Taking experts out of the court
The US courts set out four criteria for determining whether
expert testimony meets the requirement that it constitutes
‘scientific knowledge’. Is this the way forward?
The number of cases in which the decision of the court has been challenged on the basis of allegedly ‘unsafe’ expert evidence continues unabated. The recent high-profile case of Sion Jenkins is simply one more to be added to a growing catalogue. We have now reached a position where the mere fact that an expert has been involved previously in such a case appears to be sufficient grounds for launching a challenge.
Previous expert criticism counts
In R -v- Kenneth Konrad Pedder (2004) CA (Crim Div) (Latham LJ, Gray J) 15/7/2004, the defendant appealed against a conviction of murder. Dr Alan Williams, medical expert for the Crown, gave evidence that the death was caused by asphyxia, probably due to pressure from a pillow, the scene of death being entirely consistent with the cause of death. The medical expert for Mr Pedder did not discount asphyxia as one of the possible causes of death. However, the defence expert could not exclude the possibility that the death was an unexpected death caused by alcohol or an epileptic seizure.
On appeal, the defendant contended that the conviction was unsafe, and one of the grounds put forward was the court’s criticisms of Dr Williams in R -v- Sally Clark (2003) EWCA Crim 1020. The appeal was dismissed. The Court of Appeal considered the criticism of Dr Williams in R -v- Sally Clark but found that this was not directed to the professional quality of his diagnosis but to the fact that he had withheld material that could have undermined his diagnosis. It was not the case here: the medical expert for Mr Pedder had been given access to all the material Dr Williams had used. Whilst it could be accepted that criticism of Dr Williams might affect a court’s consideration of his credibility, the medical evidence in this case formed only part of the prosecution’s case. The surrounding circumstances, said the court, pointed firmly to the conclusion that the death was caused by asphyxia as a result of the appellant’s actions.
New scientific tests questioned
There have also been a number of challenges recently to convictions in cases where doubt has been cast on the credibility of expert evidence because of insufficient scientific evidence to support the techniques used. In R -v- Gerrard Francis Luttrell and others (2004) EWCA Crim 1344 the appellants appealed against a decision allowing lip-reading evidence to be admitted at their trials. They had been convicted of conspiracy to handle stolen goods. At Mr Luttrell’s trial, a skilled lip-reader had given evidence for the prosecution as to what was said at a meeting between Luttrell and another co-accused and which had been recorded on CCTV. The judge had ruled that this lip-reading evidence was admissible. However, Mr Luttrell submitted that the lip-reading evidence via video footage should not have been admitted because it had not been shown to be reliable. He further submitted that a special warning had been necessary, the judge’s directions had been inadequate and the prosecution had not disclosed all material relevant to the lip-reader’s expertise and reliability.
The Court of Appeal held that lip-reading evidence from a video, like facial mapping, was a type of real evidence that was perfectly capable of passing the ordinary tests of relevance and reliability. It was, therefore, potentially admissible in evidence. Once ruled to be admissible, lip-reading evidence did, however, require a special warning from the judge as to its limitations and risks of error, because such evidence would usually be introduced through an expert who might not be completely accurate. The precise terms of the direction, said the court, would depend on the facts of the particular case.
Similarly, in the earlier case of R -v- Dallagher (2002) EWCA Crim 1903, there had been a similar appeal against a conviction for murder. The victim was a 94 year-old arthritic and deaf woman who had been killed in her bed by an intruder who had suffocated her with a pillow. The intruder had entered the woman’s bedroom through a small window above her bed. The issue at trial had been whether Mr Dallagher was the intruder.
Ear prints were found on the glass of the window. Two experts matched those ear prints with control ear prints provided by Dallagher. Two experts gave evidence at trial: one was sure that they were Mr Dallagher’s prints and the other thought it highly likely. The defendant, himself, had led no expert evidence.
Following conviction, however, some scientists had expressed misgivings as to the extent to which ear print evidence alone could be used safely to identify a suspect. Consequently, Mr Dallagher sought to bring fresh evidence on appeal from three expert witnesses casting doubt on the reliability of ear print evidence. He submitted that:
(i) the trial judge had been wrong to have admitted the ear print evidence; and, alternatively,
(ii) if it was admissible, his case could have been made more favourable by the inclusion of evidence from the experts he now sought to rely upon; and
(iii) the judge had been wrong to allow the inclusion of evidence of previous burglaries committed by the appellant.
The Court of Appeal ruled that the developments that had taken place since the trial concerning ear comparison evidence constituted a reasonable explanation for the failure of the defence to adduce their expert evidence at trial. Therefore the evidence relating to these developments could be received if it might afford a ground for allowing the appeal. It was accepted, however, that there was no basis for excluding the evidence of the ear prints found at the scene. It was difficult, said the court, to see how it was possible to exclude evidence that was part of the investigatory process. What mattered was the value of the conclusion drawn. There was no objection in principle to the criminal justice system taking into account modern methods of crime detection. The expert evidence in the instant case had not been irrelevant or so unreliable that it should have been excluded.
Where the problems lie
These cases highlight three areas of concern:
- The suitability and qualification of an individual expert and the reliability of that expert’s evidence
- The problem of frontier science or pseudo-science, and what happens when there are new developments
- Risk evaluation in relation to expert evidence that is not guaranteed to be free from error
In the past, we have looked at various proposals concerned with the early review of expert evidence. In particular, following the Sally Clark appeal, there were calls for review panels to be appointed in cases involving children to test expert opinion prior to such evidence going before the court. But is it possible to construct a system of review that will reduce the number of cases where the value or probity of the expert evidence is later challenged, either at trial or on appeal?
Burden of proof in the UK
The burden of proof requires that the evidence in a case should be rigorously tested and evaluated at every stage. This should be done from the earliest pre-trial stage onwards. In reality, however, the expert evidence does not come under real scrutiny until trial, and sometimes even later than that.
The British civil and criminal justice systems are already taking steps to bring expert evidence under greater control by the courts. Permission to call such evidence and its admissibility are already matters that the courts seek to regulate. In reality, however, once expert evidence is admitted to proceedings, it is dealt with inside an adversarial system that places the burden of testing the evidence on the parties. The courts themselves will not examine the nature of scientific knowledge at any given time, and will rarely undertake a rigorous examination of the science or techniques used.
USA sets four criteria to test expert evidence
In the United States, where the problem of partisan expert evidence was even more acute than here, the case of Daubert -v- Merrell Dow Pharmaceuticals Inc (1992) 509 US 579 laid down a four-part test to be applied to all expert evidence.
Rule 702 of the Federal Rules of Evidence (FRE) states that if scientific, technical or other specialised knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training or education may testify thereto in the form of an opinion or otherwise. Difficulty arose over how this rule should be interpreted and applied. In Daubert, the court set out four criteria for determining whether expert testimony met the requirement that it constitutes ‘scientific knowledge’. These are:
- Whether the theory or technique ‘can be (and has been) tested’.
- Whether the ‘theory or technique has been subjected to peer review and publication’.
- In the case of a particular technique, what ‘the known or potential rate of error’ is or has been.
- Whether the evidence has gained widespread acceptance within the scientific community.
In addition, the court may have to weigh the probative value of the evidence against its possible prejudice under Rule 403 FRE. This rule gives the judge a discretion to exclude expert testimony where its ‘probative value’ is substantially outweighed by the ‘danger of unfair prejudice, confusion of the issues, or misleading the jury’.
Furthermore, in making all of these determinations, the court may also have to consider whether any additional steps, such as the appointment of a court expert, are warranted.
The issue of testing was considered in Gier -v- Educational Serv. 845 F. Supp. 1342 (D.Neb. ’94). This case concerned a claim of abuse by students in a residential home for mentally retarded children. The magistrate judge considered a motion to exclude the testimony of a psychiatrist and two psychologists who had examined the claimants. The examinations (which consisted of a history, a psychological evaluation utilising standardised tests and a clinical interview) concluded that the claimants had been sexually, physically and emotionally abused. The focus of the expert evaluation was ‘to compare the behavior or symptoms of the child to that of victims of child abuse’.
The court, citing Daubert, inquired whether the methodology used was ‘capable of being falsified’, i.e. whether it actually employed the scientific method. Answering ‘No’, the magistrate found that this type of psychological evidence was both controversial and, for the most part, ‘irrefutable’, because it could only be tested by research on patients. As a result, the expert conclusions were not subject to the type of proof available in the physical sciences. In the absence of objective research to test the expert theories or to assess their conclusions, the court found that the expert techniques did not satisfy the Daubert standard.
Under the Daubert test, peer review and publication are ‘pertinent considerations’ for determining the reliability of scientific evidence. Where the evidence comes from the mainstream of its respective field, this consideration will weigh heavily in favour of admission. If the evidence is based primarily on principles that have commonly appeared in authoritative medical treatises, the court can presume it ‘has been subjected to peer review and publication and accepted by the scientific community as having significant probative value’. Unusual or dogmatic theories are, so far as they can be, excluded. In cases where an expert presents a new or marginal theory, this factor is likely to fail the Daubert test. Thus, the magistrate hearing the Gier case gave little weight to controversial literature on the symptomatology of abuse. It was not sufficiently within the mainstream of scientific consensus.
The courts in America will take into account the existence of control standards that monitor the success or error rate of a particular scientific technique. The non-existence of such standards can, in some circumstances, render the evidence inadmissible.
So the rules of evidence generally extant at the federal and state levels anticipate that relevant, unprivileged evidence should be admitted and its weight left to the factfinder, who would have the benefit of cross-examination and contrary evidence by the opposing party. However, because there is a perception that expert testimony (as opposed to the evidence of a lay witness) is more likely to mislead, Daubert arms the district courts with broader discretion to exclude expert testimony.
In so far as the expert’s past experience is concerned, the United States expects similar standards to our own. An expert in an American court is also required, however, to furnish a list of all the cases in which he or she has been involved over the preceding 4 years.
The Daubert effect
As a result of Daubert, expert evidence in America is likely to come under close scrutiny at a relatively early stage in the proceedings. The parties are aware of the requirements from the outset, and it is common for the court to hear applications in relation to the admissibility or desirability of such evidence well before any final hearing.
Would such rules applied to our own civil or criminal justice system have saved the jury from hearing Sir Roy Meadow’s notorious statistical evidence in Sally Clark? Would they have spared Alan Williams the indignity of an appeal in the Pedder case? There would certainly have been an earlier evaluation of the expert evidence in both Luttrell and Dallagher, and there is also the possibility that such evidence would altogether fail to pass a test similar to that applied in Daubert.
Daubert is certainly not without its own problems, and it may, indeed, create more than it solves. However, American lawyers have, at least, made some attempt to address the particular difficulties surrounding the nature of expert evidence and its relationship to the judicial process. Perhaps it is time for the English courts to formulate similar rules. They might do better than the American model, but we think, at least, that they could do no worse.