction, or confirmation bias, each system has looked for new ways to provide a corrective—both during the investigation process and in reviewing possible miscarriages of justice. In so doing, procedural values are placed under the microscope as their potential within different process models is tested. In England and Wales, the establishment of the Criminal Cases Review Commission (CCRC) to investigate potential miscarriages of justice produces tensions in importing a process of inquisitorial review into a broadly adversarially structured procedure. In France, the incremental strengthening of accusatorial features seeks to challenge judicial corporatism, but in a procedure that continues to resist the strengthening of the defense role.

This chapter introduces some of the themes in miscarriages of justice across jurisdictions, before going on to focus on responses to the Outreau affair, the investigation of a pedophile ring in northern France in which the weaknesses of the instruction process ultimately led to the collapse of the case and to three separate commissions of enquiry, including the first ever parliamentary enquiry into a criminal case. The limited procedure for overturning convictions in France results in few miscarriage-of-justice cases, and so the high-profile Outreau affair became a touchstone for the weaknesses of the French system and the ways that it risks producing wrongful convictions, as well as for extensive discussions of ways of reform. It provides the focal point, therefore, for my analysis of how the system malfunctions and how core procedural values are strengthened, limited, or altered through proposals for reform. The following chapter examines the CCRC, (p.253) an organization established in response to the major miscarriages of justice of the 1980s in England and Wales, as an example of a more radical reform—not in the investigation and trial of criminal cases, but in their postconviction review.5 It considers in particular the nonadversarial procedural values that underpin its functioning and its status as something of a maverick institution: a broadly inquisitorial body investigating possible wrongful convictions, sandwiched between the adversarial procedures of trial and possible further appeal. Chapter 10 explores further the implications of this twin procedural approach for the would-be appellant and the representation of her interests. It analyzes the role of the defense lawyer as her status shifts from representative of “the accused” to that of “the applicant,” highlighting the tensions created between intersecting adversarial and inquisitorial role expectations.

A. Introduction

Some might argue that miscarriages of justice present an atypical account of a criminal justice system, telling us only about the small number of usually very serious cases where processes or people malfunction. It is certainly true that miscarriages of justice should be read in context. In addition to the wrongdoing or poor practice of criminal justice actors, successful appeals may arise out of a change in law or advances in scientific or forensic knowledge, and they may concern issues that are limited to a specific time period, a single case, or a line of connected cases. But they also tend to reveal systemic problems, such as, typically, the subordination of defense rights; a lack of police, prosecutorial, or judicial independence during the investigation; police malpractice, poor forensic investigation, and overreliance on a single expert;6 the withholding of evidence by the prosecution; or inadequate defense representation. And while some miscarriage cases involve isolated wrongdoing, many reflect the daily practices of criminal justice.7 The parliamentary enquiry into the Outreau affair in France heard from lawyers and magistrats that the case was conducted like any other; the unity of approach between prosecutor and juge d’instruction, the use made of remands in custody during the investigation, and the sidelining of lawyers were (p.254) nothing unusual.8 Legal procedures were not breached, but rather, legal discretion was exercised in ways that undermined the fairness of the investigation, the spirit of the law, and the treatment of the accused.9

What is perhaps less visible is the extent to which many of these same features contribute to wrongful convictions for less serious charges.10 The mass processing of defendants through the use of plea bargains and abbreviated trial procedures can only exacerbate this and will inevitably produce wrongful convictions when justice is negotiated in order to produce the most time- and cost-efficient outcome. Courts accept guilty pleas at face value, and the accused is often ill served by her defense, lacking any real and informed understanding of the process.11 These procedures are becoming more visible and significant as they apply to increasingly serious offenses. For example, in France in 1997, Joël Pierrot admitted committing an armed robbery, and two days later, under the rapid comparution immédiate procedure, he was tried, convicted, and sentenced to four years in prison, three of them suspended. His confession was never questioned, despite his claim to have used a plastic gun, in clear contradiction to the evidence of witnesses. He maintained his admission before the court, as he had been told that this would reduce his sentence. He appealed unsuccessfully in 1998. Only when he learned that two other men had confessed and had been convicted of the crime for which he too had been convicted and imprisoned was he finally successful in having his conviction overturned.

The features present in miscarriages of justice in both adversarial and inquisitorial procedures are strikingly similar, from the Irish terrorism cases in Britain in the 1970s and murder convictions in the 1980s and 1990s, to the Outreau case in France and the Schiedammer parkmoord (murder in Schiedam park) case in the Netherlands, to the many wrongful convictions in the United States, many of them identified through innocence commissions.12 The classic profile is an excessively conviction-oriented investigation and prosecution, coupled with a (p.255) defense that is prevented from protecting, or is unwilling to protect, adequately the interests of the accused. A few examples from different jurisdictions serve to illustrate these similarities.

The instruction procedure is the strongest model of judicial investigation in French criminal procedure, where defense rights are greatest and the magistrat is required personally to conduct, rather than simply to oversee, the enquiry, but even this cannot prevent the potentially determinative impact of the initial police interrogation. Patrick Dils was arrested in 1987 for the murder of two eight-year-old boys in Montigny-lès-Metz. Under pressure, he confessed but withdrew his confession one month later. Dils was sixteen but with a mental age of eight. He was interrogated without a lawyer, and the juge d’instruction and psychiatrists later noted his suggestibility and his tendency to give the answers desired by his interrogators. They also noted the closed nature of the questions posed by the police, which were highly suggestive of the “right” answers. The same gendarme who interrogated Dils had already produced confessions to the murders from two other men—both of which had been dismissed as not being credible. Furthermore, a key witness at the scene, who reported seeing a man spattered with blood, was never asked to provide a photofit of the man he had seen. It was fifteen years before Patrick Dils’s conviction was finally overturned in 2002—a reinvestigation revealed that a convicted serial killer had been working four hundred meters from the scene. He admitted to having seen the bodies but not to killing the boys. Together with the other weaknesses in the case, this was enough, however, to raise sufficient doubts about the involvement of Dils. This case is shocking not only in the wrongful conduct of police and the negligence of the judiciary, but also in its demonstration of the system’s inability to recognize its own errors.13

France has seen a very gradual introduction of legal assistance and safeguards such as recording of interrogations, but until recently, police and gendarmes have been able to interrogate suspects unchecked. In 1998, for example, Patrice Padé was persuaded by gendarmes to confess to the rape and murder of British schoolgirl Caroline Dickinson. He was a vagrant who had been near the hostel where the girl was killed, and he had previous convictions for child sexual assault. After forty hours in garde à vue (GAV) he finally confessed. He later retracted his confession, and DNA evidence demonstrated that sperm found on Caroline Dickinson’s body could not have been his. After eighteen days in custody, he (p.256) was released. This pattern is strikingly similar to US exoneree cases, such as that of Ronald Jones, who spent more than thirteen years in prison, before he was released after DNA testing proved that he could not have been responsible for the rape and murder of a twenty-eight-year-old woman. Homeless and an alcoholic, Jones had an IQ of 80 and confessed after many hours of interrogation, during which he alleged that he was continually assaulted. The prosecution case looked to be compelling, but the details in his statements (which only the killer would have known) were in fact provided by police investigators, and the forensic evidence was flawed and misrepresented.14

In the Netherlands, the Schiedammer parkmoord case contained all of the ingredients of a classic miscarriage of justice—a pressured confession that was retracted two days later, the suppression of exculpating forensic evidence, and a failure by the police and prosecution to take seriously the major inconsistencies in their (heavily constructed) case against the accused.15 In 2000, a ten-year-old girl and eleven-year-old boy were sexually assaulted in Schiedam Park in the Netherlands. The boy was seriously injured, but the girl was killed. Another man, CB, saw the boy and called the police. Although he did not fit the boy’s description of the attacker, CB became the prime suspect—principally because he was also a known pedophile. Under pressure, he confessed, but withdrew his confession two days later. The police (backed by a child psychologist) subjected the boy victim to further interrogation to try to force him to admit that his description of the attacker was a fabrication. He stuck to his story, but was ultimately ignored by the court. DNA found on the girl’s body and on the murder weapon was from an unknown third person. Doubts about CB’s guilt were expressed by the forensic scientists,16 but the prosecution ignored these, withheld evidence of the DNA on the murder weapon, and interpreted the absence of the suspect’s DNA on the body as evidence of his guilt—claiming that he had been careful to leave no traces.

Despite the weak and contradictory evidence, CB was convicted and was unsuccessful in both the Court of Appeal and the Supreme Court. Although there was widespread criticism of the case, the Supreme Court refused leave to review it as there was no new factual evidence—a new argument based on the existing (p.257) case file was insufficient.17 In 2004, another man confessed to the murder. His DNA matched that found at the scene and CB was released.

In England and Wales, between 1989 and 1991, the Court of Appeal quashed the convictions of a number of those serving sentences for terrorist-related offenses, including murder, in the 1970s.18 The police were found to have fabricated the notes of interview in the Birmingham Six and Guildford Four cases, and the scientific evidence was found to be unreliable, as it was in the Maguire Seven case. The scale of the wrongdoing on the part of those charged with investigating terrorism and serious crime triggered a response that went beyond the appeal courts, and a Royal Commission on Criminal Justice (RCCJ) was established to address (some might say appease) the crisis of public confidence in criminal justice at that time.19 Although these high-profile miscarriages of justice concerned some of the most serious cases, it does not necessarily follow that malpractice was limited to these. The cases were not isolated and atypical but, rather, exemplified the broader culture of policing as demonstrated through research, including that carried out for the commission itself: an overly broad discretion at all stages; a “canteen culture” based on individualized concepts of law and order and social discipline; and a tunnel vision approach to investigations in which cases are “constructed” against those identified as suspects early on in the process. Many also demonstrated serious failures in the conduct of both the prosecution and the defense cases at trial.20

More than thirty people were also wrongly convicted as a result of the nefarious activities of Birmingham’s West Midlands Serious Crime Squad, which operated (p.258) between 1974 and 1989, including the Birmingham Six, the Bridgewater Four,21 and Derek Treadaway.22 The squad fabricated evidence and, it was alleged, tortured suspects in order to obtain confessions.23

Miscarriages of justice were not limited to terrorist cases, nor to those investigated prior to the enactment of PACE with its attendant safeguards such as the recording of police interrogations and the right of suspects to have a defense lawyer present before and during police questioning. Yusef Abdullahi, Stephen Miller, and Anthony Paris (known as the Cardiff Three) were convicted in 1990 of the murder of Lynette White. One of the accused, Stephen Miller, was interrogated for thirteen hours over five days in police custody. After denying involvement more than three hundred times, he finally broke down and confessed. After hearing part of the confession, the trial judge was satisfied that it was voluntary and admitted it in evidence. The Court of Appeal took a very different view and concluded that the confession had been obtained by oppression and so unlawfully, despite the presence of Miller’s solicitor. The Court of Appeal judges were clearly shocked at the treatment of Miller at the hands of the police.

We are bound to say that on hearing tape 7, each member of this Court was horrified. Miller was bullied and hectored. The officers . . . were not questioning him so much as shouting at him what they wanted him to say. Short of physical violence, it is hard to conceive of a more hostile and intimidating approach by officers to a suspect. It is impossible to convey on the printed page the pace, force and menace of the officer's delivery . . . although we do not know what instructions he had, the solicitor appears to have been gravely at fault for sitting passively through this travesty of an interview. . . . It is clear on listening to the tapes that for extended periods, Miller was crying and sobbing, yet he was not given any respite. . . . The context was that he was being led to believe the officers were seeking to eliminate him from participation in the attack and he wanted to get to the end of the questioning.24

(p.259) As with so many miscarriages of justice, it is not during the visible and public setting of the trial that things tend to go wrong—though the trial judge’s failure to recognize the oppressive nature of the interrogation practices in Paris, Abdullahi and Miller is to be deplored. Rather, across jurisdictions, it is in the gathering of evidence, in the overly narrow focus of the investigation during which there is no opportunity for defense (or even prosecution) lawyers to challenge the case thesis or encourage an alternative perspective. This is further aggravated by failures to disclose evidence, especially relating to forensic and other expert witnesses. In the absence of any formal or effective mechanism to exercise a check on the police construction of a case against the suspect, the evidence produced risks being tainted by confirmation bias and the police investigation exists as an inquisitorial enquiry in England and Wales as well as in jurisdictions such as France or the Netherlands.

The production and nondisclosure of expert evidence, as described in the Schiedammer parkmoord case, can raise further problems. Not only does the presumed objectivity of the appointed expert mean that more credibility is likely to attach to her testimony than that of other witnesses,25 but also, as those first on the scene, the evidence of prosecution experts is often relied on by those working for the defense.26 The now-discredited tests of Dr. Frank Skuse contributed to the wrongful convictions of the Birmingham Six and the Maguire Seven. More recently, the assertions of Professor Sir Roy Meadow formed a central plank in the tragic case of Sally Clark, who was wrongly convicted of murdering her two baby sons. Meadow testified that the chances of two babies dying from cot death in the same family were seventy-three million to one—an assertion that was subsequently discredited and should never have gone before the jury.27 This proved to be doubly misleading, as there was evidence of infection in relation to one of the children, Harry, which raised the possibility of death by natural causes. Although known to the prosecution’s pathologist, Dr. Williams, this was not disclosed to other police, lawyers, or medical witnesses, nor was it mentioned at trial. This new evidence was passed to the CCRC, which referred the case back to the Court of Appeal. Experts providing evidence on appeal found it “astonishing”28 and “a (p.260) matter of great concern that this wealth of laboratory data was not disclosed.”29 The conviction was quashed after Sally Clark had spent three years in prison, but she never recovered from her ordeal and was found dead four years later.30

Failure to investigate the case objectively and to disclose evidence relevant to the defense case continues to result in wrongful convictions. Sam Hallam was one of the youngest victims of a miscarriage of justice, released at the age of twenty-four, having serving more than seven years in prison after his wrongful conviction for the murder of Essayas Kassahun. The evidence against him was from two witnesses, one of whom gave varying accounts and the other who retracted his evidence at trial. There was no forensic evidence linking him to the killing, and when an outside police force reviewed the case for the CCRC, they found that the witness evidence was so manifestly unreliable that it should never have been put before a jury. There was also evidence from his mobile phone that Hallam was in the pub with his father at the time of the murder. The police have a legal duty to pursue reasonable lines of enquiry that point away from the suspect’s guilt as well as toward it, but this did not happen.31 Instead, the enquiry was characterized by tunnel vision, ignoring fourteen witnesses who supported Hallam’s account, including the intended victim. The police investigation was flawed, convicting an innocent man and leaving the real killer free: Hallam was “the victim of a serious miscarriage of justice brought about by a combination of manifestly unreliable identification evidence . . . failure by police properly to investigate his alibi and non-disclosure by the prosecution of material that could have supported his case.”32

These high-profile appeals also tell us something about the institutional culture of criminal justice at particular moments in time, including the nature of legal and political trust.33 A small number of appeals against conviction may just as likely be an indicator of a closed and nonreflexive system, as of a well-functioning one. Dongois concludes that the small number of convictions overturned in France, for example, is the result of the very limited procedure for challenging convictions, rather than the inherent reliability of the French (p.261) criminal justice process.34 Brants argues that a significant obstacle in achieving justice in the Schiedammer parkmoord case was the fact that the trust enjoyed by criminal justice professionals prevented them from contemplating their own fallibility and that of the system itself.35 The Dutch system remains reluctant to open up to external scrutiny, or to contemplate anything that challenges the internal hierarchical systems of review.36 In France too, the failure of magistrats (who occupy different judicial positions but have trained together and share a strong common identity) to contradict their colleagues is often seen as a form of judicial corporatism. And in England and Wales, it should not be forgotten that despite the horrendous scale of wrongdoing that was finally uncovered, the earlier appeals of the Birmingham Six and others were turned down in the most definitive terms and it took a decade and a half before their convictions were finally overturned.37 Despite the experience of the high-profile miscarriages of justice in the 1980s and 1990s, the culture of adversarial investigation and prosecution (p.262) remains problematic, with nondisclosure in particular continuing to be a major cause of wrongful convictions.38

B. The Outreau Affair: The Familiar Hallmarks of a Miscarriage

The so-called Outreau affair concerns a major investigation into accusations of child sexual abuse made by a number of children and adults in the town of Outreau in Northern France. When the case came to trial in July 2004, two of the accused retracted their statements against their co-accused and, as a result, a large part of the prosecution case collapsed. Four of the seventeen accused were sentenced to between fifteen and twenty years in prison for offenses of sexual violence against their own and their neighbors’ children,39 but the remaining thirteen defendants were acquitted, seven by the cour d’assises, and a further six by the cour d’appel the following year. While this was not a classic case of wrongful conviction, the length of the investigation meant that those accused had spent several years in pretrial detention: between them, the defendants had served almost twenty-six years in détention provisoire while the juge d’instruction carried out his investigation. One suspect, François Mourmand, committed suicide while in custody.40 The repercussions of the case were enormous, resulting in three major commissions of enquiry: an investigation by the Ministry of Justice,41 one by the Inspection générale des services judiciaires (p.263) (IGSJ),42 and the first-ever parliamentary enquiry into a criminal case, the Vallini Commission.43

The gravity of the accusations in the Outreau case meant that, rather than a police enquiry overseen by the public prosecutor (the procureur), it was investigated through the more rigorous instruction, carried out under the direction of the more independent juge d’instruction, who has wide powers to question witnesses and to instruct experts. The role of the juge d’instruction is to investigate the case (not the person), and her function is defined in statute: she must follow both inculpating and exculpating lines of enquiry, that is, work à charge and à décharge. The defense also enjoys greater rights to participate in the enquiry than is the case in the 97 percent of cases that are overseen by the procureur: the suspect may have her lawyer present during all significant acts of investigation, including when questioned by the juge; the lawyer has access to the case dossier; and the defense may challenge evidence and propose new lines of enquiry, including investigative acts and the commissioning of experts. The prosecutor and the victim enjoy the same rights to participate, and the victim may be legally represented. If the juge d’instruction refuses a request to undertake a particular act of investigation, appeal lies to the chambre de l’instruction. Thus, although the juge d’instruction is the locus of power, the defense, victim, and prosecutor have the right to participate in the investigation, and they can challenge decisions of the juge, triggering review by a hierarchy of magistrats independent of the enquiry. This is understood to act as a set of checks and balances on the conduct and scope of the enquiry.

However, the parliamentary report (the most detailed of the three, with the testimony of those called televised nationally) was critical of the single case viewpoint offered by the procureur and juge d’instruction (whose roles are, in theory, quite separate and distinct) and the absence of reflexivity among magistrats: although the case passed through the hands of some sixty different magistrats, none challenged either the central case thesis or the methods of investigation.44 This is described as a form of judicial corporatism, in which there is insufficient distance between what are supposed to be different and even competing judicial functions. The parliamentary enquiry found that the procureur dominated the investigation rather than acting as any form of check, and both he (p.264) and the juge d’instruction adopted a wholly prosecution-oriented perspective to the case—they worked only à charge. Even the suspects’ declarations of innocence were seen as justifying repressive measures; the commission of enquiry was shocked to hear that such assertions were considered good reason to keep suspects in detention during the investigation.45 The proximity of the procureur and the juge d’instruction meant that they were able to discuss the case frequently and the procureur was able to influence the enquiry informally and so invisibly: investigations carried out on the prosecutor’s suggestion appeared to be on the juge’s own initiative and so to be more “neutral” and the result of judicial consideration. The unity of perspective of the juge d’instruction and the prosecutor was underlined by the fact that the final case conclusions communicated to the cour d’assises were simply copied and pasted by the juge d’instruction from those prepared by the procureur.46

In all this the defense was systematically prevented from participating in the enquiry and so from challenging the findings and methods of the juge d’instruction and stimulating investigation à décharge in the way envisaged by the text of the law. Statements were taken from the children making the accusations in ways that produced unreliable evidence, but the defense was not permitted to examine these witnesses on the grounds that it might traumatize the children further.47 There were major contradictions between the findings of the various experts, but every defense request to instruct a new psychologist or psychiatrist, or to obtain additional information or clarification, was refused first by the juge d’instruction and again by the chambre de l’instruction. The defense was prevented from introducing any element of contradictoire into this part of the investigation.48 The quality of the expert evidence was judged by the juge d’instruction alone, with no testing of what proved to be unreliable and inconsistent assertions on the part of experts through the instruction of additional witnesses.

We might expect some degree of partisanship from police investigators, but the behavior of the juge d’instruction in excluding the defence is perhaps more surprising. After related accusations of murder connected with a pedophile ring (p.265) in Belgium were discovered to be without foundation, the testimony of those witnesses who had lied was simply cut from the main dossier, and their overall credibility in relation to the other evidence does not seem to have been put in doubt. By removing their false testimony from the dossier, the defense was placed outside the procedure and deprived of any opportunity to question the witnesses’ character. The cour d’assises was also deprived of important evidence. Yet, as a simple administrative measure, this action was not subject to appeal. This deception was agreed by both the Parquet and the juge d’instruction.49

In other instances, the juge overplayed his hand, implying that he was in possession of evidence that he was not and instructing the police to question a suspect on this basis.50 Instead of being corrected, these problems were compounded by the pretrial appeal court, the chambre de l’instruction, which conducted only paper reviews and demonstrated a clear tendency to reinforce the position taken by the juge and the Parquet.51 Quite simply, the defense lawyer was described as an unwelcome outsider whose counterarguments were not well received by the investigating judge.52

This case demonstrates the array of systemic problems in concentrating power in the hands of one individual during the investigation—especially a judicial officer whose findings are afforded considerable credibility as the product of a judicial investigation—and in a system of checks that fail to scrutinize and challenge, rather than confirm, the exercise of this power. Just as the police have been shown to focus prematurely upon a suspect and then to construct a case against her, evidence was filtered by the juge in order to exclude that which cast doubt on the case he had constructed against the accused. Procedures were not followed, and key investigations were not carried out that might have challenged this case construction. At least one accused was first placed in GAV in order that he could be interrogated by the police (at that point without the presence of a lawyer), when, as the Vallini Commission noted, he should have been mis en examen in the instruction and questioned by the juge with the safeguard of legal representation.53 Contradictions in the evidence of witnesses were not followed up, and rather than the juge investigating à décharge, as he is legally required to do, it was left to the defense to try to uncover crucial gaps. Key witnesses, including the (p.266) police officer heading up the enquiry, were not called, as their evidence did not confirm the juge’s own case thesis.54

The findings of the Outreau review bodies also call into question the effectiveness of recent reforms and the ability of the defense to pierce the culture of judicial corporatism. In the 1990s and notably in 2000, a variety of due process safeguards were introduced into French criminal procedure, strengthening the role of the defense and introducing greater debate and transparency into the criminal process. In particular, provision is made for defense lawyers to participate in the instruction enquiry and for key acts and measures to be debated. This is further enshrined in the principle of contradictoire set out in the preliminary article of the code of criminal procedure, which opens: “Criminal procedure must be fair and accusatorial and it must ensure equality of arms.” The Outreau case therefore, in theory, offered the opportunity for the defense to participate, to challenge acts of judicial investigation, to request that specific enquiries be made and experts instructed, and to have the defense case considered during the investigation. However, the enquiry documented the overwhelming dominance of the prosecution perspective of the magistrats in the criminal investigation and the marginalization of the defense at all stages, leading them ultimately to question whether, given the negative experiences of defense lawyers and their treatment by magistrats, it can really be said that French criminal justice is either contradictoire or fair, as anticipated by the procedural code.55

The excessive amounts of time spent in custody by those under investigation was also highlighted in the parliamentary report. The 2000 reform had introduced a new procedure, whereby the juge d’instruction can release a suspect on bail, but detention of the mise en examen can only be decided by the juge des libertés et de la détention (JLD) after the parties have had an opportunity to make representations. Until that time, the juge d’instruction was empowered to remand in custody those she was investigating. The hope was that this new procedure would reduce the number of suspects held in custody, as well as respect better the phases of investigation and judgment, but in practice it made little difference, with 90 percent of detention requests being granted by the JLD.56 Although the idea was to have the decision made by someone more objective and distanced from the investigation, ironically, lawyers in the Outreau case were critical of the JLD’s lack of knowledge of the case and so inability to engage with the arguments (p.267) of the defense.57 Four of the accused were kept in custody for two and a half years, and three others for more than three years before the case was sent to trial.

In their reports on the investigation of the Outreau case, all three bodies of enquiry identified the kinds of role malfunctioning and external pressures that are often present in miscarriage cases whatever the procedural tradition—notably a failure to collect evidence in reliable ways; an uncritical reliance on (ultimately discredited) expert evidence; a culture of disbelieving the defense; a failure of system checks and balances; and pressure from the media.58 However, the fault here lay not only with overly conviction-oriented police officers, or even the prosecutors who were too dependent on them, but more significantly, with judicial officers who are understood to act in the public interest, protect the rights of the accused, and maintain a neutral stance in their conduct of the investigation. This might cause us to interrogate our understanding of the strengths of different models of procedure. When wrongful convictions are overturned in England and Wales, some have called for more inquisitorial procedures such as the instruction, as a means of avoiding the premature narrowing of suspicion by the police and the resulting construction of a case against an accused identified early in the investigation. Yet we see similar dangers in both procedural models—a confirmation bias that extends across police, prosecutors, and judicial investigators.59 Neither the professional ideology of the magistrat nor the checks and balances of judicial colleagues and lawyers are sufficient to ensure that the juge avoids being captive to the most plausible view of the case and investigates properly the evidence that points away from the suspect’s guilt.

L’affaire d’Outreau had a profound impact on the legal system, in parliament, and across French society. This was not the result of technical legal error or reasonable, but ultimately unproved, accusation. Neither was it the product of rogue police officers or incompetent lawyers. The case went to the heart of the inquisitorial ideal, the functioning of the judicial investigation through the iconic instruction. It also recognized a new category of “victims of the judicial system”—a system that was acknowledged to be at fault, to have ruined the lives of innocent citizens, and for which judges and politicians alike apologized in the strongest of terms.60 The case demonstrated the deep fault lines running through inquisitorial (p.268) procedure: the enormous discretion vested in the judiciary and how, without exceeding the limits of the law, they were able to present a case built on lies and unreliable accounts as an objective judicial investigation uncovering the truth. The experience of those accused in the Outreau cases was not unusual: one-third of those in French prisons at that time were held in détention provisoire and one-tenth of those were likely to be found innocent of all charges.61 It was as if the everyday practices of criminal justice were being exposed for the first time.62

C. Reforming Inquisitorial Procedure

After six months’ work, during which over two hundred witnesses were interviewed, the parliamentary enquiry produced a six-hundred-page report, containing eighty recommendations for major reforms across French criminal procedure. The failings of the criminal process and their catastrophic consequences for those placed under investigation were acknowledged throughout the report, as each phase of the instruction was examined. The proposals for change touched most aspects of the pretrial process, from police detention and questioning to the role of the chambre de l’instruction in overseeing the conduct of the instruction.63 The recommendations sought to dilute the worst excesses of inquisitorial procedure by making it less secret, more accountable, and more accusatorial—taking more seriously the principle of contradictoire as an organizing feature of the criminal process.

The commission proposed stronger defense rights throughout the criminal process and during the GAV in particular, including allowing the suspect’s lawyer to be present during interrogations and to have access to the dossier when police detention is prolonged beyond twenty-four hours, as well as video recording of all interrogations. It proposed introducing more contradictoire features into the mass of investigations overseen by the procureur, and more collegiality of (p.269) decision-making to address the isolation of the magistrat. Decisions to place those mis en examen in custody would shift from the JLD to a group of three judges, as would the process of instruction. This was commendable in its recognition of systemic problems, rather than attributing fault to isolated errors of human judgment. Yet adding some checks and balances to what remained a magistrat-controlled enquiry fell short of the kind of root-and-branch reform needed, and the basic structures of the criminal investigation were left unchanged.

Of the eighty recommendations, only a small number were taken forward by the minister, and even fewer have resulted in any concrete change. The judiciary has shown itself to be open to change, and the training of magistrats has been adapted to take account of some of the lessons of Outreau. Judges must now spend a period of time at the start of their career working alongside defense lawyers to ensure that they understand the importance of this perspective within the justice process, and the professional training and recruitment of magistrats takes more seriously the human and psychological dimension of being an investigator and the need for good social skills, as well as technical legal competence. However, recommendations that have required legislation and which challenge the centrality of the police and judicial investigation have proved much more difficult to implement.

(i) Strengthening the Principle of Contradictoire

Many of the proposals following the review of the Outreau affair addressed the broader dysfunctional working cultures and underlying systemic problems demonstrated through individual error or wrongdoing. The driving rationale of these proposed reforms was to strengthen the accusatorial nature of French criminal procedure, to decentralize roles, and to introduce more checks and balances—in Damaška’s terms, moving further away from the vertical model of hierarchical control and authority toward features of the coordinate horizontal model of cross-checks and balances.64 The reforms of the 1990s and 2000 had gradually increased the role of the defense lawyer by including her within the structure of criminal procedure and the conduct of the most serious investigations. However, (p.270) there was resistance to increasing defense rights, even in these relatively modest ways: the defense role was seen as the antithesis to an effective investigation and so particularly inapt within an inquisitorially rooted procedure. Mistrusted within the professional and legal cultures of French inquisitorialism, the role of lawyers continued to be sidelined in practice.

However, building on the principle of contradictoire, the parliamentary proposals sought to allow the defense to play a more effective role and to be involved earlier on in the investigation process. The recommendations are interesting in several respects. First, they recognize the limited impact of these earlier reforms in practice, which allowed the continuing dominance of the police and procureur and the resistance to defense engagement in the investigation on the part of both police and magistrats. Second, they recognize the determining influence of the pretrial phase and the various ways that the concentration of investigative power, even in the hands of a judicial officer, leads not to a neutral enquiry but risks confirmation bias and so to the construction of a case against the accused.65 In particular, they take much more seriously the importance of the GAV in shaping the case investigation, rather than seeing it as simply a preliminary police enquiry. Without proper information provided to the suspect and the possibility for the defense to provide assistance, the case quickly becomes constructed around a single prosecution perspective. Third, they recognize the centrality of the procureur in all criminal cases—including those passed to the juge d’instruction. It is interesting to see how these themes are repeated in later reviews of the criminal process, in decisions of the Conseil constitutionnel, and in the transnational principles developed through the ECtHR and EU legislation.

The traditional approach to the production of evidence within French criminal justice is to maximize the information available on the grounds that this assists in the discovery of the truth (the overarching objective of the criminal process), and to trust in the magistrat, as a public-interest-centered judicial officer, to investigate, scrutinize, and assess the reliability of evidence to this end, that is, to investigate evidence pointing toward and away from the guilt of the suspect—à charge and à décharge. This, of course, becomes a dangerous exercise if narrowed to evidence gathering that points only to the guilt of the suspect and is not subject to robust evaluation in terms of its credibility and relative weight within the enquiry. These risks were all too apparent in the Outreau affair: suspects were processed through the GAV, with little opportunity to contest the accusations against them, a task that became all the more difficult as the enquiry gained momentum and accusations were repeated in the instruction. Neither the procureur nor the juge d’instruction was concerned to test out the (p.271) credibility of the evidence gained during police detention and interrogation. The duty to investigate à décharge was ignored, and the field of evidence gathered was narrowed prematurely in the process of constructing a case against those considered suspects.

The recommendations of the parliamentary enquiry sought to address this bias by providing the suspect and defense lawyer with more case-related information and a greater role during the GAV. It recommended that more information be provided to suspects, setting out the precise detail of the charges; that a higher threshold of suspicion be required before placing someone in GAV; and that, from the extension of detention after twenty-four hours, the lawyer be present and have access to the dossier.

Although stemming from the malfunctions revealed in the Outreau affair, these recommendations foreshadowed the eventual, more far-reaching reforms that were required in order to comply with ECHR requirements, notably the right to effective custodial legal assistance following the landmark case of Salduz v. Turkey, discussed in preceding chapters. Similar proposals were also put forward by the Léger Commission report in 2009. Set up at the behest of President Nicolas Sarkozy, the Léger Commission was charged with examining the French criminal process and proposing ways of making it more coherent, while ensuring that it was responsive to increasing the effectiveness of crime fighting and provided greater rights to accused persons and to victims. The proposed abolition of the juge d’instruction was the commission’s headline recommendation, but it also proposed that more information be provided to suspects and that the lawyer be able to intervene after twelve hours of GAV and be permitted access to any statements made by the suspect at this time. If the GAV was extended for a second twenty-four hours, the commission also proposed allowing the lawyer to be present in the suspect’s interrogation by the police.

Although an improvement, the role of the lawyer envisaged in both these recommendations and those of the various Outreau commissions was still somewhat marginal to the police investigation—suspects would continue to have no lawyer present during the police interrogation in the first twelve or twenty-four hours of detention, and even when a lawyer was present, there were no plans to allow her to participate in any meaningful way. This was a tinkering at the edges, rather than confronting the shortcomings and dysfunction of what purports to be a procedure of judicially supervised investigation.

A second strand of recommendations designed to strengthen the principle of contradictoire, following Outreau, sought to enlarge the role of the defense lawyer in challenging the pretrial investigation and so to weaken the procedural reliance on police and magistrats alone. At the time of the Outreau recommendations, the lawyer was allowed only a thirty-minute consultation with the suspect, and even now, although she may also be present during the police interrogation (p.272) of the suspect, she still has no role beyond the GAV in cases overseen by the procureur. As a result, more than 95 percent of cases proceed without the opportunity for any defense participation or any element that might be termed contradictoire. This has implications beyond those cases dealt with only by the procureur. Even in the tiny minority of cases passed to the juge d’instruction, crucial initial investigations will have been carried out, including questioning of witnesses and the suspect. The Outreau case, as well as earlier miscarriages of justice, demonstrates the extent to which the initial imprint of the case provided by the procureur-supervised police investigation can influence the subsequent enquiry.66

In order to provide greater opportunity for cross-checking and challenging the direction of the enquiry, the parliamentary enquiry recommended that, where possible, the defense have the opportunity to participate in cases supervised by the procureur in a fashion similar to those dealt with under instruction—that the defense be able to consult the dossier, make observations, and request that specific further investigations be carried out. The commission recognized, however, that this would have a limited impact, as there would not be sufficient time for this to happen in many cases where the accused is sent straight to court.

This is not to say that the defense role during the instruction is unproblematic. In theory, the defense lawyer is able to participate in the instruction investigation, making her own investigative requests as well as monitoring those of the enquiry. In practice, however, the Outreau case demonstrated how this role can be frustrated by the juge and then compounded by those responsible for investigative review, the chambre de l’instruction, whose tendency is to confirm, rather than to challenge, the decisions of fellow magistrats.67 This goes to the heart of what underlies miscarriages of justice—the absence of any effective mechanism for testing out the evidence or challenging the case theory during the investigative stage. The trial stage (if indeed there is a trial) is too late—this is about processes that determine which evidence is brought before the court or others determining the outcome of the case, ensuring that relevant leads are pursued and important evidence is not omitted.68 A major flaw in the inquisitorial model is entrusting the discovery of the truth to a procedure that is conducted or dominated by the public prosecutor—which in many ways is not dissimilar to England and (p.273) Wales and the investigative monopoly of the police and prosecution, other than the greater impact of (the more numerous and detailed) rules of evidence that govern investigations and the admissibility of evidence in adversarial procedure. The procureur’s judicial status means that there is an assumption that evidence is reliable, without the need for scrutiny of the process of evidence gathering. Post Outreau, this has been shown not to be the case, and the various commissions of enquiry recognized that a major cultural shift was required not only to enable the lawyer to carry out the role assigned to her, but also to strengthen her function as a check on the investigation.

The problem of judicial corporatism during the instruction was also recognized by the reviews. They sought to address this by strengthening opportunities for the defense to contribute to and to challenge the investigation, from the outset through to the decision on how to proceed at the close of the instruction—what Brants and Field describe as the necessity for a culture of institutionalized dialogue.69 For example, the IGSJ review proposed allowing the defense to challenge the accused’s status as mise en examen throughout the instruction, rather than just at the outset, in order to maintain an ongoing review of the evidence against her. And where a challenge is brought to the chambre de l’instruction, it recommended that the accused be permitted to appear in person: currently the procedure is a paper exercise. The parliamentary commission heard that lawyers had only obtained copies of the case dossier after protesting to the head of the court circuit, and so in order to enable them to work more effectively, it advocated for better resourcing to make copies available to the defense. It also recommended allowing the defence lawyer to be present for all witness statements and confrontations as well as introducing the possibility of challenge in the selection of experts and the framing of their instructions.

In order to ensure the adoption of a more contradictoire approach to the instruction, a change in the approach and mindset of the juge is also needed. The commission recommended that the juge d’instruction be required to sum up the evidence that had been examined over the course of the enquiry both for and against the suspect—à charge and à décharge—in order to make explicit how this duty had been fulfilled. Once the parties have had a chance to respond, the commission also recommended that any decision, whether to discontinue the case or send it to trial, should first be debated by all parties in public. While this last recommendation has not been taken up, Article 184 CPP now requires the juge d’instruction to invite comments from the parties on her final decision, and to specify the evidence for and against the accused. This was effective for a short time, but gradually judges reverted to their previous habits, and the legislation (p.274) has been undermined by the courts’ decision not to enforce the requirements of Article 184 CPP.70

(ii) Regulating the Conduct and Recording of Interviews

As an inquisitorial-type process, the place of written rather than oral evidence is often emphasized as a feature of French criminal procedure. This is perhaps most evident during the instruction and the juge is assisted in this by a professional legal secretary, the greffier, who helps to administer the procedure and records witness statements taken by the juge. However, much of the evidence is gathered by the police—either in the first days of the investigation, or once the instruction is opened, through the delegated powers of the commission rogatoire. The taking and recording of evidence by the police has not been well regulated. Failing to recognize the centrality of evidence gathered at this early stage, the statements of witnesses and suspects, until very recently, have not been tape-recorded, nor written down verbatim, but recorded as a formal statement, lacking the spontaneity and so the authenticity of the interviewee’s own voice.71 With no guarantees as to the circumstances in which individuals are questioned, nor of the accuracy of the interview record, the weight attached to these statements once incorporated into the dossier and so “judicialized” seems misplaced.72

These dangers are amplified when those questioned are children and so more vulnerable. In 1998, a requirement to videorecord the testimony of child victims was introduced in cases of sexual assault and other serious offenses, and this was extended to child suspects in 2000.73 However, in the Outreau affair this was ignored, sometimes because of lack of resources, sometimes because this was the child’s preference: only seven of the one hundred interviews conducted were videorecorded. Those reviewing the conduct of the case found that it was impossible to evaluate the credibility of the children’s accounts, (p.275) when some had been interviewed seven times and up to ten years after the events were alleged to have taken place. The written accounts appeared unrealistic with names and dates reeled off and no indication as to the questions posed, nor whether information was spontaneously provided. The number of accused grew as the children exchanged information among themselves, and those responsible for the children’s protection appeared to bolster and encourage their accounts. This was described as a proliferation of increasingly fantastical accusations: children accused those they saw at the police station while in the waiting room, in a sports shop, in a cafe, and each time, social workers made a report and passed these to the procureur.

Another focus of post-Outreau reform, therefore, was to improve the reliability of children’s evidence: ensuring that the accounts of children were taken in an appropriate setting and by officers and social workers who were properly trained. Recommendations were made to ensure that the testimony of children is always videorecorded, providing as complete an account as possible, enabling observation of the child’s behavior, and avoiding the child having to repeat the account.74 Perhaps unsurprisingly, just as with the introduction of tape recording in England and Wales, evidence suggests that these recordings are rarely watched. This may be an indication that there is little to contest (indeed suspects’ interviews may only be consulted if there is a challenge) or the result of inadequate resources—though evidence suggests child interviews tend to be no more than thirty minutes in order to accommodate children’s difficulty in concentrating for longer periods of time.75

The concerns of the Outreau enquiry went beyond the difficulties in obtaining reliable testimony from children, however, and the commission recommended the audiovisual recording of all interrogations conducted during the GAV. The eventual reform, however, extended only to those suspects questioned by the police or by the juge d’instruction, when under investigation for the most serious offenses, crimes. This means that the overwhelming majority of suspects questioned will have neither video nor audio recording; the interrogation will be recorded only in writing. The commission’s reforms sought to bring real improvements to the process of police detention and interrogation: to provide suspects with legal assistance; to provide details of the case against them; to record all interrogations; to strengthen the oversight of the procureur; and to improve the material conditions of detention. Yet little has changed. Suspects have no access to the dossier of evidence, the procureur (p.276) does not exercise any greater control over the GAV, and just a small minority of suspects have their interrogation recorded. Lawyers are now present during the GAV, but this was not a response to the Outreau recommendations, but rather the result of France’s hand being forced by the ECtHR—and even then the lawyer’s role has been tightly circumscribed by legislation. And finally, the contrôleur général des lieux de privation de liberté (the inspector of places of detention) continues to highlight the appalling conditions in which those placed in GAV are held. Despite the commissions and reports and the public emotion following the “judicial fiasco” of Outreau, the failure to recognize the importance of ensuring the credibility and reliability of evidence gathered by the police continues; defense rights are still weakest where the suspect is most vulnerable—during police, rather than judicial, detention and interrogation.

(iii) Dispersing the Pretrial Power of the Magistrat

In addition to strengthening accusatorial measures and regulating the conduct and recording of testimony from suspects and witnesses, the commissions of enquiry suggested a range of measures that dilute the inquisitorial nature of the pretrial investigation by dispersing the power of the juge d’instruction. The Outreau case demonstrated how isolated both the juge d’instruction and the juge des libertés et de la détention are in their operation, and this was considered a major factor in the dysfunctional nature of their decisions in the case. The Vallini Commission in particular considered the abolition of the juge d’instruction, but rejected such a radical reform. Unlike the Léger Commission that followed, the Vallini review recognized that simply shifting power to the police and prosecutor would not have produced a better outcome. The abolition of the juge would require, as the Delmas-Marty Commission set out comprehensively fifteen years before, strengthening both the investigative power and the independence of the procureur. The commission decided that a reform on this scale was both unnecessary and unlikely to resolve the issues before it.

While wishing to avoid the worst excesses of inquisitorialism, a principal concern of the commission was that abandoning the model of pretrial judicial investigation would require accused persons to conduct their own investigations, the quality of which would depend on the financial resources available to them. This is a familiar argument. The adversarial model of defense participation as a party to the case, supplanting the judicial enquirer during the pretrial, is generally regarded as weakening rather than strengthening the equality of arms, as well as undermining equal treatment. The uniform competence of the magistrat is assumed, while the performance of the defense lawyer is seen to depend on (p.277) access to resources.76 Instead, the commission proposed maintaining the juge, but reducing her isolation by making the role more collegial. By working with others, young juges would in effect be mentored, and in all instances there would be less opportunity for highly individualistic or inappropriate approaches to investigation.77 Sharing responsibility for the investigation (rather than just its review) would encourage a more objective and evidence-based approach, as reasoning would be open to discussion and challenge in the normal course of decision-making.

This proposal looked unlikely to make it to the statute book after the subsequent Léger Commission’s recommendation to abolish the juge d’instruction role completely.78 However, the Léger proposal was not adopted, in part because the corresponding shifting of investigative judicial authority to the procureur that this would require became untenable after the ECtHR questioned the judicial status of the procureur.79 In 2007, a law was passed to introduce collegial decision-making in the instruction, with three judges working on each case, but without the corresponding budget required to recruit the additional 314 magistrats needed to implement the reform. The legislation was a lame duck. The number of magistrat posts continued to fall, and the reform was delayed repeatedly until May 2017, when it was finally abandoned before it had ever been implemented. Informally, magistrats have reacted by increasing the practice of co-saisine, so that there are two judges in a case to share information and decision-making. However, because no extra resource has ever been placed into this initiative, there are insufficient personnel to truly share the work, and one magistrat will effectively take charge.80

Rather than focus on strengthening the defense role or separating out the functions of the juge d’instruction, the IGSJ recommendations tend toward ensuring the proper functioning of structures already in place. For example, many of the errors in Outreau might have been avoided if the procureur had been encouraged to work more proactively in the instruction, challenging weak (p.278) evidence and protecting the public interest by ensuring that excessive charges were not brought. The IGSJ also recommended greater cooperation between the Parquet and the juges du siège in cases involving juveniles and sexual offenses, in order to maximize experience and expertise.81 These proposals are unlikely to be successful, however, without parallel changes addressing the professional culture of judicial corporatism in which as fellow magistrats, the prosecutor, juge d’instruction, JLD, and judges in the chamber de l’instruction are aligned.

The excessive use made of détention provisoire was also a concern of the parliamentary commission, given the lengthy periods of time suspects spent in custody during the instruction. The recommendation was for détention provisoire to be determined by a bench of three judges, rather than a single JLD, and that it be used only exceptionally and on clear grounds of necessity. It proposed placing limits on the time that a person under investigation could be held in custody, but this was rejected, and instead, a system of reviews every six months was put in place. The number of those detained dropped initially following this reform, but it has now risen again to the levels it was prior to Outreau, and lawyers complain that nothing has changed and that their procedural challenges continue to be rejected systematically.

D. Conclusion

The extent of the response to the Outreau affair suggests that it struck at the core of French criminal justice. Cherished at home as a truth-seeking investigator whose judicial status and ideology ensure the protection of individual rights and freedoms, the juge d’instruction is admired abroad as a model of enquiry that avoids the tunnel vision and confirmation bias of police investigations. Yet three separate commissions of enquiry into the Outreau affair showed the many dangers of concentrating investigating power in the hands of a single legal actor and of placing unswerving trust in the actions and ideology of that person. The response was a tempering of the inquisitorial ideal, in some respects continuing the trend to make French criminal procedure more contradictoire. The recommendations touched on many features of the pretrial process, but neither the proposals nor their limited implementation posed any radical challenge to the centrality of a model of investigation understood to be carried out under the authority and oversight of a judicial officer. The participation of lawyers and the means of questioning and recording witnesses and suspects changed little. This is a long way from any procedural turn, but it is perhaps a gradual acceptance (p.279) of the need for dialogue and evidential testing and challenge, in place of a more closed and centralized enquiry. This middle ground sits comfortably within the trajectory of recent decades, rejecting the need for root-and-branch procedural reform in favor of an analysis of procedural traditions, not as entirely separate, but as being more or less contradictoire.82

Some might argue that this is not enough, however.83 It fails to address the central problem of dualism contained within the instruction, or the broader issue of judicial corporatism, which prevents the defense from playing an effective role within the pretrial process. Justice malfunctions less through outright rule breaking than through the ways that discretion is exercised and the development of working cultures that contradict the stated aims of the law. Change requires both legal and structural reform. To achieve this would require the more radical solution of separating out the training and career paths of the Parquet and the juges du siège, changing what it is to be a magistrat. There is also a risk that if a more collegial approach were to be adopted, it would fail. We have seen the inability and unwillingness of magistrats to challenge the decisions of their colleagues, and a committee of three may serve to reinforce decisions, as we saw with the chambre de l’instruction, making it harder, not easier, for the defense to engage with, and to challenge, the conduct of the enquiry. The nature of the magistrature means that, despite the separation of functions, there is a continuity that is hard to breach. Lawyers are poorly resourced and regarded as outsiders, unsuitable to contribute to a public-interest-centered investigation. A more effective reform might be to open up the pretrial phase to greater external scrutiny, to move away from internal debate and make the procedure more truly open and contradictoire.


(1) The “term miscarriage of justice” is used to describe wrongful convictions. The vexed question of factual guilt or innocence is discussed separately in the following chapter. For a discussion of how the rhetoric of innocence plays out in different parts of the legal process, see Jenkins (2013), who was himself wrongly convicted and imprisoned.

(2) See, e.g., the classic debate between Langbein and Weinreb (1978) and Goldstein and Marcus (1978a, 1978b).

(3) In the Netherlands, the Schiedammer parkmoord (murder in Schiedam park), in which a man falsely confessed to the rape and murder of a ten-year-old girl in 2000, then retracting that confession, was a catalyst for debate around system change. See Franken (2008). In India too, the Malimath Commission considered very briefly the French criminal process. See Amnesty International (2003).

(4) Roach (2010b) discusses the role of innocence commissions in the United States, both in advocating reform to prevent future miscarriages of justice and in investigating wrongful convictions. He argues for greater clarity in distinguishing these two functions in order that commissions have the resources and expertise appropriate to their function. See also Roach (2015) for a comparison between Canadian and Australian approaches to miscarriages of justice.

(5) Chapter 2 considers the wider criminal justice reform following these miscarriages of justice. Although the criteria for appeal were widened and a new process for investigating possible miscarriages of justice was established through the CCRC, other reforms such as new defense disclosure obligations and the attenuation of the right to silence had played no part in the miscarriage cases and weakened the position of the accused.

(6) For discussion of some of the problems of expert testimony and the approach of the courts and of the CCRC, see O’Brian (2011).

(7) Belloni and Hodgson (2000). Garrett (2011) notes that many US exonerees come to light by chance, describing them as the tip of the iceberg.

(8) It also received more than seven hundred letters describing cases investigated in the same ways as the Outreau affair.

(9) For discussion of the exercise of discretion as a key determinant of the functioning of processes of criminal justice from the decision to arrest, through to leave to appeal and the admissibility of evidence, see Hodgson and Mou (2019).

(11) See McConville et al. (1994) for a discussion of the processing of guilty pleas by defense lawyers in England and Wales; Horne (2017) for a broader discussion of the nature and implications of guilty pleas; Hodgson (2005) and (2012) for an account of the abbreviated procedures in France; Killias (2008: 150–151) for a description of wrongful convictions in Switzerland for minor offenses dealt with by a penal order (ordonnance pénale or Strafbefehl).

(12) See Garrett (2011) for a thoroughgoing analysis of US wrongful convictions. They too are characterized by “corrupted evidence, shoddy investigative practices, unsound science and poor lawyering” (7).

(13) The process of révision is rare, with only eleven people having their convictions quashed since 1945—six of these have been since 1999. Convicted in 2009, the former mayor of Vence, Christian Iacono, was the most recent person to have his conviction overturned, in March 2015. The Cour de révision requires a new fact or evidence that was not available to the court and which could establish innocence or cast doubt on the conviction. Since 2014 the procedure has been simplified and extended to cover cases condemned by a judgment of the ECtHR.

(14) In his analysis of 250 exoneree cases, Garrett (2011) found that 40 (16 percent) made full confessions, all but two of them including details that only the killer or rapist would have known. These were most likely improperly disclosed to the suspect by the police. More than half of the exonerees were juveniles or had some mental impairment; 76 percent were misidentified by an eyewitness (many of whom were pressured by the police and then expressed their lack of certainty at trial); 61 percent of the cases contained invalid forensic evidence, with few benefiting from defense experts to challenge them; 21 percent included claims from informants, cellmates who said that they had overheard admissions; and many exonerees received poor representation.

(16) These were expressed to the trial and appeal prosecutors, but were not in their final report.

(17) As a result of this case, new legislation in force from 1 October 2012 broadens the grounds of appeal. A new expert report will allow for the opening of a review, as will a new source of evidence if it reveals a serious suspicion that the accused would not have been convicted if the information had been presented to the judge. However, a parallel law has been introduced that would allow an acquittal on a serious charge (e.g., murder, manslaughter, rape) to be overturned if it appears that there is new evidence to prove guilt.

(18) Paul Hill, Gerry Conlon, Patrick Armstrong, and Carole Richardson, the so-called Guildford Four, were convicted in 1975 of the murders of five people in pub bombings in Guildford in 1974. Their convictions were quashed in 1989 when the police were found to have fabricated notes of interviews. The Maguire Seven, Anne Maguire (Gerry Conlon’s aunt), her husband Patrick Maguire, two of her sons Vincent and Patrick, her brother William Smyth, her brother-in-law Patrick “Giuseppe” Conlon, and a family friend, Patrick O’Neill, were all convicted of possessing explosive substances in 1976. Their convictions were also overturned when the scientific evidence (the mainstay of the case) was found to be wholly unreliable. Prime Minister Tony Blair issued a public apology for the injustice that both the Guildford Four and the Maguire Seven had suffered. The so-called Birmingham Six served sixteen years in prison and had their two earlier appeals dismissed, before their convictions were finally quashed in 1991. The police had fabricated interview notes, and the scientific evidence was again unreliable. The six, Paddy Hill, Richard McIlkenny, Johnny Walker, Billy Power, Hugh Callaghan, and Gerry Hunter, had been convicted in 1975 of the murder of twenty-one people in two pub bombings in Birmingham in 1974.

(19) On the crisis in public confidence in criminal justice, see generally McConville and Bridges (1994); Field and Thomas (1994); Walker and Starmer (1999).

(20) See discussion of cases such as Ivan Fergus, Stefan Kiszko, and the Cardiff Three in Belloni and Hodgson (2000: chap. 1).

(21) See I. Burrell and J. Bennetto, “Police Unit to Blame for ‘Dozens More Injustices’: Miscarriages of Justice Emerge 10 Years after Break-Up of Group That Tortured Suspects,” The Independent, 11 November 1999, 8. Michael Hickey, Vincent Hickey, James Robinson, and Patrick Molloy were wrongly convicted in 1979 of the murder of schoolboy Carl Bridgewater. Patrick Molloy died in prison in 1981; the other three men spent seventeen years in prison before their convictions were quashed.

(22) Derek Treadaway was jailed in 1983 but had his robbery conviction overturned and went on to successfully sue West Midlands police for his treatment at the hands of the West Midlands Serious Crime Squad.

(23) Kaye (1991: 50–56). The number of those found to have been wrongly convicted as a result of the squad’s activities continues to grow, while the conviction of officers concerned has proved impossible—other than one former officer, Lawrence Shaw, who was jailed for twelve years for armed robberies in 2001 (http://www.telegraph.co.uk/news/uknews/1311129/Ex-detective-who-turned-to-crime-jailed-for-12-years.html) and then again for eleven years in 2010 (http://www.bbc.co.uk/news/uk-england-12053559).

(24) Paris, Abdullahi and Miller (1993) 97 Crim. App. R. 99, 103–105.

(25) The objectivity of expert evidence, as well as juries’ ability to make good judgments as to their reliability where there may be genuine disagreements on the science, is contested. See Redmayne (2001: 116–139). In relation to the CCRC’s approach and the value of a third expert to assist in evaluating the competing arguments, see O’Brian (2011: 20).

(26) In the Sally Clark case, criticizing the Crown’s pathologist for failing to disclose information about important and relevant tests that he had carried out, Professor Berry, a pediatric pathologist, explained that it was best practice to include all test results “to provide a complete record for anyone else who might need to interpret the post mortem report later. . . . Defence experts are completely dependent on the original pathologist and the Crown to supply all relevant results and materials.” R v. Sally Clark [2003] EWCA Crim. 1020, para. 148.

(27) R v. Sally Clark, paras. 172–180.

(28) Professor Morris, a consultant pathologist with particular expertise in the role of bacteria and bacterial toxins in SIDS—sudden infant death syndrome. R v. Sally Clark, para. 150.

(29) Professor David, a consultant pediatrician instructed by the Family Court as an independent expert. R v. Sally Clark, para. 149.

(30) A member of her defense team described her wrongful conviction and imprisonment for the deaths of her own children as “state-sponsored torture.” See Thair Shaikh, “Sally Clark, Mother Wrongly Convicted of Killing Her Sons, Found Dead at Home,” The Guardian, 17 March 2007, available at https://www.theguardian.com/society/2007/mar/17/childrensservices.uknews.

(31) Section 23(1) Criminal Procedure and Investigations Act 1996 and Code of Practice 3.5.

(32) Henry Blaxland, QC, quoted in Sandra Laville, “Sam Hallam Released after Seven Years in Prison,” The Guardian, 17 May 2012.

(33) Trust tells us much about the design and function of criminal justice systems and the structures of accountability. See Nelken (1994). The tensions around prosecutorial independence and accountability to the executive discussed in chapters 4 and 5 also reflect this. See, e.g., Guarnieri (1997) re low levels of trust in the executive and the strengthened independence of the Italian prosecutor.

(34) Dongois (2008). New procedures were introduced in 2001, allowing for appeals from the cour d’assises. The convicted person or the procureur may appeal. Between 2003 and 2005, 95 percent of convictions appealed were upheld, but only 43 percent of acquittals that were appealed. Over the three-year period of the study, 10,478 people were convicted in the cour d’assises; 1,262 appealed their conviction and 64 people were successful in having their conviction overturned. Of the 652 acquitted at trial, the procureur appealed the decision in 76 cases, and 43 people were then convicted on appeal. (See Le Monde, 21 June 2001.) The procedures for re-examining and revising convictions was simplified and expanded slightly in 2014. A conviction that has already been appealed may be overturned if there is new evidence that suggests innocence or casts doubt on the guilt of the person convicted. New investigations can be conducted by the Commission de révision des condamnations pénales, or, at the request of the convicted person prior to bringing the case to the commission, by the procureur. Christian Iacono was only the eleventh person since 1945 to have his conviction overturned and his innocence confirmed under this procedure. His conviction for the rape of his grandson was quashed in March 2015, two years after his grandson admitted to having lied and fabricated the accusations. He had served sixteen months in prison on remand at various times, and it took fifteen years for him to clear his name. Other relatively recent réhabilitations are those of Abdelkader Azzimani and Abderrahim El-Jabri in 2013 after serving twenty years for the murder of a drug dealer; Marc Machin in 2012 after another man admitted to the murder for which Machin had been convicted; Loïc Sécher in 2010 after serving seven and a half years in prison, after the woman he was alleged to have raped admitted to having lied; Patrick Dils, sentenced to life imprisonment at the age of nineteen, was finally released some twelve years later; and Rida Daalouche was finally acquitted of the murder of a drug dealer in 1999 after being convicted in 1994. (See Le Monde, 25 March 2015.)

(37) The self-protecting nature of the criminal justice system was nowhere more evident than in the comments of Lord Denning, then Master of the Rolls. In rejecting the Birmingham Six’s civil action against the police for assault in 1980, he said: “If the six men win, it will mean that the police were guilty of perjury, that they were guilty of violence and threats and that the convictions were erroneous. That would mean that the Home Secretary would have to recommend they should be pardoned or remit the case to the Court of Appeal. This is such an appalling vista that every person in the land would say: ‘It cannot be right that these actions should go any further.’ ” McIlkenny v Chief Constable of the West Midlands [1980] 2 WLR 689 at 706. He went further still in an interview for the Spectator magazine, arguing that it would have been better if the Birmingham Six had been hanged, so as to avoid all the damaging campaigns in support of guilty men and against the criminal justice system.

(38) See, e.g., House of Commons Justice Committee, Disclosure of Evidence in Criminal Cases (2018); CCRC Annual Report 2016/17, p. 3. The commission also notes that decreasing police resource aggravates this further, as officers lack the time to check material thoroughly. See also, e.g., the case of Warren Blackwell, who spent more than three years in prison for a sexual assault conviction. The police failed to disclose crucial information about the unreliability of the complainant, including previous false allegations similar to that made against Blackwell. The CCRC’s written evidence to the House of Commons Justice Committee (13 December 2018, paras. 41–42) lists the “disappointingly familiar” causes of miscarriages in the cases it reviews as instances of nondisclosure of material to the defense, police tunnel vision, failure to make adequate background checks on witnesses, and new understandings of scientific evidence. Offenses relating to the nondocumentation of refugees and asylum seekers have also produced a seam of cases.

(39) The offenses against children aged between four and fourteen years took place at the offenders’ home and were filmed. The couple Myriam Badaoui and Thierry Delay abused and prostituted their own children as well as those of their friends and neighbors (there were twelve victims recognized in the case). Another couple, David Delplanque and Aurélie Grenon, were also convicted. The locality is characterized by economic and social deprivation, many residents being unable to name the French president, for example. See blog post http://anti-ogre.over-blog.com/2016/05/les-victimes-de-l-affaire-d-outreau.html.

(40) The true extent of the abuse remains in doubt in the eyes of some. See, e.g., “Innocents ou coupables? Outreau, le poison du doute,” Le Nouvel Observateur, 9 May 2011.

(41) Viout (2005). The minister of justice instructed this working party to examine what lessons could be learned from the judiciary’s handling of the case and to make appropriate recommendations.

(42) Raysseguier (2006). The minister of justice instructed the inspectorate to carry out an administrative enquiry: to examine all of the conditions and circumstances in which the case was handled by the judiciary; to investigate the impact on the case of the malfunctioning of the justice system and the professional errors of individual magistrats at each stage of the procedure; to propose reforms that would prove to be helpful.

(43) Vallini (2006). This commission of enquiry was required to investigate the cause of the malfunctions of the justice system in the Outreau affair, and to formulate recommendations to avoid their recurrence.

(46) Like many cases of plagiarism, this included factual and typographic errors (Vallini [2006: 148]).

(47) Vallini (2006: 98–112). The proximity in time of this investigation and the Belgian pedophile case of Marc Dutroux meant that public feeling around such cases was running high, and the word of a child accuser was unlikely to be challenged. Julian Licourt, “Outreau, l séisme qui aurait dû bouleverser la justice,” Le Figaro, 24 August 2018.

(48) Psychologists and psychiatrists differed significantly in their assessment of the traits of sexual abusers, for example. The commission noted the determination of the juge d’instruction to block any defense request—even interpreting requests for a new expert as requests for clarification, thus enabling him to refuse them as being out of time. There were also delays of up to a year in making reports available, preventing the accused from knowing the extent of the evidence being built up against them, or from challenging the findings in a timely manner. Vallini (2006: 186–189).

(50) Vallini (2006: 194). The report was also critical of the juge’s decision to have the police question the individual as a mere suspect, when there was clearly enough evidence to make him mise en examen and so subject to the rights and protections of being an accused within the instruction (2006: 195).

(51) See similar observations in Hodgson (2006a).

(53) Vallini (2006: 195). See also Hodgson (2005: 223–224) for accounts of how the juge d’instruction sought to keep information from the defense.

(55) Vallini (2006: 191). Resources were also identified as undermining the effectiveness of the defense. Obtaining copies of the dossier was difficult, and the poor rates of pay were also found to have made more difficult the availability of lawyers.

(56) It may be, of course, that this is not any form of malfunction, but that the juge d’instruction requests are, on the whole, well grounded.

(57) Vallini (2006: 223–227). The JLD noted that lawyers failed to make full written arguments as required. The rather vague ground of trouble à l’ordre public—maintaining public order—was also found to have been used in this case and, more generally, too frequently and as a means of avoiding properly evidenced and reasoned decisions to remand in custody. Vallini (2006: 255–256).

(59) On confirmation bias see Nickerson (1998). In relation to criminal investigations, see Kassin et al. (2003); O’Brien (2009a, 2009b); Judson (2017).

(60) Besnier and Salas (2008). The parliamentary enquiry was hugely oversubscribed, with four applicants per place, and nearly half of the group were from a legal background. In some respects the harm suffered by the child victims was overshadowed by the shock that such an injustice could have happened in France.

(61) Parliamentary report (Vallini [2006: avant-propos by André Vallini]). This compared with a pretrial detention rate of 19 percent in Germany and England and Wales and 20 percent in the United States.

(62) This was also the case for those engaged with the criminal process. One psychology expert expressed surprise at the weight attached to his opinion and that a decision to detain a suspect in custody might be based on something that he had said, taken out of context.

(63) The reforms included stronger safeguards during the GAV; making the procureur-supervised enquiry more contradictoire; limiting the length and grounds of detention during investigation; making the instruction collegial, rather than conducted by a single judge; removing the secrecy of the instruction at the defense’s request; strengthening defense rights throughout the investigation, including access to the dossier; reforming the chambre de l’instruction; improving the quality of expert evidence; protecting the interests and improving the evidence of children; reforming the training of magistrats and the professional ethics of judges and the media; increasing parliamentary accountability for criminal justice police; and, finally, making sufficient resources available to bring about the necessary reforms.

(64) See Damaška (1975). The vertical model is typified by inquisitorial procedures, where legitimacy resides in the hierarchical structure of authority, such as that of the Parquet headed up by the minister of justice. In contrast, the horizontal model, typified by adversarial systems, is a coordinate structure of checks and balances, rather than the concentration of authority in a single office or function. Damaška argued that these structures of legal authority tended to reflect the structures and cultures of political authority. For a discussion of the models in an empirical, French context, see Hodgson (2001).

(65) The various commissions made clear that the difficulties and malfunctioning in this case were rooted firmly in the conduct of the investigation, e.g., Viout (2005: 140).

(66) For example, cases such as that of Patrick Dils, described at the start of this chapter, where the suspect’s false confession to the police was impossible to displace even when before the juge d’instruction.

(67) The Vallini review characterized the chambre de l’instruction as the chambre de confirmation.

(68) That is not to say that the trial can never influence the pretrial in, for example, setting procedural standards of fairness, such that investigations are conducted in the shadow of evidence admissibility rules should the case be contested and go to trial.

(70) Julian Licourt, “Outreau, le séisme qui aurait dû bouleverser la justice,” Le Figaro, 24 August 2018.

(71) Hodgson (2004, 2005).

(72) Treating the contents of the instruction dossier as the product of a judicial investigation is problematic when statements have been taken by the police or others prior to the judge’s involvement. This is especially pointed in terrorism investigations, where unsourced intelligence becomes evidence once “judicialized” and placed in the dossier. See Hodgson (2006a).

(73) The vulnerability of children to sexual violence and exploitation had been highlighted in cases such as the Dutroux case in Belgium in the mid-1990s, as well as through a number of international conventions such as the United Nations Convention on the Rights of the Child and the European Convention on the Exercise of Children’s Rights. The success of video recording children’s testimony in England and Wales was also cited in support of the reform. The law does not apply to all child witnesses, only to victims.

(74) Having to retell their story also risks giving the impression that the child is not believed, or that what the child says is unimportant because the adults have already forgotten it and need to hear it again.

(75) Report of the Défenseur des droits (2013: 66).

(76) See discussion in Hodgson (2005: 28–29). An effective defense will require equal access to resources, but the professional ideology of legal actors is also important. No matter how well resourced, if the juge (or indeed the defense lawyer) maintains a conviction orientation, the accused’s interests will be ignored and relevant evidence not put before the court.

(77) The juge d’instruction in the Outreau investigation was twenty-nine years old and inexperienced in such a serious, complex, and large enquiry.

(78) See discussion in Hodgson (2010). Given the centrality of the juge d’instruction to the Outreau enquiries, it was ironic that their long and considered recommendations were swept aside and contrary proposals put forward by the Léger Commission. It is widely accepted that this was one of the main objectives in establishing the Léger Commission; Nicolas Sarkozy made no secret of his desire to abolish the juge d’instruction and transfer all pretrial investigative power to the procureur.

(79) Medvedyev v. France (3394/03) 29 March 2010.

(80) Julian Licourt, “Outreau, le séisme qui aurait dû bouleverser la justice,” Le Figaro, 24 August 2018.

(82) See, e.g., Létocart (2012), who suggests that the French system is a hybrid of inquisitorial (inquisitoire) and accusatorial (accusatoire), such that it might be described as “incusatoire.”

(83) Besnier and Salas (2008) argue that the various proposed reforms were no more than a sticking plaster, a temporary solution that failed to go to the heart of the inquisitorial system that has failed those it has investigated.