The Law Commission

 


PART XIV THE CO-DEFENDANT EXCEPTION

INTRODUCTION

14.1      Evidence of a co-defendant's bad character may presently be adduced by two different routes, under the common law provisions and under section 1(f)(iii) of the 1898 Act. Under the common law, if the evidence of the bad character of a co-defendant is relevant to the defendant's defence it may be adduced. The reason for this is that a defendant should not be inhibited in the presentation of his or her defence. The test of relevance is strictly applied but there is no discretion to exclude the evidence.^[1] Under statute, section 1(f)(iii) of the 1898 Act allows evidence of a co-defendant's bad character to be adduced in crossexamination if the co-defendant has given evidence against the defendant. That evidence is only admitted as relevant on the issue of the defendant's credibility and not on the issue of his or her propensity to commit the offence charged. If the case falls within the subsection the co-defendant has a right to cross-examine the defendant on his or her character. There is no discretion in the court to decline to admit such evidence.

14.2      In this Part, for clarity we shall refer to the defendant who gives evidence against his co-defendant as D1 (male) and the co-defendant as D2 (female). Typically this rule comes into play where two defendants run "cut-throat" defences, that is, each in their evidence blames the other for the offence or where, falling short of the full "cut-throat" defence, evidence is given by D1 which supports the prosecution case against D2 or undermines D2's defence. In either of these circumstances D2 has the right to cross-examine D1 on his bad character for the purpose of suggesting that D1's evidence should not be relied on.^[2] The court has a discretion to order separate trials of the defendants to prevent the prejudice that might otherwise result.

14.3      The central problems with the present law are: that the court has no discretion to refuse to admit bad character evidence which may be of little probative value but significant prejudicial effect; the difficulty that fact-finders, magistrates and juries alike, inevitably experience in separating the evaluation of a defendant's truthfulness from that of the defendant's propensity to commit crimes; section 1(f)(iii), which only takes effect in cross-examination, may inhibit defendants from giving evidence, for fear of exposing themselves to crossexamination on their character; finally, it can produce an unbalanced and misleading picture to the fact-finders by revealing the bad character evidence of one defendant but not the other.^[3]

OUR PROVISIONAL PROPOSALS

14.4      In the consultation paper we made several provisional proposals for reform of the admissibility of the bad character evidence of a co-defendant.

Bad character evidence adduced by a co-defendant

14.5      To address the problem of lack of discretion to refuse bad character evidence on the ground of its prejudicial effect, we sought views on the question whether one defendant should always be entitled to call or elicit evidence of previous misconduct by a co-defendant if it had some relevance, or whether the question of admissibility should require the balancing of the evidence's probative value against its likely prejudicial effect.^[4]

Cross-examination of a co-defendant

14.6      We discussed seven possible options for the reform of cross-examination of a codefendant. Six of them, which we provisionally rejected, were: (i) replicating the existing law; (ii) introducing an unstructured judicial discretion to permit crossexamination under section 1(f)(iii); (iii) repealing section 1(f)(iii) so neither defendant would lose the shield; (iv) introducing a rule that both defendants lose the shield if one attacks the other; (v) a presumption that evidence of D1's character may be admitted if he attacks D2 but a discretion to exclude this evidence; and (vi) a presumption against the lost of shield coupled with a discretion to admit evidence of bad character.^[5]

14.7      We provisionally proposed our seventh option^[6] – that, where D1, in the course of his evidence, or through his witness or representative, undermines the defence of D2 charged in the same proceedings,

(i) if the challenge to D2's account concerns her conduct in the incident in question, or in the investigation of it, the shield should not be lost; but

(ii) if this is not the case, any party to the proceedings, that is D2 or the prosecution or any other defendant, should be entitled to apply to the court for leave to adduce evidence of D1's character.

14.8      We provisionally proposed that once D1 had lost his shield in this manner, evidence of his bad character should only be admitted with the leave of the court. If D2 applied for leave we proposed a presumption in favour of leave unless it would be contrary to the interests of justice. Conversely if another party to the case applied for leave, the presumption should be against granting it unless it was in the interests of justice to do so.

14.9      In deciding whether to grant leave, we proposed a court should have regard, amongst any other relevant considerations, to

(1) the degree to which D2's defence had been undermined;

(2) how unavoidable it was for D1 to undermine D2's defence;

(3) the nature, number and age of the matters of bad character which it is sought to adduce; and

(4) the relative characters of all the accused as they would appear to the factfinders if the evidence of bad character were allowed.

14.10      We also provisionally proposed that where leave was given for one co-defendant to adduce evidence of the other's bad character, the fact-finders should be permitted to use that character evidence on the issue of guilt as well as on the issue of credibility, and that, in the Crown Court, the jury should be guided by the judge in assessing its probative value, as well as being warned about its potential for prejudice.^[7]

14.11      This option would result in D1 only losing his shield if the evidence adduced or statement made against D2 did not concern the central facts of the case. This would enable D1 to put forward his defence freely without fearing crossexamination as to character. Where, however, D1 chooses to put in issue D2's character by raising matters outside the central facts, D1 would potentially expose his own previous bad character. The court would retain a structured discretion whether to give leave to adduce evidence of D1's character.

Attacks on a co-defendant by a defendant who does not testify

14.12      Under this option it was not necessary for D1 to have given evidence about D2 in order to trigger the entitlement to adduce evidence about his character, nor was it necessary for D1 to give evidence in order for the evidence of his character to be placed before the fact-finders.

THE VIEWS OF RESPONDENTS

A question of balance

14.13      Forty-one respondents addressed this issue, on which we had simply invited views.^[8] Opinions were divided. Five were clear that there should be some balancing of probative value and prejudicial effect, and a further five thought that there should be a balancing test in some circumstances. Sixteen thought there should be no balancing test and a further three held that same view subject to certain conditions. Four respondents saw severance as the only solution. The remainder either argued that the current law involves an element of balancing, or submitted ambiguous or split responses.

14.14      Not all the respondents who supported a balancing test gave reasons for their support. Two respondents recognised that there was a strong argument against fettering the defendant's right to adduce relevant evidence, but argued that the interests of justice nonetheless demanded that a balancing exercise be carried out. One practitioners' organisation conceded that the basic rule should be that an accused can adduce all evidence relevant to his or her defence, but argued that this could lead to injustice: not only by the wrongful conviction of D1, but also by the wrongful acquittal of D2. Another view was that the same standard of admissibility ought to apply to a co-defendant as to the prosecution.

14.15      Of the sixteen respondents who thought there should be no balancing test, seven respondents specifically argued that it would be unfair to fetter the defendant's right to adduce relevant evidence. The response of Professor McEwan was typical:

… there are more factors to balance than that of prejudice against probative value. It must be unfair to deprive the co-defendant of evidence relevant to his defence, whatever the effect on the defendant might be. Hence it seems that the current rule is the right one, that if a co-defendant wants to adduce evidence of previous misconduct by the defendant, the only test for admissibility is one of relevance.^[9]

14.16      Another respondent commented that the balancing of probative value and prejudicial effect was an inappropriate consideration in assessing the position between defendants. A further six thought that the discretion to exclude evidence should be exercised only where the evidence was irrelevant, as is prescribed by the current law, although they did not base this specifically on an argument about fairness to the defendant.

Cross-examination of a co-defendant

14.17      Of the options we rejected the only one to attract significant comment (42 respondents) was that of retaining the current law. The vast majority of these respondents advocated change. But there was no real consensus about exactly what the problems, or their solutions, were.

14.18      The main argument for reform was that the current law is unfair to D1 who gives evidence against D2. The response from a practitioners' organisation, for example, said:

We believe that the unfettered right of a co-accused to cross-examine under this subsection can lead to wholly disproportionate damage being done to the case of the accused whose character has been adduced.

14.19      Three respondents were concerned that the lack of discretion resulted in irrelevant or barely relevant evidence of bad character being adduced by D2. These respondents held differing views on whether this was because the issue of relevance was rarely raised with the judge or because judges were too willing to allow D1 to undermine D2's defence in the first place, since such evidence itself was often insufficiently relevant.

14.20      Those respondents who disagreed that reform was desirable thought that the introduction of an element of discretion would inhibit D2 in the presentation of her defence, arguing that it was only fair that, in the capacity of a witness, D1 should be cross-examined as to his previous convictions. A prominent practitioner said

In the practical world of the hard fights in contested criminal cases with, in so very many instances, much hard lying taking place, it is improvident to put a premium upon the invention of cunning false stories giving rise to immunity from credit being properly tested.

14.21      However, this argument can be turned on its head and applied to D1. If D1 automatically loses the shield by undermining D2's, his defence might be equally inhibited. This point was endorsed by one judicial respondent.^[10]

Our proposal

14.22      Thirty-five respondents addressed the option we had provisionally proposed in the consultation paper.^[11] Twelve respondents agreed with the proposal. A further seven respondents had some reservations, but generally agreed with our proposal. Three respondents disagreed but suggested a scheme which was similar to that proposed.^[12] Eight respondents did not agree with our proposal.

14.23      Most of those who agreed with our preferred option did so without further substantial comment. One respondent who agreed with the entire scheme of proposals in the consultation paper thought it was logical that the same test should be applied to the undermining of the co-accused. One academic stressed that

Defendants should not generally be allowed to undermine the defence of a co-accused except where the challenge concerns the coaccused's conduct in the incident or investigation in question, in which case the shield would not be lost under the proposal. A few other respondents agreed but had reservations about the rigidity of the proposal.

14.24      There were two main reasons for not agreeing with our proposals: dislike of the immunity afforded defendants who confine their attacks to the co-defendant's behaviour in the incident in question or the investigation, and dissatisfaction with the operation of the discretion.

14.25      The responses in favour of our proposals tended also to support the reasoning behind it. Respondents agreed that when putting forward a defence D1 should be able to challenge evidence concerning the incident in question, but that attacks on D2's character outside that parameter should open up the issue of D1's own credibility.

WHO SHOULD BE ABLE TO APPLY?

14.26      We proposed that any party to the proceedings ought to be entitled to apply to the court for leave to adduce evidence of D1's character.^[13]

14.27      Three respondents addressed this issue and they all argued that the prosecution ought not to be able to adduce evidence of D1's character in these circumstances. All three believed that that the prosecution should not be able to take advantage of an exception aimed at allowing one defendant to cast doubt on the credibility of the other. We find this argument convincing. If the evidence would not have been sufficiently relevant to be adduced by the prosecution were the defendants tried separately, they should not be able to adduce the evidence merely because the defendant has undermined the co-defendant's defence. The rationale of the exception is that, because D1 has undermined D2's defence, D2 has a right to adduce evidence relevant to D1's credibility. This should not allow another co-defendant, D3, or the prosecution to adduce evidence in D2's stead where D2 has not sought to do so.

ATTACKS ON D2 BY D1 WHERE D1 DOES NOT TESTIFY

14.28      In the consultation paper we proposed that an application might be made to adduce evidence of D1's bad character on the ground that he had undermined the defence of D2, even though he had not given evidence, provided the nature of his defence was such as to put his own credibility in issue.^[14] We concluded provisionally that the attack has a similar impact on the fact-finders whether or not it comes directly from D1. Whilst there is some logic in the traditional view that if the relevance of such bad character is to D1's credibility then if D1 does not give evidence, there is nothing to which the bad character evidence can be relevant, it was, and remains, our view that this understanding of credibility is too limited. The reality is that a defendant's credibility might be put in issue by outof- court statements by the defendant or by the nature of the defence.^[15]

14.29      Twenty-four respondents addressed this proposal. Seventeen agreed with our view. A further three agreed to a large extent. Only two disagreed with the proposal.^[16] Most respondents who agreed did so without further comment. The only reason given for agreement was consistency with the proposals regarding prosecution witnesses. The respondents who agreed apparently accepted that the credibility of a non-testifying defendant may be put in issue.

14.30      The responses give strong support for our view that, as far as D2 is concerned, the effect on her defence is the same whether the attack comes directly or indirectly from D1. We maintain our view that D1's credibility may be brought into issue even though he does not testify. Thus evidence should be admissible through cross-examination of D1 or otherwise where D1 undermines D2's defence, whether in the course of his evidence, or through his representative or witness.

RELEVANCE OF BAD CHARACTER TO GUILT AS WELL AS CREDIBILITY

14.31      In the consultation paper we considered the use to which previous misconduct adduced under section 1(f)(iii) could be put.^[17] The traditional justification for loss of the shield under section 1(f)(iii) is that a co-defendant should be able to "discredit someone who has given evidence against him".^[18] However, where a comparative assessment is being made of the truthfulness of two defendants, it is difficult to distinguish guilt and credibility. Both factors are in issue where the person against whom the imputations are being made is on trial. We therefore proposed the abolition of the common law rule that, where evidence of an accused's bad character is admitted under section 1(f)(iii), it is directly relevant only to the accused's credibility.^[19]

14.32      Thirty-two respondents addressed this issue, the majority of whom were strongly in favour of the provisional proposal. Twenty-five respondents thought that evidence admitted under section 1(f)(iii) should not be relevant only to credibility. Only two respondents disagreed with the proposal.

14.33      Four thought that the distinction between evidence relevant to guilt and evidence relevant to credibility was illogical, since evidence of previous misconduct must always be relevant to guilt. Seven thought that evidence might be relevant to guilt depending on the circumstances. One judicial respondent gave an example of scenario he had come across as a judge where two defendants were jointly charged with robbery. There was little evidence against D1 and rather more against D2. D1 had several previous convictions for robberies of a similar nature. D1's defence necessarily involved the imputation that D2, who was of good character, committed the offence rather than him. D1's previous convictions were, in consequence, revealed. The jury duly convicted D1 and acquitted D2. The only reason our respondent could put forward for this outcome was "that in the majority of cases the practical effect of previous misconduct has little or nothing to do with credibility and everything to do with propensity".

14.34      One of the respondents who disagreed with our proposal said that "the manner in which the defendant conducts his case should not call into question anything other than his credibility or the credibility of the defence case". The Crown Prosecution Service said that the conclusions reached in the consultation paper did a "disservice to juries", whom it felt were entirely capable of distinguishing questions such as "Is D1 telling the truth when he says that D2 did it?" from the question "Is D1 guilty?"

14.35      We can quite see that there is a distinction between saying that evidence is relevant only for the purpose of assessing credibility and not propensity. We also understand that in some cases, principally where the offence charged is of a similar kind to the bad character evidence to be revealed, it is asking a great deal of fact-finders to disregard that evidence completely on the question of guilt but to regard it solely on the question of credibility. This will be particularly so where the question of guilt and credibility may be intimately bound up. Indeed it may well be regarded by juries as bizarre to direct them to ignore certain evidence (which may plainly go to propensity) for purposes other than testing the credibility of D1. In our view, giving juries directions which they will find bizarre or incomprehensible, or which require them to engage in mental gymnastics, does no credit to the law. In our view the way forward is to seek to ensure that the evidence which is adduced is limited to that which is substantially relevant to the issues of importance between the defendants and not marginal to the issues in the trial. Where, as inevitably will sometimes be the case, evidence is admitted on the issue of truthfulness but is also likely to impinge on the question of propensity, the judge should address the problem in a straightforward manner by warning the jury of the dangers of inferring from past conduct that the defendant is guilty of the offence charged.^[20]

OUR RECOMMENDATION

14.36      Our provisional proposal embodied a number of elements:

(1) that D1 should be able to adduce evidence about D2's bad character in respect of the central set of facts (as we then defined them) without triggering any right of D2 to adduce evidence of D1's bad character which went outside those facts;

(2) if D1's character became an issue because of an attack made on D2 then any party to the proceedings could attack his character;

(3) no one would have the right to adduce evidence of D1's character, but it would always be a matter of discretion;

(4) the discretion was to be structured so that there was a presumption in favour of admitting evidence if the application was by D2 but against admissibility if the application to admit was by any other party. In each case the presumption was to be displaced on an interests of justice basis;

(5) there were a series of factors to be considered in exercising the discretion involving consideration of the damage done to D2's case, how gratuitous the attack had been, the number and nature of D1's convictions sought to be adduced, and the impact of admitting the evidence.

14.37      We have retained a number of features of our provisional proposals but have recalibrated them in order to achieve a uniformity of approach in the overall scheme we propose. We regard such uniformity of approach as an important feature of our recommendations.

14.38      We have, however, abandoned certain elements of the provisional proposals as we have been persuaded by the responses to the consultation paper that they were erroneous as a matter of principle.

14.39      In particular we think it right that D2 should be entitled as of right to adduce evidence of D1's bad character if the requirements are satisfied. We accept that each defendant should be permitted to pursue his or her defence legitimately without regard to the impact on the other defendants. On the other hand, we accept that the only party who should, on this ground, be free to adduce evidence of the bad character of D1 should be D2. If D2 does not wish to do so then we do not now think that the prosecution, or other defendants, should be able to jump on that bandwagon and adduce such evidence where they have no independent entitlement to do so.

14.40      Our first recommendation is that each defendant should be free to make attacks on the character of co-defendants in respect of matters falling within the central set of facts, as we have defined them,^[21] without seeking leave of the court and without thereby automatically entitling co-defendants to adduce evidence of bad character which goes outside the central set of facts.^[22] In this respect their freedom to adduce evidence will be no different from their freedom to adduce evidence attacking prosecution witnesses or others.

14.41      In common with our other recommendations, if a defendant wishes to adduce evidence about another defendant which goes outside the central set of facts then the leave of the court must be obtained.

14.42      Consistent with our recommendations in respect of adducing evidence of the defendant's character by the prosecution, or of a non defendant's character by any party, such leave will only be given if the evidence is of substantial probative value to an issue between the defendants, which issue is itself of substantial importance in the context of the case as a whole.

14.43      This recommendation mirrors our other recommendations in respect of evidence sought to be adduced about bad character evidence outside the central set of facts. The enhanced relevance test is designed to prevent evidence of bad character extraneous to the events in question being admitted where it is of only minimal relevance to a central issue, or where the issue to which it is relevant is itself marginal to the case.^[23]

14.44      In this way the court will require each defendant who wishes to attack the character of a co-defendant, by drawing attention to character evidence which is extraneous to the central set of facts, specifically to justify its introduction. In so doing we aim to avoid material of great potential prejudice to another defendant being placed before the fact-finders for no reason other than that it may be of some marginal relevance, or relevant to some marginal issue or, as at present, because a defendant has been unwise enough to step across a line in the evidence given against that co-defendant.

14.45      If, however, the material does satisfy the test of enhanced relevance, then, because the defendant has the right to have his or her case presented, there is a right to have it adduced. There is no exclusionary discretion, such as we propose in the case of the prosecution seeking to introduce evidence of the defendant's bad character, requiring the court to have regard to the prejudicial impact of the evidence.

14.46      Where, however, the issue between D1 and D2 is whether D1 has a propensity to be untruthful, leave may only be given if, in addition to the other requirements, the nature or the conduct of D1's defence is such as to undermine D2's defence.

14.47      In assessing the probative value of evidence for the purposes of these provisions we require the court to have regard to the same factors as where it is considering the admissibility of other bad character evidence which is outside the central set of facts.^[24]

14.48      Where the evidence of D1's bad character, outside the central set of facts, is admitted at the behest of D2, on whatever basis, we recommend no specific limit to the use to which the fact-finders may put that evidence. We envisage that the judge in summing up will routinely place such evidence in context, explaining the issue in the case to which it is relevant. Where it is admitted only in respect of the question of D1's propensity to untruthfulness, we envisage that the judge will make that clear to the jury. In every case, but in particular in relation to evidence which is admitted only on the issue of propensity to be untruthful, we would expect the judge to warn the jury that they must be very careful in the use to which they put such evidence and that it can be very dangerous to assume that a person has done something on this occasion merely because in the past he may have done something similar. We believe that reminding the jury that they should take this common sense approach to such evidence is much more likely to be understood and complied with than requiring them, as a matter of technical legal rules, to perform bizarre mental gymnastics.

The operation of the exception

14.49      We can illustrate the effect of the new test by returning to the scenario described by one of our respondents.^[25]

D1 and D2 are jointly charged with robbery. D1's defence is that D2 did it on her own. In order to get D1's criminal record admitted under the co-defendant exception on the basis that D1 has undermined D2's case, D2 must show that his convictions show he is likely to lie on oath. What is in issue is D1's propensity to tell the truth not his propensity to rob.

14.50      The evidence is less likely to be admitted in relation to credibility than under the current law. It would be clearer than it is under the current law that only evidence of D1's bad character which was relevant to his credibility would be admissible, and there would be no danger of bad character evidence of little probative value (but significant prejudicial effect) being admitted.

D2 might seek to have D1's convictions admitted on the basis that they are substantially relevant directly to the issue of who committed the robbery. If D1 has recent convictions for robbery, they are more likely to be admissible on this basis, whereas other kinds of dishonesty offence might not be sufficiently relevant.

14.51      It may be that the evidence of D1's previous convictions is directly relevant to the central issue of who committed the robbery even though it could have a similar prejudicial effect to that described by the respondent. However, the evidence would have to be substantially probative to be adduced, rather than creep in under the guise of credibility.

Severance

14.52      We believe that the need for severance of defendants would arise less frequently under our recommendation simply because evidence of the bad character of one defendant will be admitted at the behest of a co-defendant less easily than under the present law.

14.53      We recommend that leave may be given to a co-defendant to adduce evidence of the bad character of a defendant where it has substantial probative value in relation to a matter in issue between the co-defendant and the defendant where that issue is itself of substantial importance in the context of the case as a whole – except that, if it has probative value only in showing that the defendant has a propensity to be untruthful, leave may not be given unless, in addition, the defendant's case is such as to undermine that of the co-defendant.^[26]

Ý
Ü   Þ

Note 1   See para 2.40 above.    [Back]

Note 2   Although see n 201 in Part II above.    [Back]

Note 3   See paras 4.70 – 4.78 above.    [Back]

Note 4   At para 10.118 of the consultation paper.    [Back]

Note 5   These are detailed in the consultation paper at paras 13.18 – 13.40.    [Back]

Note 6   See paras 13.41 – 13.47 of the consultation paper.    [Back]

Note 7   See paras 13.51 – 13.53 of the consultation paper.    [Back]

Note 8   See para 14.5 above and paras 10.115 – 10.118 of the consultation paper.    [Back]

Note 9   Jenny McEwan “Law Commission Dodges the Nettles in Consultation Paper No. 141” [1997] Crim LR 93, 97.    [Back]

Note 10   The point was also referred to in the consultation paper: see paras 13.16 and 13.45.    [Back]

Note 11   Provisional proposal 38. See para 13.46 in the consultation paper.    [Back]

Note 12   All three thought that there should not be an immunity for defendants who confined their attacks to the co–accused’s behaviour in the incident in question.    [Back]

Note 13   Para 13.46 of the consultation paper.    [Back]

Note 14   See paras 13.48 – 13.50 of the consultation paper.    [Back]

Note 15   The case for maintaining the traditional view of when credibility is in issue is set out at para 4.62, and our response, at para 4.65 above.    [Back]

Note 16   One of these seemed to disagree only to a certain extent.    [Back]

Note 17   See paras 13.51 – 13.53.    [Back]

Note 18   Murdoch v Taylor [1965] AC 574, 585D, per Lord Morris of Borth-y-Gest.    [Back]

Note 19   See para 14.2 above.    [Back]

Note 20   See paras 17.18 – 17.19 below.    [Back]

Note 21   See para 8.31 above.    [Back]

Note 22   Though the co-defendant may be granted leave to adduce such evidence: see paras 14.44 – 14.46 below.    [Back]

Note 23   See para 9.35 above.     [Back]

Note 24   See para 7.19 above and cl 5(2) of the Bill.    [Back]

Note 25   As described at para 14.33.    [Back]

Note 26   This recommendation is given effect by cl 11 of the draft Bill.    [Back]