Abstract
The law's reliance on forensic science is an inevitable necessity. The rapid evolution of new forensic methods, allied with a too-easy reliance on expert forensic opinion does, however, expose the justice system to a range of dangers. The ever-growing list of miscarriages of justice involving misplaced reliance on flawed forensic and diagnostic techniques and misguided expert witnesses has led to numerous reviews of criminal convictions by the Criminal Cases Review Commission, as well as attracting the attention of the Law Reform Commission of England and Wales. In its latest report, Expert Evidence in Criminal Proceedings in England and Wales, 1 the Commission notes (at para 1.17) that “expert evidence of doubtful unreliability is being proffered for admission, and placed before the jury, too readily. … It has even been suggested that there may be a ‘culture of acceptance’ on the part of some trial judges”. The Commission has, consequently, proposed a statutory consolidation of the various common law rules governing the admissibility of expert opinion evidence in criminal cases, together with proposals for a new “reliability test”, allied with further guidelines for trial judges. In doing so, the Commission has sought to introduce legislative changes which go far beyond anything suggested in any other Anglo-American legal system and are likely to have a profound effect on the use of expert witnesses in the courts of England and Wales. It may be noted at this stage that the main thrust of the Law Commission's proposals relate to scientific and medical expert opinion, although these will naturally overlap with expert opinion in other areas (for instance, lip-reading techniques and forensic accountancy).
The miscarriages of justice which have been reviewed by the appellate courts in recent decades fall into a number of categories. The first of these have, regrettably, involved cases where the expert witness has deliberately misled the court, as demonstrated by cases such as Maguire and Ward. 2 In the former case, the Court of Appeal noted that the forensic scientist in question, employed by the Royal Armament and Research Development Establishment Laboratory, had deliberately misled the court as to the accuracy of a test known as thin-layer chromatography, used to test the presence of nitro-glycerine. In the latter case, the Court of Appeal identified the major cause of the injustice as arising from the fact that three government forensic scientists had failed to disclose material evidence in their pro-prosecution partisan approach adopted at trial.
The second category of cases involves expert witnesses who stray outside their discipline and opine on matters where their expertise is limited. A case in point is Clarke, whose conviction for the murder of her two infant sons was finally quashed in 2003. 3 The chief prosecution witness, who was a professor of paediatrics and child health (i.e. not a statistician), nonetheless provided expert opinion based on unreliable statistical evidence. His opinion, given to the trial court in a dogmatic manner, was that there was only a one in 73 million chance of two natural cot deaths (SIDS – sudden infant death syndrome), in the circumstances of this particular kind of case. The Court of Appeal, in its criticism of the manner in which the witness had strayed outside his own discipline, noted that it was “unfortunate that the trial did not feature any consideration as to whether the statistical evidence should be admitted in evidence”. The Law Commission, in this context, paid particular attention to the causes for concern identified in paediatric forensic pathology. 4
The third category of miscarriages of justice involves those instances where an expert witness relies on flawed or faulty forensic or diagnostic techniques. A case in point is that of Harris. 5 This case dealt with what is commonly referred to as shaken baby syndrome. In such cases, the prosecution had relied (as it had done in a number of previous cases) on expert testimony to the effect that deliberate head injuries to an infant could always be inferred once a particular triad of intra-cranial injuries had occurred. These were: acute encephalopathy (a disorder of the brain), subdural haemorrhage (bleeding around the brain) and retinal haemorrhage (bleeding in the retina). The Court of Appeal ruled, however, that while the presence of this triad of injuries could be a strong indicator of a non-accidental injury, it could not be relied on as positive proof. This point was repeated in the recent case of Butler where the Court of Appeal reiterated its caution against relying on expert medical evidence that this triad of injuries leads inexorably to a finding of guilt. 6 In recognition of these rulings the Crown Prosecution Service guidelines now make it clear that criminal charges would not be justified if the sole evidence against a defendant is the afore-mentioned triad of injuries. 7
The Law Commission's proposals focus on the second and third categories above. The Commission's Report provides a full and ample exploration of the major miscarriages of justice in this area and culminates in a draft Bill to provide both a number of statutory tests as well as guidance to trial judges. It is not possible to consider the full Report here. However, an overview of the following issues would be pertinent.
As far as the basic rules of admissibility of expert evidence are concerned, the Commission's draft Bill reiterates the common law position. Under clause 1(1) expert testimony is admissible only if the court is satisfied that the testimony would equip the court with “information which is likely to be outside a judge or jury's experience and knowledge, and, which would give them help they need in arriving at their conclusions”. In addition, clause 1(1) goes on to make it clear that admissibility depends on the prerequisite that the expert is properly qualified and that the evidence is not rendered inadmissible on the grounds of bias.
The issue of proper qualifications has long been treated by the common law with a certain amount of latitude; the Commission goes so far as to describe the current judicial approach (at paras 1.8 and 1.17) as laissez faire. A more stringent regime applicable towards expert witnesses is mandated by a number of factors, including the particular danger that, because of the experts' education and experience, a lay jury will accord the opinion expressed a probative value it would not otherwise possess. In their responses to the Law Commission's consultation paper, the General Medical Council had noted “a robust assessment is necessary”, while the Criminal Bar Association agreed that “rightly or wrongly, [expert evidence] is often ‘trusted’ like no other category of evidence”. 8 The Commission now proposes a new test in clause 2(1): “a person may be qualified to give expert evidence by virtue of study, training, experience or any other appropriate means”. This is not merely a matter for individual judicial discretion as clause 2(2) makes it clear that the court “must be satisfied on the balance of probabilities that the person is so qualified”.
The Commission goes on to explain (at paras 4.18–4.24) that the standard set out provides only a minimum acceptable level of expertise and that the proposals were in accordance with the views, among others, of the Society of Expert Witnesses. In the vast majority of cases (involving, for instance, surgeons and psychiatrists) this would be merely a matter for record and would not put the court (and the expert) to any undue delay or inconvenience. In other matters, the expert witness concerned would be required “to prove in a more direct way that he or she has the skill. In other words, the individual, or the party calling the individual, might need to provide the court with the results of a relevant test or demonstration undertaken in controlled conditions which show that he or she is skilled and therefore qualified to provide evidence as an expert witness”. The Commission acknowledges a number of potential adverse consequences but rebuts these. First, there is the objection that this would lead to unnecessary delay and costs. Their response is that this would be a small price to pay given the importance of the matters at stake. Moreover, the courts should already be requiring proof of expertise whenever it is sought to adduce unusual or novel skills or techniques. In any case, the Commission notes that due to the statutory nature of the requirements imposed, any expert witness would be sufficiently aware, in advance, that they would be required to provide a sufficient explanation of their qualifications as the basis for their expert opinion. Finally, the Commission points out that their proposals are not intended to impose an obligation of periodic testing: “We are merely recommending that any individual who wishes to give expert evidence in criminal proceedings should have to prove that he or she is in fact an expert.” 9
It is submitted that the proposed test, given its wide generality, is in fact extremely generous to the party wishing to rely on the particular expert in question. Admissibility requirements have, perforce, to be of this nature. Critics of such a generalised test would have to acknowledge that the only possible alternative would have been to create multiple registers for accredited witnesses, which would be neither practical nor feasible.
In relation to the misplaced reliance on flawed scientific techniques (especially diagnostic techniques), the Law Commission's proposals highlight the problems caused by an expert's reliance on a hypothesis that is not sufficiently tested. The Court of Appeal had commented on this issue in Henderson. 10 Here the Court noted that where the prosecution, through the use of expert opinion, identified a non-accidental injury for which the defence could not provide an alternative cause, there was the temptation of an easy progression to the unwarranted conclusion that the defendant was guilty. “Such a temptation must be resisted. In this, as in so many fields of medicine, the evidence may be insufficient to exclude, beyond reasonable doubt, an unknown cause.” Under clause 4(1) of the draft Bill, therefore, if reliance is to be placed on a hypothesis (for instance that an infant's head injuries are non-accidental), then the expert relying on that hypothesis can only do so if (a) the opinion is “soundly based”, and (b) the “strength of the opinion is warranted having regard to the grounds on which it is based”. Clause 4(2) then goes on to provide a list of the factors that may be considered by the court in its determination that the expert opinion evidence was not sufficiently reliable: (a) the opinion is based on a hypothesis which has not been subjected to sufficient scrutiny (including, where appropriate, experimental or other testing), or which has failed to live up to scrutiny; (b) the opinion is based on unjustifiable assumptions; (c) the opinion is based on flawed data; (d) the opinion relies on an examination, technique, method or process which was not properly carried out or applied, or was not appropriate for use in the particular case; and (e) the opinion relies on an inference or conclusion which has not been properly reached.
In a further elaboration, a Schedule to the draft Bill, under clause 4(3), provides a list of eight generic factors affecting reliability to be taken into consideration. These range from a consideration of the extent and quality of the data on which the opinion is based, together with the validity of the methods by which the data was obtained, to a consideration of whether the expert had complied with the established practice in that area of expertise. Most of these involve the classic safeguards of valid science allied with a sufficiently stringent process of publication and peer review. The Commission had also been urged to draw up a similar list of “reliability factors” for specific fields of scientific expertise. The British Psychological Society, for instance, had favoured specific guidelines for psychologists testifying as expert witnesses (see para 3.55 of the Report). The Commission, wisely, it is submitted, chose not to go down this route, commenting (at para 3.56) that providing subject- and discipline-specific guidelines “must remain a long-term goal”. Instead the Schedule merely confers power for such a compilation in the future.
The Commission, in justifying their proposals, applied these retrospectively to the Harris case (para 5.67 of the Report). The evidence base for the hypothesis of shaken baby syndrome could be described as an “inverted pyramid”. The database on which the hypothesis was based was small and was allied with poor-quality research. Moreover, there was a considerable divergence of expert opinion on the subject. 11
The Commission notes that their proposals vis-à-vis the Harris case would have had the following consequences: First, the experts, because of the need to demonstrate the reliability of their hypothesis would have relied on reliable research, either of their own or from reliable sources, including the identification of flaws. This would have included the publication of peer-reviewed papers. Second, in the light of the greater judicial scrutiny required, the experts would have been aware of the need to properly assess reliability and to moderate their opinion according to the limitations of the research. In Harris itself, the expert witnesses had expressed their opinions with a degree of certainty that was not warranted. Third, the greater onus on the trial judge would have meant that the experts would have been permitted to give their opinions only to the extent that was justified on the available research data. Fourth, given the paucity of the available research, it would have been highly unlikely that the prosecution would have been allowed to advance a case based solely on that opinion evidence. Fifth, this would not have prevented a conviction if there had been other relevant and cogent evidence, which, taken together with expert opinion (even if weak), would have satisfied the burden of proof on the prosecution.
In the Clark case (above), curial scrutiny along the lines suggested would have resulted in the conclusion that the expert's hypothesis (that there was only a one in 73 million chance of two natural cot deaths in the same family), was not based on sufficiently reliable data. The Court of Appeal held (at para 173 in its judgement) that this figure grossly misrepresented the likelihood of two sudden deaths within the same family. On this point the Commission notes (at para 8.18), “We believe an investigation into the expert's hypothesis would have revealed little if any evidence to support it, and indeed would in all likelihood have revealed evidence that SIDS deaths are more likely to occur in families where there is a history of SIDS”.
It is difficult to take issue with the Law Commission's conclusions that the vast majority of the recent miscarriages of justice would not have occurred if similar considerations had been applied. Two other cases may be mentioned. In Dallagher, the expert testimony relating to the alleged unique nature of ear prints and their use for purposes of identification was clearly a hypothesis that fell squarely within the factors enumerated in clause 4(2) of the draft Bill, above. 12 Similarly, in Gilfoyle one issue for the court related to the admissibility of the evidence of a forensic pathologist. This expert witness had opined that it was unlikely that the deceased committed suicide as she was within the final trimester of her pregnancy, leading to an inference that the defendant had stage-managed her death so that it appeared that she had hanged herself. The Court of Appeal concluded that there were, in fact, no established criteria by which this opinion could be tested, there was no substantial body of academic writing which supported the hypothesis and the expert witness had made “speculative conclusions that were not the stuff of which expert evidence was made”. 13
Among the cases that had led to public and judicial disquiet were those where it was abundantly clear that the testimony of the expert witness had not been wholly impartial. Far from being neutral and independent, the witness had adopted (in the main) a pro-prosecution stance, which inevitably raised concerns as to reliability. A number of cases had set out the relevant common law principles. 14 These principles were further refined in the Criminal Procedure Rules 2010 where the three essentials of the expert's duty to the court are set out in Rule 33.2. It is made explicit that, first, the expert is to give objective, unbiased opinion on matters within his expertise; second, that this duty overrides any obligation to the party who instructs or pays him; and third, that he is obliged to inform the court, and all parties, if he changes his opinion. In a further reinforcement of these cardinal principles the draft Bill repeats these in clause 3(1) and (2) but also goes on at clause 3(3) to state that if the court concludes that the expert has not complied with the above requirements, the court may go so far as to rule that the evidence is inadmissible. 15 Under clause 3(4), however, the fact that the expert has an association (for instance, an employment relationship; particularly important in cases involving employees of investigatory bodies) does not, by itself, indicate a risk to impartiality. The Commission explains the intended manner of operation of its proposals (at para 4.28) as follows: first, the court is to adopt a presumption of impartiality as a starting point; second, this presumption may be rebutted if there is sufficient evidence indicating a significant risk of the expert's non-compliance with the duty of impartiality; third, if the court considers that there is such a significant risk, a presumption of inadmissibility arises; fourth, despite this, the evidence may be admitted if it is in the interests of justice to do so (for instance, where, despite the risk of bias, the risk is adjudged to be relatively low).
The Law Commission's Report deals with much else besides the points above. For instance, there is a re-statement of the importance of prior disclosure of expert opinion as well as a consideration of pre-trial hearings to determine the admissibility of novel or contested scientific techniques or diagnoses. This is based on the US case of Daubert v Merrell Dow Pharmaceuticals 16 and itself derived from the much earlier case of Frye v US. 17 There are also recommendations, as existing in civil cases, for court-appointed experts. Full consideration of these is outside the scope of the present paper. However, two final points may be raised in conclusion. The Commission declined to adopt a proposal that had been put to it which would have allowed the trial judge to stop the trial and discharge the jury in situations where there has been unwarranted reliance on expert evidence. While such a power had been included in other statutes, it was felt (see para 5.86–5.88) that there were sufficient existing general safeguards not to require such a specific power to be included in the draft Bill. In addition, the Commission also dealt (at paras 5.89–5.94) with the issue of appellate reviewability of any decision at first instance to admit expert opinion evidence. In brief, this is dealt with on the basis that the new admissibility and reliability tests deal with questions of law (which are reviewable), rather than questions of fact (which generally are not).
While the substance of the recommendations put forward already exist under various common rules of evidence and procedure, these have sometimes been honoured more in the breach than in practice. Any re-statement of these rules in a statutory codified form is, therefore, to be welcomed. As the General Medical Council, and others, had noted “together with statutory guidelines, the [proposed] test would tighten the criteria and clarify the powers of the court, thereby reducing the risk of unreliable evidence being placed before a jury”. 18 In a final note of caution, however, it must be noted that the Law Commission can do no more than submit its recommendations and its draft Bill to Parliament. Whether there is a political will to take up this draft Bill, and to carve out the necessary parliamentary time for debate and enactment, is another matter. 19