202401911 B3 - [2025] EWCA Crim 1317

Mr. Bentley accepted that the judge, in his reasons for allowing D2’s bad character application, had correctly directed himself on the relevant principles by reference to Braithwaite, Phillips and other well-known authorities. There is no need for us to re-traverse all of that territory. It was common ground that s.101(1)(e) may apply where the important issue between two co-defendants is raised through one defendant’s case being put by his advocate in cross-examination of the other (R v Jones [2007] EWCA Crim 2741 at [23]).

Leaving aside cases where as a matter of logic bad character evidence has no probative value in relation to an issue, in other words it is irrelevant, the issue of whether that evidence has substantial probative value or “force” on that point is a matter of judgment. Where a judge has made no error as to the legal principles to be applied, this court will not interfere with his or her decision on the admissibility of that evidence unless the judgment made in the application of those principles was plainly wrong or Wednesbury unreasonable. The assessment made by a trial judge is highly fact-sensitive and his or her feel for the case is often crucial (see R v Hanson [2005] EWCA Crim 824; [2005] 1 WLR 3169; R v Renda [2005] EWCA Crim 2826; [2006] 1 Cr. App. R.24 at [3]; R v Lawson [2006] EWCA Crim 2572; [2007] 1 WLR 1191 at [39] and [44]; Braithwaite at [12]).

Section 109(1) of the CJA 2003 requires (subject to subsection (2)) an assessment of the relevance or probative value of bad character evidence to be made on the assumption that that evidence is true. It was not suggested that s.109(2) applied in this case. But the assumption in s.109(1) is not dispositive of the question of admissibility; instead it provides the context in which the decision on admissibility falls to be made (R v Dizaei [2013] EWCA Crim 88; [2013] 1 WLR 2257 at [36]). In other words, although the bad character evidence is assumed to be true, all the legal tests for admissibility under the gateway relied upon still have to be applied to that evidence.

In R v Mitchell [2010] EWCA Crim 783 this court upheld a decision to admit under s.101(1)(e) evidence referred to in a prison report of a co-defendant’s alleged misconduct and his allegedly false explanation which was similar to the one given for the index offence [12]. The court held that s.109(1) had required the judge to assume the truth of the bad character evidence sought to be relied upon. In this instance, that referred to both the giving of the explanation and to its falsity ([17]-[18]). That then provided the context for the application of the substantial probative value test. Similarly, in Musone the court held that although the trial judge had “grave doubts” about the reliability of an alleged confession by a co-defendant, the test for admissibility under s.101(1)(e) had to be applied on the assumption that the confession was true, given that s.109(2) did not apply ([44]-[45] and [60]).

By contrast in Braithwaite, where the trial judge had refused to admit bad character evidence relating to a prosecution witness under s.100(1) of the CJA 2003, this court held that s.109(1) was not engaged [17]. That was because of the wholly inadequate nature of the particular material contained in two CRIS reports [7]. One incident involved objects being thrown at a house from a passing car. The witness was in the area and was arrested. But no person identified him. The police merely suspected links between the witness and the car. The CPS advised that there was insufficient evidence to proceed. The other report referred to an attack by three men on a complainant who did not wish to pursue the matter. Someone gave the police first names for the three men and, relying upon “intelligence checks”, the police suspected the witness of being involved. As this court said, this material was “… no more than evidence that a complaint or allegation had been made. It was not evidence that the witness had done what was alleged.” “The CRIS reports did not even contain any accusation by anyone identifying [the witness] as responsible for the bad character conduct alleged” [20]. Therefore, it would have been inappropriate for the jury to be told, that “police officers suspected that the witness had committed the offence, or that they had investigated him for it, because neither of those is evidence at all that he committed it” [22]. That conclusion would also have applied if there had been an attempt to adduce the same material by way of admission or direct evidence from a witness. Section 109 does not come into operation or bite where the material in question does not amount to evidence that the witness did the act alleged [17].

Not surprisingly, this court held that the trial judge in Braithwaite had been entitled to rule that the material in the CRIS reports did not satisfy the substantial probative value test. Similarly, it would have been inappropriate for counsel for the prosecution to cross-examine the witness by putting allegations to him based on the CRIS reports, with counsel wanting “… additionally to be seen to derive what he said [in questioning] from police material, with a view to showing that it was likely to be true. That however would be to make the CRIS reports apparently evidence that the witness committed the offence when they were nothing of the kind” [22].

We refer back to the judge’s reasoning in the present case. Here, the circumstances were very different from Braithwaite. Instagram messages could be proved from material downloaded from mobile phones. The other allegations did not depend upon disputed statements of another pupil who had been unwilling to support a prosecution. The incident in October 2020 related to the confiscation by a named teacher of a bread knife brought into the school by the appellant. The CRIS report recorded the appellant’s acceptance that the knife had been seized from him and his explanation that he had it with him for self-protection. The matter was dealt with by the school and the youth offending team, rather than by the police. Likewise, in November 2020 the appellant was said to have made the threat to stab another boy to an identified teacher on a school bus. This was also dealt with by the school and a safeguarding team. As Ms Beattie pointed out, the appellant’s counsel did not suggest to the judge that the CRIS reports were inaccurate in so far as D2 sought to rely upon their content.All this formed part of the relevant context for the judge’s determination as to whether the material was admissible.

We conclude that, on the facts and in the circumstances of this case, the judge was entitled to reach the judgment that the three items of bad character evidence taken together did have substantial probative value in relation to the issue between the appellant and D2, namely which of the two men carried the chrome flick-knife and stabbed C1. The evidence was capable of showing a propensity on the appellant’s part to carry knives.That judgment was not Wednesbury unreasonable. In reaching this conclusion we have kept in mind the reasons set out in Phillips at [40] as to why the substantial probative value test is important.

The next issue is whether the judge erred in allowing D2 to cross-examine the appellant about the three items of bad character evidence without establishing whether D2 could and would call evidence to substantiate the allegations in the CRIS reports if the appellant were to deny them. Here again, the context we have summarised above is relevant.

In R v Miller [2010] EWCA Crim 1153; [2010] 2 Cr. App. R.19 the prosecution, relying upon s.100(1) of the CJA 2003, wished to attack the credibility of a defence witness by putting to him allegations that he had been involved in a conspiracy to supply drugs and had supplied a shotgun. However, the prosecution said that if the witness denied the allegations it did not intend to call evidence to prove them and would be bound by the answers he gave ([11]-[13]).

This court said that although the prosecution had sound material to prove the allegations, unless they were in a position to prove them and intended to do so, they should not have been allowed to embark upon the exercise because of the unfair prejudice it was capable of generating ([20] and [23]). But the court went on to say that this was not an absolute principle. There might be circumstances in which such cross-examination might take place, if it was limited in scope. However, in the event that an accusation is denied and otherwise unproved, the jury would have to be directed that the questions did not in any way support the prosecution case. They were only questions, not evidence ([21]).

In this case D2’s counsel made it plain that he was proposing either to draft admissions for the appellant to consider or to put the matters in cross-examination. The appellant’s counsel stated that there would be no agreed admissions, but he never suggested to the judge that the appellant denied the allegations. Nor did he ask the judge to rule that cross-examination should not take place unless D2 called evidence to support the material in the CRIS reports. He had a sufficient opportunity to consider the implications of taking that stance. We do not consider that, in the particular circumstances of this case, the judge is to be criticised for having allowed the cross-examination which did take place. This conclusion is, of course, fact sensitive.

The cross-examination on behalf of D2 was not unfair, either in terms of its content or its effect on the appellant’s trial. It was limited in duration and scope. The appellant admitted the essence of the allegation in relation to the first incident and repeated his explanation that he had taken the knife into his school in order to defend himself. The appellant did not deny the second allegation. He simply said that he did not remember the conversation with a teacher alleged to have taken place in November 2020. In those circumstances it was reasonable for D2’s counsel to have asked him one question about whether he remembered the involvement of the police and the school.

In view of the answers given by the appellant to the questions about the alleged incident in November 2020, it would have been preferable for the judge to have given a direction at that point in the trial that there was no evidence to support the allegation, it being raised only by questioning, and they should therefore ignore it. In addition, the effect of the evidence was that there were only two items of bad character evidence remaining, the October 2020 incident and the Instagram messages. If D2’s application under s.101(1)(e) had been based solely upon that material we doubt whether it would have been granted. However, the appellant did not ask the judge to revisit the issue of admissibility and to rule that those two remaining matters should be disregarded altogether by the jury.

The question is whether the jury’s verdicts on counts 1 and 3 are unsafe. We have reached the firm conclusion that they are not.

The judge’s legal direction to the jury was clear and sufficient. The jury was told unambiguously that the prosecution had not relied upon the bad character material in their case against the appellant and the jury should ignore it when considering that case. They were also told that the appellant’s good character remained unaffected. The jury was also directed separately that they could only consider that material when dealing with the cases of any defendant whose counsel had suggested that they were not involved in carrying or using knives and that the appellant had committed the offences.

The written submissions of counsel for the appellant did not seek a direction expressly referring to the incident in November 2020. That is unsurprising. The trial lasted about 10 weeks in all and the short cross-examination of the appellant on bad character had taken place about one month before the summing up. We also bear in mind the observations in [70] above. In any event, we are satisfied that the direction given to the jury was sufficient to indicate that they should not take the November 2020 allegation into account as part of the prosecution’s case against the appellant.

The issue raised by D2 was straightforward and the trial did not involve complexities or mental gymnastics of the kind discussed in Robinson ([2006] 1 Cr. App. R. 32 at p 488). Indeed, the effect of Robinson is that a direction of the kind given in the circumstances of the present case is appropriate (see [81]). There is no reason to think that the jury would not have followed such a direction (Miller at [24]). The direction properly told the jury to ignore the bad character evidence in the prosecution’s case against the appellant. Neither the cross-examination nor the legal directions render the appellant’s trial or convictions unsafe.

For these reasons the appeal must be dismissed.