fact evidence: the origins and erosion of the modern similar fact
The presumption of innocence
1 The fundamental safeguard in our system of justice has always been the presumption that defendants are innocent until they are proved guilty. It is precisely in order to protect this presumption of innocence that defendants are not generally required to face evidence which, while it may be highly prejudicial, does not actually prove the particular case against them. For this reason �similar fact evidence�, which suggests that the defendant may on other occasions have committed crimes similar to those now alleged against him, or may be inclined to do so, is generally excluded from criminal trials.
2 It follows from the very nature of the common law system that any move to relax the restrictions on the admission of such prejudicial evidence might, if not made with the utmost care, lead to the erosion, or even the destruction of the presumption of innocence.
3 The reason why dispensing with the presumption of innocence is rarely, if ever, seriously contemplated is that there has long been a recognition that justice, whose evident purpose is to protect against social ills and to promote moral good, can, if it is not wisely administered, inflict forms of harm on innocent citizens which are incalculably more damaging than the mere failure to punish a particular crime. By convicting innocent defendants of offences they have not committed, justice can itself become a form inadvertent crime, more terrible for the fact that it is committed, or abetted, by servants of the state � by judges themselves.
4 However, in the last century, and particularly in the latter part of that century, the law on similar fact evidence was subjected to a series of changes. As a result the safeguards which once protected innocent defendants against the introduction of prejudicial evidence were progressively weakened. In one particular respect they disappeared almost completely. It was this development in the law, which took place in 1991, which made it for the first time relatively easy to obtain convictions on the basis of multiple uncorroborated allegations, all of which might be false.
5 The spectacular growth in police trawling operations which took place in the last decade of the twentieth century was a direct result of these legal changes. It is therefore not possible to understand the nature of the problem posed by trawling operations without considering the changes in the law relating to similar fact evidence which were the precondition of their emergence.
The sinilar fact principle
6 The modern similar fact principle was formulated in the case of Makin v. Attorney General for New South Wales,1894 (AC 57, 65). In this case Lord Herschell reaffirmed the presumption that similar fact evidence would not normally be admitted. He went on to outline the exceptional circumstances in which this exclusionary principle could be overridden. Evidence of similar facts could not be admitted merely because it seemed relevant to the count on the indictment, but only if it was both relevant and probative.
7 In Makin (1894) the two defendants, a husband and wife, were accused of murdering a baby they had fostered. In their trial, evidence was offered of a number of other murders they had apparently committed. The admissibility of the similar fact evidence in this case turned on the improbability of there being any innocent explanation for the presence of the bodies of twelve other fostered infants buried in the gardens of premises previously occupied by the defendants.
8 In R v Smith, 1915, (11 Cr App R, 229), the �brides in the bath� case, the defendant was accused of one murder but evidence was offered of two more. The admission of this evidence followed similarly from the improbability that three different women with whom he had gone through a form of marriage, and who had made financial arrangements from which he would benefit, had all drowned in the bath by accident shortly afterwards.
9 In these two classic cases the probative power of the similar fact evidence, which made it just to admit it in spite of its prejudicial effect, derived from the improbability of the strikingly similar facts having any rational explanation other than the guilt of the accused.
10 Although the leading cases in which the similar fact principle was established concerned the crime of murder and the admission of similar facts which were not disputed, from the beginning of the twentieth century onwards there was growing pressure on the judiciary to lower the threshold for the admission of such evidence in order to obtain convictions in sexual cases. This pressure was particularly strong in relation to cases involving allegations of homosexuality.
The role of prejudice
11 The most decisive move in this direction was made in an appeal court judgment drafted by Denning J, later to become Lord Denning. In the case of R v Sims, (1946, 31 Cr App R 158) the Court considered an appeal which turned on whether allegations of buggery made by three different men against the accused could be admitted within the same trial in support of one another.
12 In the judgment drafted by Denning J, the Court ruled that the principle applied to strikingly similar facts in such cases as Makin and Smith should be extended and applied to similar allegations. The judgment failed to present any sound rationale for changing the law in this manner. In place of such a rationale, however, it offered Denning J�s own view that homosexuality was an abomination, quoting with approval the words of Lord Sumner to the effect that �Persons . . . who commit the offences now under consideration, seek the habitual gratification of a particular perverted lust, which not only takes them out of the class of ordinary men gone wrong but stamps them with the hallmark of a special and extraordinary class, as much as if they carried on their bodies some physical peculiarity� (see R v Thompson, 1918, 13 Cr. App R, p. 80).
13 In the case of DPP v Boardman (1975), which concerned allegations made by three adolescent boys against their male teacher, the House of Lords lent its own authority to the judgment in Sims, while rejecting its view of homosexuality. At the same time that it did this, however, it drew attention to the grave dangers which might follow from this change in the law. Lord Cross pointed out that that there was a marked difference between cases which involved allegations and the cases (such as Makin and Smith) which involved counts of murder and undisputed evidence of previous deaths. In the original cases there was, said Lord Cross, �no question of any witness for the prosecution telling lies�.
14 Lord Cross went on to note that the crucial consideration in cases like Sims and the case then before their lordships, was that the similar fact evidence was disputed and that a series of allegations were all denied by the accused: �In such circumstances the first question which arises is obviously whether his accusers may not have put their heads together to concoct false evidence and if there is any real chance of this having occurred the similar fact evidence must be excluded� (DPP v Boardman, 1975 AC 457).
15 Lord Wilberforce underlined this view and warned against the danger that, as a result of the extension of the similar fact principle, innocent defendants might find themselves facing a series of grave allegations, all of which were false.
16 He clearly states his own view that the courts should be on their guard against the possibility that a series of false allegations might arise either from collusion or from a process of contamination. If there was any real possibility of this having occurred there should be no question of the similar fact evidence being admitted. Instead the judge should order separate trials.
17 The general test prescribed by Boardman was a cautious one. In deciding the question of admissibility, the judge should weigh the probative value of the evidence against its prejudicial effect. If the similar fact evidence was so weak, so unreliable or so contaminated that its probative value was outweighed by its capacity to prejudice a jury, then it should be excluded.
18 In their judgment their Lordships reinforced the caution embodied in this test by placing particular stress on the need (already acknowledged in Sims) for there to be �striking similarities� between allegations before the possibility of their being admissible could even be considered.
19 In Boardman their Lordships in effect accepted the highly dangerous precedent created by Denning J, but did so only after insisting on two vital safeguards against the injustices it might lead to.
20 However, in two crucial House of Lords judgments, delivered in 1991 and 1995, the two safeguards which had been put in place by Boardman were both removed.
The erosion of Boardman
21 In 1991 in a judgment given by the Lord Chancellor, Lord Mackay, in the case of DPP v P, no reference was made to the manner in which the judgment in Sims had transformed the original similar fact principle. Moreover Lord Mackay went on to reject the requirement that allegations, in order to be admissible, should be �strikingly similar�. In doing so he wrongly claimed that this requirement applied only to cases where identification was at issue.
22 The effect of the judgment in P was dramatically to lower the threshold for the admissibility of multiple allegations and make it much more likely that innocent defendants would find themselves facing a series of false allegations.
23 In 1995, in R v H, the second safeguard which had been put in place by Boardman was also removed. Lord Mackay, who once again gave judgment, made explicit what had already been implicit in his earlier judgment and held that, in ruling on the admissibility of a series of similar allegations, the judge should generally assume that the allegations in question were true.
24 The main objection to this approach has been made pointedly by Professor Sir John Smith: �But the judge has to weigh the probative value [of a series of allegations] against the prejudicial effect. How can he assess the probative value of the statements without taking into account their reliability? If he has to assume that they are true for the purpose of this exercise, they will always be admissible because if they are true, they are conclusive proof of the defendant�s guilt � the probative value then obviously outweighs the prejudicial effect. [Yet] if the witnesses have put their heads together to concoct false stories the evidence is obviously worthless (Criminal Law Review, 1999, Commentary on �Severance� by John C. Smith, pp. 859-60).
25 The objections which Professor Smith so trenchantly makes draw attention to the existence of a state of affairs which is little less than astonishing. For what Smith points out is that judges, bound as they are to follow the precedents created by P and H, have little choice but to renounce their judicial responsibilities in this matter. Under the guise of prescribing an exercise whose purpose is to protect innocent defendants against having to confront dangerous, unreliable and highly prejudicial evidence, the House of Lords has in fact given its authority to a principle which ensures that just such dangerous evidence will regularly be allowed to become the very basis on which criminal trials are conducted.
The emergence of trawling
26 It so happened that the decision of the House of Lords to dispense with the two vital safeguards which had been ratified in Boardman coincided with the emergence among social workers and police officers of the paedophile ring theory of sexual abuse in children�s homes. As was almost inevitable, the newly created weakness in the law was almost immediately seized on by police forces in order to successfully push through a number of highly dangerous prosecutions which could never have been brought prior to the decision in P.
27 The massive growth of police trawling operations has followed inexorably from these early successes. As trawling has grown exponentially, the legal channels which have enabled it to do so have become, if anything, deeper and more dangerous as a result of constant use. The similar fact principle, once zealously guarded by the Defence as an essential means of keeping prejudicial evidence out, is now the favoured device of the Prosecution, valued as an almost ever-open conduit for letting prejudicial evidence in.
28 In any society where the crime of child sexual abuse has been demonised to the extent that it has in our own society, the presumption of innocence in relation to those who are accused of this crime is seriously threatened. When defendants are compelled to face large numbers of allegations of sexual abuse made by different complainants in a single trial, the presumption of innocence is almost inevitably destroyed.
29 In such cases the effective destruction of the presumption of innocence has been brought about by the courts themselves and in particular by the decisions made by the House of Lords in P and H. Because these decisions have permitted, and indeed encouraged, the admission of evidence which is both highly unreliable and massively prejudicial, innocent defendants have again and again found themselves facing large numbers of allegations, all of which are false. Again and again, as could readily have been predicted (and as was foreseen in the original exclusionary similar fact principle), juries have convicted on the basis of such evidence.
30 The result over the past eleven years has been a systematic and continuing perversion of justice on a scale which would once have been unimaginable. Unlike previous systematic injustices in the twentieth-century, the original authors of these injustices are the courts themselves.
31 Because, ultimately, it is the state of the law which has led to the gravest series of miscarriages of justice in British legal history, it is the law which must now, as a matter of urgency, be changed and reformed.
In its report , published on 31 October 2002, the Home Affairs Committee responded to this memorandum and to other similar submissions by recommending that the law on similar fact evidence should be reformed. Having considered the in trials involving multiple allegations of sexual abuse, it that the requirement for similar fact evidence to be linked by 'striking similarities', as insisted on by their lordships in Boardman (1975), should be restored in cases involving allegations of historical child abuse. In this respect, as it acknowledged in its report (at paragraph 83) the Home Affairs Committee's recommendations run directly counter to those in the Government's White Paper Justice for All (2002), which proposes lowering the threshhold for the admission of similar fact evidence still further.
 This point has been made by Professor Colin Tapper who also notes that, of the panel of five law lords, there were three whose only experience before their elevation was in Scottish Law and one whose experience was exclusively in Chancery. (�The erosion of Boardman v DPP�, New Law Journal, August 11 1995, p. 1224, note 15).
In his opening address to the North
Wales Tribunal Gerard Elias QC implicitly acknowledged this: �In relation
to allegations of sexual abuse . . . questions of corroboration clouded
the issue for much of the period, but at least since the House of Lords
decision in DPP v P (1991), the prosecution of those against whom
more than one similar offence (or type of offence) is alleged has been
made procedurally and evidentially easier� (Opening Statement, Press
� Richard Webster, 2002