Criminal law has considered this question in an attempt to provide safeguards within the criminal justice system to avoid miscarriages of justice. The burden of proof is ‘beyond all reasonable doubt’, a much higher standard than the civil court’s ‘balance of probabilities’. There is the option at Crown Court to have a jury, an impartial collection of people from a variety of demographics and backgrounds, hear your case. There are legal tests and defences available for a variety of offences. A range of admissible evidence may be provided, which must be sufficient to convince the jury of the prosecution’s case, to impose criminal liability onto the defendant. The reason for this is compelling; a defendant stands to lose his liberty, reputation, job, family or financial assets, depending on the conviction. The courts have developed the way they have to avoid miscarriages of justice and ensure certainty and consistency within the legal system, an important part of the Rule of Law.
The Innocence Project (IP) provides research on cases that have been exonerated through DNA in the United States. Out of 350 DNA exonerations, 71% of the original convictions were due to eyewitness misidentification. Surprisingly 32% involved multiple misidentifications of the same person. In fact, IP have gone so far as to state that eyewitness misidentification is the leading cause of wrongful conviction as proven by DNA testing. Such revelations have triggered a need for change. A case that heard several appeals of wrongful conviction is Turnbull. Turnbull was a five-member, landmark Court of Appeal case, in which there were four appeals where defendants had contested identification evidence that made up the only, or majority of evidence against them. This is problematic as there is a ‘special risk of wrong conviction’ in cases which depend wholly or mainly on eyewitness identification evidence. This essay will consider the guidelines put in place by Turnbull, as well as other rules and guidelines which have been implemented to uphold justice in cases such as these. It will then consider the effectiveness of the operation of these rules and guidelines in practice.
What is Pre-Trial Identification Evidence?
Pre-trial identification evidence is given by eyewitnesses who saw the accused either at the scene of the crime, or close by in location and time. This evidence may support the prosecution’s case and strengthen their ability to secure a conviction. However it may vary in quality, duration and reliability. Ideally, any witness evidence would be corroborated by the presence of scientific evidence. However, on occasions it is unavoidable that an entire case’s outcome hinges on the evidence of an eyewitness, making it vital that the jury are confident in the reliability of the evidence before deciding on a verdict. In fact it has been discussed that even recognition evidence – an eyewitness who identifies someone they know at the scene – can be mistaken, no matter how earnestly they believe it. A Report to the Secretary of State regarding this matter remarked that ‘mistakes can be made even by the best witnesses in the best conditions.’ Where the witnesses only caught a ‘fleeting glance’ of the offender, the evidence can obviously prove ‘extremely unreliable’.
There are a variety of ways pre-trial identification evidence is provided. Where a witness believes they can identify the offender, the police organise confirmation of the identification; an identity parade (or other procedure) where a line of individuals that look similar to the defendant are presented to the witness, who then identifies the defendant from memory.
The Judge’s Role
Turnbull provided guidelines as to the way in which the judge was required to sum up the jury in identification cases to reduce the risk of wrong convictions. It was held that visual identification evidence was shown to be unreliable and the courts must take steps to prevent and reduce injustice. The judge must warn the jury of the ‘special need for caution’ before considering a conviction of the accused in reliance upon the correctness of the identification evidence. Turnbull is said to have ‘changed the practice, not the law’. However, the consequence of disregarding the need for a Turnbull instruction is quashing the conviction on appeal. The Turnbull guidelines are aimed at assessing the quality of the evidence:
In our judgement when the quality is good…the jury can safely be left to assess the value of it even though there is no other evidence to support it…When in the judgement of the trial judge, the quality of the identification evidence is poor…the judge should then withdraw the case from the jury and direct an acquittal unless there is other evidence which goes to support the correctness of the identification.
First then, the judge must consider whether the evidence is of such poor quality that it is anticipated to be unreliable, and it would be unwise to leave the jury with solely that information on which to make a potentially unsafe conviction. If the evidence is particularly poor, the judge should order an acquittal and withdraw the case from the jury, unless there is other evidence which supports the accuracy of the identification provided. The judge must objectively consider whether the quality of the identification evidence alone justifies leaving the case to the jury. This is an unusual amount of power in the hands of the judge, which conflicts with the general principle that the functions of the judge and jury should be kept apart. Identification cases are an exception to this, as the risk may be ‘seen as serious enough to outweigh’ the principle in identification cases, where there is a ‘ghastly risk’ connected with low-quality evidence. If the judge does see fit to continue with the case, the jury must closely examine the individual circumstances of identification evidence submitted.
Turnbull acknowledges ‘recognition’ as a type of identification evidence. Recognition is often more reliable than evidence given by a stranger, but honest mistakes may still be made. It was discussed that a warning was still necessary, even in the case of recognition by close friends or family. The Court of Appeal case of Bentley elaborated on this: ‘A recognition, which was the type of identification evidence here, could not be regarded as trouble free because many people had experienced seeing someone on the street that they knew and later discovered that they were wrong.’ This suggests that it is relatively common to be mistaken even where an honest witness is convinced that they recognised the defendant committing the offence.
The ‘Turnbull Direction’
Turnbull theorises that the risk of a miscarriage of justice can be much reduced if judges sum up their juries in the way that Turnbull instructed: whenever a case depends either wholly or substantially on the correctness of one or more witness identifications of the accused, which the defence alleges to be mistaken, this triggers the need for a Turnbull jury instruction. The judge must therefore instruct the jury as to the reason for the need for such a warning, acquainting them to the possibility that a mistaken witness and a convincing witness are not mutually exclusive. Additionally, even numerous witnesses may be mistaken, such as a group of people on a bus at night who pass a crime being carried out and all catch a fleeting glance of the offender.
Before leaving the jury with the evidence, the judge must also direct them to closely consider the circumstances under which the identification was made. There are many factors the jury should be encouraged to examine: How long did the witness observe the offender for? It makes a large difference if they caught a quick glance, or watched the situation play out; How good was the lighting? Clearly a dark environment makes an identification more dubious than a brightly lit one; What was the distance between the witness and the offender during the observation? Was there any impediment? Traffic, a crowd or even thick foliage may impede the quality; Has the witness ever seen the accused before? Familiarity may add weight to the identification’s reliability, compared to that of a complete stranger. Building on that idea, how often had they seen the accused? The more familiar the accused is to the witness, the more confidence this may suggest, although caution should still be used, to avoid the jury from settling into a false sense of confidence. Did the witness have any special reason for remembering the accused? Perhaps they have a distinctive feature that stood out and made the witness more confident about their identification. There are some factors that connect with the PACE police rules and may require some police evidence: How long elapsed between the witness’ observation of the accused, and giving the description to the police? Were there any discrepancies between the description provided, and the defendant’s appearance? The defence may request this information specifically from the prosecution, who must provide it.
The judge should remind the jury of specific and relevant weaknesses which may have appeared in the identification evidence, as well as any corroborating evidence which supports the identification. The jury also needs to be informed of any evidence at risk of misinterpretation of the jury (such as opting out of answering a question or giving evidence). The overall ramifications of this should be presented to the jury.
Flexibility of Approach
The factors discussed above, which the jury may consider, are not a binding and exhaustive list. Indeed a ‘formulaic recital of possible dangers of relying on identification evidence’ may cause the jury more confusion and be counterproductive. Rather, the guidance given should be related directly to the case at hand and be of ‘practical’ application, not ‘hypothetical’. It must be kept in mind that the purpose of a Turnbull direction is to draw the jury’s attention to any potential dangers in relation to identification evidence. This cannot be achieved by rehearsing difficulties that might affect evidence on a purely theoretical basis. A judge should be conscious of the need to relate difficulties in relying on this type of evidence to the actual circumstances of the case, on which they have to reach a verdict.
In this way it has been held as important for the judge to remain flexible with the approach to the directions given to the jury. There is no precise wording that must be followed in order to meet the criteria of Turnbull; it merely requires the important aspects of the warning. In fact, subject to the judge’s discretion, not every small issue of identification needs a full Turnbull direction. Breslin stated that the direction should be given in every identification case, however it may be worded more or less strongly depending on the case itself.
In some cases the offence itself is denied but the presence of the defendant is not disputed. One such case is that of Slater, where there was an assault in a nightclub that the defendant admitted to attending at the time. However they denied any involvement. There was an appeal as the trial judge had not given the jury a full Turnbull direction. This was not considered strictly to be an identification case and the appeal was dismissed. The court held:
The need for a direction arises only where there is the possibility of mistaken identification. Where there is no issue as to the accused’s presence at, or near the scene, but the issue is as to what he was doing, it does not automatically follow that the direction must be given. It will be necessary where the possibility exists that a witness may have mistaken one person for another.
In fact the court went so far as to say that ‘”it would be contrary to common sense” to require the direction in all cases where presence is admitted but conduct disputed.’ The need for the judge’s discretion becomes apparent in such instances as the purpose of a Turnbull guidance must be kept in mind.
Code D of the Police and Criminal Evidence Act 2003
Before considering the effectiveness of the Turnbull directions in a court setting, it would remiss to fail to consider the second layer of risk reduction, with regards to miscarriages of justice due to identification evidence. Before the matter ever reaches the court, there must first be police interaction, an arrest, and an identification. The implication of ‘pre-trial’ identification evidence is that the identification occurred prior to the trial proceedings. The Police and Criminal Evidence Act (PACE) Core D, which came into effect on 1 February 2008, provides detailed procedures for what is expected of the police throughout the identification of suspects by witnesses during crime investigations. It is the ‘Code of practice for the exercise by police of statutory powers to identify persons’. The police must comply with the Code, as this protects not only the accused, but the officers themselves. If they do not they could ‘destabilise what would otherwise be a proper prosecution and a safe conviction’.
Where the defendant disputes the witness’ identification of their person, the prosecution is put to proof that the defendant is the person who committed the offence, beyond all reasonable doubt in the eyes of the jury.
Procedures continue to evolve with new technology and law. At the current time, police use identification parades, video identification, and group identification. The police must consider the type of evidence provided, whether the quality is sufficient to secure a conviction and whether the issue of identification is relevant. In the case of Dance v DPP, the witness’ description only detailed the clothing and ages of the accused. In a case such as this, identification procedures would be unhelpful with no description of the facial features, colouring, height or anything concrete that could be used to identify an individual. The Court of Appeal ruled that procedures such as a parade are unnecessary if the witness identifies distinguishing features such as clothing rather that the offender themselves. If there is no reasonable likelihood of the witness being able to pick the suspect out in a parade, there is no reason to hold one, despite the fact that the defendant may have requested one.
Paragraphs 1.2 and 3.0 of PACE specifies that identification evidence arises when an eye-witness is given an opportunity to identify a person they have seen ‘committing a crime or in any other circumstances which tend to prove or disprove the involvement of the person they saw in a crime’. This process has a two-fold purpose: it works to test whether they can identify the suspect as the person they saw committing the offence, and provides initial safeguards against the dangers of mistaken identification.
The process includes taking an accurate record of the first description provided to police by the witness before any identification procedures take place. This avoids the witness’ memory from being adapted to those presented in the procedure, and to act as a record to check against for future reference. This record is also presented to the suspect or their solicitor for use in building a defence.
Identification Procedures
The most commonly used identification procedures is video identification as set out in paragraph 3.5 PACE. This is due to the speed, cost-effectiveness and efficiency of video identification compared with parades, and has been directed as the first option in paragraph 3.14. This procedure involves the witness being shown (usually moving) images of a known suspect alongside other similar images of people who resemble the suspect. This is similar to an identity parade where the witness is shown a line of people who resemble and include the suspect.
An identification procedure must be held in the circumstances set out by paragraph 3.12: the suspect disputes being the person the witness claims to have seen committing the crime, an eye-witness has identified a suspect, the available witness indicates that they are able to identify the suspect, and there is a reasonable chance that the witness can make the identification. The procedure must then be held unless it is either impracticable, or the procedure useless in proving the suspect’s involvement.
The identification procedures aim to provide an unbiased and decisive test as to the reliability of the identification evidence. In addition to the initial description of the suspect, the procedures are all documented and a record is kept of the procedure used, the reasons for any part of the procedure which is omitted, any lack of co-operation, and ‘the circumstances and conditions under which the person is given an opportunity to recognise an individual must be made’. This includes whether the witness was given any information concerning the suspect, what the witness was told about the offence before the procedure, how and by whom the witness was directing during the procedure, who else was in attendance of the viewing and and other relevant details. This information will help a judge in assessing the quality of the evidence at court.
Where the Code is not adhered to, a breach may lead to a conviction being quashed on appeal. In Rutherford and Palmer, the defendant had a right to request a parade. This was ignored notwithstanding the witnesses’ indication that they had a reasonable prospect of making an identification. This was held to be a breach despite the otherwise persuasive evidence. Regardless of the breach, the case was strong and this resulted in the appeal being dismissed. This suggests that where the same conviction would have been made despite the failure to follow the procedure correctly, and in the presence of other compelling evidence the court will uphold the conviction instead of quashing based on an effectively inconsequential oversight. Failure to follow the procedure was still held to be a breach, and officers are wise not to risk this sort of failure as it is quite possible that such a breach could upset an otherwise safe conviction. In fact the Crown Prosecution Service’s website states the following:
If the provisions of a relevant Code of Practice are not followed, the consequences of the failure to comply may result in evidence being excluded. However, not all breaches will render evidence inadmissible. The Court will look for a significant and substantial breach before going on to consider whether or not to exclude evidence obtained as a result of the breach.
Operation of the Rules and Guidelines
Turnbull and PACE Code D work together with the aim of reducing the risk of wrong convictions, and provide more certainty and confidence in the criminal justice system. Mistaken witnesses are an inevitable part of the fallibility of humankind, and this eventuality must be provided for and protected against. Of course, the system is not perfect; the law by its nature is both rigid and flexible.
Characteristics of Roles within the Justice System
Judges must follow the law, however they are also in a position to interpret the law with their discretion. This gives more freedom to consider the individual circumstances and facts of the case, however this can be a double-edged sword; in some cases the facts may be unusual and warrant a different approach, much in the way that some identification cases need an extensive Turnbull direction from the judge and others simply need a brief warning. Many tests and precedents have been set out in common law, such as the way Turnbull strives to protect from wrongful conviction; ‘nonetheless the end result of both this common law and statutory law is frequently a general rule open to the judge’s discretion and interpretation. Consequently judges’ discretion, albeit often within complex rules, permeates innumerable areas of the proceedings.’ Discretion can also be a good thing as it allows judges to use common sense and take a more nuanced approach instead of a rigid one-size-fits-all presumption. In addition, as provided for in Turnbull, if the evidence is not of a good enough quality, the judge has the power to order an acquittal, therefore acting as a safeguard against wrongful conviction. In addition, if this is not considered and handled appropriately by the judge, then the conviction will be quashed upon appeal by the defendant. In other cases, it is possible that judges can be biased in how they conduct a trial or instruct the jury. They may have been trying to maintain public confidence in the criminal justice system, or acting to preserve the reputation of the prosecution or police officers. This can clearly have a detrimental effect on the defendant’s situation. In most cases, the appeal system should provide justice for a wrongly convicted individual.
The jury has the guidance of an experienced judge, and they must make a unanimous decision which provides a safeguard against the whim or difference in opinion of individuals. In theory this means that there should be certainty in every conviction. When the full responsibility that rests on the jury is accounted for however, it becomes clear that their task is a delicate one. There are many factors they must consider: complex evidence, conflicting expert opinions, the adversarial nature of the arguments of the prosecution and the defence and so on. This means that the conclusion is not usually clear cut. When witness testimony, by its nature fallible, is added into the equation (as it most often is) the jury’s work can become even more complicated. The burden of proof in criminal courts is ‘beyond all reasonable doubt’. This high standard of proof required surely works to protect the liberty of all defendants but those the law is most certain have committed the offences they were accused of. However for a jury the term ‘reasonable doubt’ may appear vague. There may be issues with regards to the jury’s interpretation of ‘reasonable doubt’ leading to problems in the verdict. Oftentimes, juries are inclined to believe that if a witness has taken an oath before the court promising to tell only the truth, that any statements given by the witness in such a situation must be true. This can lead to convictions in cases where, without the witness testimony, there would be nothing but circumstantial evidence to rely upon, based on which no reasonable jury would find a guilty verdict. There is always the chance that a particularly assertive member of the jury may be able to persuade the group, or even that the judge’s summing up could confuse or mislead them. In addition to these issues, there is the ‘Friday afternoon effect’, a phenomenon causing rushed decisions, where members of a hung jury would ‘vote either way in order to get home’.
Police officers are also in a position of responsibility and are capable of making omissions and mistakes. As seen in Rutherford and Palmer, oversights and breaches can occur. In that case it was not substantial enough to impact the outcome of the case, but this is always a real possibility. Officers may be under pressure to secure convictions or make arrests, especially in high profile rape or murder cases. They may lose, overlook or misinterpret evidence that could have supported the claim that a suspect is innocent. Police ‘can easily come to see their role as achieving a conviction and if this is done beyond an overarching concern for the actual truth, the presumption of innocence is rapidly eroded.’ So it is important to keep in mind their search for truth instead of simply making arrests.
All of the above roles can have a serious impact on the cases they deal with and in a criminal court, the impact of this on the suspect’s liberty, reputation and faith in the justice system is substantial. This is why the law has taken such a serious and careful approach by putting the rules and guidance discussed in place.
How Effective are the Rules and Guidelines?
It is difficult to confidently state how effective the safeguards are in their operation. The secret nature of jury meetings makes it impossible to conclusively evaluate the value given to judge’s warnings with regards to identification evidence, however the ability of the judge to order an acquittal on the basis of low quality evidence with no corroboration is surely an effective way to mitigate the most tenuous of accusations. Academic sources have stated that ‘wrongful convictions are an inevitable risk…while there are certain safeguards that protect from some of the problems of the past, there remains a high potential for such events to occur.’
However, various academic sources agree that these changes in law have improved the outlook for reliability of convictions in identification cases. Even so, it is still seen as an unavoidable inevitability that some cases may slip through the system. In those cases, the appeal process can offer some justice. On the the other hand there is the danger that higher courts could be concerned with protecting the decisions of lower courts over protecting the potentially innocent. It has been suggested that the Court of Appeal should stay vigilant when considering such cases with an attitude of uncertainty rather than finality, as it is recognised that the ‘special risk’ of wrongful conviction certainly exists in identification cases.
Conclusion
It is clear that there are many factors that could cause an honest eye-witness to also be an unreliable witness: memory issues, stereotypes and social perception of events and people, the power of subconscious suggestion, incorrectly carried out identification procedures and many more. The presumption of innocence and the role of the defence offers the accused a strong starting point. The PACE rules attempt to ensure that the procedure and events leading up to the trial are carried out fairly and accurately. the extensive and accurate documentation kept throughout the process aims to keep the court informed and fair. Breaches are grounds for an appeal and in serious cases, quashing. In addition, the quality of the identification evidence (when heavily relied upon) is assessed before the jury are allowed to consider it, and the acquittal power of the judge is another safeguard for the defendant. Finally, the Turnbull direction from the judge is a vital warning to the jury or any weaknesses of the evidence relevant to the case, as well as a somber reminder that honest and convincing witnesses can be mistaken. Although it is far from a definitively perfect system, the standard required from those working together in the justice system is very high. A conviction should not happen unless everything was carried out fully, the witness was able to make a good quality identification and the judge and jury are satisfied beyond all reasonable doubt that the information is reliable and accurate. Overall the Turnbull guidelines and the rules set out in PACE work together to try and provide as much certainty as possible to eliminate to majority of wrongful convictions. Although it appears to be a work in progress, it could be described as a successful endeavour for the most part. Of course there is an inevitable risk as discussed, but in general this seems to be addressed and diminishing, therefore it is the exception and no longer such a common occurrence.