Bad character evidence as evidence of identity

Facts

On 14 November 2013 the appellant, Simon Peter Richardson, was convicted of wounding with intent and intimidation at Nottingham Crown Court. The salient facts of the offence were as follows: On 29 June 2013, a taxi picked up a male passenger from the appellant’s address and set off with a view to taking that passenger to Longbow Public House. This taxi, whilst en route to the public house, mounted a kerb, prompting the complainant, Mr Dixon, to shout at the vehicle. The taxi moved off, turned and parked in a lay-by, at which point the passenger (wearing sunglasses and a ‘distinctive shirt’ at [4]) then vacated the taxi and racially abused the complainant, threatening to ‘take him and his family out’ (at [4]). The passenger also attempted to punch the complainant, held him in a headlock and bit a 2 cm section from the complainant’s left ear. When this was over, the taxi resumed its journey towards the Longbow Public House, the complainant having taken its registration number. It is also worthy of note that a number of days after the offence, a friend of the appellant threatened the complainant and his family at his home address with violence if the prosecution were not abandoned.

The most probative evidence for the prosecution was listed as follows: The appellant lived alone at the address from which a sole male was collected by the taxi. Further, the appellant matched a description provided by the complainant, who also picked out the appellant during an identification procedure. The complainant also picked out another male, identifying both as potentially having been his attacker. The attacker also wore a polo shirt and sunglasses that were identical to those found at the appellant’s home.

Cell site evidence put the carrier of phones associated with the appellant at the relevant address, and showed the carrier move towards the scene of the crime, and then away from it in the direction of the public house. The most significant evidential category for the present commentary was a previous conviction for Assault Occasioning Actual Bodily Harm, contrary to s. 47 of the Offences Against the Person Act 1861 (OAPA 1861), which was admitted as evidence of bad character, capable of supporting identification of the appellant as the assailant on the night of 14 November 2013. The conviction, for an offence under s. 47 OAPA 1861, was held to demonstrate a disposition towards violence and a similarity in the geographical area in which both offences were carried out. The court heard that the appellant, ‘assaulted a traffic warden, who had told his partner to move from a disabled parking bay, in the same street as the instant offence, by punching, head-butting and pushing him into a shop window’ (at [8]). It is the submission of this conviction that formed the ground for appeal.

Held, dismissing the appeal

The court ultimately decided that: ‘On the facts here, the line between evidence of propensity and evidence capable of supporting identification may appear to the layman to be a fine one, but it is one that the trial judge was entitled to make and did make…This was a prosecution case built on circumstantial evidence, but it was none the worse for that…the admission of the previous conviction was not used to bolster a weak case and the various strands made a strong case. Accordingly…we were not persuaded that the judge was wrong to admit the evidence of a previous assault…No criticism is made or could be made of the judge’s directions. In our view, the judge approached both the application and his directions to the jury with scrupulous care’ (at [17], [19] and [20]).

Commentary

Evidence of the accused’s bad character is admissible under s. 101(1)(d) of the Criminal Justice Act 2003 if the evidence is ‘relevant to an important matter in issue between the defendant and the prosecution’. Admissibility is (upon a defence application) subject to the operation of s. 101(3) Criminal Justice Act 2003, according to which the court must not admit such evidence if ‘it appears to the court that the admission of the evidence would have such an adverse effect on the fairness of the proceedings that the court ought not to admit it’. Section 103(1)(a) provides that, for the purposes of s. 101(1)(d), matters in issue include ‘the question whether the defendant has a propensity to commit offences of the kind with which he is charged, except where his having such a propensity makes it no more likely that he is guilty of the offence’. In Hanson [2005] 1 WLR 3169 (at [9]), the Court of Appeal indicated that:

a single previous conviction…will often not show propensity. But it may do so where, for example, it shows a tendency to unusual behaviour or where its circumstances demonstrate probative force in relation to the offence charged…

The ABH conviction that the present case concerned was a single conviction, but the judge did not admit it as evidence of propensity. Rather, it was admitted as evidence that possessed probative value in identifying the accused as the offender by showing that he ‘had previously launched a similarly unprovoked attack upon a stranger in the same area, again using disproportionate force’ (at [11]).

It is clear that evidence of the accused’s bad character may be admissible under s. 101(1)(d) for purposes other than establishing the accused’s propensity (i.e. that the important matter in issue to which such evidence is relevant may be a matter other than propensity). Examples of such purposes include the use of evidence of the accused’s bad character as evidence of identification (see, for example, Dossett [2013] EWCA Crim 710), in rebuttal of the defence of innocent association (see, as an example, Vo [2013] EWCA Crim 2292), as evidence that allegations made against the accused by several complainants in the context of a multi-count indictment were all true (see, for example, Freeman [2009] 1 Cr App R 15) and as circumstantial evidence that the same person committed all of the counts with which the accused was charged in the context of such an indictment and that the accused is that person (see, for example, Wallace [2008] 1 WLR 572).

In Richardson, evidence of the accused’s ABH conviction was properly admitted, not as evidence of the accused’s propensity to commit offences of the kind with which he was charged but, rather, as circumstantial evidence of the offender’s identity. As the court clarified in the present case, and in R v LN [2014] EWCA Crim 506 (at [13]), the earlier offences do not ‘properly sustain’ such an inference towards propensity. This is not a post-2003 Act development. Rather, even under the much more restrictive common law regime that governed the admissibility of evidence of bad character before the bad character provisions of the Criminal Justice Act 2003 came into force, under which evidence of the accused’s bad character was not admissible merely as evidence of propensity, such evidence was still potentially admissible as evidence of identification (see, for example, Singh [2001] EWCA Crim 2884).

As the Court of Appeal put it in Richardson:

On the facts here, the line between evidence of propensity and evidence capable of supporting identification may appear to the layman to be a fine one, but it is one that the trial judge was entitled to make and did make (at [20]).

More specifically, the value of the bad character evidence can seemingly be attributed to the similarity in geographical location and the violent, unprovoked nature of the attack. If the probative value of the previous ABH conviction does rest upon similarity to the degree iterated above, then the obvious reliance upon a ‘lack of coincidence’ coupled with the unreliability of the remaining elements of the prosecution case (identification evidence; cell site evidence and other circumstantial evidence) necessitated further analysis of the judgment in light of the guidelines provided in Hanson [2005] 1 WLR 3169. Whilst the court in the present case maintained that the ABH conviction did not serve to bolster a weak case (at [19]), an unprovoked violent attack arguably lacks exceptionality and perhaps should not be considered sufficiently uncommon as a method of commission to support the assertion that the accused is more likely to have committed the offence charged on this basis, even when coupled with the geographical connection between the ABH conviction and the offence charged. Given the ‘prohibition’ on the use of character evidence to ‘bolster a weak case’ or to ‘prejudice the minds of a jury against a defendant’ (Hanson at [3174]), it is questionable whether there was sufficient strength in the evidence of bad character to provide sufficient foundation for the identification of the appellant as the perpetrator of the offence. When one considers that when applying the s. 101(3) fairness test, the court is likely to consider whether the probative value of evidence outweighs its prejudicial effect, it is unclear whether, on the facts of the case as a whole, the ABH conviction should have surmounted this hurdle imposed by s. 101(3) (see, for an example of the application of this test in the context of s. 101(1)(3): R v Norris [2013] EWCA Crim 712).

The Court of Appeal in Richardson asserted that the separate, circumstantial elements of the prosecution case, when read together, made a strong case against the appellant ([19]). This is not wholly disputed, but it is also eminently arguable, upon reading the judgment of the Court of Appeal, that the bad character evidence did not serve to support the evidence surrounding the identification parade, so as to make it fair to admit it. The court drew parallels with the case of R v LN [2014] EWCA Crim 506, in which the court adopted a comparable approach, ultimately concluding that three previous convictions for street robberies were admissible as evidence to support identification of the appellant as the perpetrator of an offence of wounding with intent. It is noteworthy that in LN, despite the victim and his friend independently identifying the appellant during an identification parade, the court recognised the need for a Turnbull [1976] 3 WLR 445 direction as a result of the poor lighting conditions at the time the offence was committed (amongst other factors). In Richardson the appellate court noted that the victim correctly identified the appellant within the identity group, but not unequivocally or independently, as one other male was identified during this identification parade. The result of the identification parade in Richardson was that the victim was only able to say that his attacker was one of two men, one of whom was the appellant. The level of uncertainty surrounding the identification of the appellant in Richardson would surely have necessitated a Turnbull direction? It is stated by the court in Richardson that the judicial direction at first instance could not be faulted (at [20]), so presumably such a direction was given as its absence was not raised on appeal.

Assuming that a Turnbull direction was delivered at first instance, this should have necessitated an emphasis on evidence capable of providing support to the weak identification of the appellant. It is possible that the evidence of bad character provided could have served a dual function and could be capable of having provided such support, given that the initial reason for admission turned upon the similarities across the two offences. This would seem a little implausible, however, when one considers the strength of the bad character evidence in the context of Hanson. If evidence of bad character should never be used to bolster a weak case, then one must question how appropriate it would be for evidence that is insufficient to support a propensity inference be used to support weak or uncertain identification evidence, or indeed identity more generally. The admission of bad character evidence in this instance is also likely to have brought about a complex summing-up from the trial judge, who would need to balance a thorough direction on the use of the ABH conviction to support identity (and not propensity) on that basis that it makes it more likely than not that the appellant committed the offence, whilst also recognising the weakness in the identity parade evidence and, perhaps, the consequent need for the jury to consider the character evidence in support.

Overall, whilst the Court of Appeal was correct to avoid the making of propensity inferences based upon the evidence of bad character in Richardson, the previous conviction does not sit comfortably alongside the evidence arising out of the identity parade, as each seems to serve to ‘bolster’ the other. The court may have over-estimated the capability of the ABH conviction to render it more likely the appellant perpetrated the offence, and when this is considered alongside the somewhat uncertain results of the identity parade, one wonders whether the evidence of bad character was indeed safely admitted in this case.