- The risks of mistaken identification are well accepted. There are many factors which may affect the reliability of identification evidence. A witness who gives evidence identifying the accused may be honest and convincing, but may be wrong in her identification of the accused. See The Modern Law of Evidence (‘MLE’) 13th ed at pp277-278.
- The risk of miscarriages of justice arising from mistaken identification can never be eliminated entirely, but the guidance given in R v Turnbull goes a long way towards reducing the risk, as do robust pre-trial procedures which apply to how identification evidence is obtained during criminal investigation.
- The Turnbull guidance has two general aspects. The first aspect concerns cases which rely on identification evidence which is of poor quality and is unsupported by any other evidence, the judge must withdraw the case from the jury. This might include cases where there was only a ‘fleeting glance’ or where there was a longer observation but in difficult circumstances. See MLE 13th ed at pp285-286. Where the evidence is of good quality (see MLE 13th ed at p 285) or is supported by other evidence, it may safely go to the jury. The second aspect concerns the direction which the jury must receive where the case depends wholly or substantially on the correctness of identification evidence. See MLE 13thth ed at pp 279-280.
- A crucial part of the direction is that the judge must warn the jury of the special need for caution, explaining the reason for the need for caution, i.e. that a mistaken witness can be a convincing witness and a number of such witnesses can all be mistaken). See MLE 13th ed at pp279-280.
- The judge must also direct the jury to examine closely the circumstances in which the identifying witness observed the perpetrator. For example, the light, the distance, the duration of the observation, any obstructions during the observation, whether the witness had seen the person he identified before, how long between the observation and the identification, any breaches of pre-trial procedure (for example, breaches of the Police and Criminal Evidence Act 1984, Code D), and any other specific weaknesses. See MLE 13th ed at pp281-282.
- A Turnbull warning should also be given in recognition cases, i.e. cases where the witness recognises the perpetrator as someone he knew (see R v Beckford). See MLE 13th ed at pp 282-284.
- Similar directions are given in the United States (Telfaire instructions). However, empirical research indicates that Telfaire instructions are of dubious efficacy. See the analysis of Cutler and Penrod, Mistaken Identification: The Eyewitness, Psychology and the Law (1995).
- A factor which may affect the accuracy and reliability of identifications is the nature of human memory. There have been a number of scientific findings on human memory some of which are general common sense, such as the fact that memory deteriorates with time, so that the longer the time between the observation of the perpetrator and the identification, the greater the risk that memory will weaken and the identification will be less accurate and reliable. The Turnbull direction already draws attention to lapse of time. However, perhaps judicial guidance on other findings might be useful to jurors when deciding on the reliability of an identification, such as the finding that memories of experienced events are always incomplete or that memories typically contain only a few highly specific details. See Keane, The use at trial of scientific findings relating to human memory  Crim LR 19.
- Informal identifications by social networking sites carry significant risks because the identification may be prompted (see R v Alexander) or otherwise contaminated. The judge will need to direct the jury carefully on whether the witness already knew the name of the person he identified or had other information about that person and whether he was prompted. See MLE 13th ed at p292.
- Pre-trial procedures held in accordance with Code D of the Police and Criminal Evidence Act 1984 also help to reduce the risk of miscarriages of justice arising from mistaken visual identifications because they provide a good test of the witness’s accuracy in remembering the perpetrator and make positive identifications more reliable. Code D now covers identification by viewing films, photographs and images formally and informally put into circulation for the purposes of identification. See MLE 13th ed at p292. This is to be welcomed because the use of social media is ubiquitous, and the number of informal visual identifications from by sites such as Facebook are likely to continue to increase. Procedural safeguards may lessen the risk of a miscarriage of justice where identifications from images on social networking sites are made informally.
- Voice identification evidence is more difficult than visual identification, so the risks of mistake are greater. R v Roberts. See MLE 13th ed at pp278 & 295.
- Voice identification evidence is admissible, but in R v Chenia it was held that the jury should not be asked to compare the accused’s voice to a recording of a voice alleged to be the accused without the assistance of an expert in phonetics. Although this is still the position in Northern Ireland (see R v O’Doherty), in England and Wales Chenia was superseded by R v Flynn which stated that there was no authority to support the position that a voice comparison exercise by the jury could only take place with expert assistance, and the assistance of ‘lay listeners’ was acceptable. The evidence could still be excluded though, if its prejudicial effect outweighed its probative value. See MLE 13th ed p295.
- Given the difficulties with voice identification and the increased risk of mistaken identification, great care should be taken when considering whether to admit the evidence. The Turnbull Guidelines apply to voice identification when it comes to directing the jury and presumably apply also in respect of when to withdraw a case from the jury, where the evidence of voice identification is of poor quality.
- When it comes to deciding whether the evidence is admissible, it is submitted that it should be scrutinized carefully, considering factors listed in the Crown Court Compendium (July 2019), Part 1, 15 -29 (see MLE 13th ed at pp295-297). These are factors to be addressed in any direction to the jury, but they are relevant to admissibility too, because they bear on the probative value of the evidence.
- Although voice identification evidence is more hazardous than visual identification, pre-trial procedure is limited to minimal requirements stated in R v Flynn (see MLE 13th ed at p297-298). Code D does not apply, beyond simply permitting the police to hold aural identification procedure if they think it is appropriate. In R v Hersey, a dubious parade led to an identification. The Court of Appeal held that the judge was right to admit the evidence of identification and also right to exclude expert evidence casting doubt on it. Given the risks associated with identification evidence, the decision is questionable and the kind of live parade held was deemed unacceptable in a Home Office Circular (see below).
- The limited nature of pre-trial procedure increases the risk that dubious evidence of voice identification will be admitted. Although the Crown Court Compendium, Part 1 (July 2019), provides a list of considerations for directing juries, many of which are relevant to admissibility, robust pre-trial procedure is needed. Code D should be developed to cover voice identification. The recommendations for good practice contained in a Home Office Circular of 2003 should be considered for inclusion in Code D. One of its recommendations was prohibiting live voice parades (see parade in R v Hersey, which the Court of Appeal approved). See MLE 13th ed at pp297-298.
- The key issue in this question is whether the identification evidence is admissible. Jim has identified Steve from the image Andy has shown him on his mobile phone (an informal identification from a photograph on social media). In the law of evidence, identifications are not invalidated because they occur in an informal setting. However, the circumstances and facts of the identification will be scrutinized carefully and in a so-called ‘Facebook identification’, the courts will be particularly concerned about whether the identifying witness has been prompted by another person present when the identification takes place (R v Alexander  1 Cr App R 334 (26), CA). ‘Prompting’ can make an identification unreliable because the witness is not making the identification independently, but rather is being influenced. See MLE 13th ed at p292.
- In R v LT  4 WLR 51, CA , the identifying witness made an identification in circumstances which were broadly similar to the circumstances of Jim’s identification. The Court of Appeal held the identification was admissible mainly because nothing was said to the witness before he was shown the image and identified the accused. In the question, Andy simply shows the image to Jim, who makes an identification; nothing, other than ‘Look at this’, is said by Andy just before Jim makes his identification. See MLE 13th ed at p292 (R v LT referred to in footnote 123).
- However, this must be seen in the context of Jim telling Andy what had happened the day before. So, Jim might well assume the images Andy was showing related to the assault and were possibly of the perpetrator. Further, the identification is contaminated by what occurs afterwards. Andy appears to endorse Jim’s identification of Steve, and also states that the assault on Jim is the kind of thing Steve would do (i.e. Steve is of bad character and has done other similar acts). The effect, we can see in the facts of the question, is that Jim almost immediately firms up his identification. He moves straight from saying, ‘I think it is him, it looks quite like him’, to saying, ‘…it’s him alright, I’m 100% sure’.
- The contamination of the identification provides a basis for the exclusion of the evidence under s 78 of the Police and Criminal Evidence Act, 1984. Section 78 states that the court may refuse to allow prosecution evidence to be given if it appears to the court that having regard to all the circumstances, including the circumstances in which the evidence was obtained, the admission of the evidence would have such an adverse effect on the fairness of the proceedings that the evidence ought not to be admitted. One of the principles of exclusion under s 78 is that the evidence is unreliable, and Steve’s advocate will argue this in respect of Jim’s identification. See MLE 13th Ed at pp293-294.
- Another issue concerns the police investigation. After the officers were sent the images, they seem to have made efforts to comply with the guidance on ‘Facebook identifications’ contained in R v Alexander  1 Cr App R 334 (26), CA. They sought to obtain details from the witness of the circumstances of the identification including details about searches and images viewed. Such details enable the jury to properly assess how the identification occurred and the defence are better able to test its reliability. These are arguments which might be made by the prosecution in support of admissibility: the ability of the defence to test the identification evidence and the jury to assess it means that it would not be unfair to admit it. See MLE 13th ed at p292.
- However, in refusing Jim a formal identification procedure, the police are in breach of Code D3.12. According to R v Forbes  1 AC 473, the obligation to hold a formal identification applies even where there has been a prior informal identification (see the reasoning in the text). However, Roberts in his ‘Commentary’  Crim LR 801 doubts the forensic value of a formal identification after an informal identification and argues that one should only be held if the suspect requests it. (In the facts of the question, Jim has requested it).
- A breach of Code D provides grounds for exclusion, but does not necessarily mean evidence will be excluded. The focus will be on whether it renders the trial unfair. Steve may argue that he has been deprived of an opportunity to have the identification tested in a formal procedure and, to show the importance of this, can point to the discrepancies between Jim’s first description and what he (Steve) actually looks like. The prosecution may argue that any unfairness can be addressed by cross-examination about the breach and a judicial direction in respect of it (directions will be mentioned shortly in this note- form answer). See MLE 13th ed at pp293-294.
- It is hard to know whether Jim’s identification evidence would be excluded by the judge. The defence certainly have a strong argument that the identification was flawed by Andy’s comments endorsing the identification and implying that Steve had perpetrated other similar acts. This, combined with the refusal by the police to hold a formal identification procedure in breach of Code D, could well be enough to justify exclusion.
- If the evidence is admitted, the jury would need clear directions about the circumstances of the identification including, in particular, the possibility that Jim was influenced in his identification (see R v Alexander and the Crown Court Compendium (July 2019), Part 1, 15-1, para 6). The jury would also need to be directed that the breach of Code D was a breach of a positive obligation and that the accused had been deprived of the opportunity to have the identification tested in a formal procedure (see R v Gojra  EWCA Crim 1939 and The Crown Court Compendium (July 2019), Part 1, 15-1, Example). See MLE 13th ed at pp293-294.