MOST cases concerning admissibility of evidence of a defendant's bad character involve prosecution applications. There, the law's traditional stance has been protective, with considerable caution exercised about it being adduced. The principal fears are of jury misuse of, or over-reliance upon, such evidence. Though section101 and connected provisions of the Criminal Justice Act 2003 significantly qualify that stance, it still cuts considerable judicial ice. More difficult, in a sense, are cases where a co-defendant seeks to adduce such evidence, either as itself tending to exculpate the co-defendant (issue relevance) or as undermining the truthfulness of an attack mounted by the defendant upon the co-defendant (credibility relevance). Whilst maintaining a proper interest in protecting the defendant, the law seeks not to confine the co-defendant in mounting their defence. Clearly, these two policies conflict, so must be reconciled. The burden of this note is that, in Simpson and Benzahi [2019] EWCA Crim 1144, the Court of Appeal (Criminal Division) failed to accord a defendant a proper degree of protection.
In early 2018, Junior Simpson, Daniel Luke and others stood charged with the murder of Jermain Goupall. That murder had been born of a rivalry between two London gangs, though, by chance, Goupall belonged to neither of them. The prosecution alleged that the murder, effected by a knife, had been carried out by three men who had emerged from a car driven to and from the scene by Simpson. Simpson testified that he had been the driver but had not been a party to the murder, as well as that Luke had been one of those that had emerged from the car. Luke himself denied ever having been in the car.
Simpson was convicted but Luke acquitted. Simpson appealed, on the basis that the judge had been wrong to allow counsel for Luke to adduce, under section 101(1)(e) of the 2003 Act, evidence about the “Goat incident”.
The Goat incident issue was first raised by a section 101(1)(d) prosecution application to adduce evidence thereof as relevant to an important issue between it and Simpson, by showing him prepared to use knives to stab others. That incident had taken place outside “The Goat” public house, and a CCTV recording thereof did show someone drawing a large knife, then stabbing another person, but Simpson denied that it was him. At his youth court trial for possession of a bladed article, D.C. Davies testified that he had viewed the CCTV footage, and that it was Simpson, whom he had seen on several previous occasions, that it showed. However, he was unable to name any particular features of the person shown that led him to that positive identification. In light of that, and following a successful submission of no case to answer, the judge had acquitted Simpson.
When Simpson was interviewed at length about the Goupall murder, D.C. Gibson was present. He was then recruited to view the Goat CCTV footage, but before doing so, was told, presumably by a colleague, that Simpson appeared on it. Unsurprisingly, that was in clear breach of rules governing the identification process, and, in particular, of Code D: 3.35 made under the Police and Criminal Evidence Act 1984. Anyway, having described certain physical features of the person carrying out the stabbing, he had positively identified that person as Simpson.
The judge refused the prosecution leave to adduce Gibson's evidence; there was “no safe basis” for granting it. Though the acquittal of the earlier offence provided no bar to it being adduced, it would be “dangerous to allow the evidence to go before the jury” (see at [19]). Lack of detail in Gibson's description of the culprit provided a further reason for refusing leave.
Then, once Simpson had testified to Luke's involvement, counsel invoked section 101(1)(e). This time, the judge held a voir dire, at which he asked Gibson for further detail about the circumstances of the identification. He then ruled that the jury should hear Gibson's evidence.
The first aspect of the court's decision to uphold that ruling was straightforward, but the second both complex and surprising. As to the former, a co-defendant may adduce evidence of a defendant's bad character showing a propensity to be untruthful “only if the nature or conduct of his defence is such as to undermine the co-defendant's defence” (see s. 104(1) of the Act). Here, as the court held, by naming Luke as having emerged from the car, Simpson had clearly undermined his defence.
The second aspect was whether or not the Goat incident evidence cleared the section 101(1)(e) hurdle itself. Here, two things are absolutely clear. First, that hurdle is higher than the section 101(1)(d) one. Whilst it suffices for the latter that the evidence is “relevant to an important matter in issue”, for the former it must have “substantial probative value in relation to an important matter in issue”. Furthermore, it was held in the leading case, Phillips [2011] EWCA Crim 2935, [2012] 1 Cr. App. R. 25, at [39], that “the term ‘substantial probative value’ must mean that the evidence has an enhanced capability of proving or disproving a matter in issue” (emphasis added).
Second, while section 101(3) grants discretion to exclude evidence satisfying the section 101(1)(d) test where it has the potential to render the proceedings unfair, neither that provision nor the similar one in section 78(1) of the 1984 Act applies to evidence satisfying section 101(1)(e). Yet Phillips itself gives us important context for the latter proposition. As Pitchford L.J. put it (at [40], and see also at [53]): “The purpose of the requirement [of substantial probative value] is to ensure so far as possible that the probative strength of the evidence removes the risk of unfair prejudice” (emphasis in original).
The court, without the advantage of being referred to those two important Phillips points, refused to overturn the judge's ruling, stressing four things. First, the judge had questioned Gibson in detail about the identification, had allowed the Code D breaches to be explored fully in cross-examination, and had given proper jury directions about both the identification and the breaches. Second, under section 109(1) of the act, a reference to the probative value of bad character evidence must be treated as a reference thereto on the assumption that it is true. Though section 109(2) allowed the judge not so to assume if “no court or jury could reasonably find it to be true”, he had decided otherwise. Third, both courses of action fell within his area of discretionary judgment, which the cases, in particular, Renda [2005] EWCA Crim 2826, [2006] 1 W.L.R. 2948, at [3], and Lawson [2006] EWCA Crim 2572, [2007] 1 W.L.R. 1191, at [39], recognise is unlikely to be interfered with on appeal “unless it is demonstrated that [the judge] is plainly wrong or Wednesbury unreasonable” (Lawson at [39], citing Associated Provincial Picture Houses Ltd. v Wednesbury Corp [1948] 1 K.B. 223). Finally, there was no discretion to exclude evidence satisfying the section 101(1)(e) test.
What is to be said about the court's reasoning? In effect, a judge having rejected one officer's identification of Simpson as a wholly insufficient basis for conviction of the Goat incident offence, another officer, who had attended Simpson's lengthy interrogation about the later murder, and had been told that Simpson did appear on the CCTV footage, then viewed it and testified that it indeed showed Simpson holding the knife. Though the possibility of “confirmation bias” was mentioned (at [34]), both judge and appeal court thought that any such issue could properly be dealt with via cross-examination. Yet it surely went to the very essence of the trustworthiness of Gibson's evidence, not in the sense that he might be consciously lying, but that one could have no confidence whatsoever in his honest judgment. If ever there were a strong case for applying section 109(2), this was surely it. Then, returning to the judge's decision not to admit that very evidence under section 101(1)(d), because it was dangerous to allow the jury to consider it, it is hard to see how those dangers were attenuated for section 101(1)(e) purposes. And, as already pointed out, for the court in Phillips [2011] EWCA Crim 2935, at [40] and [53], effect could be given to that concern without recourse to a section 101(3) equivalent. Phillips also tells us (at [40]) that the danger of satellite issues about bad character evidence diverting the jury from its proper task can properly be obviated by excluding such evidence under the enhanced probative value test. Did not such issues abound as regards the Goat incident identification evidence? In short, the judge's decision does seem to have been “plainly wrong or Wednesbury unreasonable”, so should have been overturned.
There is a broader point to be made. Very often, in this area no less than in others, there is legal authority supporting the common-sense outcome. Here, it was present in the leading case of Phillips. Ironically, the court's attention was drawn to that case in Simpson and Benzahi, but only for a minor point made there (at [33]). So, whilst recognising the real practical constraints upon both counsel and the court, had the former hit the spot, perhaps the latter would have too.
