Law Commission - 2011 - EXPERT EVIDENCE IN CRIMINAL PROCEEDINGS IN ENGLAND AND WALES

Download as pdf or txt
Download as pdf or txt
You are on page 1of 222

Law

Commission
Reforming the law

Expert Evidence in Criminal Proceedings in


England and Wales

Law Com No 325


The Law Commission
(LAW COM No 325)

EXPERT EVIDENCE IN CRIMINAL


PROCEEDINGS IN ENGLAND AND
WALES

Presented to Parliament pursuant to section 3(2) of the Law


Commissions Act 1965

Ordered by The House of Commons to be printed


21 March 2011

HC 829 London: The Stationery Office £37.00


© Crown copyright 2011

You may re-use this information (not including logos) free of charge in any format or medium,
under the terms of the Open Government Licence. To view this licence, visit
http://www.nationalarchives.gov.uk/doc/open-government-licence/ or write to the Information
Policy Team, The National Archives, Kew, London
TW9 4DU, or e-mail: psi@nationalarchives.gsi.gov.uk.

Any enquiries regarding this publication should be sent to us at The Law Commission, Steel
House, 11 Tothill Street, London SW1H 9LJ.

This publication is also available on http://www.official-documents.gov.uk/

ISBN: 9780102971170

Printed in the UK for The Stationery Office Limited


on behalf of the Controller of Her Majesty’s Stationery Office

ID 2419408 03/11

Printed on paper containing 75% recycled fibre content minimum.

ii
THE LAW COMMISSION
The Law Commission was set up by the Law Commissions Act 1965 for the purpose of
promoting the reform of the law.

The Law Commissioners are:


The Right Honourable Lord Justice Munby, Chairman
Professor Elizabeth Cooke
Mr David Hertzell
Professor David Ormerod
Miss Frances Patterson QC

The Chief Executive of the Law Commission is Mr Mark Ormerod CB.

The Law Commission is located at Steel House, 11 Tothill Street, London SW1H 9LJ.

The terms of this report were agreed on 21 February 2011.

The text of this report is available on the Internet at:


http://www.lawcom.gov.uk/expert_evidence.htm

iii
THE LAW COMMISSION

EXPERT EVIDENCE IN CRIMINAL PROCEEDINGS


IN ENGLAND AND WALES

CONTENTS

Paragraph Page

PART 1: INTRODUCTION

Background to this project 1.1 1

Why special rules for expert evidence? 1.13 3

Our provisional proposals 1.31 8

A summary of our recommendations 1.36 9

Acknowledgments 1.44 10

The structure of this report 1.46 10

PART 2: THE CURRENT LAW

The common law admissibility test 2.1 13

Assistance 2.3 13

Relevant expertise 2.6 13

Impartiality 2.8 14

Evidentiary reliability 2.12 15

The relationship between the four admissibility tests 2.17 16

Opinion evidence and evidence of fact 2.19 17

PART 3: CONSULTATION

Introduction 3.1 18

Evidentiary reliability 3.3 18

Our provisional proposal 3.3 18

Comments on our provisional proposal 3.12 20

iv
Paragraph Page

Our recommendation on the reliability test 3.35 25

Guidance for the judiciary 3.40 26

A limited power to disapply the reliability test 3.65 32

The onus of persuasion 3.79 34

Codification of the admissibility test generally 3.125 45

PART 4: CODIFICATION

Introduction 4.1 48

The Turner test 4.12 49

The qualification (expertise) test 4.15 50

The impartiality test 4.25 52

The scope of clauses 1 to 3 4.37 55

Monitoring compliance with clauses 1(1)(b) and 2 4.38 55

PART 5: EVIDENTIARY RELIABILITY

Introduction 5.1 58

The reliability test 5.3 58

Specific factors in the Schedule 5.19 62

A limited power to disapply the reliability test 5.42 69

The onus of persuasion 5.62 74

Applying the reliability test in practice 5.63 74

Scientific (medical) evidence 5.64 74

Non-scientific evidence 5.71 78

Summary 5.83 81

A power to stop the trial? 5.86 82

Addressing the judge’s ruling on appeal 5.89 82

Closing comments 5.99 84

PART 6: COURT-APPOINTED EXPERTS

Introduction 6.1 89

v
Paragraph Page

The pre-trial hearing 6.15 91

The view of our consultees 6.25 93

Our revised approach 6.41 97

The efficacy of our proposed scheme 6.59 101

Costs, inconvenience and possible delays 6.69 103

Recommendation 6.74 105

PART 7: PROCEDURAL MATTERS

Introduction 7.1 107

Clause 8 7.10 109

Amending the Criminal Procedure Rules 7.17 111

Expert reports 7.21 112

Other pre-trial disclosure 7.27 115

Developing rule 33.6 of the Criminal Procedure 7.43 118


Rules

Directing the jury 7.53 121

PART 8: THE NEW TEST IN PRACTICE

Introduction 8.1 124

The cases in our consultation paper 8.9 126

Ear-print identification evidence – Dallagher 8.10 126

Statistical evidence on SIDS – Clark (Sally) (No 2) 8.13 127

Inferring murder from unexplained infant deaths – 8.21 129


Cannings

Inferring a non-accidental cause from intra-cranial 8.24 130


injuries – Harris

A hypothetical case 8.31 132

PART 9: SUMMARY OF RECOMMENDATIONS 137

APPENDIX A: DRAFT CRIMINAL EVIDENCE (EXPERTS) BILL 144


WITH EXPLANATORY NOTE AND FLOW DIAGRAM

vi
Page

APPENDIX B: PART 33 OF THE CRIMINAL PROCEDURE 159


RULES 2010

APPENDIX C: IMPACT ASSESSMENT 164

APPENDIX D: ACKNOWLEDGEMENTS 20

vii
viii
THE LAW COMMISSION

EXPERT EVIDENCE IN CRIMINAL


PROCEEDINGS IN ENGLAND AND WALES
To the Right Honourable Kenneth Clarke QC, MP, Lord Chancellor and Secretary
of State for Justice

PART 1
INTRODUCTION

BACKGROUND TO THIS PROJECT


1.1 This report follows the publication of our recent consultation paper, The
Admissibility of Expert Evidence in Criminal Proceedings in England and Wales1
and makes recommendations in the light of the comments we received on the
provisional proposals made in that paper. We now set out and explain our
recommendations for reforming the law relating to expert evidence in criminal
proceedings. We also provide a draft Criminal Evidence (Experts) Bill which, if
enacted, would give effect to our principal recommendations.

1.2 Our decision to address the law on expert evidence was prompted by a call for
reform from the House of Commons’ Science and Technology Committee.2 We
shared the Committee’s concern that expert opinion evidence was being admitted
in criminal proceedings too readily, with insufficient scrutiny.

1.3 In our consultation paper, we provided some examples of wrongful convictions in


cases involving unreliable expert opinion evidence adduced by the prosecution.
We believe that if the relevant provisions of our draft Bill had been in force at the
time of those proceedings, the problems we identified in those cases, which we
summarise below, would almost certainly not have occurred. We explain why in
Part 8.

1.4 In the case of Dallagher,3 D’s conviction for murder was based on unreliable
expert opinion evidence relating to the comparison of an ear-print made by D with
a latent ear-print found on a window. At D’s trial, one of the experts opined that
he was “absolutely convinced” that D had left the latent print, and a second
prosecution expert was willing to countenance only a “remote possibility” that the
latent print had been left by someone else. Notwithstanding the strength of these
opinions, DNA evidence taken from the latent print subsequently established that
it had not been left by D, demonstrating the unreliable nature of the evidence
used to secure his conviction.4

1
Law Commission Consultation Paper No 190 (2009). References in this report to a
“consultation paper” are references to this paper.
2
Consultation Paper No 190, paras 3.15 to 3.17.
3
[2002] EWCA Crim 1903, [2005] 1 Cr App R 12.
4
D’s conviction was quashed (and a retrial ordered) before the DNA evidence became
available; see The Guardian, 23 January 2004.

1
1.5 In Clark,5 an expert paediatrician gave unreliable opinion evidence. This expert,
who was not a statistician, had formulated his opinion on the assumption that
there were no genetic or environmental factors affecting the likelihood of naturally
occurring cot deaths,6 opining that there was only a one in 73 million chance of
two such deaths in the same family. The Court of Appeal took the view that the
figure grossly misrepresented the chance of two sudden deaths within a family
from unexplained but natural causes, and added that if the issue of the statistical
evidence had been fully argued it would probably have provided a distinct basis
upon which to allow C’s appeal.7 The court also noted that the way the expert
had presented his evidence could have had a major impact on the jury’s
deliberations.

1.6 In Cannings,8 C’s convictions for the murder of her two infant sons had been
based on the dogmatic expert view (that is, a view based on a hypothesis which
had not been sufficiently scrutinised or supported by empirical research)9 that the
mere fact of two or more unexplained infant deaths in the same family meant that
murder had been committed. The Court of Appeal quashed C’s convictions.
Fresh evidence suggested that multiple cot deaths in the same family could have
an underlying genetic cause; and a report relating to the largest follow-up study of
cot-death families concluded that “the occurrence of a second unexpected infant
death within a family is … usually from natural causes”.10

1.7 Until the judgment of the Court of Appeal in Harris and others,11 the prosecution
had been allowed to rely on a hypothesis that a non-accidental head injury to a
young child could confidently be inferred from nothing more than the presence of
a particular triad of intra-cranial injuries. The prosecution had in effect been able
to rely on nothing more than expert opinion evidence based on the triad to secure
convictions for very serious offences against the person, including murder.12 This
was the case even though the diagnosis of a violent assault was predicated on
empirical research which has been criticised as comprising only a small, poor-
quality database.13 In other words, the hypothesis underpinning the diagnosis had

5
[2003] EWCA Crim 1020, [2003] 2 FCR 447 (second appeal).
6
Or Sudden Infant Death Syndrome (“SIDS”).
7
[2003] EWCA Crim 1020, [2003] 2 FCR 447 at [178] to [180]. The appeal was allowed for
unrelated reasons; see Clark [2003] EWCA Crim 1020 at [164] and Consultation Paper
No 190, para 2.16. It is noteworthy that the report containing the data the expert relied on
was accompanied by explanatory text which warned that the data did “not take account of
possible familial incidence of factors other than those included”.
8
[2004] EWCA Crim 1, [2004] 1 WLR 2607.
9
[2004] EWCA Crim 1, [2004] 1 WLR 2607 at [18] to [20].
10
[2004] EWCA Crim 1, [2004] 1 WLR 2607 at [141].
11
[2005] EWCA Crim 1980, [2006] 1 Cr App R 5.
12
See Editorial, British Medical Journal 29 July 2010 (issue 2771): “For 40 years,
mainstream medical experts who give evidence in court have largely agreed that shaken
baby syndrome can be unambiguously diagnosed by a triad of symptoms at post-mortem
… . Murder convictions are often secured on the basis of these alone, even in the absence
of other signs of abuse … .”
13
See M Donohoe, “Evidence-based Medicine and Shaken Baby Syndrome” (2003) 24
American Journal of Forensic Medicine and Pathology 239, 241. See also D Tuerkheimer,
“The Next Innocence Project: Shaken Baby Syndrome and the Criminal Courts” (2009) 87
Washington University Law Review 1, 12 to 14 and 17 to 18.

2
been insufficiently scrutinised or supported by empirical research to justify the
strong opinion evidence founded on it.

1.8 In our consultation paper we explained that the common law approach to the
admissibility of expert opinion evidence is one of laissez-faire, with such evidence
being admitted without sufficient regard to whether or not it is sufficiently reliable
to be considered by a jury. We concluded that this is unsatisfactory and proposed
that the common law approach should be replaced by a new admissibility test set
out in primary legislation.

1.9 We expressed particular concern about expert opinion evidence which is


presented as scientific. We explained that, for evidence of this sort, there is a
danger that juries will abdicate their duty to ascertain and weigh the facts and
simply accept the experts’ own opinion evidence, particularly if the evidence is
complex and difficult for a non-specialist to understand and evaluate.14

1.10 However, our proposals were not limited to scientific or purportedly scientific
evidence. We also addressed other types of expert evidence: non-scientific
expert evidence such as the opinion evidence of lip-readers and forensic
accountants.

1.11 The provisional conclusion we reached in our consultation paper was that special
rules are required for assessing the reliability of expert evidence as a factor
bearing on admissibility, and that opinion evidence with insufficient indicia of
reliability (that is, pointers to reliability) ought not to be admitted in criminal
proceedings.15 This is still our view.

1.12 We also believe, as we explain in Part 7, that there should be further disclosure
obligations in relation to all expert evidence, whether the evidence is relied on by
the prosecution or by the defence.

WHY SPECIAL RULES FOR EXPERT EVIDENCE?


1.13 There are several reasons why we believe special rules on admissibility and
disclosure are needed for expert evidence in criminal proceedings.

1.14 First of all, expert witnesses are quite different from other witnesses
(conventional witnesses of fact). Expert witnesses stand in the very privileged
position of being able to provide the jury with opinion evidence on matters within
their area of expertise and outside most jurors’ knowledge and experience.16

14
Concerns regarding the reliability of expert opinion evidence primarily relate to cases tried
before a judge and jury in the Crown Court. We should stress at the outset, however, that
similar problems may arise in other criminal proceedings, and our use of the term “jury”
should be taken to encompass lay magistrates and professional judges who sit as the fact-
finding tribunal in magistrates’ courts (or in the Crown Court on appeal against a conviction
in a magistrates’ court).
15
We say “special rules” because, as a general rule, factors bearing on the reliability of
evidence go to weight rather than admissibility.
16
Experts will occasionally provide evidence of fact, such as how a particular piece of
machinery works, but they are usually called to provide an opinion based on their special
knowledge and experience. Non-expert witnesses are prohibited from providing opinion
evidence save for the concession which permits any witness to present his or her oral
evidence of what he or she perceived in a natural way.

3
Moreover, following the demise of the so-called “ultimate issue rule”, expert
witnesses can even provide opinion evidence on the disputed issues the jury has
been empanelled to resolve.17

1.15 A related point, touched on already, is that a jury, comprised as it is of lay


persons, may not be properly equipped in terms of education or experience to be
able to address the reliability of technical or complex expert opinion evidence,
particularly evidence of a scientific nature.18 This being the case, there is a real
danger that juries may simply defer to the opinion of the specialist who has been
called to provide expert evidence, or that juries may focus on perceived pointers
to reliability (such as the expert’s demeanour or professional status).19

1.16 As the UK Register of Expert Witnesses accepted in their response to our


consultation paper, because expert evidence, or much of it, is heavily based in
opinion, special rules are required to ensure that it “is to inform rather than
mislead, particularly in criminal trials dominated by expert evidence”. Similarly,
the General Medical Council said: “it is because juries and other lay tribunals
tend to afford a special status to [scientific medical] evidence that a robust
assessment of its admissibility prior to trial is critical”. The Criminal Bar
Association noted in its response to our consultation paper that, “rightly or
wrongly, [expert evidence] is often ‘trusted’ like no other category of evidence”.20

1.17 Secondly, as explained above, a number of recent criminal cases suggest that
expert opinion evidence of doubtful reliability is being proffered for admission,
and placed before the jury, too readily. This follows from the current laissez-faire
approach to admissibility.21 It has even been suggested that there may be a
“culture of acceptance” on the part of some trial judges, particularly in relation to
evidence of a scientific nature.22

1.18 The Criminal Bar Association, agreeing with the proposals in our consultation
paper, commented “that the current treatment of expert evidence in criminal
proceedings has contributed to a significant number of miscarriages of justice,
risks continuing to do so, and requires urgent reform”. And, in line with a point we
17
The “ultimate issue rule” was the common law rule which prevented experts from giving an
opinion on the disputed facts in issue.
18
In his response to our consultation paper, Lord Justice Aikens noted the increasing
technicality of expert evidence, scientific or otherwise, the length of time needed to present
it to the jury and the difficulty for the jury in being able to cope with some expert evidence
or being able to assess it rationally.
19
Consultation Paper No 190, paras 2.3 to 2.11 and 2.28.
20
Similarly, the London Criminal Court Solicitors’ Association accepted that expert evidence
“has an effect on the fact-finding tribunal … like no other type of evidence”; and the
Association of Forensic Science Providers accepted that scientific expert evidence can
have a disproportionate effect on juries.
21
See Consultation Paper No 190, Part 3. See also para 2.16 below.
22
Andrew Campbell-Tiech QC, told us that, in his experience, “there is a culture of
acceptance [of expert medical evidence] that needs to change. We need judges whose
approach is one of engaged enquiry”. Similarly, the Criminal Cases Review Commission,
commenting on the proposals in our consultation paper, opined that judges “need to guard
against complacency” and “ensure that they are prepared to question and probe”
assertions made by expert witnesses which may sound impressive at face value. The
General Medical Council argued that trial judges should have a more proactive role in
scrutinising and assessing expert medical evidence.

4
made in our consultation paper,23 Associate Professor William O’Brian (University
of Warwick) commented that “virtually all of the areas of ‘forensic science’, with
the exception of DNA evidence, have quite dubious scientific pedigrees”.24

1.19 In a similar vein, Judge Andrew Gilbart QC, the Honorary Recorder of
Manchester, told us that he is often struck by “how poor some suggested
scientific evidence is in criminal trials”, adding that he is also frequently struck by
“how ill equipped advocates are to challenge it when they have no experts of their
own to advise them”.

1.20 Cross-examination, the adduction of contrary expert evidence and judicial


guidance at the end of the trial are currently assumed to provide sufficient
safeguards in relation to expert evidence, by revealing to the jury factors
adversely affecting reliability and weight.25 However, as we explained in our
consultation paper, and repeat below, it is doubtful whether these are valid
assumptions.26 A more credible assumption, at least in relation to complex
scientific or technical fields, is that juries will often defer to the expert providing
the opinion. If such an expert’s opinion evidence is unreliable, the dangers
associated with deference are obvious, particularly if the opinion forms a critical
link in the prosecution’s case.

1.21 Thirdly, even if we are willing to assume that lay triers of fact are sufficiently well-
informed to be able to address the reliability of technical or complex expert
opinion evidence, there is a basis for believing that, where expert evidence of
questionable reliability is admitted, it is not effectively challenged in cross-
examination.27 Confirmation on this point was provided by the UK Register of
Expert Witnesses, who told us that there was a sense among the respondents to
its own internal consultation that cross-examining advocates tend not to probe,
test or challenge the underlying basis of an expert’s opinion evidence but instead
adopt the simpler approach of trying to undermine the expert’s credibility.28 Of
course, an advocate may cross-examine as to credit in this way for sound tactical
reasons; but it may be that advocates do not feel confident or equipped to
challenge the material underpinning expert opinion evidence. Either way, juries
may be provided with insufficient evidence to be able to come to a proper
assessment of the reliability of such evidence. To put it another way, while cross-
examination can be an effective forensic tool in the right hands for challenging
many types of evidence, it would appear to be an insufficient safeguard, at least
generally speaking, for expert opinion evidence adduced under a laissez-faire
approach to admissibility.
23
Consultation Paper No 190, para 2.26.
24
In addition, in the specific context of summary proceedings where they act as prosecutors,
the RSPCA criticised what they saw as a lax approach to the screening of defence experts,
referring to bias, the selective interpretation of scientific evidence and experts acting
outside their areas of specialisation.
25
Consultation Paper No 190, paras 3.12 to 3.14. Mr Justice Treacy, Presiding Judge of the
Midland Circuit, agreed with our view that the current practice “tends towards letting the
trial process sort the matter out”.
26
Consultation Paper No 190, para 2.9.
27
Consultation Paper No 190, para 2.9.
28
The Association of Forensic Science Providers also felt that “cross-examination is not
necessarily an effective tool” for challenging scientific expert evidence.

5
1.22 A fourth reason for special rules for experts and their evidence is that all experts
owe an overriding duty to provide the court with impartial evidence within their
area of expertise.29 We acknowledge, of course, that all witnesses are under a
duty to provide truthful evidence; but only experts are under an explicit overriding
obligation set out in rules of court. Expert witnesses therefore owe a unique,
elevated duty to the court, with a concomitant duty to ensure that they do not
mislead the court, regardless of the impact this may have on the party for whom
they have been called. There is, therefore, a further principled justification for
special rules for experts and, in particular, for requiring that all experts,
regardless of their client, disclose matters which may have a bearing on the
reliability of their evidence.

1.23 In this context it is pertinent to note a comment provided to us by Bruce Houlder


QC, Director of Service Prosecutions:

My own practitioner’s experience … is that some charlatans or


certainly biased and even incompetent experts still exist in the field of
science and also in forensic accountancy. The decisions of the Court
of Appeal that have underlined the independent role of the expert,
and where their prime duty lies, have not always been heeded, and a
“market” still exists for opinions that assist the cause of the
paymaster, and insufficiently scrutinise the value of the evidence that
points away from the conclusions contended for at trial.30

1.24 In short, given the special nature of expert opinion evidence, the likelihood that
the current safeguards associated with the trial process are insufficient, and the
risk that juries may simply defer to ostensibly reputable experts and accept their
opinion evidence at face value, it is difficult to disagree with the view of the
Criminal Bar Association that it “must be in the interests of justice to ensure that
only expert evidence which has been properly scrutinised and has confirmed
validity goes before the jury”.

1.25 Lord Justice Leveson made a similar point in a recent speech for the Forensic
Science Society and King’s College, London:31

It is, in my opinion, perfectly clear that expert evidence of doubtful


reliability may be admitted too freely with insufficient explanation of
the basis for reaching specific conclusions, be challenged too weakly
by the opposing advocate and be accepted too readily by the judge or
jury at the end of the trial. In that regard, therefore, the law of England
and Wales is not satisfactory and reform is undoubtedly required.

1.26 We also endorse the following comment provided by Lord Justice Aikens in his
response to our consultation paper:

29
Criminal Procedure Rules 2010, r 33.2(1) and (2), following Harris [2005] EWCA Crim
1980, [2006] 1 Cr App R 5 at [271] and Bowman [2006] EWCA Crim 417, [2006] 2 Cr App
R 3 at [176].
30
See also Consultation Paper No 190, para 1.16 and fn 20.
31
16 November 2010, available at www.judiciary.gov.uk/media/speeches/2010/speech-lj-
leveson-expert-evidence-16112010 (last visited 3 February 2011).

6
There has to be some check to ensure that the “expert evidence” is
truly a discipline based on proper principles of research and
evaluation, whether the subject is a scientific one, or an area … such
as accountancy.

1.27 But a check of this sort can only be part of a broader solution to the problems
associated with expert opinion evidence. As we intimated above, there must be
greater scrutiny of expert evidence at the admissibility stage more generally, and
the parties and judiciary should be provided with the information they need to
challenge and assess the trustworthiness of such evidence (and the individuals
called to provide it) before it is placed before a jury in a criminal trial. It is for this
reason that we recommend in Part 7 a strengthened disclosure regime with
respect to expert witnesses and their evidence.

1.28 We should also repeat here the important point we made in our consultation
paper that a more enquiring approach to expert evidence in the criminal courts
should encourage higher standards amongst expert witnesses and the wider
expert communities.32 This should result in expert evidence of higher quality
being tendered for admission in all criminal proceedings and therefore reduce the
risk that unreliable evidence will be placed before juries.

1.29 In its response to the provisional proposals in our consultation paper, the Criminal
Cases Review Commission agreed that a new statutory test of the type we
proposed would bring a number of benefits, two of which were that:

(1) the parties seeking to adduce expert evidence would have the statutory
criteria in mind from the outset, encouraging a more considered
approach and so bringing a measure of quality control; and

(2) quality control would be encouraged amongst experts themselves as


they would need to prepare their opinions in the knowledge that they
would be scrutinised with reference to a statutory test.33

1.30 What we have said above should not, however, be taken as a suggestion that
only poor quality expert evidence is currently being admitted in criminal trials. As
Gary Pugh (Director of Forensic Services for the Metropolitan Police) argued in
his response to our consultation paper, there will often be organisational
structures in place that go some way towards ensuring that reliable expert
evidence is tendered for admission. He felt that this ought to be more clearly
recognised.34

32
Consultation Paper No 190, paras 6.14 to 6.16.
33
Some individual consultees (eg Dr Keith JB Rix) also noted that putting experts on notice
as to what would be expected of them would result in higher standards.
34
A similar point was made by some other consultees. For example: Professor Wesley
Vernon, a podiatrist with a particular interest in forensic identification, said that the
standards and processes of the institution within which the work has been undertaken is an
important factor; the UK Accreditation Service suggested that “judges should take account
of the increased confidence that can be derived from the fact that an expert works within
the context of an accredited organisation, which is regularly assessed by an independent,
impartial national accreditation body”; and Skills for Justice argued that that “the ongoing
assessment of competence in the workplace of the expert witness” is an important criterion
for determining reliability.

7
OUR PROVISIONAL PROPOSALS
1.31 In our consultation paper we set out a number of proposals for reform along with
some questions seeking views on related matters.

1.32 Our central proposal was that there should be a new reliability-based admissibility
test for expert opinion evidence which would need to be applied in relation to
most expert opinion evidence tendered for admission in criminal proceedings. We
proposed that there should be a rule along the following lines:35

(1) The opinion evidence of an expert witness is admissible only if the court
is satisfied that it is sufficiently reliable to be admitted.

(2) The opinion evidence of an expert witness is sufficiently reliable to be


admitted if:–

(a) the evidence is predicated on sound principles, techniques and


assumptions;36

(b) those principles, techniques and assumptions have been


properly applied to the facts of the case; and

(c) the evidence is supported by [that is, logically in keeping with]


those principles, techniques and assumptions as applied to the
facts of the case.

1.33 In tandem with this new rule, we also proposed that the trial judge should have a
number of guidelines to assist him or her in the determination of evidentiary
reliability, with one set of guidelines for scientific (or purportedly scientific)
evidence,37 and a separate set of guidelines for experience-based, non-scientific
expertise.38 We explained that the party proffering the expert evidence would
bear the onus of demonstrating its reliability.39 We also suggested, however, that
it would be open to the court to take “judicial notice” of some assumptions or well-
established theories about which there was no meaningful dispute.40

1.34 We also suggested that the new reliability test should be incorporated into a
broader test governing the admissibility of expert evidence generally, including
the separate common law requirements relating to assistance, expertise and
impartiality.41

1.35 In addition, we asked our consultees to consider whether the trial judge should, in
exceptional cases, have the power to call upon the services of an independent

35
Consultation Paper No 190, paras 6.10 and 6.78.
36
That is, principles, techniques and assumptions which are not only well founded, but also
appropriate for the type of evidence in question.
37
Consultation Paper No 190, paras 6.26 and 6.79.
38
Consultation Paper No 190, paras 6.35 and 6.80.
39
Consultation Paper No 190, paras 6.57 and 6.81.
40
The doctrine of judicial notice allows certain facts to be regarded as proved if the facts are
so well known or accepted that it would be pointless to adduce evidence to establish them.
41
Consultation Paper No 190, paras 1.2, 1.3 and 1.8 with para 6.82.

8
expert to help him or her apply the reliability test to particularly complex
evidence.42

A SUMMARY OF OUR RECOMMENDATIONS


1.36 We explain later in this report that there was broad consensus amongst our
consultees that there should be a special statutory admissibility test for expert
opinion evidence along the lines we proposed.

1.37 There was also broad support for our view that the party seeking to rely on the
evidence should bear the burden of demonstrating its reliability, even if that party
is the accused, and for our view that the reliability test should be incorporated into
a broader admissibility test. This broader test would encompass the current
common law requirements relating to assistance, expertise and impartiality.

1.38 In this report we therefore take forward our central proposal regarding a new
reliability test for expert opinion evidence. We also recommend that this reliability
test should be incorporated into a broader test in primary legislation
encompassing all aspects of the current common law admissibility test, albeit with
some refinements.

1.39 As we explain in Part 3, however, we now believe that it would be better if the
courts did not have to rely on the doctrine of judicial notice as the justification for
not applying the test (in relation to underlying matters which are not case-
specific).43 It will be seen, therefore, that we recommend a further requirement
which would have the effect of ensuring that the reliability test will be applied only
if it appears to the court that the evidence might be insufficiently reliable to be
admitted.

1.40 We also now believe that it would be better to have a single list of guidelines (or
factors) to help trial judges in their application of the new test, rather than the two
sets of guidelines we provisionally proposed in the consultation paper.

1.41 In addition, for the principled reasons to which we have already alluded, but also
to ensure that our proposed scheme would function effectively in practice, we
make recommendations on pre-trial disclosure and court-appointed experts and
recommend some amendments to the Criminal Procedure Rules 2010.

1.42 However, in line with what we said in our consultation paper, these
recommendations are unlikely to provide a panacea.44 It is imperative that there
be a broader context of change in tandem with the reforms we recommend, with
safeguards and appropriate regulatory schemes designed to ensure minimum
standards (particularly for forensic scientific evidence) and a more critical
approach on the part of some judges to the evidence placed before them.

42
Consultation Paper No 190, paras 6.67 and 6.83.
43
See paras 3.65 to 3.78.
44
Consultation Paper No 190, paras 1.13 to 1.20.

9
1.43 It is also important that appropriate training on how to determine evidentiary
reliability, particularly in relation to evidence of a scientific nature, should be
undertaken by all judges and lawyers involved in criminal proceedings.45

ACKNOWLEDGEMENTS
1.44 Our consultation paper was published on 7 April 2009 and our consultation period
ran until 7 July 2009. We also set up an on-line forum for the duration of our
consultation period. We would like to offer our thanks to all the individuals and
bodies, listed in Appendix D, who provided comments on our provisional
proposals, doubly so for the individuals and bodies who commented on an earlier
draft of our report or on certain aspects of our draft Bill. We also wish to express
our thanks to the various officials within Government departments who engaged
with us during the consultation process and beyond.

1.45 Following the end of our consultation period, and during the formulation of the
policy which informs our present recommendations, we sought the views of, and
exchanged ideas with, other individuals who are also listed in Appendix D. These
individuals were either members of our working party for the project or persons
whose practical experience provided us with crucial information on the workability
of our proposals. We are extremely grateful to all these individuals, for their time
and their invaluable assistance.

THE STRUCTURE OF THIS REPORT


1.46 In Part 2 we summarise the current law on the admissibility of expert evidence in
criminal proceedings.

1.47 In Part 3 we explain and analyse our consultees’ comments on the central
proposal in our consultation paper that there should be a new reliability test for
expert opinion evidence in criminal proceedings and on the guidelines we
proposed for judges when applying the test. We also explain consultees’ views
on our suggestion that the current common law admissibility requirements
relating to assistance, expertise and impartiality are satisfactory and ought to be
codified alongside our proposed reliability test.

1.48 In Part 3 we also introduce our recommendations:

(1) that there should be a new test in primary legislation which would prevent
the admission of expert opinion evidence which is not sufficiently reliable
to be admitted;

(2) that the legislation should permit the trial judge to presume evidentiary
reliability (as a matter bearing on admissibility) if there is no appearance
of unreliability;

45
Training should also be provided to prospective lawyers, newly-qualified lawyers and
experienced practitioners. Ideally, law students would in due course receive instruction on
scientific methodology and statistics as part of their undergraduate courses, and the CPD
requirements for practising solicitors and barristers who undertake work in criminal law
would be amended to require attendance at approved lectures covering the same areas (in
the context of criminal proceedings). The Criminal Bar Association has told us that relevant
training is an area which it could be proactive in addressing and which could perhaps be
incorporated into its seminars and lectures.

10
(3) that the legislation should set out the factors the court should take into
consideration when applying the reliability test; and

(4) that the legislation should be a new statutory code for the admissibility of
expert evidence in criminal proceedings generally, supplanting the
various common law admissibility limbs.

1.49 Appendix A is our draft Criminal Evidence (Experts) Bill which, if taken forward,
would become this primary legislation. The admissibility test in our Bill would
apply only to those parts of a criminal process in England and Wales to which the
strict rules of evidence apply (including criminal trials and “Newton hearings”).46
However, because this test, including the new reliability limb, is founded on
universal principles, the Government may in due course wish to consider
extending its application to hearings involving risk assessments for sentencing, to
service courts (for the armed forces) and to other proceedings, particularly family
proceedings involving serious allegations and disputed medical evidence.47

1.50 In Part 4 we set out and explain the provisions of our draft Bill which, save for
some relatively minor refinements, would restate the common law admissibility
tests relating to assistance, expertise and impartiality.

1.51 In Part 5 we set out and explain the limb in our new admissibility test for
determining whether expert opinion evidence is sufficiently reliable to be admitted
and the factors which the judge should take into consideration when addressing
this issue. We also set out and explain our recommendations, first, that there
should be a power which would permit the trial judge to presume evidentiary
reliability in most cases where expert evidence is tendered for admission and,
secondly, that the factors relevant to the determination of evidentiary reliability
should be set out alongside the admissibility test. We also make reference to the
provisions in our draft Bill which, if implemented, would give effect to these
recommendations.

1.52 In Part 6 we recommend a new statutory power which, in exceptional


circumstances, would allow a trial judge to call upon a further expert witness –
one who has been independently screened for expertise and impartiality – to
provide the judge with additional expert assistance when applying the reliability
test.

1.53 In Part 7 we set out a number of further recommendations, principally relating to


pre-trial disclosure and expert reports.

1.54 In Part 8 we examine how our statutory test would work in practice, measured
against the cases we described in Part 2 of our consultation paper and in
paragraphs 1.4 to 1.7 above. We also consider a hypothetical case involving
disputed defence expert evidence.

46
A Newton hearing is a trial to determine the facts if D pleads guilty, where there is a
dispute as to the facts relevant to sentencing.
47
Medical expert evidence is routinely admitted in family cases at the behest of local
authorities to prove on the balance of probabilities a non-accidental injury. For a useful
guide to issues relating to expert evidence in family proceedings, see Lord Justice Wall, A
Handbook for Expert Witnesses in Children Act Cases (2nd ed 2007).

11
1.55 Part 9 provides a summary of our recommendations.

1.56 Appendix A is our draft Criminal Evidence (Experts) Bill and explanatory note.

1.57 Appendix B sets out Part 33 of the Criminal Procedure Rules 2010.

1.58 Appendix C is our impact assessment.

1.59 Appendix D lists the individuals and bodies who responded to our consultation
paper, the individuals on our working group for this project and the individuals
whom we consulted on specific proposals after the consultation period.

12
PART 2
THE CURRENT LAW

THE COMMON LAW ADMISSIBILITY TEST


2.1 Four requirements relating to the admissibility of expert evidence in criminal
proceedings have developed at common law, principally with reference to expert
opinion evidence.

2.2 In this Part we first provide a summary of these requirements (“assistance”,


“relevant expertise”, “impartiality” and “evidentiary reliability”) and then set out our
view on whether they apply, and whether they should apply, to expert evidence of
fact.

Assistance
2.3 According to the leading case of Turner,1 an expert’s opinion:

is admissible to furnish the court with ... information which is likely to


be outside the experience and knowledge of a judge or jury. If on the
proven facts a judge or jury can form their own conclusions without
help, then the opinion of an expert is unnecessary.2

2.4 In other words, for expert opinion evidence to be admissible it must be able to
provide the court with information which is likely to be outside a judge or jury’s
knowledge and experience, but it must also be evidence which gives the court
the help it needs in forming its conclusions.

2.5 The Court of Appeal’s judgment suggests that an expert’s evidence is


inadmissible if it is “unnecessary”. It should be noted, however, that to be
admissible an expert’s evidence is “necessary” only in the limited sense that it
has to provide helpful information which is likely to be outside a judge or jury’s
knowledge and experience.3

Relevant expertise
2.6 The individual claiming expertise must be an expert in the relevant field. This was
described in the South Australian case of Bonython4 as a requirement that the
individual “has acquired by study or experience sufficient knowledge of the
subject to render his [or her] opinion of value”,5 a description which has found
favour in England and Wales.6

1
[1975] QB 834.
2
[1975] QB 834, 841.
3
See Mohan [1994] 2 SCR 9, 10f (Canadian Supreme Court).
4
[1984] 38 SASR 45.
5
[1984] 38 SASR 45, 47.
6
Stubbs [2006] EWCA Crim 2312, [2006] All ER (D) 133; Leo Sawrij v North Cumbria
Magistrates’ Court [2009] EWHC 2823 (Admin), [2010] 1 Cr App R 22.

13
2.7 Generally speaking, it is the expertise itself which determines whether this
admissibility requirement is satisfied, not the route by which the expert came to
have it.7 A recent judicial comment suggests, moreover, that the threshold for
demonstrating expertise is quite low.8 Against those points, however, it should be
noted: first, that the threshold cannot (we suggest) be any lower than a
requirement of proof on the balance of probabilities; secondly, that amateurs are
not qualified to give some types of expert evidence;9 and, thirdly, that explicit
guidelines for determining expertise are now being formulated for certain
scientific fields.10

Impartiality
2.8 The expert must be able to provide impartial, objective evidence on the matters
within his or her field of expertise. In the civil case of Field v Leeds City Council,11
Lord Woolf, the Master of the Rolls, said that for an expert to be “qualified to give
evidence as an expert” he or she must be able to provide an objective, unbiased
opinion on the matters to which his or her evidence relates.12 More recently, in
the case of Toth v Jarman,13 the Court of Appeal (Civil Division) recognised that
an expert witness “should provide independent assistance to the court by way of
objective unbiased opinion”14 and that where an expert witness “has a material or
significant conflict of interest, the court is likely to decline to act on his [or her]
evidence, or indeed to give permission for his [or her] evidence to be adduced”.15

2.9 This common law admissibility requirement has been reinforced for criminal
proceedings by an explicit provision in secondary legislation. Rule 33.2 of the
Criminal Procedure Rules 2010 provides that an expert has an overriding duty to
give opinion evidence which is objective and unbiased.

2.10 It has been held in civil proceedings that an expert’s evidence is inadmissible if it
might appear to a reasonable person that the expert could be biased in favour of
the party who has called him or her to testify,16 although this was thought to be
the wrong test in R (Factortame Ltd) v Secretary of State for Transport, Local

7
See, eg, Silverlock [1894] 2 QB 766.
8
See R (Doughty) v Ely Magistrates’ Court [2008] EWHC 522 (Admin) at [24]: “Whether the
claimant is a good expert or not is neither here nor there. The quality of his report is neither
here nor there. … These matters are not a sufficient basis for having ruled the claimant to be
simply not competent to give expert evidence at all.”
9
See Robb (1991) 93 Cr App R 161, 164, where Bingham LJ said that the opinion evidence of
an amateur psychologist would be inadmissible.
10
See Henderson and others [2010] EWCA Crim 1269, [2010] 2 Cr App R 24 at [207] and
[208] (medical experts) and Weller [2010] EWCA Crim 1085 at [49] (experts on DNA).
11
[2000] 1 EGLR 54.
12
Above, at [15] and [16]. Waller LJ simply referred to the need to demonstrate that the expert
is aware of what his Lordship called the expert’s “primary duty to the court”.
13
[2006] EWCA Civ 1028, [2006] 4 All ER 1276.
14
[2006] EWCA Civ 1028, [2006] 4 All ER 1276 at [100], citing Polivitte Ltd v Commercial
Union Assurance Co Plc (1987) 1 Lloyd’s Rep 379, 386.
15
[2006] EWCA Civ 1028, [2006] 4 All ER 1276 at [102].
16
Liverpool Roman Catholic Archdiocesan Trustees Inc v Goldberg (No 3) [2001] 1 WLR 2337.

14
Government and the Regions (No 8)17 and in Morgan v Hinton Organics
(Wessex) Ltd.18

2.11 Whatever the position in civil proceedings, it is now clear that apparent bias does
not render an expert’s evidence inadmissible in criminal proceedings.19

Evidentiary reliability
2.12 The expert’s opinion evidence must in other respects satisfy a threshold of
acceptable reliability.20 The existence of a further common law admissibility
requirement of some sort can be discerned from the cases where the Court of
Appeal (Criminal Division) has:

(1) held that the field of expertise must at least be “sufficiently well-
established to pass the ordinary tests of relevance and reliability;”21

(2) cited the admissibility test for expert opinion evidence in Bonython22
which has a reliability component, albeit one which has never been
properly analysed in England and Wales;23 and

(3) suggested a particular type of reliability test.24

2.13 In Bonython29 this admissibility requirement was described as being “whether the
subject matter of the [expert’s] opinion forms part of a body of knowledge or
experience which is sufficiently organized or recognized to be accepted as a
reliable body of knowledge or experience”.30 In our consultation paper we took
the view that, if this aspect of the Bonython test is part of the law of England and

17
[2002] EWCA Civ 932, [2003] QB 381 at [70].
18
[2009] EWCA Civ 107 at [67] and [68].
19
Stubbs [2006] EWCA Crim 2312, [2006] All ER (D) 133; Leo Sawrij v North Cumbria
Magistrates’ Court [2009] EWHC 2823 (Admin), [2010] 1 Cr App R 22.
20
See Consultation Paper No 190, para 3.1.
21
Dallagher [2002] EWCA Crim 1903, [2003] 1 Cr App R 12 at [29]; Luttrell [2004] EWCA Crim
1344, [2004] 2 Cr App R 31 at [37]; see also Reed [2009] EWCA Crim 2698, [2010] 1 Cr App
R 23 at [111] and Broughton [2010] EWCA Crim 549 at [32].
22
[1984] 38 SASR 45.
23
The reliability limb of the Bonython test is set out in para 2.13 below. In his recent speech for
the Forensic Science Society and King’s College, London (16 November 2010), Lord Justice
Leveson suggested that this limb could not yet be said to represent the current state of the
law in England and Wales. It should be noted, however, that in Reed [2009] EWCA
Crim 2698, [2010] 1 Cr App R 23 at [111] the Court of Appeal indicated that it is part of the
law; and in Broughton [2010] EWCA Crim 549 at [32] the Court of Appeal expressly stated
that it is part of the law which a criminal court “must consider”.
24
Gilfoyle (No 2) [2001] 2 Cr App R 5 at [25].
29
[1984] 38 SASR 45.
30
[1984] 38 SASR 45, 47.

15
Wales, the question is whether the body of knowledge or experience is accepted
as reliable by the courts rather than by a relevant community of experts.

2.14 Following the publication of our consultation paper, the existence of a common
law reliability test was confirmed by the Court of Appeal in Reed,31 at least for
“expert evidence of a scientific nature”; but it is to be noted that the court did not
demur from the established position that there is no enhanced reliability test for
such evidence.32

2.15 The existence of a common law reliability test for evidence of a scientific nature
was also recently recognised in Weller,33 where the Court of Appeal referred to
the trial judge’s function “in determining whether there is a sufficiently reliable
scientific basis for [scientific] expert evidence to be given”.34

2.16 In our consultation paper35 we took the view that this reliability requirement in the
common law admissibility test was insufficiently robust, reflecting a generally
laissez-faire approach to the admissibility of expert evidence in England and
Wales.36 As mentioned above, the Court of Appeal has held that this requirement
is satisfied if the field of expertise is “sufficiently well-established to pass the
ordinary tests of relevance and reliability”.37

The relationship between the four admissibility tests


2.17 The first limb of the common law admissibility test (“the Turner test”) ensures that
expert evidence is admitted only when it has sufficient probative value, in the
sense that the evidence is likely to help the court resolve a disputed issue. The
purpose of the other limbs is to ensure that such expert evidence is admitted in
criminal proceedings only when it satisfies a minimum threshold of general
reliability, what might be called “reliability in the round”.

31
[2009] EWCA Crim 2698, [2010] 1 Cr App R 23.
32
The Court of Appeal held at [111] that while “expert evidence of a scientific nature is not
admissible where the scientific basis on which it is advanced is insufficiently reliable for it to
be put before the jury” there is “no enhanced test of admissibility for such evidence”. The
court did not explain how the trial judge is to determine whether the scientific basis of an
expert’s evidence is or is not sufficiently reliable to be admissible, save for mentioning the
“enhanced test for admissibility used in the United States as set out in the decision of the US
Supreme Court in Daubert v Merrell Dow Pharmaceuticals 509 US 579”. See also Broughton
[2010] EWCA Crim 549 at [32].
33
[2010] EWCA Crim 1085.
34
[2010] EWCA Crim 1085 at [48]. See also Henderson and others [2010] EWCA Crim 1269,
[2010] 2 Cr App R 24 at [206] where the court noted that Reed “is concerned with DNA
evidence but the observations of the court in relation to the admissibility of expert evidence
apply with equal force to cases concerning baby shaking”. The same point was made in T
[2010] EWCA Crim 2439 at [70]: “the principles for the admissibility of expert evidence were
summarised recently in Reed & Reed at paragraphs 111 to 112: the court will consider
whether there is a sufficiently reliable scientific basis for the evidence to be admitted … .”
35
Consultation Paper No 190, para 3.14.
36
For recent confirmation, see Henderson and others [2010] EWCA Crim 1269, [2010] 2 Cr
App R 24 at [206]: “We shall say no more about admissibility since the unsatisfactory state
of the law has been the subject of the Law Commission Consultation Paper No 190 … and is
likely to lead to changes in the current approach of laissez-faire … .”
37
Paragraph 2.12(1).

16
2.18 The fourth limb (evidentiary reliability) was the principal issue in our consultation
paper and, equally, it is the principal issue in this report. It is concerned with the
reliability of opinion evidence provided by an impartial, properly-qualified expert. It
is therefore the reliability requirement which addresses, or at least ought to
address, matters underpinning the expert’s opinion, such as the soundness of his
or her field of expertise and methodology and the validity of any assumptions
relied on. Where we refer in this report to the common law reliability test, or to a
new reliability test to replace it, we are referring to this specific aspect of
“reliability in the round”.

OPINION EVIDENCE AND EVIDENCE OF FACT


2.19 The case law on expert evidence focuses almost exclusively on expert opinion
evidence, the reason being that expert witnesses are usually called to provide
such evidence.38 It is important to understand, however, that expert witnesses
may be called to give expert evidence of fact. For example, an expert may be
called to give an explanation of how an unusual piece of machinery operates, or
evidence of a reading provided by an instrument or a symptom which was
observed when a patient was examined.

2.20 If such factual evidence is proffered for admission, logic demands that the first
three limbs of the common law test be applied in the same way that these limbs
apply to expert opinion evidence. The witness giving expert evidence of fact
should be able to do so only if the court is likely to need such assistance, the
witness is an expert in the relevant field and the witness will provide impartial,
objective evidence (that is, a sufficiently complete account as opposed to a partial
picture distorted by bias).

2.21 So, although there is authority – the case of Meads39 – to suggest that expert
evidence of fact is not covered by the common law rules summarised above, we
prefer the view in Phipson on Evidence40 that it would have been preferable to
treat the evidence in that case “as expert evidence [governed by the common law
test], where the level of expertise required was of a very low order”.41

2.22 It will be seen, therefore, that our recommendation in Part 4 that the first three
limbs of the common law test should be codified does not distinguish between
expert evidence of fact and expert evidence of opinion.

2.23 We appreciate, however, that it would be very difficult to see how the fourth limb
of the common law admissibility test, or any new test for determining evidentiary
reliability, could be meaningfully applied to expert evidence of fact. Accordingly, it
will be seen in Part 3 that our proposed statutory alternative to the fourth common
law requirement would apply only to expert opinion evidence.

38
See para 1.16 above.
39
[1996] Criminal Law Review 519 (a case involving a reconstruction).
40
(17th ed 2010) para 33-19.
41
We note that in Meads [1996] Criminal Law Review 519 no authority was cited beyond a
comment that the South Australian case of Bonython [1984] 38 SASR 45 refers solely to
expert opinion evidence.

17
PART 3
CONSULTATION

INTRODUCTION
3.1 In this Part we summarise the responses we received from our consultees on:

(1) the principal proposal in our consultation paper that there should be a
new reliability test in primary legislation for expert opinion evidence;

(2) the proposal in our consultation paper that the burden of demonstrating
reliability should be borne by the party seeking to adduce the evidence;
and

(3) the suggestion we made in our consultation paper that this legislation
should also codify the other limbs of the common law admissibility test for
expert evidence.

3.2 In the light of our consultees’ comments, we then introduce our recommendations
for these issues, which we explain more fully in Part 4 and Part 5.

EVIDENTIARY RELIABILITY

Our provisional proposal


3.3 The view we expressed in our consultation paper was that the courts have
adopted a policy of laissez-faire to the admissibility of expert opinion evidence in
criminal proceedings. So, although there is at present a rudimentary common law
reliability test for such evidence, its practical effect is largely illusory.1 Moreover,
trial judges have been given little assistance on how to assess reliability in
practice.

3.4 Criminal courts in England and Wales therefore only rarely rule expert opinion
evidence inadmissible on the ground of evidentiary unreliability.2 The courts tend
to allow expert evidence to be admitted on the assumption that its reliability will
be effectively challenged during the trial by cross-examination or by the adduction
of contrary expert evidence by another party, or both. However, as we stated in
our consultation paper,3 and explained again in Part 1 of this report, cross-
examination would seem to be an insufficient safeguard against unreliability for
expert opinion evidence adduced under a laissez-faire approach to admissibility.

3.5 The central proposal in our consultation paper, therefore, was that there should
be a new, more stringent reliability test for expert opinion evidence in criminal

1
For the common law position, see paras 2.12 to 2.16 above.
2
That is to say, the courts freely admit expert opinion evidence if the other limbs of the
common law admissibility test described in Part 2 are satisfied.
3
Consultation Paper No 190, paras 2.8 to 2.28.

18
proceedings.4 This new test would replace the fourth limb of the common law
admissibility test we described in Part 2.5

3.6 We proposed that there should be a new admissibility requirement along the
following lines:6

(1) The opinion evidence of an expert witness is admissible only if the court
is satisfied that it is sufficiently reliable to be admitted.

(2) The opinion evidence of an expert witness is sufficiently reliable to be


admitted if:–

(a) the evidence is predicated on sound principles, techniques and


assumptions;7

(b) those principles, techniques and assumptions have been


properly applied to the facts of the case; and

(c) the evidence is supported by [that is, logically in keeping with]


those principles, techniques and assumptions as applied to the
facts of the case.

3.7 Limb (a) of this test would require the courts to assess the soundness of the
principles and methodology underpinning the expert witnesses’ opinion evidence.
If sound, limb (b) would require the judge to consider whether the general
conclusions drawn from the methodology had been properly applied to the facts
of the case by the expert.8

3.8 Limb (c) would require the judge to focus specifically on the expert witness’s
reasoning, to ensure that the expert’s final conclusions (including the strength of
any opinion he or she might wish to give) were logically in keeping with the
proper application of the general conclusions drawn from the underlying
methodology to the facts of the case.9

3.9 We also proposed that this admissibility test should provide, in line with the
general position for the admission of evidence in criminal proceedings, and
4
We took the view that it would be possible to take “judicial notice” of the reliability of some
aspects of scientific expert evidence, a point on which a number of our consultees agreed.
We accept, however, that because there may be two or more views on many aspects of
scientific understanding, particularly for novel or nascent fields, it will often be difficult to
use the doctrine in the context of scientific opinion evidence. On this point generally, see:
C Onstott, “Judicial Notice and the Law’s ‘Scientific’ Search for Truth” (2007) 40 Akron Law
Review 465.
5
Paragraphs 2.12 to 2.16 above.
6
Consultation Paper No 190, paras 6.10 and 6.78.
7
That is, principles, techniques and assumptions which are not only well founded, but also
appropriate for the type of evidence in question.
8
Consultation Paper No 190, paras 6.38 to 6.40.
9
Consultation Paper No 190, paras 6.41 to 6.43. For a recent example of this point being
addressed at common law, see Weller [2010] EWCA Crim 1085 (consideration of the
underlying science relating to the transfer of DNA and the question whether the expert
called by the prosecution should have been permitted to provide an evaluative opinion on
the likely provenance of DNA found on the accused’s fingernails).

19
indeed civil proceedings, that the party tendering the opinion evidence for
admission should bear the onus of demonstrating to the court that it is sufficiently
reliable to be admitted.10

3.10 In tandem with our central proposal that there should be a new statutory
requirement of evidentiary reliability, we also proposed that the trial judge should
have a number of guidelines to help him or her determine whether or not the test
was satisfied, with one set of guidelines for scientific (or purportedly scientific)
evidence,11 and another set of guidelines for experience-based, non-scientific
expertise.12 We suggested that guidelines of this sort could be incorporated into
legislation.

3.11 There was very broad (but not universal) support for a new reliability test for
expert opinion evidence along the lines proposed in our consultation paper, as
we explain in the following paragraphs.

Comments on our provisional proposal


3.12 We received considerable support for our proposed reliability test from individuals
and bodies with a particular interest in the workings of the criminal justice system.
Nevertheless, some of this support was couched with reservations regarding the
possibility of increased complexity, costs and delays; and some of our consultees
suggested slight revisions to our proposal.

3.13 Conversely, some of our consultees suggested that the case for reform was more
compelling than we had made out in our consultation paper, or that more radical
reform was needed than the measures we had proposed.13 An alternative
approach has also recently been suggested in an academic legal journal.14

10
We did not propose any particular standard of proof because the question would not be
whether a particular past or present fact is established, but whether there are sufficient
indicia of reliability to justify the admission of the evidence. See Consultation Paper No
190, paras 6.57 to 6.61.
11
Consultation Paper No 190, paras 6.26 and 6.79.
12
Consultation Paper No 190, paras 6.35 and 6.80.
13
One suggestion was that there should be specially trained judges sitting without a jury for
trials involving complex expert evidence. A similar proposal was suggested by the Science
and Technology Committee of the House of Commons in Forensic Science on Trial –
Report of the House of Commons Science and Technology Committee (2004-05) HC 96-I.
The then Government’s response to that proposal was that the Criminal Justice Bill (2002–
2003) originally included provision for the prosecution to apply, on the grounds of length or
complexity, for a trial on indictment to take place without a jury, but it was amended by
Parliament to apply only to fraud cases. The Government added that it had “no plans to seek
to extend the ambit of this provision” (Forensic Science on Trial: Government Response to the
Committee’s Seventh Report of Session 2004-05 (2005-06) HC 427, p 16).
14
J Hartshorne and J Miola, “Expert evidence: difficulties and solutions in prosecutions for
infant harm” (2010) 30 Legal Studies 279, 293 to 294, suggesting that there should be a
panel of three Lord Justices of Appeal who would screen “new theories and techniques” for
reliability before they could be relied on in a criminal trial. Leaving aside the problematic
question of costs and delays, we have difficulty with this proposal because, first, it is
predicated on the assumption that an underlying hypothesis or technique can always be
divorced from the expert opinion evidence founded on it and, secondly, it does not protect
against the fact that unreliable expert opinion evidence may be founded on a hypothesis or
technique which is not “new” (see Consultation Paper No 190, paras 2.26 and 4.35).

20
3.14 Amongst the various bodies expressing their support for our central proposal
were: the Rose Committee of the Senior Judiciary, the Council of HM Circuit
Judges,15 the Bar Law Reform Committee, the Criminal Bar Association, the Law
Society, the Crown Prosecution Service, the Justices’ Clerks’ Society and the
Criminal Cases Review Commission.

3.15 In their written response, the Better Trials Unit of (what was then) the Office of
Criminal Justice Reform said they were broadly supportive of our proposal, save
that they were concerned about the possibility of disruptions to the trial process
and additional costs. In a similar vein, the judges of the Central Criminal Court
(whom we met) and the Rose Committee of the Senior Judiciary (who provided a
written response) emphasised the importance of flexibility so that the reliability
test would not have to be applied unnecessarily. The RSPCA made a similar
point.16

3.16 We also received positive responses from the General Medical Council, the
British Medical Association, the Police Superintendents’ Association, the Legal
Services Commission, the Association of Forensic Science Providers, Forensic
Access Ltd, the Forensic Institute, the Expert Witness Institute, the Forensic
Science Society, the Society of Expert Witnesses, the Academy of Experts,17 the
Royal Statistical Society, the UK Register of Expert Witnesses, the Royal College
of Psychiatrists,18 the Royal College of Paediatrics and Child Health, the UK
Forensic Speech Science Community and the British Association for Shooting
and Conservation.

3.17 Individual consultees who expressed their support for our proposal included:
members of the judiciary (Lord Justice Aikens,19 Mr Justice Treacy,20 Judge
Andrew Gilbart QC21); senior practitioners (Bruce Houlder QC,22 Andrew
Campbell-Tiech QC), the Forensic Science Regulator,23 various scientists and a
number of academic lawyers (albeit some expressing reservations regarding
detail).24

15
Albeit preferring the test to be in procedural rules.
16
By contrast, some of our academic consultees counselled against giving the judiciary a
discretion or power to disapply the test; see para 3.70 below.
17
Albeit with some reservations.
18
Support “in principle”.
19
His support was qualified, however, with a suggestion that we should propose something
more radical.
20
Presiding Judge of the Midland Circuit.
21
The Hon Recorder of Manchester.
22
Director of Service Prosecutions.
23
The Forensic Science Regulator did, however, stress the need for precedents to ensure
that the same questions would not be addressed again and again in subsequent trials.
24
For example: Professor Paul Roberts (University of Nottingham) argued in favour of
simplicity and felt that the reliability test should apply to evidence of fact as well as to
opinion evidence; and Professor Pierre Margot (University of Lausanne) emphasised the
need for the forensic scientists who interpret evidence to be present at the place where the
evidence is found, to direct the investigations and have a full appreciation of the relevant
context.

21
3.18 There were, however, some organisations and individuals who provided only
lukewarm or equivocal support for our proposal or who opposed any such reform
measure.

3.19 The UK Accreditation Service supported our proposal, but felt that trial judges
should also take into account the “increased confidence that can be derived from
the fact that an expert works within the context of an accredited organisation”.
The same point was made by Gary Pugh (Director of Forensic Services at the
Metropolitan Police) who felt that the focus should be on the importance of
“organisational structures” when assessing whether or not expert opinion
evidence was reliable. In a similar vein, LGC Forensics agreed that the reliability
of expert evidence should be properly tested before it reached court, but were not
convinced that our proposal offered the most efficient or cost-effective way of
addressing the problem. They suggested that there should be a greater
appreciation of “current and developing arrangements for validating and
accrediting forensic science and scientists, and particularly the work of the
Forensic [Science] Regulator”. And although the Centre for Criminal and Civil
Evidence and Procedure (at Northumbria University) accepted that there was a
“clear need” for a reliability test of the sort we proposed, they suggested in
conclusion that a more cost-effective approach to the problems associated with
expert evidence would be to insist on better training and greater accreditation.

3.20 Practical objections were raised by the Forensic Science Service. The FSS did
not support our proposal because of the possibility of repeated challenges
against evidence and, in their view, the difficulty of providing the judiciary “with
sufficient scientific knowledge” for judges to be able properly to assess the
reliability of forensic scientific evidence.25 Similarly, the British Psychological
Society, while accepting the value of our proposals in principle, doubted whether
it would be practicable for judges to acquire sufficient knowledge to make
informed rulings on the reliability of expert opinion evidence.26 The Academy of
Experts argued that, instead of creating new legislation, the existing common law
admissibility rules should be more effectively enforced.

3.21 Adam Wilson (Sheffield Hallam University) told us that we had provided
insufficient evidence of the workability of our proposals.27 We should therefore
pause here to explain: first, that we devote much of Part 5, and all of Part 8, to a
discussion of the practical application of our recommendations; and, secondly,
that following an internal consultation on our provisional proposals by the UK
Register of Expert Witnesses, their respondents’ view was that the scheme we
proposed would work.

3.22 Finally, the Medical Defence Union accepted the need for high-quality expert
evidence, but argued that there were no “hard data” to justify the reforms we

25
They also felt that we had paid insufficient regard to the “Forensic Science Regulator’s
standards and validation framework in providing assurance that processes and techniques
are fit for purpose”.
26
On the question whether judges should be able to call upon further expert assistance when
assessing evidentiary reliability, see Part 6.
27
He instead proposed that there should be a “working party, with cross-discipline
membership” which could analyse forensic scientific disciplines to determine both
admissibility and codes of practice.

22
proposed; and Dr Déirdre Dwyer (Oxford University) felt that there should be no
special admissibility rules for expert evidence over and above the rules for
evidence generally.28

3.23 Nevertheless, notwithstanding the arguments raised against what we said in our
consultation paper, which we have summarised above, it is fair to say that our
central proposal received broad support from our consultees including,
significantly, the bodies listed in paragraphs 3.14 to 3.16 above. Some
consultees said that our decision to get to grips with this area of the law was
“most welcome,”29 “particularly valuable”30 and “long overdue”.31

3.24 Lord Justice Aikens expressed the view that “the way expert evidence is dealt
with in jury trials is one of the system’s weaker links”; the Criminal Bar
Association opined that the “current treatment of expert evidence in criminal
proceedings has contributed to a significant number of miscarriages of justice,
risks continuing to do so, and requires urgent reform”; and the Bar Law Reform
Committee felt that the common law approach to expert evidence was “deeply
unsatisfactory”.

3.25 The General Medical Council felt that “together with statutory guidelines, the
[proposed] test would tighten the criteria for and clarify the powers of the court,
thereby reducing the risk of unreliable evidence being placed before a jury”. Mr
Justice Treacy opined, in line with comments we made in our consultation
paper,32 that our proposed test would “introduce a welcome additional spur to the
judiciary, practitioners and experts in their scrutiny of expert evidence”.

3.26 Importantly, several of the consultees who opposed our proposed reform
measure nevertheless accepted that it was right in principle. Their opposition was
based on perceived practical difficulties.

3.27 There were very few comments or questions on the specific limbs of our
proposed reliability test (paragraph 3.6(2) above). One consultee asked whether
limb (a) would compel the judge to choose between competing views. On this
point the answer is that it would not. In particular, expert opinion evidence based
on a reliable foundation of sound scientific methodology would not be rendered
inadmissible just because the relevant data could equally be explained by a
plausible alternative hypothesis.

3.28 If two plausible hypotheses are properly applied to the facts of the case, and
each expert witness’s opinion is logically in keeping with his or her preferred
hypothesis as applied to the facts, it would be permissible for these experts to
give their alternative opinions. The determination of the accused’s guilt or
innocence would depend on all the admissible evidence and on the burden and
standard of proof for criminal proceedings: the prosecution must prove the
accused’s guilt beyond any reasonable doubt.

28
We do not agree with Dr Dwyer’s objection, for the reasons given in Part 1.
29
Andrew Campbell-Tiech QC.
30
The Criminal Cases Review Commission.
31
The Bar Law Reform Committee.
32
Consultation Paper No 190, paras 6.12 to 6.16.

23
3.29 Nevertheless, the fact that there are plausible alternative hypotheses would limit
the nature of any opinion evidence given for the prosecution. Most obviously,
assuming the case to be one where an opinion of the following sort might be
properly advanced, a prosecution expert witness relying on one of two alternative
hypotheses would not be able to opine that a particular item of evidence of itself
provides certain proof of an aspect of the prosecution case. If the same item of
evidence could be explained in a way which is consistent with the accused’s
innocence by the application of the alternative (defence) hypothesis, such a
strong opinion would be inadmissible.

3.30 We considered limb (c) to be an important aspect of our proposed test even
though it would encroach on matters which have traditionally been left to the jury.
Our view was – and remains – that a particular expert opinion (including the
strength of that opinion) should be considered by the jury in criminal proceedings
only if it is sufficiently reliable to be placed before a jury, bearing in mind that the
safeguards provided by the trial process may be insufficient for such evidence.33
Furthermore, the parties and their experts would have this limb at the forefront of
their minds when preparing for the trial. They would expect to have to justify their
opinion evidence and would therefore ensure that the opinion evidence they
proffer for admission is likely to stand up to judicial scrutiny.

3.31 One of our academic consultees34 suggested, rightly in our view, that limb (b)
would need to be broadened to accommodate what North American
commentators usually call “social framework evidence” but is perhaps best
described as “background expert evidence”. Expert opinion evidence is
occasionally given in criminal proceedings on background matters.35 We accept
that any new admissibility test would need to be broad enough to encompass
such evidence insofar as it is otherwise admissible.

3.32 The feedback our consultees provided on the provisional reliability test described
in our consultation paper may be summarised as follows:

(1) we received considerable support for the test from individual consultees
and the principal bodies concerned with criminal justice in England and
Wales;

(2) our consultees generally felt that a statutory test of the sort we proposed
would encourage a more considered approach to expert evidence and so
bring a measure of quality control to criminal proceedings (although,
understandably, concerns were also raised about possible cost and time
implications);

33
See paras 1.20 and 1.21 above.
34
Professor Mike Redmayne (London School of Economics), the author of Expert Evidence
and Criminal Justice (2001).
35
See, for example, S(VJ) [2006] EWCA Crim 2389. In that case an expert on autism was
permitted to give evidence that a person of the complainant’s age, with the complainant’s
autistic condition, would find it difficult to create and maintain a false allegation.

24
(3) despite the broad support for our proposed test, a number of consultees
wondered whether our proposed approach would have excluded the
unreliable expert opinion evidence we referred to in our consultation
paper;36 and some consultees were understandably concerned about the
possibility of an investigation being conducted whenever expert opinion
evidence is proffered for admission;37

(4) some bodies connected with forensic science favoured a more ‘grass
roots’ approach, with better quality control in forensic laboratories (rather
than new admissibility rules in the criminal courts).

3.33 On the last point, this was the approach recommended by the United States
National Research Council of the National Academies in their 2009 report,
Strengthening Forensic Science in the United States: A Path Forward. The
fundamental difference between the position in the USA and that in England and
Wales, however, is that rule 702 of the US Federal Rules of Evidence already
sets out a reliability test for expert evidence. Indeed we drew on rule 702 when
formulating the equivalent test we proposed in our consultation paper.

3.34 Our view is that there should be a new admissibility test for expert opinion
evidence alongside better quality control in forensic scientific laboratories. We
would encourage and endorse all reasonable measures designed to enhance the
reliability of expert opinion evidence in criminal proceedings; and we believe that
a new statutory admissibility test with an explicit reliability limb would act as a
further incentive to expert communities. Expert opinion evidence would then be
properly assessed for reliability at an early stage in the proceedings and be
tendered for admission only when it is likely to withstand judicial scrutiny.

Our recommendation on the reliability test


3.35 Given the broadly positive responses to what we said in our consultation paper,
particularly the support provided by the bodies listed in paragraphs 3.14 to 3.16
above, we believe that our provisional proposal should be taken forward into
legislation, albeit with some refinements.

3.36 We therefore recommend that there should be a statutory admissibility test


which would provide that an expert’s opinion evidence is admissible in
criminal proceedings only if it is sufficiently reliable to be admitted.38

3.37 In line with our provisional proposal, we focus exclusively on expert opinion
evidence for this test. We have adopted this approach for three reasons:

(1) if the expert evidence would ordinarily be classified as “factual” (for


example, an explanation of how an unfamiliar piece of machinery
operates) it is unlikely there will be any issue as to evidentiary reliability;

36
Consultation Paper No 190, paras 2.14 to 2.24. We endeavour to address this question in
Part 8.
37
Some consultees did not support our proposal for this reason.
38
That is, sufficiently reliable to be taken into consideration by the fact-finding tribunal.

25
(2) there is unlikely to be any dispute as to whether expert evidence is
evidence of fact or evidence of opinion in the sort of case where the
reliability of an expert’s evidence is being challenged;39 and

(3) it is far easier to formulate a generally-applicable reliability test and


complementary guidelines if we limit the application of the test to opinion
evidence.

3.38 However, on the basis that there may be borderline areas, where an expert
purports to be presenting evidence of unfamiliar fact but the evidence ought
properly to be regarded as opinion evidence, and the reliability of that evidence is
being challenged, we believe it would be prudent to include a provision requiring
that the evidence be treated as opinion evidence if there is any doubt on the
matter.40

3.39 We therefore recommend a rule which would provide, for the reliability test,
that if there is any doubt on the matter expert evidence presented as
evidence of fact should be treated as expert opinion evidence.

Guidance for the judiciary


3.40 In our consultation paper we provisionally proposed that, in addition to a new
statutory test for determining the reliability, and therefore the admissibility, of
expert opinion evidence in criminal proceedings, there should be guidelines to
help trial judges apply the test in practice.41 We proposed that there should be
two sets of guidelines, one set for expert evidence of a scientific nature – that is,
scientific or purportedly scientific evidence – and a separate set for experience-
based expert evidence.

3.41 Our first set of guidelines listed the indicia of reliability traditionally associated
with sound scientific methodology, that is, the type of methodology demonstrating
the classic hallmarks of valid science, including properly conducted experiments
and observations, the revision of hypotheses in the light of new data, publication
and peer-review.42 We proposed that our second set of guidelines should be
used for other types of expert opinion evidence: the non-scientific, experience-
based evidence for which there were likely to be fewer objective indicia of
reliability.43 We added, however, that for the areas of professional but non-
scientific expertise, where there are well-accepted practices and methodologies
(for example, the field of accountancy) the court would be able to admit the

39
The very fact that experts disagree would of course suggest that the evidence should be
treated as opinion evidence.
40
Dr Glyn Walters (a retired consultant chemical pathologist) told us that “conjecture is an
important feature of diagnostic medicine … but all too often pathologists present what are
speculative ideas as though they are established fact”. Possible fact / opinion borderline
areas mentioned by consultees are blood groupings (Bar Law Reform Committee) and lie-
detector evidence (Professor Paul Roberts).
41
Consultation Paper No 190, paras 6.21 to 6.37.
42
Consultation Paper No 190, para 6.26.
43
Consultation Paper No 190, para 6.35.

26
evidence if the expert had followed accepted practices and provided a sufficient
explanation of what he or she had done.44

3.42 The guidelines we published in our consultation paper also included factors
unrelated to methodology which might nevertheless have a bearing on the
reliability of the expert’s opinion evidence in the round, that is, factors relating to
the credibility of the witness as a provider of expert opinion evidence.45 The two
sets of guidelines we proposed shared a number of features in common.

3.43 There was very broad support for guidelines of the sort we proposed – in tandem
with appropriate training46 – to help the judiciary apply our proposed reliability
test.

3.44 However, although some consultees expressly favoured separate guidelines for
scientific and non-scientific expert evidence, and many did not criticise this
aspect of our proposals, a significant number of consultees either queried or
opposed our dichotomy. The opposition to separate guidelines was largely based
on the need to avoid what was thought to be unnecessary complexity and the
desirability of reducing the potential for arguments in court.

3.45 Some consultees envisaged disputes about the nature of some expert opinion
evidence on the ground that there is no clear line separating scientific and non-
scientific evidence. One consultee, Mr Justice Treacy, told us that he could
foresee our guidelines “leading to argument as to which category some witnesses
fall into”; Professor Paul Roberts (University of Nottingham) opined that the
distinction we had drawn between scientific and experience-based evidence was
the sort of taxonomic issue the courts should avoid; and Associate Professor
Andrew Roberts (University of Warwick) suggested that:

definitional problems that are likely to beset attempts to categorise


expert evidence could be avoided by providing a consolidated list of
criteria and leaving trial judges to apply those that are relevant to
evaluation of the evidence in the case.47

3.46 A number of consultees made the important point that forensic scientific evidence
– the types of scientific evidence proffered for admission in criminal trials –
usually involves a scientific underpinning and an experience-based interpretive

44
Consultation Paper No 190, para 6.37. The Rose Committee of the Senior Judiciary
endorsed this comment.
45
The guidelines therefore included factors relevant to impartiality and expertise.
46
We emphasised the importance of proper training for lawyers and the judiciary; see
Consultation Paper No 190, paras 1.15(3) and 6.72 to 6.74.
47
A Roberts, “Rejecting general acceptance, confounding the gate-keeper: the Law
Commission and expert evidence” [2009] Criminal Law Review 551, 561.

27
element.48 The UK Forensic Speech Science Community provided a useful
explanation of this overlap from their perspective. They said the following:

We consider there to be a continuum between experience-based


evidence and narrowly scientific evidence. … For example, in our
own field certain methods for analysing speech samples derive from
the physics of sound and are clearly very much at the narrowly
scientific end of the continuum. However, the conclusion one arrives
at does not arise algorithmically or automatically from applying these
methods. Rather, it relies on experience and bringing to bear
knowledge of the likely effects of factors such as the speaking
situation … , the range of variation encountered in a particular dialect,
the speaking style used, the state of the speaker and the recording
characteristics (eg direct conversation or telephone).

In view of this, it must be recognised that evidence arising from the


analysis of speech samples will, inevitably, involve both narrow
scientific and experience-based elements. [Similarly], a forensic
pathologist in carrying out an autopsy will draw on scientific principles
and tests derived from such fields as histology, physiology and
biochemistry in attempting to determine cause of death. But ultimately
the outcome of the autopsy will involve interpretation of the results of
these specific tests, and this judgment will be crucially dependent on
the experience of the pathologist.

3.47 A similar point was made by several other consultees in relation to disciplines
such as fingerprint and ear-print analysis.49 It is difficult to disagree with these
objections to our original dichotomy.50

3.48 Bearing in mind the generally positive support for our proposal that there should
be guidelines to assist the judiciary, and given the well-founded criticisms of our
original dichotomy, we now believe that the evidentiary reliability limb of our new
admissibility test should be read with supplementary guidelines or factors, in line
with our original proposal, but that there should be just one set of generic factors.

3.49 The trial judge would select from these factors as appropriate, depending on the
type of expert opinion evidence being proffered for admission. We therefore

48
According to Professor Paul Roberts, forensic sciences are “typically disciplinary hybrids
applied to practical problem-solving” bearing the hallmarks or classical sciences but also
incorporating other matters. For a recent example, relating to evidence of fibre analysis,
see Hall [2011] EWCA Crim 4 at [48]: “the judgement whether two or more textile fibres of
similar dimensions and similarly dyed are distinguishable is, first and foremost, a matter for
the experienced and expert examiner. There is no measurement which, by itself, is
capable of making that judgement.”
49
For example, the UK Register of Expert Witnesses, the Bar Law Reform Committee, the
Forensic Science Service and the Forensic Institute.
50
Interestingly, however, the UK Register of Expert Witnesses told us, following their internal
consultation on our proposals, that our two sets of guidelines had drawn broad support and
that the division we had proposed as between scientific and non-scientific evidence would
not give rise to problems in practice.

28
endorse a point made (tacitly) by Andrew Roberts51 that if we are to expect
trained judges to be able to apply our reliability test to the wide range of
disciplines and fields on which expert evidence may be given, we should also
trust the judges to be able to select for themselves, from a single body of factors,
what is relevant to this exercise.

3.50 This approach also brings with it the advantage of enhanced flexibility. A
recurring theme in the responses we received, particularly from the judiciary, was
that any guidelines we recommend should not be too prescriptive. The Rose
Committee of the Senior Judiciary pointed out that “in certain circumstances,
some of the identified factors may not be applicable and flexibility may be
required”. Similarly, Lord Justice Aikens commented that the trial judge’s
“flexibility to deal with unexpected areas and situations must be preserved”.

3.51 To be fair, the guidelines we proposed in our consultation paper contained a


great deal of flexibility: they permitted the trial judge to take into account other
factors not specifically listed;52 and the guidelines for experience-based evidence
directed that the judge should not take into consideration any inapplicable
factors.53 Nevertheless we agree with the comments set out in the preceding
paragraph.

3.52 If we are to have a single set of generic factors, it follows that some of them may
not be relevant in a particular case. Accordingly, the judge should be directed to
take into account only what is relevant; and, equally, in line with the position we
proposed in our consultation paper, the judge should be able to take into account
relevant matters which are not expressly set out.54 This approach accords with
the position for other aspects of the law of criminal evidence.55 It also reflects the
view of Mr Justice Treacy, who suggested that there should be a single body of
guidelines from which the judge would select relevant factors but which would
also permit the judge to consider other relevant factors.

3.53 Before leaving this area we also need to address one further suggestion made by
some of our consultees, no doubt prompted by what we said in paragraph 6.74 of
our consultation paper.56 Their suggestion was that there should be an
authoritative compendium of different guidelines for the various scientific
disciplines and that these should be published in a format which could be easily
updated.

51
A Roberts, “Rejecting general acceptance, confounding the gate-keeper: the Law
Commission and expert evidence” [2009] Criminal Law Review 551, 561.
52
Consultation Paper No 190, paras 6.26(1) and 6.35(1).
53
Consultation Paper No 190, para 6.35(1).
54
For example, accreditation or organisational structures, if such factors are considered to be
relevant to the evidentiary reliability of the expert’s opinion evidence in the particular case.
55
It is not uncommon for the judge to be directed by primary legislation to take into account
what he or she considers to be relevant in the factual context of the case; see, for
example, s 30(3)(d) of the Criminal Justice Act 1988 and ss 100(3), 114(2) and 116(4)(d)
of the Criminal Justice Act 2003.
56
We opined that the Judicial Studies Board might wish to work with relevant professional
bodies with a view to producing for Crown Court judges a practical guide for assessing
expert evidence in criminal proceedings, perhaps using parts of the US Federal Judicial
Center’s Reference Manual on Scientific Evidence as a framework.

29
3.54 The most enthusiastic proponent of this idea was Professor Paul Roberts, who
felt that a better way of reforming the law on expert evidence would be to have:

generic criteria of validity specified in primary legislation, reinforced


by detailed, non-statutory guidance in judicial bench books … and
associated training programmes, which can easily be modified in the
light of experience and updated to keep pace with ongoing scientific
and technical innovations.57

3.55 Similarly, the British Psychological Society told us that they favoured specific
guidelines for psychologists who give expert opinion evidence. They explained
that if psychologists give evidence about a particular individual, they have to
exercise their clinical skill, first, in the choice of which technique should be used
to assess a particular factor and, secondly, in the interpretation of the results. The
Society stressed that “any clinical opinion should not simply be supported by the
literature, but also be based upon current practice and knowledge” which means
taking into consideration the “choices and interpretations which would reasonably
be made by the majority of experts working in the speciality at that moment in
time”. They also suggested that if a psychologist is called to provide more general
background information on a topic, the choice of technique or approach and the
interpretation of results may be the major issues at stake, rather than the basic
soundness of the experimental design.58

3.56 We accept that there is a sound argument for providing trial judges with detailed,
up-to-date information and guidance on the various types of complex expert
evidence, particularly scientific evidence, which may be proffered for admission in
criminal proceedings. It would clearly make sense if the judiciary had access to
the specific matters which have a bearing on the evidentiary reliability of expert
opinions derived from fields such as psychology, psychiatry and statistics (to take
some obvious examples). Nevertheless, this must remain a long-term goal; and
in any event a bench book of the type suggested by Professor Roberts would
provide only a partial replacement for a list of generic factors. We say this for two
reasons.

3.57 First, we have discussed the matter with the Judicial Studies Board (JSB) – the
independent body responsible for training the judiciary and for publishing the
Crown Court Bench Book59 – and have been told by their Director of Studies60
that they are unlikely in the near future to have the resources to be able to liaise

57
Dr Keith JB Rix (a consultant forensic psychiatrist) and Adam Wilson also offered
proposals along these lines. Mr Wilson proposed that there should be a “working party,
with cross-discipline membership” which could analyse forensic scientific disciplines to
determine both admissibility and codes of good practice; and he expressed the hope “that,
if the Commission’s proposals are implemented, recourse is not simply made to these
guidelines but, rather, specialist bodies continue to review forensic testimony and propose
improvements and good practice”.
58
The Society therefore suggested that psychology might be an area where a court-
appointed expert would be able to provide the trial judge with useful assistance in
determining reliability, on the ground that “unaided, judges may be hard pressed to make
informed decisions on such matters”. We set out our recommendations for court-appointed
experts in Part 6.
59
Directing the Jury (March 2010).
60
Judge John Phillips.

30
with relevant professional bodies to produce a compendium of specific non-
statutory guidelines. It is true that the JSB have recently been consulted in
relation to a project which will provide judges and practitioners with information
on the field of statistics, but this has been funded by a specific grant and we are
told that separate funding for other fields of expertise is extremely unlikely to be
forthcoming.61

3.58 Secondly, even if an authoritative compendium of different guidelines could


eventually be formulated by the JSB or another body in association with the
various professional bodies, it is highly unlikely that judges would ever have the
benefit of specific guidelines for all the various types of expertise which come
before them.

3.59 We therefore believe that a list of generic factors offers the only practicable
advance in the short to medium term, and possibly the only advance in the long
term for some fields of expertise. However, on the basis that specific factors may
eventually be produced and approved for various fields of expertise, we have
included a provision in our draft Bill directing the trial judge to consider such
factors if they have been published for the type of evidence in question.62

3.60 The only remaining issue addressed by some of our consultees was whether the
guidelines we originally proposed should be set out in legislation or in a code of
practice. There were mixed views on this. The Rose Committee of the Senior
Judiciary said: “Yes. Guidelines contained in statute would be helpful as long as
they remain guidelines rather than a mandatory scheme”; the Bar Law Reform
Committee could “see great benefit in a catalogue of guidelines being set down in
statute”; and the Better Trials Unit of (what was then) the Office of Criminal
Justice Reform expressed the view that generic guidelines could be set out in
primary legislation and that it would be unusual to set out such matters in
secondary legislation. However, some consultees, including judges, suggested
that the guidelines should be included in secondary legislation or in a code of
practice to ensure flexibility and rapid revision where necessary.63

3.61 On this point we have come to the firm conclusion the generic factors should be
set out in the primary legislation which contains the new admissibility test, so we
have included them in a Schedule to our draft Bill. The factors are generic, and
so would need to be amended only very rarely, if ever; the list is relatively short;
by including them in the Bill they can be read alongside the reliability test; and we
believe they deserve to be set out in the Bill.

61
The project is being run by the Royal Statistical Society’s Working Group on Statistics and
Law, chaired by Professor Colin Aitken, with the support of the Nuffield Foundation. The
Working Group has recently published its Practitioner Guide No 1: Fundamentals of
Probability and Statistical Evidence in Criminal Proceedings.
62
Clause 4(3)(b).
63
These consultees included Bruce Houlder QC (Director of Service Prosecutions) and HM
Council of Circuit Judges. One consultee, Dr Cedric Gilson (University of Westminster),
suggested that the guidelines should simply be provided in the form of training.

31
3.62 We therefore recommend that trial judges64 should be provided with a
single list of generic factors to help them apply the reliability test and that
these factors should be set out in the primary legislation containing the
test.

3.63 We recommend that the trial judge should be directed to take into
consideration the factors which are relevant to the expert opinion evidence
under consideration and any other factors he or she considers to be
relevant.

3.64 We set out and explain the generic factors we recommend in Part 5.

A limited power to disapply the reliability test


3.65 In our consultation paper we suggested that it would not be necessary to
question assumptions or well-established theories about which there was no
meaningful dispute. Our view was that the trial judge would be able to take
judicial notice of such matters (for example, the validity of the scientific
knowledge underpinning expert opinion evidence on DNA and the extreme
unlikelihood that two persons will ever have the same fingerprints).65 A further
limitation was provided by our decision to restrict our proposed test to expert
opinion evidence, so it would not need to be applied in relation to expert evidence
of fact.

3.66 Nevertheless, the senior judiciary (and some other consultees) suggested that
there would need to be a further restriction, that is, a degree of flexibility which
would allow the judge to disapply the test in cases where it was potentially
applicable, judicial notice could not be taken and it would be counter-productive
to apply the test.

3.67 The Council of HM Circuit Judges agreed with the concept of a new reliability
test, and believed our proposed test to be sound, but they felt that it should be
refined so as not to require the issue to be addressed in every case where expert
opinion evidence is tendered. This need for flexibility was particularly emphasised
by the judges of the Central Criminal Court.

64
We explain in Part 5 that, if evidentiary reliability has to be investigated in a magistrates’
court, the matter should ordinarily be addressed by a District Judge (Magistrates’ Courts).
65
Even identical twins have different fingerprints. It therefore seems that fingerprints are
generated by a combination of genetic and environmental factors in the womb, meaning
that it is extremely unlikely that two individuals will share a complete fingerprint. However,
this does not mean that an individual will always be correctly identified from a crime-scene
print, given the greater possibility that two individuals will share part of a print and, more
importantly, the difficulty of discerning whether or not a partial or smudged crime-scene
print matches a print taken from the accused in controlled circumstances. Judicial notice
has been taken of the uniqueness and permanence of fingerprints in the United States of
America (see, for example, USA v Plaza (7 January 2002) Cr No. 98-362-10 and USA v
Plaza (No 2) (13 March 2002) Cr No. 98-362-10 (USDC, ED Pa)). In United States v
Mitchell (2004) 365 F3d 215, however, the United States Court of Appeals (Third Circuit)
took the view that it is not possible to take judicial notice of such facts under the Federal
Rules of Evidence because, for the purposes of the test in the Rules, it could not be said
that they were “not subject to reasonable dispute” (as the defence had challenged the
assumptions). Nevertheless, the court also recognised that the prosecution had presented
overwhelming evidence that fingerprints are unique and permanent, so the point is perhaps
of academic interest only.

32
3.68 Likewise, the Rose Committee of the Senior Judiciary agreed with our proposals
and endorsed them save that they felt there should be a statutory power for the
court, of its own motion, or following a successful application by one of the
parties, to require the party proffering the evidence to demonstrate that the test
was satisfied, so that in other situations it would not be necessary to apply the
test. This would be a desirable restriction, the judges said, because “it would be
burdensome and unnecessary to require the parties to show that an admissibility
threshold has been met in … every case”.

3.69 A number of other consultees expressed similar concerns about the delay and
complexity our proposed test might bring if it had to be applied routinely. The
Better Trials Unit of (what was then) the Office of Criminal Justice Reform said
that while they were “broadly supportive” of our proposal, they were concerned
about the possibility of delays and disruptions to the trial process and unsure
whether the application of the doctrine of judicial notice would be sufficient
safeguard against routine challenges.66 The Bar Law Reform Committee feared
that, notwithstanding our assurances to the contrary, the language of our test
suggested an admissibility investigation in every case, which they felt would be
undesirable. The Forensic Science Regulator expressed concern that the criminal
justice system might “be flooded with applications to exclude evidence and may
be disadvantaged”. Similarly, the RSPCA felt that there should be mechanisms to
prevent meritless challenges.

3.70 However, some of our academic consultees with a particular interest in expert
evidence expressed concern that the judiciary might even be able to rely on the
doctrine of judicial notice in relation to fields of forensic science. Dr Tony Ward
(University of Hull) felt that reliance on judicial notice would raise “the danger that
some forms of forensic expertise will continue to be relied upon because they
have always been relied upon”. He said, in relation to forensic scientific evidence:
“Insisting that all fields of expertise must establish their validity would provide
[experts with] a salutary incentive to ensure that their scientific houses were in
order.” Similarly, Associate Professor William O’Brian (University of Warwick)
suggested, first, that we “should be robust in drafting [our Bill] to prevent …
evasion by [the] courts” and, secondly, that we should “disavow [our] suggestion
that the courts should take judicial notice of the validity of some techniques”.67

3.71 The comments in the previous paragraph are not without merit. We certainly
agree that an important policy objective underpinning the statutory measures we
recommend must be the encouragement of better practices, with standards being
raised and expert opinion evidence being tendered for admission only when it is
likely to withstand judicial scrutiny.

3.72 However, we also believe that if we are to recommend a legislative solution to the
problems associated with expert opinion evidence in criminal proceedings it must
be a proportionate response. In other words, we agree with the judiciary and
other consultees that the trial judge should not have to investigate the reliability of
the evidentiary foundation of an expert’s opinion evidence unless it is appropriate
66
The Better Trial Unit’s response was qualified with the caveat that it may not represent the
ultimate view of the Government.
67
In addition, Professor Mike Redmayne (London School of Economics) suggested that
judges should be required to apply our proposed reliability test seriously.

33
to do so. We also accept that the doctrine of judicial notice may be an insufficient
basis for disapplying our proposed test. Much reliable expert opinion evidence is
tendered for admission in the criminal courts every day, and it would clearly not
serve the interests of justice to have evidence about which there are no concerns
routinely assessed against our reliability test with the consequence that criminal
trials generally become longer and more costly.

3.73 In short, we believe that a balance needs to be struck as between what may be
desirable in principle and what is in fact practicable and cost-effective.

3.74 In this context it is worth reflecting on what the Court of Appeal said in Jisl:68

The funding for courts and judges, for prosecuting and the vast
majority of defence lawyers is dependent on public money, for which
there are many competing demands. Time itself is a resource. Every
day unnecessarily used, while the trial meanders sluggishly to its
eventual conclusion, represents another day’s stressful waiting for the
remaining witnesses and the jurors in that particular trial, and no less
important, continuing and increasing tension and worry for another
defendant or defendants, some of whom are remanded in custody,
and the witnesses in trials which are waiting their turn to be listed.

3.75 It follows that there is unlikely to be any appetite for a new reliability test, whether
in the Government, in Parliament or amongst the judiciary, if it is thought that
substantial costs and unwarranted delays would be generated as a result.

3.76 A proportionate legislative response cannot, however, mean recommending a


scheme which would provide only an ostensible barrier to the admission of
unreliable expert opinion evidence in criminal proceedings.

3.77 We therefore make the following recommendation:

(1) criminal courts should have a limited power to disapply the


reliability test so that it does not have to be applied routinely and
unnecessarily;

(2) but, equally, the power to disapply must not be such that the
reliability test becomes only a nominal barrier to the adduction of
unreliable expert opinion evidence.

3.78 We explain this recommendation more fully in Part 5.

The onus of persuasion


3.79 In our consultation paper we did not propose any standard of proof for the issue
of evidentiary reliability, but we did propose that the party wishing to adduce the
evidence should have to show that it is sufficiently reliable to be admitted. We
took the view that the reliability of much expert opinion evidence may not be
capable of being proved in the way that it is possible to prove a past fact or that
an event occurred.69 We therefore suggested that the general question should be
68
[2004] EWCA Crim 696 at [114].
69
A point with which the Council of HM Circuit Judges agreed.

34
whether there are sufficient indicia of reliability to justify the admission of the
expert opinion evidence before a jury.70

3.80 There was very broad consensus amongst our consultees, often very firmly
expressed, that the burden of establishing reliability should lie with the party
tendering the evidence.

3.81 However, some consultees expressed concern about the possibility of the burden
shifting to the accused to prove his or her innocence in a case dominated by
disputed medical evidence, or the difficulty the defence would face in having to
demonstrate reliability given its limited resources. The judges of the Central
Criminal Court felt that there might need to be some flexibility for trial judges
when considering the admissibility of defence evidence.

3.82 Professor Paul Roberts queried whether it was right as a matter of principle to
expect the accused to establish the reliability of his or her expert opinion
evidence in the way we suggested. He said that

it should not be assumed, without argument, that the appropriate


validity threshold for accepting prosecution evidence … will
necessarily be the same threshold test of validity that should be
applied to potentially exculpatory evidence tendered by the defence

3.83 Professor Roberts added that it

might be possible to construct a principled argument, rooted in the


presumption of innocence, for affording the methodological
credentials of expert evidence tendered by the accused a somewhat
more generous benefit of the doubt.

3.84 Indeed, Professor Roberts went so far as to suggest that defence expert
evidence with “validity … demonstrated to a 0.49 level of probability”71 should
nevertheless be admissible as it could prevent the prosecution from proving its
case beyond reasonable doubt.

3.85 It is of course quite right that the prosecution must prove the accused’s guilt
beyond reasonable doubt, and the admissibility of expert opinion evidence must
always be viewed against this fundamental requirement. Other important
contextual points are that, as a general rule, the accused is entitled to adduce
any admissible evidence which might place in the jury’s collective mind a
reasonable doubt as to his or her guilt, and that the prosecution must establish a
“prima facie case” before the accused has to present any evidence in support of
his or her defence.72

70
Consultation Paper No 190, paras 6.53 to 6.63.
71
That is, evidence which is probably invalid, assuming of course that it would ever be
possible to assess validity in this precise way.
72
That is to say, by the close of the prosecution case there must be sufficient evidence of D’s
guilt for a reasonable jury, properly directed, to be able to conclude beyond reasonable
doubt that D is guilty.

35
3.86 However, we believe that the presumption of innocence, crucially important
though it is, does not give the accused the right to adduce unreliable expert
evidence in order to mislead the jury or distract the jury from reliable evidence
which points to his or her guilt. A number of cases suggest that unreliable
defence evidence having little probative value and (amongst other things) the
potential to cloud the issues and mislead the jury can be excluded at common
law on the ground that it is “irrelevant” and therefore inadmissible.73 In a similar
vein, Parliament has in recent years enacted legislation to prevent the accused
from being able to put logically relevant but misleading or distracting evidence
before the jury.74 And Crown Court trial judges are now beginning to provide
juries with directions designed to counterbalance defence suggestions which are
unsupported by empirical research and therefore unreliable and potentially
misleading.75

3.87 We explained in Part 1 why expert opinion evidence is a special type of evidence
which deserves special treatment, and our reasons apply equally to defence
evidence. In particular, expert opinion evidence of a scientific nature, whether
tendered for admission by the prosecution or the defence, should always be
founded on the principles of good science. The Court of Appeal recently
acknowledged as much in Reed,76 following the publication of our consultation
paper, where it held that “expert evidence of a scientific nature is not admissible
where the scientific basis on which it is advanced is insufficiently reliable for it to
be put before the jury”. 77

3.88 We therefore recommend for our proposed reliability test that, where the
test is applied, the party wishing to adduce the expert opinion evidence
should bear the burden of demonstrating that it is sufficiently reliable to be
admitted.

3.89 The party wishing to adduce the expert opinion evidence would therefore need to
provide the necessary evidence and arguments to support a submission that the
opinion evidence the expert wishes to give is sufficiently reliable to be taken into
consideration by a jury.

73
Consultation Paper No 190, Appendix A.
74
See, for example: s 100 of the Criminal Justice Act 2003 (limits on the admissibility of the
extraneous misconduct of prosecution and defence witnesses other than the accused);
s 126 of the Criminal Justice Act 2003 (a discretion to exclude admissible hearsay
tendered by the defence or the prosecution); and s 41 of the Youth Justice and Criminal
Evidence Act 1999 (limits on the admissibility of the extraneous sexual behaviour of
complainants in sexual offence cases).
75
See MM [2007] EWCA Crim 1558, [2007] All ER (D) 196 (Jun) and D(JA) [2008] EWCA
Crim 2557, [2009] Criminal Law Review 591 (directions on possible reasons why sexual
offence complainants do not immediately complain about the alleged assaults, to
counterbalance the defence suggestion that a delayed complaint is evidence of a false
allegation). See also the recent observations in Miller [2010] EWCA Crim 1578, [2010] All
ER (D) 170 (Jul) and the guidance provided in Chapter 17 of the Judicial Studies Board’s
Crown Court Bench Book, Directing the Jury (March 2010).
76
[2009] EWCA Crim 2698, [2010] 1 Cr App R 23.
77
[2009] EWCA Crim 2698, [2010] 1 Cr App R 23 at [111]. However, in line with the
traditional laissez-faire approach to admissibility which has developed at common law, the
court added that there is “no enhanced test of admissibility” for evidence of a scientific
nature.

36
3.90 We are confident that the application of this recommendation to defence expert
opinion evidence would not place the accused in the position of having to prove
his or her innocence.

3.91 It is for the prosecution to prove the accused’s guilt beyond reasonable doubt on
the basis of the admissible evidence presented in the trial. Save for the
exceptional situation where the accused has to prove a specific defence, the
accused bears nothing more than a tactical burden to place in the jury’s collective
mind a reasonable doubt as to his or her guilt. Under our recommendations, it
would be for the prosecution to demonstrate the reliability of its expert opinion
evidence (when required to do so) because such evidence would be supporting
an affirmative proposition relating to the accused’s guilt.

3.92 To take an example drawn from Harris and others,78 if the basis of the
prosecution case is that D murdered his or her child on account of the presence
of certain intra-cranial injuries, it would be for the experts called by the
prosecution to demonstrate that their opinion evidence, including the strength of
their conclusions, is fully justified. They would have to establish the soundness of
their hypothesis that such injuries demonstrate a non-accidental injury (as
opposed to an accidental or congenital cause) to the extent necessitated by the
strength of their opinions. If the prosecution has no other evidence of D’s guilt,
save for what the prosecution experts regard as D’s implausible exculpatory
explanation, the prosecution’s experts would need to show that their hypothesis
has been established through empirical research to the extent necessary to
justify their opinion that they are sure that the child’s death resulted from a non-
accidental injury.79 Moreover, the jury would need to be sure – satisfied beyond
reasonable doubt – that the prosecution experts’ opinion in this respect is correct.

3.93 However, if a defence expert has simply been called to draw attention to possible
problems with the prosecution expert witness’s methodology, data, inferences,
assumptions, reasoning and so on, and if the defence expert is not putting
forward an alternative proposition (other than the claim that the prosecution
expert is wrong), so he or she is not relying on any hypothesis or empirical
research (and so forth) of his or her own, then the reliability test would be
inapplicable to that expert’s evidence. The defence will always be able to call
impartial expert witnesses to reveal flaws in the methodology and reasoning of
prosecution expert evidence.

3.94 An analogy can be drawn with the type of case where the defence advocate,
acting on the guidance of an expert or undisclosed expert report, probes the

78
[2005] EWCA Crim 1980, [2006] 1 Cr App R 5, para 1.7 above.
79
There would need to be properly conducted research to establish a sound correlation
between the intra-cranial injuries and a non-accidental cause (from independent evidence)
and the absence of such injuries where there have been accidents or congenital
conditions. The stronger the expert’s opinion, the greater would need to be the data
consistent with it.

37
evidence of a prosecution expert witness during cross-examination.80 The
defence advocate may not be positing any suggestion of his or her own, but will
be seeking to challenge and undermine the opinion evidence of the prosecution
expert by revealing flaws in reasoning, research methods or the inferences drawn
from data.81

3.95 In this context, it is worth setting out what the Court of Appeal recently said in
Henderson and others:82

We must recognise the limits of medical science and in particular that


there may be events, deaths or symptoms which are unexplained and
unforeseen. Further, any conclusion must acknowledge the
importance of the burden of proof in the context of cases such as
these.83 It is not for the defence to provide any explanation; the mere
fact that it is unable to do so is not of itself a sound basis for
concluding that the prosecution’s evidence is correct.

3.96 But if an expert called by the defence is not limiting his or her evidence to
pointing out weaknesses or inaccuracies in the prosecution expert witnesses’
evidence, and relies on his or her own hypothesis or database, or provides an
explicit or tacit opinion based on a different interpretive model from that relied on
by the prosecution experts, then our test would apply, potentially, in relation to
those aspects of the defence evidence.84 In such cases, the obligation on the
defence under rule 33.4(1) of the Criminal Procedure Rules 2010 should mean
that the prosecution and trial judge are aware of the defence’s expert evidence
before the trial, and, if necessary, the question of its reliability would be
addressed at a pre-trial hearing, as we explain in Part 5.

3.97 Although the defence can be expected to disclose its expert opinion evidence
before the trial – if only because deliberate non-disclosure for tactical reasons
could lead to the evidence being excluded under rule 33.4(2) of the 2010 Rules85
– it is possible that a defence expert called simply to reveal flaws in prosecution
expert evidence might wish to develop his or her evidence during the trial and go
beyond giving expert evidence of fact. We recognise that it might be difficult for

80
Rule 33.4(1) of the Criminal Procedure Rules 2010 provides that the parties must serve
their expert evidence (as expert reports) on the other parties and the court “as soon as
practicable”. However, by virtue of r 33.4(2), the defence may use an expert’s report,
without disclosing it, as the basis for cross-examining prosecution expert witnesses; but if a
defence submission of “no case to answer” fails, and the defence would then like to
adduce its expert evidence to counter the prosecution evidence, the defence will be able to
do so only if the trial judge gives his or her permission (assuming the prosecution does not
agree that the evidence should be admitted). The trial judge’s power under r 33.4(2) to
prevent the defence from relying on undisclosed expert evidence should not be
overlooked; see Ensor [2009] EWCA Crim 2519, [2010] 1 Cr App R 18.
81
One of our consultees, a forensic accountant, told us that in cases where he had not had
to disclose his report (presumably because he was not called to testify) it was used during
cross-examination and “destroyed the prosecution case”.
82
[2010] EWCA Crim 1269 at [44], [2010] 2 Cr App R 24.
83
Non-accidental head injury allegations.
84
We say “potentially” because the judge would have the power to disapply the test; see
paras 3.65 to 3.78 above and paras 5.42 to 5.61 below.
85
See Ensor [2009] EWCA Crim 2519, [2010] 1 Cr App R 18.

38
such an expert to limit his or her evidence solely to factual matters. Where a
defence expert wishes to provide a previously undisclosed opinion in the course
of his or her testimony, and there has been no deliberate ploy to ambush the
prosecution, the question whether the opinion evidence should be assessed for
evidentiary reliability during the trial would depend on factors such as the nature
of the opinion evidence, whether or not the prosecution wishes to challenge the
evidence and the strength of the prosecution case. In the situation just described,
it is unlikely that the defence expert would be positing an opinion of the sort which
would need to be assessed for evidentiary reliability, so it is likely that the judge
would disapply the reliability test and allow the trial to proceed. We do not foresee
trials being routinely disrupted by prosecution challenges to such opinion
evidence, and where a challenge is made we would expect the trial judge to
adopt a sensible, proportionate response to the problem and so avoid
unnecessary disruptions to the trial process. In the situation described, the judge
might simply allow the prosecution experts to be recalled to comment on the
undisclosed opinion evidence.

3.98 As intimated above, the type of situation our recommendations are designed to
address is where a defence expert puts forward an alternative, affirmative
proposition which is central to the accused’s defence – for example, that a child’s
death resulted from scurvy or a particular congenital condition rather than a non-
accidental injury. For such evidence, the court may require the reliability test to
be satisfied in respect of the expert’s underlying hypothesis to the extent
necessitated by the strength of that expert’s opinion evidence. As explained
already, where expert evidence of this sort is relied on by the defence it will
almost always be disclosed in advance of the trial, and this would permit a pre-
trial assessment of evidentiary reliability.

3.99 The qualification in the last paragraph – that the reliability test would need to be
satisfied to the extent necessitated by the strength of the defence expert’s
opinion – is important, and explains why the accused would never have to prove
his or her innocence in the type of case which turns on disputed medical or other
scientific evidence. Given the obligation on the prosecution to prove the
accused’s guilt beyond reasonable doubt, the defence expert would merely need
to show that his or her alternative explanation for the child’s death is a
reasonable (that is, a realistic, sound) alternative, either on the basis of the
empirical research relied on by the prosecution experts or on separate research.
We expect it will often be the former, with the parties’ respective experts
suggesting that different inferences can legitimately be drawn from the same
data.

3.100 If the prosecution has called impartial expert witnesses relying on research
findings published in reputable scientific journals, those experts would be aware
of, and would have taken into consideration, the research data supporting and
the research data undermining their opinion evidence. The defence is likely to
rely on the same data and research findings, and possibly on other published
data, but suggest a different interpretation from that advanced by the prosecution
experts. It is possible, however, that the defence experts would have conducted
or relied on other research which has not been published in a reputable journal or
been peer-reviewed in some other manner and which is not known to the
prosecution witnesses; but this may well give rise to legitimate concerns about
the reliability of the defence experts’ data and findings.

39
3.101 Either way, the accused would merely have to persuade the jury that there is a
realistic, sound alternative to the proposition advanced by the prosecution
experts. The defence would not have to prove to the judge or to the jury that the
inference it asks the jury to accept is established on the balance of probabilities,
but (if the reliability test is applied) the defence would have to demonstrate to the
judge that that inference is a legitimate one to draw before it could be placed
before the jury. The defence expert would have to show that it is an inference
based on sound scientific principles and data obtained from properly conducted
scientific research.

3.102 As the Forensic Science Service suggested in relation to scientific evidence –


though the same principles apply equally to other expert opinion evidence –
regardless of which party is seeking to adduce the evidence, an expert’s
evaluative opinion should always demonstrate “balance, logic, transparency and
robustness”.

3.103 But it should always be borne in mind that a scientific hypothesis may ultimately
be shown to be wrong, for scientific understanding is constantly evolving and
developing, as the Court of Appeal recently acknowledged:

Conclusions of … experts … necessarily involve a process of


induction, that is inferring conclusions from given facts based on other
knowledge and experience. But particular caution is needed where
the scientific knowledge of the process or processes involved is or
may be incomplete. As knowledge increases, today’s orthodoxy may
become tomorrow’s outdated learning. Special caution is also needed
where expert opinion evidence is not just relied upon as additional
material to support a prosecution but is fundamental to it.86

3.104 The fact that a defence scientific expert’s opinion may be incorrect, or that it is
based on a hypothesis supported by only a small minority of scientists in the field,
should not render the opinion inadmissible if it is based on the application of
sound scientific principles. In scientific disciplines there will often be two or more
competing but legitimate hypotheses which are supported, or not discredited, by
the results of properly conducted research. Following further research, a minority
view may subsequently become the generally accepted position, and what was
once orthodoxy may subsequently come to be regarded as outdated learning.

3.105 Because a defence scientific expert’s opinion should always be founded on


sound scientific principles, we are not persuaded by the argument, rooted in the
presumption of innocence, for affording the methodological credentials of expert
evidence tendered by the accused a somewhat more generous benefit of the
doubt. Under our proposals, in a case which turns on conflicting expert medical
(scientific) evidence, the defence experts would be expected to demonstrate that

86
Holdsworth [2008] EWCA Crim 971 at [57]. A similar point was made in civil proceedings in
Re U (A Child), Re B (A Child) [2004] EWCA Civ 567, [2005] Fam 134 at [23]. See also
Henderson and others [2010] EWCA Crim 1269, [2010] 2 Cr App R 24 at [1]: “Where the
prosecution is able, by advancing an array of experts, to identify a non-accidental injury
and the defence can identify no alternative cause, it is tempting to conclude that the
prosecution has proved its case. 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.”

40
their opinion evidence is a legitimate, scientifically valid conclusion, based on
sound principles and properly conducted research.

3.106 To put it another way, and with reference to the “hard case” example provided by
Professor Roberts, defence expert opinion evidence based on a methodological
foundation with “validity … demonstrated to a 0.49 level of probability”87 should
not be admissible and it would not be admissible under our proposed test,
assuming of course that it would ever be possible to quantify validity – or rather
invalidity – with such mathematical precision. This is because an expert opinion
based on a methodological foundation which is more likely to be invalid than valid
would not be sufficiently reliable to be admitted.

3.107 The unlikelihood that such precise invalidity could ever be demonstrated means
that Professor Roberts’ “0.49 level of probability” situation is unlikely ever to arise
in practice. In reality, the trial judge facing a case of indeterminable, borderline
validity would err in favour of the accused and address the admissibility of the
expert’s opinion evidence on the assumption that its foundation is valid.
Nevertheless, Professor Roberts’ example is an important one and we are
grateful to him for raising it. It allows us to reaffirm our view that defence expert
opinion evidence based on a methodological foundation which is as likely to be
invalid as valid should not be admitted. If there were to be no requirement of
underlying validity established above the 0.50 level of probability, and the
accused were to be entitled to adduce expert opinion evidence founded on invalid
methodology, it would be open to the accused to adduce all sorts of “junk
science” in support of his or her defence.

3.108 To give a practical example, if D is charged with the murder of his or her child on
account of the presence of a triad of injuries associated with shaken baby
syndrome in tandem with other circumstantial evidence of non-accidental injury, it
would be quite wrong if D’s experts were to be permitted to adduce evidence of a
discredited hypothesis that the triad of injuries could arise from non-traumatic
natural causes.88 As we said in our consultation paper, “the defence should not
be able to divert the jury’s attention from reliable prosecution evidence by being
allowed to adduce inherently unreliable expert evidence which might give rise to
an unwarranted (as opposed to a reasonable) doubt as to the guilt of the
accused”.89

3.109 It bears repeating, however, that this does not mean that the defence would be
prevented from relying on an expert opinion held by only a small number of
experts in the field. Insofar as such theoretical matters can ever be meaningfully
quantified, defence expert opinion evidence based on a foundation with validity
demonstrated to a “0.51 level of probability” should be admissible – even if the
inference to be drawn from the underlying research data consistent with D’s
defence is one which very few other experts in the field would support and so is
generally thought to be incorrect. It is the minimum requirement of 51 per cent
standard of scientific validity which would allow the jury properly to be able to
infer that there is a reasonable doubt as to D’s guilt, and D should not be denied

87
Paragraph 3.84 above.
88
See paras 8.31 to 8.41 below.
89
Consultation Paper No 190, para 6.63.

41
the opportunity of placing that opinion evidence before the jury just because only
a small minority of experts in the field would support the drawing of that
inference.90

3.110 The opinion of a scientific expert would be admissible for the defence even if it
would place in the jury’s collective mind only a small possibility that D is not
guilty, so long as it is a reasonable possibility. But for that possibility to be a
reasonable possibility (generating a reasonable doubt as to D’s being guilty) it
would need to be founded on a hypothesis which is consistent with the available
observational data; and, if it is founded on new experimental data, the underlying
research should have been properly conducted and scrutinised in accordance
with the principles of sound scientific methodology. Again, an expert opinion of a
scientific nature should be admissible for the accused in criminal proceedings if it
was reached by the application of valid scientific methodology and reasoning,
even if the opinion is generally thought to be wrong.

3.111 To summarise, if the accused wishes to rely on expert opinion evidence


amounting to an affirmative proposition predicated on a particular scientific
hypothesis to show there is a reasonable doubt as to his or her guilt, it would be
for the defence to demonstrate (in effect, to prove on the balance of probabilities)
that the opinion is underpinned by a foundation of valid scientific methodology or,
in the words of clause 4(1)(a) of our draft Bill, to show that the opinion is “soundly
based”.91 An expert opinion will not be “soundly based” if it is underpinned by
methodology which is no more likely to be valid than invalid. However, the
question whether or not the expert’s opinion evidence is sufficiently reliable to be
admissible is not so much a fact which is capable of being proved to a particular
standard and, under our proposed test, it would not need to be proved to any
standard, whether the party tendering it for admission is the prosecution or the
defence. Determining the evidentiary reliability of an expert’s opinion is more like
a judgment call based on all the available evidence.92

3.112 Lord Justice Leveson appeared to make this point in his speech to the 2009 Bond
Solon Expert Witness Conference,93 when he said:

a court examining expert evidence is more like an English tutor


considering an essay, rather than the maths tutor looking for the right
numbers.
90
One of the reasons why we do not support an admissibility test based on scientific
consensus – a so-called Frye test – is because such a test could prevent the adduction of
scientifically sound opinion evidence supported by only a minority of scientists in the field.
See Consultation Paper No 190, para 4.31.
91
Although a defence proposition giving rise to a reasonable doubt as to the accused’s guilt
should (we suggest) simply require underlying soundness to be established on the balance
of probabilities, the position for prosecution evidence is less straightforward. As explained
in para 3.92 above, the prosecution experts would need to establish the soundness of their
underlying material (including any hypothesis relied on) to the extent required by the
strength of their expert opinions.
92
For the view that there should be explicit standards of proof for the reliability of expert
evidence in criminal proceedings – see J Hartshorne and J Miola, “Expert evidence:
difficulties and solutions in prosecutions for infant harm” (2010) 30 Legal Studies 279, 294
to 296.
93
6 November 2009, http://www.judiciary.gov.uk/media/speeches/2009/speech-wall-lj-
06112009 (last visited 3 February 2011).

42
3.113 So, although we believe the same reliability test (and guiding factors) should
apply whether the expert opinion evidence is tendered for admission by the
prosecution or defence, and although it would be for the party calling the expert
witness to demonstrate that the witness’s opinion evidence is sufficiently reliable
to be admitted, it does not follow that the defence would have to prove to the
judge on the balance of probabilities that its expert opinion evidence is correct.
The criteria for assessing evidentiary reliability would always be the same, but the
strength of the expert’s opinion evidence, together with the burden and standard
of proof to be applied by the jury (or other fact-finding tribunal) in a criminal trial,
would determine the foundation of knowledge and research data needed to
demonstrate that that opinion evidence is sufficiently reliable to be admitted.

3.114 To illustrate this, consider again a trial where the prosecution’s case is that D
murdered a young child (V) by shaking but there is no evidence of D’s guilt other
than (1) D’s being alone with V at the time when V suffered the fatal injury and (2)
the opinion of a scientific expert that V’s fatal injury was non-accidental.

3.115 In such a case, in order for the prosecution to secure a conviction for murder, the
expert called by the prosecution would have to give evidence that he or she was
sure that the injury was non-accidental (that is, caused by shaking) and the jury
would have to be sure that his or her opinion evidence was correct. There would
therefore need to be an extremely cogent (and broad) evidential foundation
supporting the hypothesis underpinning the prosecution expert’s opinion, with
nothing in that large database to suggest that the hypothesis might be incorrect.
So, in the absence of a sufficiently broad, corroborative foundation of research
data, the expert’s opinion would be inadmissible on the ground of insufficient
reliability.94

3.116 But if in a similar type of case there was other cogent evidence of D’s guilt (for
example, evidence that D had previously committed acts of violence against V or
that D had tacitly admitted his or her culpability to E) the prosecution might be
able to rely on weaker expert opinion evidence (for example, an opinion that V’s
condition was unlikely to have resulted from an accident or congenital condition,
although an accident or congenital condition could not be ruled out). An opinion
of this sort would of course need to be based on sound scientific principles, but
the empirical research underpinning it would not need to be as extensive as that
supporting an opinion of the type described in the previous paragraph.

3.117 The evidential foundation supporting the shaken-infant hypothesis underpinning


the expert’s opinion in the second type of case would not need to be as cogent as
that in the first type of case. But in each case, in determining whether the expert
opinion evidence was sufficiently reliable to be admitted, the criteria relating to
scientific validity applied by the trial judge would be the same. That is to say, for
scientific opinion evidence, the underlying evidence supporting the hypothesis
and the chain of reasoning underpinning the opinion would always need to be
scientifically valid; but the required extent to which there has been scientific
research and the required extent of the corroborative data supporting a

94
In this example, if the prosecution expert was not giving evidence in terms of being sure
that an accident could be ruled out, but only making accident unlikely, then while his or her
opinion could be admissible, there would be no case for D to answer (given the absence of
any other evidence of D’s guilt).

43
hypothesis will depend on the nature and strength of the opinion and the extent to
which it is qualified.95

3.118 Ear-print comparisons relied on by the prosecution provide another helpful


example. In Dallagher96 a prosecution expert on ear-prints testified that, following
a comparison of control ear-prints provided by D and latent ear-prints lifted from a
window, he was “sure” and “absolutely convinced” that D left the latent prints. In
our consultation paper we suggested that this opinion evidence was insufficiently
reliable to be admissible.97

3.119 The reason for our view was that there was an insufficient body of research data
to support the assumption as to the uniqueness of ear shapes or, if uniqueness is
accepted, the assumption that ears leave unique prints, assumptions which
underpinned the strength of this expert’s opinion.98

3.120 In addition, the expert’s opinion relied heavily on subjective factors, that is,
experience rather than objectively verifiable measuring techniques.99

3.121 But in another case, a weak opinion based on ear-prints may well be sufficiently
reliable to be admitted (under our proposed test), if the prosecution relies on the

95
Associate Professor William O’Brian said in his response to our consultation paper that an
expert’s evidence might be insufficiently reliable to found a conviction but be sufficiently
reliable to establish a reasonable doubt as to D’s guilt; and Dr Tony Ward also provided a
comment broadly consistent with our analysis. In addition, the General Medical Council
said in relation to medical evidence that “the more significant the evidence is to the issues
in a case, the greater the scrutiny of its admissibility should be”.
96
[2002] EWCA Crim 1903, [2003] 1 Cr App R 12.
97
Consultation Paper No 190, paras 2.14 and 2.15.
98
The Court of Appeal noted (at [9]) that the expert “had simply become interested in ear
print identification and read what was available on the topic. He had built up a portfolio of
about 600 photographs and 300 ear prints and from his experience and what he had read
he was satisfied that no two ear prints are alike in every particular.” It was also noted that
the prosecution experts agreed that “it would be very useful if further research was done to
see whether it were possible for prints from two separate ears to be produced showing
apparent similarities”. On appeal, D relied on fresh expert evidence to the effect that there
“is no empirical research, and no peer review to support the conclusion that robust
decisions can be founded on comparisons which in turn are critically dependent on the
examiner’s judgment in circumstances where there are no criteria for testing that judgment”
(at [11]).
99
Unfortunately the Court of Appeal did not consider the strength of the prosecution
witness’s opinion when addressing admissibility. Adopting the traditional laissez-faire
approach to admissibility, whereby expert evidence is admissible if “it is sufficiently well-
established to pass the ordinary tests of relevance and reliability”, the court held that the
expert opinion evidence of ear-print analysts is admissible, and then went on to consider
(at [30]) whether the jury had been “properly equipped to assess the weight to be attached”
to the prosecution experts’ conclusions, given that the witnesses had accepted in evidence
that their opinions were based on assumptions “supported by relatively limited information”
(at [34]).

44
expert’s opinion merely to provide additional support for other cogent evidence of
the accused’s guilt.100

3.122 So, if a prosecution expert’s opinion is that the latent print from the crime scene
and the control print taken from the accused show consistencies and no
inconsistencies, the expert could opine that this is the case (assuming that the
Turner test is satisfied and there is no other reason for excluding the
evidence).101 If there is a sufficient body of data on similarities and differences
between individuals’ ear-prints, the expert might even be able to give an opinion
as to the probable number of persons (including the accused) who could have left
the latent print.

3.123 But under our proposed admissibility test, the body of data in support of a
hypothesis of uniqueness would need to be very strong indeed before any such
expert would be permitted to opine that ear-print evidence standing alone
establishes the accused’s guilt beyond reasonable doubt.

3.124 It should not be thought, however, that different considerations necessarily apply
if there is additional evidence of the accused’s guilt independent of the opinion
evidence provided by a prosecution expert. The important point is that if an
expert interpreter of ear-prints wishes to give an opinion, based on such prints,
on the probability that the accused left a latent print at the scene of the crime,
then his or her opinion must be sufficiently reliable to be admitted, and it is the
particular opinion, including its strength, which will need to assessed against our
proposed test.

CODIFICATION OF THE ADMISSIBILITY TEST GENERALLY


3.125 In Part 2 we explained that there are four common law admissibility requirements
for expert evidence in criminal proceedings, and in this Part we have reaffirmed
our view that the reliability limb should be replaced by a new statutory test. We
now turn to the other three limbs of the common law test, set out in paragraphs
2.3 to 2.11 above.

3.126 In our consultation paper we expressed the view that these three limbs (relating
to assistance, expertise and impartiality) are fundamentally sound and relatively

100
In Dallagher [2002] EWCA Crim 1903, [2003] 1 Cr App R 12 the expert witness called by D
(on appeal) opined, according to the Court of Appeal’s summary, that an “ear print
comparison can help to narrow the field … but cannot alone be regarded as a safe basis
on which to identify a particular individual as being the person who left one or more prints
at the scene of a crime” (at [12]). See also Consultation Paper No 190, fn 14 to para 2.15,
where we address the subsequent analysis in Kempster (No 2) [2008] EWCA Crim 975,
[2008] 2 Cr App R 19.
101
The Turner test (para 2.3 above) would not be satisfied in this context if the expert’s
evidence would not furnish the jury with information which is likely to be outside its
experience and knowledge. In addition, admissible prosecution evidence may be excluded
by the judge at common law or under s 78(1) of the Police and Criminal Evidence Act 1984
if its probative value would be outweighed by the unduly prejudicial effect it would have on
the jury.

45
uncontroversial, although we acknowledged that they could occasionally give rise
to problems when applied in practice.102

3.127 We asked our consultees whether they agreed with us. More to the point, we
asked whether we should incorporate these aspects of the common law test into
the legislation we envisaged for our proposed reliability test so that all the
admissibility requirements would stand together in a single statutory framework or
code.103

3.128 There was very broad agreement with our view that the rest of the common law
admissibility test is satisfactory.104 Indeed there was very little disagreement at
all.105 There was also broad support for codification of the admissibility test for
expert evidence, the suggestion being that by incorporating all aspects of the test
into a single Act of Parliament we would bring certainty, stability, and uniformity to
the law.106

3.129 However, some consultees could see no good purpose in codifying the law. In
particular, the Council of HM Circuit Judges argued that the present rules are
“workable and flexible” and “should remain so”. We note, however, that this
comment would appear to have been based on a misunderstanding of what we
intended by codification. Our view was that the present requirements should be
placed on a statutory footing with no loss of flexibility.

3.130 In any event, we do not disagree with the argument that codification in isolation
would bring little benefit. Our suggestion was predicated on the provisional
proposal that there should be a new test relating to evidentiary reliability set out in
primary legislation. Codification would be beneficial, therefore, because it would
bring all the admissibility requirements together.

3.131 Mr Justice Treacy supported codification for this reason. He noted that the
common law rules are “well-established and accepted” and it would be
“desirable” to incorporate them into the new legislation so that “all relevant tests
and materials are in a single place and carry equal authority”. Similarly the
Criminal Cases Review Commission felt there was a strong argument for
incorporating all the common law rules relating to expert evidence in a single
code.

102
Consultation Paper No 190, para 1.8. In truth it is the Turner test which has occasionally
given rise to difficulties, not the other requirements.
103
Consultation Paper No 190, paras 1.2, 1.3 and 1.8, with para 6.82.
104
Some consultees recognised, in line with the view we expressed in our consultation paper,
that, while the test itself is satisfactory, there may be problems in its application. The same
point was also recently made by Lord Justice Leveson in his speech for the Forensic
Science Society and King’s College, London, 16 November 2010, available at
www.judiciary.gov.uk/media/speeches/2010/speech-lj-leveson-expert-evidence-16112010
(last visited 3 February 2011).
105
Professor Mike Redmayne and Associate Professor William O’Brian were unenthusiastic
about the Turner test. Professor Redmayne felt that the test is too vague, whereas
Associate Professor O’Brian felt that some (unspecified) aspects of the test were “deeply
troublesome”.
106
A point made by the Police Superintendents’ Association and the Crown Prosecution
Service.

46
3.132 Representing the Bar, the Criminal Bar Association felt that the other aspects of
the common law admissibility test were satisfactory and should be codified in
primary legislation, incorporating “all steps to admissibility … as well as the
[reliability] provision”; and the Bar Law Reform Committee suggested that there
should be “comprehensive codification of the law relating to expert witnesses in
the criminal trial to be contained in primary legislation”. The Committee added
“that there is no reason why statutory reform in this area ought not also to provide
for more rigorous examination of the sufficiency of an individual’s suitability to
appear as an expert witness”. The Law Society, representing solicitors, also
supported codification.

3.133 The Rose Committee of the Senior Judiciary accepted, rightly, that codification
was not necessary, but they agreed that it “might be of assistance if the evidential
rules were codified in primary legislation, so as to provide a trial judge with a
framework, or reference point, for his [or her] determination of the issue of
admissibility”.

3.134 Other consultees who supported codification included the Crown Prosecution
Service, the Justices’ Clerks’ Society, the RSPCA, the Forensic Science Society
and the Police Superintendents’ Association. Northumbria University’s School of
Law’s Centre for Criminal and Civil Evidence and Procedure rightly
acknowledged that it would be illogical to have a statutory reliability test in
tandem with the existing common law tests.107

3.135 Given the positive responses to our suggestion that there should be general
codification of the admissibility requirements in criminal proceedings, we have
come to the conclusion that the statutory admissibility test for expert evidence
should incorporate all aspects of the common law test.

3.136 We therefore recommend that there should be a single framework in


primary legislation governing the admissibility of all expert evidence in
criminal proceedings.

3.137 We explain this recommendation more fully in Part 4.

107
Some consultees went further, suggesting that we should emphasise particular aspects of
the common law admissibility test. The Forensic Science Society stressed the importance
of the impartiality requirement.

47
PART 4
CODIFICATION

INTRODUCTION
4.1 In Part 3 we expressed our view following consultation that our proposed
statutory admissibility test for expert evidence should incorporate all aspects of
the common law test, in addition to our new reliability test. In this Part we set out
our recommendation for the codification of the three common law admissibility
rules summarised in paragraphs 2.3 to 2.11 above.

4.2 In our summary in Part 2 we explained that, at common law, expert evidence can
be admitted in criminal proceedings only if:

(1) the evidence would be likely to assist the jury or other fact-finding tribunal
(the Turner test);1

(2) the evidence is to be given by an individual who is qualified through study


or experience to give such evidence (that is, he or she is an expert in the
relevant field); and

(3) the expert is able to provide objective, impartial evidence.

4.3 On the third limb, we also explained that rule 33.2 of the Criminal Procedure
Rules 2010 now expressly provides that an expert has an overriding duty to give
opinion evidence which is objective and unbiased.

4.4 In Part 3 we explained that there was broad support amongst our consultees in
relation to both the nature of these common law requirements and the desirability
of bringing them together in a single code; and, as noted above, we introduced
our recommendation that these requirements should be codified alongside our
new reliability test.2

4.5 In the following paragraphs we therefore set out and explain the clauses in our
draft Criminal Evidence (Experts) Bill which would bring about codification.3

4.6 We recommend no change to the Turner test or to the requirement regarding the
need to demonstrate expertise. We do, however, recommend that there be an
explicit standard of proof (the balance of probabilities) in relation to the need to
demonstrate expertise. The question whether or not an individual claiming
expertise is indeed an expert is a matter which can and should be determined
according to a standard of proof. A fixed standard would provide the criminal
courts with the yardstick they require to determine the question, and it would
bring certainty, clarity and transparency to the law. Importantly, it would also put
the parties and the expert communities on notice that any individual claiming the

1
Following Turner [1975] QB 834, 841.
2
Paragraphs 3.125 to 3.137.
3
The new admissibility rule we recommend in respect of evidentiary reliability is explained in
Part 5.

48
status of an expert witness will not be able to provide expert evidence in criminal
proceedings unless and until it is established that he or she is in fact an expert.

4.7 Our recommendation for the impartiality requirement may be slightly different
from the common law position for criminal proceedings. There are few cases on
the question of expert impartiality for the law of criminal evidence, so in this
respect we have drawn on the common law as described by the Civil Division of
the Court of Appeal. We also believe that the duty currently set out in rule 33.2 of
the Criminal Procedure Rules 2010 is of such importance that it should be brought
into our draft Bill and that our impartiality requirement should be defined with
specific reference to it.

4.8 In short, we recommend that primary legislation should provide that expert
evidence is admissible in criminal proceedings only if:

(1) the court is likely to require the help of an expert witness (the
Turner test); and

(2) it is proved on the balance of probabilities that the individual


claiming expertise is qualified to give such evidence (the
qualification or expertise test).4

4.9 We also recommend that this legislation should provide that expert
evidence is inadmissible if there is a significant risk that the expert has not
complied with, or will not comply with, his or her duty to provide objective
and unbiased evidence, unless the court is nevertheless satisfied that it is
in the interests of justice to admit the evidence (the impartiality test).

4.10 It should be noted that we are not recommending a broader code which would
incorporate all aspects of the law on expert evidence in criminal proceedings.

4.11 In particular, we make no recommendation that the common law hearsay


exception described in rule 8 of section 118(1) of the Criminal Justice Act 2003
should be codified. This rule, which allows an expert witness to “draw on the body
of expertise relevant to his [or her] field”, will remain with the other hearsay
provisions of the 2003 Act.

THE TURNER TEST


4.12 Clause 1(1) of our draft Criminal Evidence (Experts) Bill provides that expert
evidence is admissible in criminal proceedings only if –

(a) the court is satisfied that it would provide 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 …

4.13 As explained above, this is a straightforward codification of the common law


Turner test. The common law authorities would therefore continue to guide the

4
Rule 33.3(1)(a) of the Criminal Procedure Rules 2010 already provides that an expert
witness’s written report must “give details of the expert’s qualifications, relevant experience
and accreditation”.

49
courts, but the courts would not be inhibited from revisiting the way the Turner
test has been applied in specific contexts if this were thought to be desirable. Our
test would not change the law, but it would permit the law to develop
incrementally as it has in the past.

4.14 The test merely requires that the court (the trial judge or magistrates’ court) be
“satisfied” that the expert evidence “is likely” to be outside a judge or jury’s
experience and knowledge and that the evidence would provide help. This is not
a requirement which would need to be proved.

THE QUALIFICATION (EXPERTISE) TEST


4.15 Clause 1(1) of our draft Criminal Evidence (Experts) Bill also provides that expert
evidence is admissible in criminal proceedings only if –

(b) the person who gives it is qualified to do so …

4.16 Clause 2(1) provides that, for the purposes of clause 1(1)(b), “a person may be
qualified to give expert evidence by virtue of study, training, experience or any
other appropriate means”. Clause 2(2) provides that, for that person’s evidence to
be admissible, the court “must be satisfied on the balance of probabilities that [he
or she] is so qualified”.

4.17 This standard of proof provides the minimum acceptable level of knowledge,
experience or skill which must be established before an individual can provide
expert evidence. The extent of the expert’s expertise beyond this minimum
threshold goes to weight. That is to say, all other things being equal, the greater
the level of expertise borne by an expert witness, the more likely it is that the fact-
finding tribunal will accept his or her evidence.5

4.18 Clause 2(1) has been drafted in this way for a number of reasons:

(1) it expressly sets out the traditional ways in which the equivalent common
law requirement has been satisfied;

5
The requirement that expertise be proved on the balance of probabilities accords with
section 54(2) of the Youth Justice and Criminal Evidence Act 1999 (on the determination of
general witness competence) and the equivalent requirement in s 123 of the Criminal
Justice Act 2003 (on a person’s capability to make a statement where such evidence is
tendered instead of live oral evidence). This standard is also consistent with our
understanding of the current law in para 2.7 above and, more generally, with what has
been described as “the default standard of proof for preliminary facts for all the parties” in a
number of common law jurisdictions; see R Pattenden, “The proof rules of pre-verdict
judicial fact-finding in criminal trials by jury” (2009) 125 Law Quarterly Review 79, 100.

50
(2) it would allow the courts to require a combination of study, training or
experience before expertise is recognised for some disciplines, in line
with the developing common law approach;6 and

(3) it directs the parties to consider adducing other evidence of expertise in


cases where the court is unlikely to be satisfied by evidence of “study,
training or experience”.

4.19 It should always be possible for a criminal court to determine on the balance of
probabilities whether or not a particular individual is qualified to provide expert
evidence of a particular type. In the vast majority of cases we would expect
expertise to be established, as it is now, by admissible evidence of study, training
or experience (or a combination of the same). For example, it is highly unlikely
that professionals with relevant experience, such as surgeons, psychiatrists,
accountants and engineers, would have to provide any more evidence than they
currently have to provide when called upon to demonstrate their expertise.

4.20 Exceptionally, however, evidence of study, training or experience may be


insufficient to prove the claim to expertise, as we intimated in paragraph 4.18(3).
If an individual claiming expertise is relying on a skill such as lip reading or ear-
print analysis which is unusual in the sense that evidence of study, training or
experience may be unavailable or, if it is available, the court may not be satisfied
that it suffices to discharge the requirement of proof in clause 2(2), the individual
concerned would need to be prepared 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.7

4.21 We therefore agree with the underlying thrust of the following comment made by
the Society of Expert Witnesses in response to our consultation paper:

6
See para 2.7 above. The trial judge may require relevant academic experience, practical
experience, forensic experience in the context of criminal investigations or proceedings, or
any combination of the same. The criminal courts should be left to develop their
requirements incrementally, depending on the nature of the evidence being tendered for
admission. The fact that (say) a forensic scientist has been certified as having expertise in
accordance with the International Organization for Standardization’s conformity
assessment standard ISO 17024 would be relevant in this context.
7
Given the clear terms of cl 2(2), the court might intimate or direct at a pre-trial hearing that
such results should be provided. Of course, if testing is undertaken, and the results provide
a quantifiable assessment of the witness’s skill, the witness’s evidence would be
inadmissible if the results show that he or she is as likely to be wrong as right, even if the
witness has been called by the defence. This follows from the fact that all parties will need
to prove their witnesses’ expertise on the balance of probabilities. Professor Mike
Redmayne expressed the point well when he suggested that if a defence expert is as likely
to be wrong as right “you might as well toss a coin”.

51
An expert’s performance in terms of giving the correct opinion can,
and perhaps should, periodically be tested against predetermined
standards. … The expert gives his or her opinion in a simulated,
blinded setting where the correct results are known and, by
comparison, the reliability of that expert’s opinion can be assessed in
terms of such measurements as the incidence of false positive and
false negative results. Predetermined standards of reliability can be
set and, if appropriate, periods of retraining and retesting can be
arranged.

4.22 This view also accords with the opinion of Associate Professor William O’Brian. In
his response to our consultation paper, he said that evidence such as handwriting
analysis and lip-reading “are methods that could be subjected to empirical testing
… Evidence that can be empirically tested should be subjected to such tests
before it is used as a basis for criminal convictions”.

4.23 There would be some cost implications, of course, in that testing would be
required in some circumstances, but this would not be an onerous burden for the
individuals concerned given the very basic nature of the testing required.8 In any
event:

(1) it would be a very small price to pay for the guarantee of reliability in the
round provided by proof that an individual claiming to have an unusual
skill actually has that skill, if called to give expert evidence in criminal
proceedings; and

(2) given what we say in paragraph 2.7 above, the courts should already be
requiring proof of expertise in cases where individuals relying on an
unusual skill are called to provide expert evidence.

4.24 Nor would an obligation to prove expertise with test results lead to any significant
delays. Because the obligation to prove expertise on the balance of probabilities
is expressly set out in clause 2(2) of our draft Bill, a witness claiming expertise
would be fully aware of the need to prove an unusual skill in advance of the trial,
if his or her “study, training [or] experience” (if any) would be unlikely to provide a
sufficient basis for determining the issue.

THE IMPARTIALITY TEST


4.25 Clause 1(1) of our draft Criminal Evidence (Experts) Bill also provides that expert
evidence is admissible in criminal proceedings only if –

(c) the evidence is not made inadmissible as a result of section 3


(impartiality).

4.26 Drawing on what is currently set out in rule 33.2(1) and (2) of the Criminal
Procedure Rules 2010, clause 3(1) of our draft Bill provides that an expert “has a
8
As the Society of Expert Witnesses pointed out, periodic testing should suffice to
demonstrate an ongoing skill. In addition, the results could be used in subsequent
proceedings. It should be noted that we are not seeking to impose an obligation that any
expert should undergo 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.

52
duty to the court to give objective and unbiased expert evidence”9 and clause 3(2)
provides that this duty “overrides any obligation to the person from whom the
expert receives instructions or by whom the expert is paid”.10

4.27 We have brought these provisions into our draft Bill because of their importance,
their relationship with our other impartiality provisions in clause 3 and, perhaps
more fundamentally, because the expert’s overriding duty does not sit easily in
rules which are concerned with procedure. Our draft provisions are, however,
slightly wider than the equivalent provisions in the Criminal Procedure Rules 2010.
Clause 3 encompasses all expert evidence, whether of fact or opinion,11 whereas
rule 33.2 of the Rules is limited to expert opinion evidence. We believe our wider
approach is desirable because it is possible to envisage a biased expert
deliberately providing misleading evidence of fact. An expert might, for example,
provide an incomplete explanation of what was observed or give only a partial
description of the data generated by an experiment.12

4.28 Clause 3(3) provides that, if it appears to the court that there is a significant risk
that an expert will not comply (or has not complied) with the duty in clause 3(1),
his or her expert evidence is inadmissible unless the court is satisfied that it is in
the interests of justice to admit it. This provision can therefore be broken down as
follows:

(1) the starting point is that there is a presumption of impartiality, based on


the common-sense assumption that expert witnesses generally tend to
comply with their duty to the court;

(2) this presumption of impartiality will be rebutted, however, if there is


sufficiently cogent evidence to suggest a significant risk that the expert
will not comply (or has not complied) with that duty;

(3) where it appears to the court that there is a significant risk of non-
compliance with that duty, there is a presumption of inadmissibility;

(4) but the expert’s evidence will be admitted if the court is satisfied that it is
in the interests of justice to admit it.

4.29 We expect that the presumption of impartiality will stand unchallenged in the vast
majority of cases, so in practice this subsection should affect only a small
minority of criminal cases.

4.30 We should explain, however, that clause 3(3) may be slightly different from the
current common law position for criminal proceedings. There are few reported
cases on biased expert witnesses in criminal proceedings, so the common law is
9
Rule 33.2(1) currently provides that an expert “must help the court … by giving objective,
unbiased opinion on matters within his [or her] expertise”.
10
Rule 33.2(2) currently provides that the expert’s duty in r 33.2(1) “overrides any obligation
to the person from whom he [or she] receives instructions or by whom he [or she] is paid”.
11
Clause 10.
12
We have not incorporated r 33.2(3) of the 2010 Rules into our draft Bill because there is no
good reason why this provision should be brought into primary legislation. Clause 3(5) of
our draft Bill provides the Criminal Procedure Rules Committee with the power to make
further provision in connection with the expert’s overriding duty to the court.

53
unclear in this context. In formulating our test we have therefore decided to build
on the duty in rule 33.2 of the Criminal Procedure Rules 2010; but we have also
included a degree of flexibility to reflect what the common law and procedural
rules require in civil proceedings, drawing on a judgment of the Court of Appeal
(Civil Division).13

4.31 This degree of flexibility – the discretion to allow expert evidence to be admitted
notwithstanding a significant risk of bias – will ensure that the interests of justice
are always satisfied, in accordance with the overriding objective of the Criminal
Procedure Rules 2010.14 It is important to understand, however, that if this
discretion is applied to allow a biased expert’s evidence to be admitted, it would
be open to the other party or parties to apply to have the evidence of bias placed
before the jury so that the expert’s evidence can be properly assessed at the end
of the trial.

4.32 An example of a situation where an expert’s evidence might be admitted, despite


a significant risk of bias, could be where the risk is relatively low (albeit
significant), the expert’s evidence would materially support the accused’s defence
(if the evidence is believed), evidentiary reliability is not disputed and there is a
dearth of alternative expert evidence for the accused to draw upon.

4.33 In accordance with the current position for criminal proceedings,15 the mere
appearance of bias would be insufficient to justify the exclusion of an expert
witness’s evidence. But of course if the facts underpinning the argument that
there is apparent bias allows an inference of actual bias to be drawn, such that it
appears to the court that there is a significant risk that the expert might not
comply with the duty in clause 3(1), then the presumption of inadmissibility would
arise.

4.34 It follows that clause 3 should not affect current practice whereby employees of a
police force or other investigatory agency (for example, drugs officers and
forensic accountants) are able to give expert evidence for the prosecution.
However, each case will turn on its own facts. If there is credible evidence of bias
on the part of a particular expert then the presumption of inadmissibility would
arise.

4.35 To ensure that the current practice referred to above is not affected, and that an
expert’s evidence will be inadmissible only if there is evidence of actual bias,
clause 3(4) provides as follows:

13
See R (Factortame Ltd) v Secretary of State for Transport, Local Government and the
Regions (No 8) [2002] EWCA Civ 932, [2003] QB 381 at [70], suggesting a similar approach in
civil proceedings governed by the Civil Procedure Rules. For the possible types of bias in civil
proceedings, see D Dwyer, “The causes and manifestations of bias in civil expert evidence”
(2007) 26 Civil Justice Quarterly 425.
14
Rule 1.1(1) provides that the “overriding objective” of the rules “is that criminal cases be
dealt with justly”.
15
Paragraph 2.11 above.

54
The fact that the expert has an association (for example, an
employment relationship) which could make a reasonable observer
think that the expert might not comply with [the duty in clause 3(1))]
does not in itself demonstrate a significant risk [of non-compliance
with that duty].

4.36 In other words, the mere fact of an association does not of itself allow an
inference to be drawn that there is a significant risk of non-compliance with the
duty. But the particular association may allow such an inference to be drawn (for
example, because a prosecution expert was heavily involved in leading the police
investigation, or a defence expert has previously been associated with a criminal
gang to which the accused belongs).

THE SCOPE OF CLAUSES 1 TO 3


4.37 These clauses would apply to all expert evidence tendered for admission in
criminal proceedings in England and Wales.16 Importantly, these provisions would
apply whether the expert evidence is to be given orally, in a written report or in
some other way.17

MONITORING COMPLIANCE WITH CLAUSES 1(1)(B) AND 2


4.38 In Part 1 of this report we summarised our concerns about the expert opinion
evidence given for the prosecution in the case of Clark.18 It will be remembered
that the expert in question, a paediatrician, provided unreliable statistical
evidence which he was not qualified to give.

4.39 In responding to our consultation paper, a number of our consultees emphasised


the need to ensure that experts stay within their area of expertise. Mr Justice
Treacy felt that this requirement should be part of the statutory test, and the
Forensic Science Regulator suggested that there should be a process to monitor
an expert’s evidence to prevent drift into other areas. Similar points were made
by the British Psychological Society, the Forensic Science Society, the Royal
Statistical Society, the Bar Law Reform Committee and an American judge.19 Our
consultees’ view in this respect also accords with a recommendation made by the

16
See paras 2.19 to 2.22 above. Consistent with the Criminal Justice Act 2003, cl 10
provides that “criminal proceedings” means “criminal proceedings to which the strict rules
of evidence apply” (which includes criminal trials and Newton hearings); see Bradley
[2005] EWCA Crim 20, [2005] 1 Cr App R 24 at [29] and [36]. A Newton hearing is a trial to
determine the facts if D pleads guilty, where there is a dispute as to the facts relevant to
sentencing. Clause 11(6) provides that the Act would extend to England and Wales only.
17
Clause 10 provides that “expert evidence” includes “all such evidence, in any form and
however given”. The admissibility requirements would apply if the parties agree to the
admission of an expert’s hearsay evidence (under s 9 of the Criminal Justice Act 1967 or
s 114(1)(c) of the Criminal Justice Act 2003), but the judge would be unlikely to rule
against the parties in such a case.
18
[2003] EWCA Crim 1020, [2003] 2 FCR 447, para 1.5 above.
19
The Hon Theodore R Essex.

55
Hon Stephen Goudge in his 2008 report following an enquiry into paediatric
forensic pathology in Ontario.20

4.40 Criminal courts in England and Wales are already under a tacit (and ongoing)
duty to monitor expert witnesses’ evidence to prevent drift because an expert
witness can provide expert evidence only insofar as the common law admissibility
requirements for such evidence are satisfied. Equally, under our proposed
alternative an expert witness would be able to provide expert evidence in criminal
proceedings only to the extent permitted by the provisions in our draft Bill.

4.41 Be that as it may, we agree with the view of our consultees, and the
recommendation of Stephen Goudge, that it would be sensible to have this tacit
duty expressly set out in legislation. An explicit provision requiring the court to
rule on the scope of an expert witness’s expertise would have the effect of
priming the trial judge (or magistrates) in advance of the expert’s testimony to do
what is necessary to ensure that the expert stays within his or her field. The judge
or magistrates would monitor the expert’s evidence and intervene to prevent the
expert from straying outside that field.

4.42 In the absence of such intervention, a Crown Court judge would nevertheless be
able to ameliorate any problem by subsequently providing a direction to the jury
to disregard expert evidence the witness was not qualified to provide.
Alternatively, if exceptional circumstances warranted a more robust approach, the
judge could discharge the jury from its obligation to return a verdict.

4.43 Importantly, however, if the judge monitors the expert’s testimony, and ensures
that the expert’s evidence relates solely to matters within his or her relevant area
of expertise, the risk of a problem and the possibility of an expensive re-trial
would be correspondingly reduced.

4.44 We therefore believe that the Criminal Procedure Rules 2010 should include a
provision to the effect that the trial judge (or magistrates) must rule on the scope
of the expert witness’s expertise before he or she testifies (in line with the
requirements of clauses 1(1)(b) and 2) and then monitor the position to ensure
that the expert witness does not give evidence on matters outside his or her area
of expertise. A rule of this sort would go some way towards ensuring that weak,
tangential evidence given by an eminent figure, and which for that reason might
be accepted at face value, would not be heard by the jury.21

4.45 In addition, we believe a further safeguard would be provided in this context if an


expert witness were to be required, while in the witness box but before testifying,

20
The Goudge Report, vol 3, p 475, Recommendation 129: “When a witness is put forward to
give expert scientific evidence, the court should clearly define the subject area of the
witness’s expertise and vigorously confine the witness’s testimony to it.”
21
Referring again to the facts of Clark [2003] EWCA Crim 1020, [2003] 2 FCR 447, the judge
would presumably have ruled that the paediatrician was qualified to provide an opinion
only on matters relating to children’s health. Accordingly, the expert would not have been
permitted to drift from his area of expertise to provide an opinion on statistical analysis.
Counsel for the prosecution would not have asked him about such matters; and if through
an oversight the expert was asked questions beyond his remit, defence counsel would
have intervened to raise the matter before the judge to prevent drift or the judge would
have raised the matter himself.

56
to read to him or herself the part of his or her report where it is stated that he or
she understands the expert’s overriding duty to the court.22

4.46 We therefore recommend that the Criminal Procedure Rules be amended to


include the following additional requirements:

(1) before giving oral evidence, an expert witness should be referred to


his or her overriding duty to give expert evidence which is

(a) objective and unbiased, and

(b) within his or her area (or areas) of expertise;

(2) the trial judge or magistrates’ court should rule on the expert
witness’s area (or areas) of expertise before he or she gives
evidence and monitor the position to ensure that he or she does not
give expert evidence on other matters.

22
See Criminal Procedure Rules 2010, r 33.3(1)(i).

57
PART 5
EVIDENTIARY RELIABILITY

INTRODUCTION
5.1 In this Part we develop the recommendations we introduced in Part 3 on our
proposed reliability test for expert opinion evidence, and on the power which
would permit a trial judge not to apply it.

5.2 We also explain the factors we believe should be in our draft Criminal Evidence
(Experts) Bill, to provide trial judges with guidance on how evidentiary reliability is
to be assessed; and we explain our view on how a ruling on evidentiary reliability
should be addressed on appeal.

THE RELIABILITY TEST


5.3 As we explained in Part 3, there was very broad (but not universal) support for a
new reliability test for expert opinion evidence along the lines proposed in our
consultation paper.

5.4 In the light of the comments we received during the consultation process, our
twofold recommendation for criminal proceedings is that:

(1) there should be a statutory admissibility test which would provide that an
expert’s opinion evidence is admissible only if it is sufficiently reliable to
be admitted;1 and

(2) there should be a statutory provision to the effect that, if there is any
doubt on the matter, expert evidence presented as evidence of fact
should be treated as expert opinion evidence.2

5.5 The test set out in our consultation paper, and repeated in paragraph 3.6(2)
above, was intended to describe in broad terms the factors which ought to be
addressed when considering evidentiary reliability, particularly for evidence of a
scientific nature, but it was not draft legislation. It is now necessary to consider
how the new test is to be formulated for an Act of Parliament, given that the test
must encompass, potentially, every conceivable type of expert opinion which
might be proffered for admission in criminal proceedings. What is required is a
genuinely universal test.

5.6 In other words, while the statutory test must be framed so that it encompasses all
forensic scientific opinion evidence, and so must make reference to the factors
set out in paragraph 3.6(2), it must also be broad enough to encompass,
potentially, all other types of expert opinion evidence. Examples of non-scientific
expert opinion evidence which might be proffered for admission in criminal
proceedings are an academic lawyer’s opinion on the legal position in a foreign
jurisdiction, a lip-reader’s opinion on what was said by another individual on a

1
Paragraph 3.36 above; see cl 1(2) of our draft Bill.
2
Paragraph 3.39 above; see cl 1(3) of our draft Bill.

58
particular occasion and a literary expert’s opinion on whether an ostensibly
obscene novel is justified “in the interests of literature, art or learning”.3

5.7 A general test requiring reference to whether an expert’s opinion evidence is


predicated on “sound principles, techniques and assumptions” (paragraph
3.6(2)(a) above) is certainly apt to cover scientific evidence, and perhaps the
opinion evidence of an expert on foreign law, but it is probably too narrow to
capture all the myriad types of expert opinion evidence which might need to be
addressed for reliability in criminal proceedings.

5.8 The practical likelihood is that most, if not all, opinion evidence scrutinised for
reliability against the test in paragraph 3.6(2) would indeed be encompassed by
it. However, that likelihood cannot justify a formulation which would not
necessarily work for the whole of its potential range of application. In short, the
reliability test must be one which can in principle be applied to anything within the
range it purports to cover, even if unlikely ever to arise in practice.

5.9 In our draft Bill we have therefore opted for a test which is truly universal, but as a
result it is also a little vaguer than the test described in our consultation paper.
Our core reliability test, which is set out in clauses 1(2) and 4(1) of the Bill, is
therefore supplemented by a number of statutory examples in clause 4(2) which
demonstrate, in broad terms, some of the types of expert opinion evidence which
would not be sufficiently reliable to be admitted. To put it another way, the
subsection lists key justifications for ruling that expert opinion evidence is
insufficiently reliable to be admitted in criminal proceedings.

5.10 The admissibility test and these examples are further supplemented by a list of
lower-order factors in Part 1 of the Schedule to the Bill, drawing on the two sets
of guidelines we originally proposed in our consultation paper.

5.11 The examples in clause 4(2) and the further factors listed in Part 1 of the
Schedule direct the trial judge to matters which have a bearing on the question of
evidentiary reliability in a particular case; but, more than that, they also explain
what the reliability test means for the type of expert evidence being proffered for
admission.

5.12 Importantly, in line with a provisional proposal in our consultation paper, our draft
Bill directs the trial judge to consider not only the information supporting the
expert’s opinion evidence (for example, whether an underlying scientific
hypothesis has been sufficiently scrutinised, and whether the expert has properly
taken into account all relevant evidence in the instant case) but also the strength
of the opinion underpinned by that information.

5.13 It will always be the particular opinion evidence proffered for admission which has
to be scrutinised for reliability. The judge must therefore look at the general
foundation material, the extent to which relevant case-specific matters were taken
into consideration by the expert, the legitimacy and logic of the expert’s reasoning
process in coming to his or her opinion and whether the sort of opinion the expert

3
Obscene Publications Act 1959, s 4(1) and (2).

59
wishes to give, including its strength, can be objectively justified, bearing in mind
the uncertainties inherent in the foundation material.

5.14 In short:

(1) we have taken forward our provisional admissibility test, and the
guidelines we originally proposed to assist in its application, but we have
reformulated the test to ensure that it is truly universal;

(2) we have drawn out from our original guidelines some key higher-order
examples of when expert opinion evidence is likely to be unreliable;

(3) we have supplemented our new admissibility test and those key
examples with further, lower-order factors which will provide trial judges
with specific guidance (where relevant); and

(4) we have ensured flexibility by ensuring that trial judges are neither bound
by nor limited to the factors expressly listed in the Schedule.4

5.15 It bears repeating that evidentiary reliability means the reliability of relevant
opinion evidence given by a qualified, impartial expert, so the court is directed to
consider factors in the Schedule which, in the main, are not related to the expert
witness him or herself.

5.16 The reliability test in clauses 1(2) and 4(1) of our draft Bill, the statutory examples
in clause 4(2) and the various lower-order factors in the Schedule describe
matters which have a bearing on reliability, but this does not mean that all
subjective matters are irrelevant. Forensic scientific opinion evidence depends on
a scientific underpinning and a subjective interpretive element. Whether the
expert in question has the general interpretive skill required to provide expert
evidence is governed by clauses 1(1)(b) and 2; but whether that skill has been
properly applied in the instant case, so as to draw reliable inferences and reach
reliable conclusions, must be addressed under the reliability test, against
objective standards.5

5.17 We therefore recommend for criminal proceedings:

(1) a statutory provision in primary legislation which would provide


that expert opinion evidence is admissible only if it is sufficiently
reliable to be admitted;6

4
Clause 4(3) of our draft Bill provides that the court must have regard to “such of the
generic factors set out in Part 1 of the Schedule as appear to the court to be relevant” and
“anything else which appears to the court to be relevant”.
5
See cl 4(2)(e) of the draft Bill and para 1(h) of Part 1 of the Schedule.
6
Draft Bill, cl 1(2). As explained already, cl 1(3) provides that if there is a doubt as to
whether an expert’s evidence is evidence of fact or opinion evidence, it is to be taken to be
opinion evidence.

60
(2) a provision7 which would provide our core test that expert opinion
evidence is sufficiently reliable to be admitted if8 –

(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;9

(3) a provision10 which would set out the following key (higher-order)
examples of reasons why an expert’s opinion evidence is not
sufficiently reliable to be admitted:11

(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 stand up to scrutiny;

(b) the opinion is based on an unjustifiable assumption;

(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;

(e) the opinion relies on an inference or conclusion which has


not been properly reached;12

(4) a provision which would direct the trial judge to consider, where
relevant, more specific (lower-order) factors in a Schedule to the
Act and to any unspecified matters which appear to be relevant.13

7
Draft Bill, cl 4(1).
8
In context this means “only if”, but it is unnecessary to spell it out explicitly.
9
For this aspect of the test (cl 4(1)(b) of our draft Bill), the expert’s opinion should be
expressed with no greater degree of precision or certainty than can be justified by the
underlying material on which it depends. The material includes relevant general matters
(such as scientific hypotheses) and relevant evidence in the particular case.
10
Draft Bill, cl 4(2).
11
This test is framed in this negative way to accord with the procedural provisions in cl 6.
12
This example (clause 4(2)(e) of our draft Bill) addresses the reasoning process of the
expert and the use of any subjective interpretive skill (see also para 1(h) in Part 1 of the
Schedule). The question whether an expert has the necessary interpretive skill to give an
opinion is governed by the requirements of cl 1(1)(b) and cl 2; see paras 4.15 to 4.24
above.
13
Draft Bill, cl 4(3).

61
5.18 The requirements set out in clauses 1 and 4 of our draft Bill would apply to
criminal proceedings in England and Wales;14 and they would apply whether the
expert evidence is to be given orally, in a written report or in some other way.15

Specific factors in the Schedule


5.19 We have already introduced our recommendation that trial judges should be
provided with a single list of generic factors to help them apply the reliability limb
of our proposed admissibility test. A judge would be directed to take into
consideration factors which are relevant to the opinion evidence under
consideration and any other factors he or she considers to be relevant.16

5.20 We now turn to the question of what the list of generic factors should include. A
large number of individual and corporate consultees endorsed our guidelines (or
guidelines of the sort we proposed); and the UK Register of Expert Witnesses
told us, following their own internal consultation, that our guidelines drew broad
support from the experts who responded. Nevertheless, a number of consultees
took issue with some factors we included, or failed to include, in the versions we
provisionally set out in our consultation paper. Five principal concerns were
expressed:

(1) some consultees queried the value of peer-reviewed publications and


literature without explicit reference to the quality of such publications;17

(2) several consultees queried or opposed the inclusion of subjective matters


relating to the witnesses’ expertise (such as his or her qualifications,
experience and standing);18

(3) some consultees thought we should have included a reference to


membership of a relevant professional body or reputable organisation;19

14
Clause 10 of our Draft Bill provides that “criminal proceedings” means criminal proceedings
to which the strict rules of evidence apply (which includes criminal trials and Newton
hearings); see Bradley [2005] EWCA Crim 20, [2005] 1 Cr App R 24 at [29] and [36].
Clause 11(6) provides that the Bill, if in force, would extend to England and Wales only.
15
Draft Bill, cl 10. The admissibility requirements in our Bill would apply in principle to an
expert report even if the parties themselves were to agree to the admission of such
evidence (under s 9 of the Criminal Justice Act 1967 or s 114(1)(c) of the Criminal Justice
Act 2003), but the judge would be unlikely to rule against the parties in such a case.
16
See paras 3.62 and 3.63, and paras 5.10 to 5.16 and 5.17(4) above.
17
These consultees included Dr Malcolm Park (University of Melbourne), the Forensic
Science Regulator and Professor Mike Redmayne. Some consultees queried the value of
any peer review.
18
Northumbria University School of Law’s Centre for Criminal and Civil Evidence and
Procedure, Gary Pugh (Director of Forensic Services, Metropolitan Police), Professor Mike
Redmayne, the Society of Expert Witnesses and the Bar Law Reform Committee.
19
The British Association for Shooting and Conservation and the Expert Witness Institute.

62
(4) some consultees wished to see a reference to organisational structures
or the principles of “balance, logic, robustness and transparency” which
underpin the operational practices of many forensic science providers;20
and

(5) one consultee favoured a weighted hierarchy of factors.21

5.21 On the first point, we explained in our consultation paper that the value of peer
review has often been called into question,22 but we also took the view that it is
one guiding factor, amongst others, which a trial judge should consider in the
context of any purportedly scientific element underpinning an expert’s opinion
evidence.

5.22 This is still our view, so long (of course) as the peer-review process is credible.
As the Criminal Bar Association said in their response, peer review:

provides for a long term and objective opportunity to test and refute
the theory and practice of the technique. It also evidences the proper
willingness of its proponents to subject their ‘project’ to outside
scrutiny and criticism.

5.23 The credibility of any particular peer review process was not expressly stated in
our original guidelines, for we assumed (as we still do) that trial judges would be
able to attach appropriate weight to this factor without explicit guidance.

5.24 On the second point, we have now expressly set out “impartiality” and being
“qualified” to provide expert evidence as separate limbs in our proposed
admissibility test for expert evidence, and have limited the reliability test in clause
4(1) of our draft Bill so that it requires an objective assessment of whether an
expert’s opinion evidence is sufficiently reliable to be admitted. This being the
case, we are now satisfied that extraneous subjective factors (that is, matters
relating to expertise and impartiality) can and should be removed from our
reliability limb and from the factors in the Schedule to our Bill. It would serve no
useful purpose if the judge were to be directed to apply separate but overlapping
tests.

5.25 The approach we have adopted in our draft Bill should avoid unnecessary
distractions (even more so if the subjective factors would be difficult to
establish);23 and it has brought the added benefit of shortening the list of factors.

20
Professor Wesley Vernon (a podiatrist with a particular interest in forensic identification),
Gary Pugh (Director of Forensic Services, Metropolitan Police), the UK Accreditation
Service, Skills for Justice and LGC Forensics.
21
Dr Geoffrey Morrison, a researcher on forensic voice comparisons (Australian National
University) said that an objectively verifiable analysis should take precedence over
subjective factors.
22
Consultation Paper No 190, paras 4.61 to 4.63.
23
As one of our consultees, Dr Keith JB Rix, mentioned in relation to our original reference to
“standing in the community”, this issue could be difficult to determine because there may
be experts in senior positions with a poor reputation amongst their peers.

63
5.26 So, when a trial judge addresses the reliability limb of our admissibility test, he or
she should focus on the validity of the material, processes and reasoning
underpinning the expert’s opinion evidence.24 Subjective factors, though relevant
to reliability in the round,25 should be addressed under clauses 2 and 3.

5.27 It follows that it would not be desirable to include references to other extraneous
matters such as the expert’s membership of a relevant professional body or the
particular organisational structures within a forensic science laboratory (the third
and fourth of the five points listed at paragraph 5.20 above). A sound
organisational framework may well be a factor the judge will wish to take into
account in an appropriate case, and the open-ended nature of the list of factors
would permit this (where relevant). However, because general organisational
structures do not necessarily enhance the reliability of a particular expert’s
opinion evidence, and because the inclusion of such a factor could give rise to an
expectation that the judge should undertake an investigation into a collateral
matter which may have little bearing on the reliability of the evidence in issue, we
are not persuaded that organisational structures should be expressly included in
the list of factors. The judge should focus on the processes and the reasoning
underpinning the particular expert opinion evidence which has been proffered for
admission.26

5.28 On the fifth point, we see no need for a hierarchy of factors, particularly now that
we have abandoned subjective factors and recognise the desirability of
flexibility.27

5.29 When formulating the factors in our Schedule to the draft Bill, we also asked
ourselves whether trial judges should be directed to consider if a judge in other
proceedings had, after due enquiry, previously:

(1) admitted or excluded an opinion provided by the expert on the ground


that his or her opinion was, or was not, sufficiently reliable to be admitted,
or

(2) ruled that the underlying basis of the expert’s opinion was, or was not,
sound (to the extent required by the opinion proffered in that case).

5.30 We initially took the view that a properly articulated ruling in a previous case,
following an investigation into evidentiary reliability, was something the judge
applying the reliability test ought to be directed to take into consideration in a
case where the same issue arises, and be given whatever weight the judge
considered appropriate. We saw this as a possible way of ensuring that the same
evidentiary basis of an expert opinion would not have to be re-assessed over and
over again at first instance until the matter was definitively resolved by an
appellate court. This approach would save time and other resources.

24
On “validity” as opposed to “reliability”, see Consultation Paper No 190, p 34, fn 51.
25
Paragraph 2.17 above.
26
Equally, there is no reference in our factors to the expert’s membership of a professional
body.
27
Paragraph 3.50 above.

64
5.31 We accept that another judicial ruling, after due enquiry, is something which a
trial judge may find useful when addressing the reliability of an expert’s evidence
proffered for admission in his or her case (depending on the facts); and we also
accept that referring to a previous ruling would be a useful way of saving
resources, particularly if it relates to the level of progress in a particular scientific
discipline.28

5.32 However, we have come to the conclusion that it is unnecessary, and potentially
problematic, to have a factor of this sort expressly included in the Schedule to our
draft Bill. We would not wish to create any expectation that a judge should have
to consider other rulings on evidentiary reliability or an expectation that a costly
new system should be established to record, transcribe, store and index such
rulings, particularly if many of the rulings would never be looked at again on
account of their case-specific content.

5.33 No doubt a ruling on general matters which other judges would find useful is likely
to be reported in a journal such as the Criminal Law Review or the Journal of
Criminal Law, regardless of whether there is an explicit factor in our guidelines.
Practitioners need no formal direction to encourage them to bring useful Crown
Court and other first instance rulings to the attention of the wider legal
community.

5.34 On balance, therefore, we have concluded that it would be best not to include an
explicit factor in our Schedule in respect of other judicial rulings. Clause 4(3)(c) of
our draft Bill provides that judges must have regard to anything not listed “which
appears to the court to be relevant”. This gives a trial judge sufficient flexibility to
decide whether to consider another judge’s ruling, should a transcript be
available and brought to his or her attention; but there would be no expectation
that the judge should have to consider another ruling, or search for any such
ruling; and there would be no expectation that judicial rulings on evidentiary
reliability should be routinely recorded, transcribed and reported.29

5.35 In the light of the foregoing, we recommend that a trial judge who has to
determine whether an expert’s opinion evidence is sufficiently reliable to be
admitted should be directed to have regard to:

(1) the following factors (insofar as they appear to be relevant):30

28
For our consultees’ concerns on this issue, see para 3.20 (and fn 23 to para 3.17) above.
29
A further problem arises from the difficulty of drafting a provision of this sort, as it would
have to sit within an existing framework of binding precedent. That is to say, the factor
would have to direct the judge to take into account other judicial rulings (where relevant)
which are not binding, but exclude judgments of other courts which would in any event bind
the judge. The resulting provision would have been unnecessarily complex.
30
Draft Bill, cl 4(3)(a) and Part 1 of the Schedule to the Bill. Some of the factors in the
Schedule are already required by r 33.3(1) of the Criminal Procedure Rules 2010 as
matters which must be included in an expert’s written report.

65
(a) the extent and quality of the data on which the expert’s
opinion is based, and the validity of the methods by which
they were obtained;31

(b) if the expert’s opinion relies on an inference from any


findings, whether the opinion properly explains how safe or
unsafe the inference is (whether by reference to statistical
significance or in other appropriate terms);

(c) if the expert’s opinion relies on the results of the use of any
method (for instance, a test, measurement or survey),
whether the opinion takes proper account of matters, such
as the degree of precision or margin of uncertainty, affecting
the accuracy or reliability of those results;

(d) the extent to which any material upon which the expert’s
opinion is based has been reviewed by others with relevant
expertise (for instance, in peer-reviewed publications), and
the views of those others on that material;

(e) the extent to which the expert’s opinion is based on material


falling outside the expert’s own field of expertise;32

(f) the completeness of the information which was available to


the expert, and whether the expert took account of all
relevant information in arriving at the opinion (including
information as to the context of any facts to which the
opinion relates);33

31
Rule 33.3(1)(b) to (e) of the Criminal Procedure Rules 2010 sets out some of the matters
which an expert witness’s written report must currently contain, including: “details of any
literature or other information which the expert has relied on”; “a statement setting out the
substance of all [material] facts given to the expert”; and a summary of the “findings”
(following an examination, measurement or test) on which the expert witness relies.
32
Rule 33.3(1)(d) of the Criminal Procedure Rules 2010 currently requires an expert witness
to “make clear” in his or her written report “which of the facts stated in the report are within
the expert’s own knowledge”.
33
Factor (f) provides an expectation that the expert should be provided with the relevant
contextual material, even if the emotive nature of some such information might give rise to
unconscious bias (as to which, see I Dror and S Cole, “The vision in ‘blind’ justice: Expert
perception, judgement and visual cognition in forensic pattern recognition” (2010) 17
Psychonomic Bulletin & Review 161). We recognise that procedures may need to be
introduced in forensic scientific laboratories to ensure that tangential information which is
likely to give rise to significant unconscious bias should be kept back from scientific
experts, or training provided to reduce the risk of such bias, but this is not a concern we
can address in our Bill. The Forensic Science Regulator has informed us that this is a
problem he will be seeking to overcome. Once a protocol has been established to minimise
the risk of unconscious bias, it may be necessary for factor (f) to be amended, for example
by incorporating “appropriate” before “information as to the context”.

66
(g) whether there is a range of expert opinion on the matter in
question; and, if there is, where in the range the expert’s
opinion lies and whether the expert’s preference for the
opinion proffered has been properly explained;34

(h) whether the expert’s methods followed established practice


in the field; and, if they did not, whether the reason for the
divergence has been properly explained; 35

(2) approved factors, if any, for assessing the reliability of the


particular type of expert evidence in question (insofar as they
appear to be relevant);36 and

(3) any other factors which appear to be relevant.37

5.36 It should be remembered that the factors in Part 1 of the Schedule do not stand
alone. They must be read in conjunction with the admissibility test in clause 4(1)
and the higher-order examples in clause 4(2); but, more than that, they must also
be read with our recommendation that the party proffering the evidence must
show that it is sufficiently reliable to be admitted in accordance with clauses 1(2)
and 4(1).38 This means that the expert wishing to give opinion evidence which is
being assessed for reliability will need to provide a reasoned explanation as to
why his or her opinion is sound. In tandem with factor (h), this obligation to
provide a reasoned explanation would be particularly useful for expert opinion
evidence which cannot be scientifically tested for reliability, or which it would not
be reasonably practicable to test in a scientific way. In addition, according to
Professor Nigel Eastman, a member of our working group for this project, this
combination summarises “particularly well what should be the approach to
medical evidence which is psychiatric in nature”.

5.37 We refer to the “trial judge” in the opening words of paragraph 5.35 because we
believe, as we explain below, that where the question arises in a magistrates’
court it should generally be determined by a District Judge (Magistrates’ Courts),
who would ordinarily be expected to try the case.39 Robust pre-trial case-

34
This information must be included in an expert witness’s written report; see r 33.3(1)(f) of
the Criminal Procedure Rules 2010. It is implicit in this factor that, where an expert
witness’s opinion is at variance with the opinions of most experts in the field, the expert
witness will be saying that the general opinion held by other experts is flawed (or, in the
context of scientific evidence, that it is not scientifically valid).
35
This factor makes the obvious point that if an expert is relying on a novel approach he or
she should explain why an opinion founded on it is sound. This factor should not in any
way be understood as a presumption against the admission of expert opinion evidence
based on new or nascent developments in science and technology.
36
Draft Bill, cl 4(3)(b) and Part 2 of the Schedule to the Bill. On this aspect of our
recommendations, see paras 3.53 to 3.59 above.
37
Draft Bill, cl 4(3)(c).
38
For our recommendation, see para 3.88 above; for the corresponding provisions in our
draft Bill, see cl 6(2) and (3).
39
Draft Bill, cl 7(1). On the importance of the pre-trial process and the desirability of having
an experienced judge who would address the pre-trial issues and try the case, see
Henderson and others [2010] EWCA Crim 1269, [2010] 2 Cr App R 24 at [204].

67
management will therefore be required in magistrates’ courts, supported by the
use of sanctions as to the payment of wasted costs, to ensure that the parties
and their legal representatives properly identify the issues and give notice of any
intention to challenge evidentiary reliability so that the case can be allocated to a
District Judge in advance of the trial and a pre-trial hearing arranged in good
time.40

5.38 The District Judge would ordinarily determine evidentiary reliability before the
trial,41 provide a ruling on admissibility – a written ruling, we suggest – and try the
case. However, we accept that magistrates’ courts should have the power to
allocate the trial to a bench of lay magistrates, if it is unnecessary for a District
Judge to continue with the case.42

5.39 Effective pre-trial case management and the real possibility of a wasted costs
order being made against a dilatory legal representative or party, or both, should
mean that in the vast majority of cases the magistrates’ court will have sufficient
notice of a prospective challenge to be able to allocate the case to a District
Judge for a pre-trial hearing. Inevitably, however, there will be some late
challenges. A challenge may be brought to a magistrates’ court’s attention for the
first time at or just before the trial, or even during the trial itself. This may be due
to an oversight, lost papers, confusion caused by a change of legal
representative, or because of fresh information having come to light at a late
stage in the proceedings. It follows that a bench of lay magistrates may on
occasion be confronted with a challenge to the evidentiary reliability of an
expert’s opinion evidence under clause 4 of our draft Bill. In this very rare
situation it may be necessary for the trial to be stopped and re-allocated to a
District Judge, but this should not be an inflexible rule. On account of the
particular bench’s wealth of legal and non-legal experience and the nature of the
evidence proffered for admission, it may be that the bench, aided by its legal
adviser, is properly able to address the issue itself.

5.40 Accordingly, clause 7(3)(b) of our draft Bill provides the Criminal Procedure Rules
Committee with the power to make rules which would permit a trial before a lay
bench to be stopped and allocated to a District Judge (the likely course of action
if a late challenge is made just before or soon after the trial starts). But clause
7(3)(a) (in tandem with clause 7(4)) also permits rules to be made which would
allow a lay bench to address evidentiary reliability and hear the trial if it would be

40
Section 19A of the Prosecution of Offences Act 1985 allows a wasted costs order to be
made against a legal representative as a result of any unreasonable act or omission.
Accordingly, an unreasonable failure to comply with the case-management process on the
part of a legal representative would give rise to the possibility of a wasted costs order
against that representative. Section 19 of the 1985 Act sets out the test for the parties
themselves, requiring an “unnecessary or improper act or omission”.
41
Draft Bill, cl 7(2) and (6).
42
The legal committee of HM Council of District Judges (Magistrates’ Courts) told us, in a
letter supporting our recommendations, that the District Judge determining evidentiary
reliability should provide his or her ruling in writing and that he or she should try the case
“where possible”, save that there should be a discretion as to the subsequent allocation of
the trial. Clause 7(3) to (5) of our Bill provides the Criminal Procedure Rules Committee
with the power to create the necessary procedural rules.

68
in the interests of justice to adopt this course of action.43 It would clearly be in the
interests of justice for a lay bench to retain jurisdiction if a late challenge is made
during the trial, most of the evidence has already been adduced and the
magistrates are competent to deal with the challenge.

5.41 It is important to explain that we recommend that the question of evidentiary


reliability be addressed in magistrates’ courts by District Judges not because we
believe lay magistrates would be unable to apply the reliability test, but simply
because it would be more cost-effective to train only the professional judges and
magistrates’ legal advisers for this particular task. A bench of magistrates,
assisted by its legal adviser, would always be expected to determine whether the
reliability test needs to be applied and would also need to be familiar with the
provisions in the Bill to the extent necessary for effective pre-trial case-
management. Magistrates would be equipped to deal with these matters from
their ongoing training. However, we believe it would be too costly to train all
magistrates on how to determine evidentiary reliability, particularly as credible
challenges to the evidentiary reliability of expert opinion evidence are unlikely to
be that common in summary proceedings.44

A limited power to disapply the reliability test


5.42 In paragraph 3.77 above we introduced the following recommendation:

(1) criminal courts should have a limited power to disapply the reliability test
so that it does not have to be applied routinely and unnecessarily;

(2) but, equally, the power to disapply must not be such that the reliability
test becomes only a nominal barrier to the adduction of unreliable expert
opinion evidence.

5.43 In line with the traditional adversarial approach which applies in criminal
proceedings, we take the view that the party opposing the admissibility of an
expert’s opinion evidence on the ground of insufficient reliability should formulate
a sound, reasoned argument to explain why there is a need to investigate
evidentiary reliability, at least as a general rule. This would not be a burden to
adduce evidence of unreliability, but it would be an obligation to explain why it
would be inappropriate to presume that the opinion evidence is sufficiently
reliable to be admitted.45 If this mere burden of reasoned objection is discharged,
the burden of establishing reliability to the required standard would be borne by
the party wishing to adduce the evidence. This is consistent with what the Court
of Appeal recently said in Reed:46 “unless the admissibility is challenged, the

43
The same approach would apply whether it is a trial following a not guilty plea or a Newton
hearing to determine the factual basis of a guilty plea (for the purposes of sentencing); see
cl 7(6).
44
The legal committee of HM Council of District Judges (Magistrates’ Courts) told us it
believed that most expert evidence would not be objected to in summary proceedings.
45
The current laissez-faire approach to admissibility may properly be regarded as the
recognition of a presumption that expert opinion evidence is sufficiently reliable to be
admitted, with an expectation that any concerns about reliability will be revealed during the
trial.
46
[2009] EWCA Crim 2698, [2010] 1 Cr App R 23 at [113].

69
judge will admit that evidence. That is the only pragmatic way in which it is
possible to conduct trials … . However, if objection to the admissibility is made,
then it is for the party proffering the evidence to prove its admissibility.”

5.44 We use the phrase “mere burden of reasoned objection” because the burden on
the challenging party to formulate an argument as to unreliability should be set
quite low. This would accord with the general principle in the law of evidence that
it is for the party tendering evidence for admission to show that the necessary
admissibility preconditions are satisfied, not for the party opposing the adduction
of evidence to show that they are not satisfied. And we say that this burden
should rest with the challenging party “as a general rule” because the trial judge
should be the person with ultimate control over the proceedings. That is to say,
the judge should have the power to require an investigation into evidentiary
reliability even if a challenge is not made by a party, as a further safeguard
against the adduction of expert opinion evidence which should not be admitted.47

5.45 So, in the absence of a reasoned objection from another party suggesting to the
court that the expert opinion evidence might not be sufficiently reliable to be
admitted, it would ordinarily be presumed that the expert opinion evidence
satisfies the threshold reliability test for admissibility.48 This weak presumption
would cease to operate, however, if a party were to discharge the burden of
reasoned objection and demonstrate to the court that the evidence might not be
sufficiently reliable to be admitted. But as a further safeguard the judge would in
all cases have a discretion to disapply the presumption, whether or not a
challenge has been made by another party.

5.46 Either way, once the presumption ceased to operate it would be for the party
seeking to adduce the evidence to demonstrate that it was sufficiently reliable to
be admitted. The judge would direct that there be a hearing to resolve the matter,
unless the matter could be properly resolved without a hearing.49 A hearing would
not need to be held, for example, if the opinion evidence was patently unreliable

5.47 The consensus amongst our consultees was that the matter should be decided at
a hearing before the trial if possible, but there must always be some flexibility to
ensure that the reliability of an expert’s evidence could be challenged during the
trial if necessary. Accordingly, a hearing before the trial, as part of the pre-trial
case-management process, would be the default position, but the judge would be

47
For similar powers in relation to the admissibility of confessions, see ss 76(3) and 76A(3)
of the Police and Criminal Evidence Act 1984.
48
If the challenging party could not discharge this burden of reasoned objection in advance
of the trial, the expert opinion evidence in question would ordinarily be admitted.
Exceptionally, however, the challenging party might be able to re-open the matter during
the trial if new information suggesting unreliability were to come to light.
49
Hearings to address evidentiary reliability would not necessarily be “preparatory hearings”
under Part III of the Criminal Procedure and Investigations Act 1996 (as to which, see
Consultation paper No 190, Appendix B).

70
able to hold a hearing during the trial, in the absence of the jury, if it was either
necessary or appropriate in the case being tried.50

5.48 A hearing during the trial would be necessary if, exceptionally, new evidence
came to light during a lengthy trial suggesting that the admissibility of expert
opinion evidence should be investigated at that late stage. As the Crown
Prosecution Service (and some other consultees) pointed out, provision must be
made for the possibility that additional experts will become available or that new
developments will come to light during a lengthy trial.

5.49 A hearing during the trial might be appropriate in a case where the prosecution
case depends critically on one or more key, but potentially unreliable, witnesses
of fact. In such a case, the trial judge might wish to see whether those witnesses
would survive cross-examination before undertaking an investigation into expert
opinion evidence, if that evidence is being tendered merely to support their oral
evidence. We note that the Rose Committee of the Senior Judiciary referred to
the fact that the relevance of expert evidence will always be fact specific, and that
the trial judge will be able to assess the relevance of such evidence only during
the trial.

5.50 The Rose Committee also expressed the view that a pre-trial admissibility
hearing, when required by the judge, could be very valuable because the judge
would have a “dry run” and, if he or she rules in favour of admitting the disputed
opinion evidence, could ensure that the significance of any disagreements
between experts is explained and that the experts articulate their evidence in a
form understandable to a jury of lay persons. The judge would also be able to
rule that only part of the expert’s evidence was admissible, “thus narrowing the
issues and ensuring that only reliable, sound, and understandable expert
evidence went before the jury”.51 Some other consultees noted that a pre-trial
hearing would provide the judge with the opportunity to analyse the evidence and
ask the experts questions which he or she might be reluctant to ask in front of the
jury.

50
One or two consultees expressed concern that a pre-trial hearing could provide expert
witnesses with the tactical advantage of a dry run. On a related point, see C [2010] EWCA
Crim 2578 at [40] where the Court of Appeal stressed that pre-trial admissibility hearings
“are not to be used for the ulterior purpose of cross-examining experts in advance of the
trial” and explicitly directed trial judges to “ensure that this does not happen”.
51
See also Henderson and others [2010] EWCA Crim 1269, [2010] 2 Cr App R 24 at [203] to
[205] on the desirability of properly marshalling and controlling the expert evidence before
the jury is sworn and on the importance of the pre-trial process and “robust pre-trial
management”. With regard to the importance of robust case management in the specific
context of expert opinion evidence on DNA, see Reed [2009] EWCA Crim 2698, [2010]
1 Cr App R 23 at [131] and C [2010] EWCA Crim 2578 at [32] and [40].

71
5.51 Given the foregoing, and what we say in paragraph 3.77, the conclusion we have
reached is that the reliability test should be applied, usually before the jury is
sworn, if it appears to the trial judge that the opinion evidence in question might
not be sufficiently reliable to be admitted. If the challenging party persuades the
trial judge to this low threshold, it would be necessary for the party seeking to
adduce the expert opinion evidence to show that it is sufficiently reliable to be
admitted. Following a challenge, the trial judge would therefore have a limited
power to decide whether or not the reliability test needs to be applied (and the
judge would always have the power to apply the reliability test in a case where no
challenge has been made). If a challenge is made, the reliability test would have
to be applied if there is an appearance of unreliability (that is, if it appears to the
court that the reliability test might not be satisfied), but otherwise the reliability
test would not need to be applied.52

5.52 We have come to the conclusion that this approach strikes the right balance
between the need on the one hand to control the admissibility of expert opinion
evidence and the desirability on the other hand of ensuring that criminal
proceedings are not unnecessarily delayed or rendered more costly. A limited
power for the judge, and an expectation that the opposing party should formulate
a reasoned argument as to unreliability, should reduce or remove concerns about
the test having to be applied in cases where it would be unnecessary to enquire
into the issue.

5.53 So, save for the situation where the trial judge acts of his or her own motion, the
onus would be on the challenging party to enquire into the matter and prepare a
reasoned argument which would put the party tendering the expert evidence to
proof. The challenging party must therefore be prepared to formulate a credible
argument that there are doubts about the reliability of the opponent’s expert
opinion evidence.53

5.54 For example, the opposing party could refer to the fact that the expert’s
underlying hypothesis has never been properly tested, or that the hypothesis has
been criticised in reputable journals, or that the data generated by observation
and testing are insufficient to justify the expert’s proffered opinion, or that the
expert’s opinion is unsubstantiated “orthodoxy”, and so on.

5.55 The party challenging the admissibility of an expert’s opinion, and the trial judge
him or herself, would of course refer to the basic criteria set out in clause 4(1) of
our draft Bill, and any relevant examples or factors in clause 4(2) and the
Schedule, when determining whether or not the evidence appears to be
sufficiently reliable to be admitted.

5.56 We therefore recommend the following for criminal proceedings:

52
See clause 6(2) and (3) of our draft Bill. These powers could come to be regarded as
limited judicial discretions, just as s 78(1) of the Police and Criminal Evidence Act 1984
and s 101(3) of the Criminal Justice Act 2003 are often referred to as “discretions”.
53
In Part 7 we set out recommendations on pre-trial disclosure which would ensure that a
party challenging admissibility is properly equipped to formulate an argument of this sort.

72
(1) there should be a presumption that expert opinion evidence
tendered for admission is sufficiently reliable to be admitted, but
this presumption would not apply if:

(a) it appears to the court, following a reasoned challenge, that


the evidence might not be sufficiently reliable to be admitted,
or

(b) the court independently rules that the presumption should


not apply;

(2) if the presumption no longer applies, the court should direct that
there be a hearing to resolve the question of evidentiary reliability,
unless the question can be properly resolved without a hearing; and

(3) for Crown Court jury trials, the reliability hearing should ordinarily
take place before the jury is sworn, but, exceptionally, it should be
possible to hold a hearing during the trial in the absence of the jury.

5.57 As explained above, the court should resolve the question of evidentiary
reliability, following a hearing, with reference to the test, examples and factors in
our draft Bill.

5.58 Clause 6(2) of our Bill provides the implicit presumption of sufficient reliability (for
the purposes of determining admissibility) and the rule that if, following a
challenge, it appears to the court that a party’s expert opinion evidence might not
be sufficiently reliable to be admitted, the party tendering the evidence for
admission must show that it is sufficiently reliable to be admitted.54 For Crown
Court cases to be tried before a jury, clause 6(4) provides a tacit presumption
that the court’s investigation into evidentiary reliability should take place before
the jury is sworn (and an explicit rule that if the investigation takes place during
the trial, the jury should not be present).55

5.59 There would be no obligation on the court to investigate evidentiary reliability


under clause 6(2) just because a representation has been made to the court that
an expert’s opinion evidence is insufficiently reliable to be admitted. It would need
to appear to the court that the evidence might not be sufficiently reliable to be
admitted. This requirement, and the fact that clauses 4 and 6 of our draft Bill do
not apply to expert evidence of fact, would limit the scope of any attempt to
disrupt the criminal process by unmeritorious objections to admissibility. If an
expert is simply presenting evidence of fact then it would not be possible to
challenge the evidentiary reliability of that evidence under clauses 1(2) and 4. If
an expert is proffering an expert opinion, then that opinion could potentially be
challenged, but only if there is a sound argument for displacing the implicit
presumption of reliability. If there is such an argument, then it is of course right in
principle that the evidence should be scrutinised for evidentiary reliability. But if

54
Clause 6(3) provides, in addition, that the court may disapply the tacit presumption of
reliability and require the party proffering the expert opinion evidence to show that it is
sufficiently reliable to be admitted.
55
For the position in magistrates’ courts, see paras 5.37 to 5.41 above.

73
there is no such argument, the presumption of reliability would stand (unless the
court chooses to act of its own motion under clause 6(3)).

5.60 As explained in paragraph 5.39, we do not envisage any significant disruption to


criminal proceedings in magistrates’ courts as a result of the enhanced
admissibility rules in our draft Bill. No doubt some unmeritorious defence
challenges will be made in the early years, but robust pre-trial case management
in tandem with the provisions in our Bill should prevent such challenges being
pursued or, if they are pursued, proving successful.

5.61 We have also explained that a trial judge would be able to rely on clause 4(3)(c)
of our draft Bill to consider a relevant judicial ruling in another case.56 According
to the legal committee of HM Council of District Judges (Magistrates’ Courts) the
written rulings of District Judges should be centrally collated so that other District
Judges would have access to them when having to “meet the prevalence at
certain times of particular arguments” so as to prevent inconsistent approaches
by differently constituted courts and the risk of delay. We agree with this
suggestion. We also agree with the following comment provided by the legal
committee:

There will be a requirement on District Judges to be alert to potential


unmeritorious and time-wasting objections to expert evidence in the
initial stages of the new legislation. But practice should settle fairly
soon and [the] admission of expert evidence in the majority of cases
will be unchallenged but with the advantage that the parties will have
had to address the issues of reliability when preparing the evidence.

The onus of persuasion


5.62 In Part 3 we explained in some detail our recommendation that, where the
reliability test is applied, the party wishing to adduce the expert opinion evidence
should bear the burden of showing that it is sufficiently reliable to be admitted.57
The party wishing to adduce the expert opinion evidence would need to provide
the evidence and explanation necessary to support a submission that the opinion
evidence the expert wishes to give is sufficiently reliable to be taken into
consideration by a jury. Clause 6(2) and (3) of our draft Bill would also give effect
to this recommendation.

Applying the reliability test in practice


5.63 It may assist understanding if we now provide an indication of how these rules
would work in practice.

Scientific (medical) evidence


5.64 In our consultation paper, and in Part 1 of this report, we referred to the case of
Harris and others58 where new evidence undermined the medical view of a
number of experts that a non-accidental head injury to an infant – shaken baby
56
Paragraph 5.34 above.
57
Paragraphs 3.79 to 3.124.
58
[2005] EWCA Crim 1980, [2006] 1 Cr App R 5.

74
syndrome – could confidently (in effect, always) be inferred from nothing more
than the presence of a particular triad of intra-cranial injuries.59

5.65 In that case, the Court of Appeal recognised that this triad of injuries could be
caused, albeit only rarely, by a minor fall or non-violent handling and held that,
without more, the mere presence of the triad could not automatically or
necessarily lead to a diagnosis of non-accidental head injury.60 Previously in
cases of this sort, the prosecution had been able to secure a conviction solely on
the basis of an expert diagnosis founded on the triad (in tandem with what the
medical experts regarded as an implausible exculpatory explanation from the
accused). And yet it seems the diagnosis of a violent assault was founded on
only a poor-quality database.61

5.66 According to the test in clause 4(1) of our draft Bill, if a party wishes to rely on a
hypothesis, and provide an expert opinion based on it, it will be necessary to
show that the opinion is “soundly based” and that the strength of the opinion is
“warranted having regard to the grounds on which it is based”. Any inference
drawn by the expert must be expressed with no greater degree of precision or
certainty than can be justified by the material supporting it. The onus will be on
the party proffering the evidence, and the party’s experts, to refer to properly
conducted empirical research (testing and observing) which substantiates the
hypothesis and does not undermine it. The court will then consider whether the
opinion evidence the expert wishes to provide (including its strength) is
sufficiently reliable to be admitted, bearing in mind the extent and quality of the
research, the margins of uncertainty in the findings, the extent of the data relied
on, any “known unknowns” and, in particular, whether there is a plausible,
alternative explanation for the findings.

59
Acute encephalopathy (a disorder of the brain), subdural haemorrhage (bleeding around
the brain) and retinal haemorrhage (bleeding in the retinas).
60
[2005] EWCA Crim 1980, [2006] 1 Cr App R 5 at [70], [152], [175] and [257]. See also [69]:
“There remains a body of medical opinion which … whilst recognising that the triad is
consistent with [non-accidental head injury], cautions against its use as a certain diagnosis
in the absence of other evidence.” The Court of Appeal did recognise, however, that the
presence of the triad is a “strong pointer” to a non-accidental head injury. This was
reaffirmed in Henderson and others [2010] EWCA Crim 1269, [2010] 2 Cr App R 24 at [6]:
“it is now commonly accepted that the triad is strong prima facie evidence of shaking”.
Importantly, however, strong prima facie evidence of shaking does not necessarily mean
proof of shaking. In Butler [2010] EWCA Crim 1269, [2010] 2 Cr App R 24 at [84] to [118],
it was recognised that a prosecution case of non-accidental injury dependent on the triad
of intra-cranial injuries was seriously undermined by the fact that the injured child had
completely recovered. This suggested that the triad had resulted from some other
(unknown) cause. The Crown Prosecution Service has recently published updated
guidance for prosecutors dealing with cases of this sort. This guidance explains that it is
unlikely a charge of murder, attempted murder or assault will be justified if the only
evidence against the accused is the triad of injuries; see CPS, Non-Accidental Head Injury
Cases (NAHI, formerly referred to as Shaken Baby Syndrome [SBS]) – Prosecution
Approach, http://www.cps.gov.uk/legal/l_to_o/non_accidental_head_injury_cases/ (last
visited 26 January 2011).
61
Paragraph 1.7 with fn 13 above. See also Consultation Paper No 190, fn 31 to para 2.24.

75
5.67 As we explained in our consultation paper,62 and in Part 1 of this report,63 the
evidence base for the hypothesis of shaken baby syndrome when the appellants
in Harris and others64 were tried has been described as an inverted pyramid “with
a very small database (most of it poor quality original research, retrospective in
nature, and without appropriate control groups) spreading to a broad body of
somewhat divergent opinion”.65 If our proposed admissibility test had been in
force at the time when the prosecution was seeking to rely on the triad of intra-
cranial injuries as proof of a non-accidental head injury, and the prosecution’s
expert opinion evidence had been challenged:

(1) the experts who wished to give opinion evidence for the prosecution
based on the hypothesis of shaken baby syndrome would have been
more mindful of the likely need to demonstrate the reliability of their
hypothesis at a pre-trial hearing, and would therefore have conducted, or
sought data from, appropriate scientific research with a view to seeking
support for the hypothesis (or identifying flaws so as to refine the
hypothesis);66

(2) the prosecution experts would have been more mindful of the need to
ensure that the opinion evidence they wished to give, underpinned by the
hypothesis of shaken baby syndrome, would stand up to judicial scrutiny
at a pre-trial hearing to assess evidentiary reliability, moderating their
opinions to the extent required by the limitations in the empirical research
and any aspects of the research data which undermined the
hypothesis;67

62
Consultation Paper No 190, fn 31 to para 2.24.
63
Paragraph 1.7.
64
[2005] EWCA Crim 1980, [2006] 1 Cr App R 5.
65
JF Geddes and J Plunkett, “The evidence base for shaken baby syndrome” (2004) 328
British Medical Journal 719, quoting the conclusion of M Donohoe, “Evidence-based
Medicine and Shaken Baby Syndrome” (2003) 24 American Journal of Forensic Medicine
and Pathology 239, 241. See also D Tuerkheimer, “The Next Innocence Project: Shaken
Baby Syndrome and the Criminal Courts” (2009) 87 Washington University Law Review 1,
12 to 14 and 17 to 18.
66
Ideally the research would have resulted in the publication of peer-reviewed papers in
reputable medical journals.
67
In this context, the hypothesis would need to be shown to be reliable by sufficient
observational data and/or simulations. There would need to be properly conducted
research showing a sound correlation between the intra-cranial injuries and a non-
accidental cause (from independent evidence) and demonstrating the absence of such
injuries in cases where there have been accidents or congenital conditions. The stronger
the expert’s opinion, the greater would need to be the observational data consistent with it
(and the absence of observational data inconsistent with it). We note, however, the
difficulties associated with using biomechanical models to simulate the complex anatomy
of an infant’s brain and that the science of biomechanics is “complex, developing and (as
yet) uncertain”; see Henderson and others [2010] EWCA Crim 1269, [2010] 2 Cr App R 24
at [180] to [182].

76
(3) the judge would have been aware of the importance of scrutinising the
experts’ proffered opinion evidence for reliability in advance of the trial,
with reference to the nature and extent of the empirical research
underpinning shaken baby syndrome, and would have permitted the
experts to give an opinion at trial only to the extent that could be justified
by the research data then available;

(4) given the limited research data supporting the hypothesis, it is highly
unlikely that the judge would have allowed the prosecution to advance a
case at trial founded solely on expert opinion evidence that the deceased
or injured infant exhibited the triad of intra-cranial injuries associated with
shaken baby syndrome (and that the accused’s exculpatory explanation
could therefore be disregarded as untrue);68

(5) however, a conviction would have been possible – as it is today – on the


basis of the triad of intra-cranial injuries in association with other
sufficiently cogent circumstantial evidence of the accused’s guilt (such as
separate injuries consistent with abuse).69

68
Professor Tim David (Professor of Child Health and Paediatrics at the University of
Manchester) pointed out that, before Harris [2005] EWCA Crim 1980, [2006] 1 Cr App R 5
a diagnosis of child abuse in non-accidental head injury cases was based on weighing up
the explanation provided by the accused against the observed injuries. We accept that the
plausibility of the accused’s exculpatory explanation has evidential value; but, equally, the
extent to which D’s explanation was considered to be plausible by medical experts no
doubt depended on how confident those experts were that the hypothesis underpinning a
diagnosis of baby shaking was correct. William E Bache, a solicitor, told us that, in his
experience: the prosecution tended to approach the same group of experts for the same
opinion; only the tests which could provide evidence of an offence were conducted soon
after the injury occurred; and by the time the defence was in a position to commence its
own investigation, it was too late to conduct further tests which might undermine the
prosecution case by suggesting an alternative cause.
69
See Henderson [2010] EWCA Crim 1269, [2010] 2 Cr App R 24 and Oyediran [2010]
EWCA Crim 1269, [2010] 2 Cr App R 24. Conversely, the cogency of the triad as “strong
prima facie evidence of shaking” would be profoundly weakened by circumstantial
evidence suggesting an unknown (innocent) cause; see Butler [2010] EWCA Crim 1269,
[2010] 2 Cr App R 24 at [84] to [118].

77
5.68 In addition, as under the law at that time, any admissible expert opinion evidence
adduced by the prosecution would have been challenged in cross-examination
and by the adduction of contrary expert evidence by the defence; and the judge
would have provided the jury with a careful direction on how the conflicting expert
evidence should be approached.70

5.69 In short, if the empirical underpinnings of shaken baby syndrome consisted of


nothing more than a very small database taken from poor-quality research, as
would seem to have been the case when the appellants in Harris and others were
tried, and if the trial judge’s attention had been drawn to this at a pre-trial
admissibility hearing, it is very difficult to believe the outcome would have been
the same. The judge would no doubt have ruled that the expert opinion evidence
in support of the prosecution assertion of a non-accidental injury could not be
admitted unless the experts concerned were willing to modify or qualify their
opinions to reflect the uncertainties associated with the hypothesis and the quality
of the research supporting it. Accordingly, the judge would almost certainly not
have allowed the prosecution experts to provide opinion evidence that the triad of
injuries permitted a certain diagnosis of non-accidental injury.

5.70 For scientific opinion evidence, the underlying evidence supporting the
hypothesis and the chain of reasoning underpinning the opinion would always
need to be scientifically valid; but the required extent to which there has been
scientific research and the required extent of the corroborative data supporting a
hypothesis will depend on the nature and strength of the opinion and the extent to
which it is qualified.

Non-scientific evidence
5.71 Our new reliability test would not, however, be limited to expert opinion evidence
which is based on evidence of a scientific nature. It may be that our test will
occasionally need to be applied to other types of evidence, such as a lip-reader’s
interpretation of what he or she has observed. So, depending on the facts, if the
prosecution wishes to call a qualified lip-reader71 who has viewed a CCTV
recording of two individuals talking to each other, to give an opinion on what was
said, the prosecution might need to demonstrate that the way the lip-reader
examined the CCTV recording for the instant case and the way the lip-reader
formulated his or her opinion from what was seen provide a sound basis for
holding that that evidence is sufficiently reliable to be admitted. If the judge were
to direct that the reliability test must be applied in this context, the prosecution
would need to show that the lip-reader’s observational methodology and
assumptions were valid.

5.72 Clearly it would not always be necessary to apply the reliability test to evidence of
this sort. Indeed, where a lip-reader is called to give an expert opinion, the only

70
The need for a careful direction relating to conflicting scientific opinion evidence was
recently emphasised in Henderson and others [2010] EWCA Crim 1269, [2010] 2 Cr App
R 24 at [203] and [217] to [219].
71
That is, qualified (skilled) as required by cl 1(1)(b) and cl 2 of our draft Bill.

78
real issue for the court in most cases is likely to be whether or not the witness
has the skill to provide such evidence.72

5.73 However, it might be appropriate to apply the reliability test in some cases.
Factors such as line-of-sight, facial hair, regional accents and lighting may have a
bearing on the reliability of a lip-reader’s interpretation. If the angle of observation
and the lighting were poor, and the fundamental issue is whether the observed
person said just one or a few key words, then the lip-reader’s evidence could be
insufficiently reliable to be admitted in a given case.73

5.74 If a valid objection to the admissibility of a lip-reader’s opinion evidence were to


be raised, so as to displace the presumption of threshold reliability, it would be for
the party calling the expert to show that the lip-reader’s methodology, or the way
the expert applied his or her skill for the instant case, provides sufficient evidence
of reliability to justify his or her opinion evidence being placed before the jury.74

5.75 The same approach would apply to the methodology of other non-scientific
experts, such as police officers called to give evidence as “ad hoc experts”.75 In
some cases it may be necessary for the prosecution to provide results from re-
enactments designed to test whether the officer’s technique and assumptions
provide a sufficiently reliable foundation for his or her opinion evidence.76

5.76 However, there are some professional, non-scientific practices which are so well
established that any kind of assessment or testing, even if possible, would
provide little if any additional guarantee of reliability in the round (or any further
justifiable guarantee, given the cost involved) beyond that provided by the
requirement to demonstrate expertise and impartiality. Within this category would
fall the evidence of professionals such as some accountants, whose academic
and professional qualifications, experience and reliance on generally-accepted
practices would provide a sufficient guarantee in most cases.

72
Above.
73
In Luttrell [2004] EWCA Crim 1344, [2004] 2 Cr App R 31 the Court of Appeal recognised
that a lip-reader was providing an interpretation (an expert opinion) on what was said and
that he or she could make errors. Reference was made to factors which can increase the
difficulties associated with applying this skill, including light and angle of observation.
Specific reference was also made to “whether the probative effect of the evidence depends
on the interpretation of a single word or phrase” (at [38]).
74
This would not be problematic or expensive. The circumstances of the particular lip-
reading incident could simply be repeated with different individuals being observed saying
different things. The results would show whether the lip-reader’s general skills provide a
reliable interpretation in the specific context of the instant case.
75
In England and Wales an ad hoc expert is generally a police officer who has spent a
substantial amount of time analysing CCTV footage and is called by the prosecution to
provide an opinion on whether the person filmed is the accused.
76
In this context there may be no particular skill beyond an eye for detail and patience, with
the issue of reliability turning on the way in which the officer applied that skill in analysing
and comparing images. There is no reason why the officer’s methods could not be
inexpensively tested for reliability when required, for example by having the officer
compare a CCTV image against several individuals, to see whether or not he or she
makes a correct identification.

79
5.77 The Rose Committee of the Senior Judiciary agreed with the comment in our
consultation paper77 that for some areas of professional non-scientific disciplines,
such as forensic accountancy, where there are well-accepted practices, there
would be no need for a minute consideration of the underlying basis of the
expert’s opinion evidence.

5.78 But of course expertise alone does not necessarily mean that an expert is
providing reliable opinion evidence. If a forensic accountant’s opinion is based on
a technique other than a well-accepted practice, the court could decide to apply
the reliability test. If evidentiary reliability can be demonstrated objectively by an
assessment of some kind, then supporting evidence of that sort is what the court
is likely to expect. If such an assessment is not possible or practicable, the court
would nevertheless need to determine whether there are sufficient indicia of
evidentiary reliability in other respects to justify the admission of the expert’s
opinion evidence. We have been told by a forensic accountant based in the
Serious Fraud Office78 that factors (a), (g) and (f) in Part 1 of the Schedule to our
draft Bill could be material in this context. With regard to factor (g), he said that it
is essential that the reports provided by forensic accountants called by the
defence “highlight where in the bounds of probability their opinion lies to prevent
them extolling theories which are on the extremes of possibility whilst ignoring
more sound hypotheses”.

5.79 Other experts whose opinion evidence (and underlying basis) might not require a
minute consideration are experience-based experts on matters such as retail
theft or industrial practices, or Trading Standards officers who have extensive
experience of the way consumers behave and are likely to behave in the future.
For such experts, we would expect the courts to focus on the admissibility
requirement in clauses 1(1)(b) and 2 (expertise) and apply the reliability test only
if the strength of the expert’s opinion demands an enquiry, for the strength of an
expert’s opinion must always be warranted. Nevertheless, for such fields of
expertise, it is fair to say that the expert’s wealth of experience is likely to provide
a sufficient guarantee of reliability in the round for much expert opinion evidence
founded on it, meaning that the evidentiary reliability test is likely to be applied to
such evidence only rarely.79 The courts would also be likely to adopt the same
approach to expert opinion evidence on matters such as foreign law or the public
good associated with allegedly obscene publications.

5.80 Our view in this respect accords with the opinion of the Bar Law Reform
Committee, who felt that the evidentiary reliability test we proposed in our
consultation paper should not have to be applied to experts whose opinion

77
Consultation Paper No 190, para 6.37.
78
Simon Daniel.
79
Again, there will be challenges in some cases; and judges will occasionally consider it
necessary to investigate the evidentiary reliability of an expert opinion which is not founded
on scientific methodology. For an interesting recent example in this context, see the
Canadian case of Abbey 2009 ONCA 624 at [119] where the Court of Appeal for Ontario
set out the factors it considered relevant to the determination of threshold evidentiary
reliability when considering background evidence provided by an acknowledged expert in
the culture of Canadian street gangs. The expert’s opinion evidence in that case related to
the possible reasons why a young male member of an urban street gang would have a
teardrop tattoo inscribed on his face.

80
evidence is based on a wealth of experience. They gave the example of an
expert on sado-masochistic relationships and police officers with extensive
experience of drug pricing.

5.81 It should always be remembered, moreover, that our reliability test would be
potentially applicable only if the expert witness wished to provide an expert
opinion. It would not apply if a police officer were called to provide expert
evidence on factual matters such as the sort of paraphernalia commonly used by
drug-dealers. To provide such evidence the officer would simply need to prove
that he or she was qualified to provide expert evidence, with reference to
information such as the number of recent cases involving drugs he or she has
worked on, the nature and extent of his or her involvement, the courses and
seminars attended and so on. The reforms we recommend in this report would
not prevent suitably-qualified police officers from providing expert evidence of fact
(or, indeed, any expert opinion evidence which is sufficiently reliable to be
admitted).

5.82 But if a police officer were to be called to give an expert opinion which has been
challenged on the ground of insufficient reliability, the officer should be prepared
to justify the admission of the opinion against the admissibility rule, examples and
factors in our draft Bill. The reliability limb would probably need to be applied, for
example, if a drugs officer wished to give an opinion on whether the number of
ecstasy tablets found in D’s possession in a nightclub was more than would be
required for personal consumption. The officer would first need to show that he or
she was qualified to provide an expert opinion on such matters; and, secondly, he
or she would need to demonstrate that his or her opinion was based on sound
empirical research and that the strength of the opinion was warranted by the data
relied on and the inferences legitimately to be drawn from the data.80

Summary
5.83 Our proposed reforms would introduce a framework for effectively challenging the
admissibility of expert opinion evidence in any appropriate case and a basis for
being able properly to investigate and determine evidentiary reliability. We
particularly have in mind the forensic sciences, of course, but it is possible that
experts in other, non-scientific disciplines would also be required to demonstrate
the reliability of their opinion evidence in some cases.

5.84 The greater the strength of the expert’s opinion, the greater the likelihood that it
would be challenged and, accordingly, the greater would be the onus on the
expert to be prepared to demonstrate that his or her opinion evidence is
warranted.81

5.85 Before closing this discussion, we should make one final point. Whilst in broad
terms we agree with the view of one of our consultees, Dr Geoffrey Morrison,82

80
Compare Hodges [2003] EWCA Crim 290, [2003] 2 Cr App R 15.
81
As the General Medical Council accepted in relation to medical evidence, “the more
significant the evidence is to the issues in a case, the greater the scrutiny of its
admissibility should be”.
82
A researcher on forensic voice comparisons (the Australian National University).

81
that “forensic analyses which are more objective and whose reliability can be
quantitatively demonstrated should be preferred over more subjective analyses
for which it is harder to quantify reliability”, we also believe that if a subjective
analysis can be tested in controlled circumstances, and opinion evidence
founded on such an approach can thereby shown to be reliable, there is no
reason why such opinion evidence should be excluded.83

A POWER TO STOP THE TRIAL?


5.86 Some consultees suggested that there should be a provision for expert evidence
similar to section 125 of the Criminal Justice Act 2003. This provides the trial
judge with the power to stop a trial if the case against the accused is based on
hearsay evidence which is so unconvincing that a conviction based on that
evidence would be unsafe.

5.87 We considered this question before the consultation paper was published and
decided that it was unnecessary to have a provision of this sort because the trial
judge has a general power to reconsider an admissibility ruling during the trial.84
The judge’s pre-trial ruling that expert evidence is admissible may therefore be
reversed during the trial and the evidence ruled inadmissible with a direction to
the jury to disregard it. If this would provide an insufficient safeguard for the
accused, given that inadmissible prosecution evidence has been heard by the
jury, the judge would be able to discharge the jury.85

5.88 Given these safeguards, and bearing in mind the judge’s general discretion to
exclude any prosecution evidence (even if it has been admitted) on the ground
that its unduly prejudicial effect outweighs its probative value, we do not believe
that a provision similar to section 125 of the Criminal Justice Act 2003 is
necessary in our proposed scheme for expert evidence.

ADDRESSING THE JUDGE’S RULING ON APPEAL


5.89 In our consultation paper we explained that the ruling on admissibility would be a
question of law and, as such, could be examined by the Court of Appeal (or the
Queen’s Bench Division of the High Court, for summary proceedings).86

5.90 Our view was that the judge’s ruling on the evidentiary reliability test, in relation to
matters which are not case-specific, should be approached by the appellate court
as the application of a rule, a legal judgment, rather than the exercise of a judicial
discretion. This would allow the appellate court itself to investigate underlying
scientific propositions and properly police the application of the reliability test, so
the court would not simply decide whether the judge had acted within the
parameters of what any reasonable judge could have done.

83
We note, in line with the current laissez-faire approach to the admissibility of expert
evidence in criminal proceedings, the Court of Appeal suggested in Flynn [2008] EWCA
Crim 970, [2008] 2 Cr App R 20 that an expert opinion based on an auditory analysis is
admissible even without the support of an acoustic (or spectrographic) analysis.
84
Watson (1980) 70 Cr App R 273, 276 (Consultation Paper No 190, Part 6, fn 54).
85
Azam [2006] EWCA Crim 161 at [48] (Consultation Paper No 190, Part 6, fn 55).
86
Consultation Paper No 190, paras 6.44 to 6.46.

82
5.91 This is still our view. We note that the equivalent reliability test in the United
States (the Daubert test)87 has been criticised as insufficiently effective for
criminal proceedings because, amongst other things, it provides the trial judge
with a wide discretion in the determination of evidentiary reliability and that
appeals in relation to the application of this test are judged against a very narrow
“abuse of discretion” standard of review.88 We believe that the assessment of
evidentiary reliability in respect of matters which are not case-specific, principally
questions of underlying scientific methodology, should be addressed anew in the
Court of Appeal (if leave to appeal is given) not according to whether the trial
judge acted within the parameters of a wide discretion.

5.92 Our policy in this respect was supported by the consultees who expressly
addressed it (the Royal Statistical Society, the Bar Law Reform Committee and
Associate Professor William O’Brian). Our proposed admissibility test with the
examples in clause 4(2) and the guiding factors in Part 1 of the Schedule, in
tandem with better training for lawyers and the judiciary and better policing by the
appellate courts, should overcome the problems identified with the Daubert test in
the USA.

5.93 But it should be remembered that the key issue on appeal will always be whether
the expert opinion evidence in question was sufficiently reliable to be placed
before a jury. Accordingly, although we refer above to a new assessment before
the appeal court of matters which are not case-specific, we believe the court
should adopt the same approach to the opinion evidence founded on those
matters. That is to say, the appeal court should adopt a holistic approach,
addressing the judge’s ruling on the admissibility of the opinion evidence in the
same way that it addresses the reliability of any hypothesis underpinning it.89

5.94 We therefore recommend that, if challenged on appeal, the trial judge’s


ruling under the reliability test should be approached by the appellate court

87
From Daubert v Merrell Dow Pharmaceuticals 509 US 579 (1993), a decision on r 702 of
the US Federal Rules of Evidence; see Consultation Paper No 190, paras 4.41 to 4.49.
88
See in particular, the National Research Council of the National Academies’ 2009 report,
Strengthening Forensic Science in the United States: A Path Forward, pp 9 to 11, 95 to 98
and 106 to 110. The report points out that appeal courts will interfere with judicial rulings
only if they are, in the language of England and Wales, plainly wrong or Wednesbury
unreasonable (see fn 90 below). Other problems identified in the report are that judges and
lawyers lack the expertise to deal with the Daubert reliability test and trial judges sitting
alone do not have the time for extensive research and reflection.
89
We now accept that it would be best to apply a unified approach, in line with the view of
Judge Jeremy Roberts QC (Consultation Paper No 190, para 6.46) and Professor Paul
Roberts (response to our consultation paper).

83
as the exercise of a legal judgment rather than the exercise of a judicial
discretion.90

5.95 The provisions which would give effect to this recommendation are set out in
clause 5 of our draft Bill.

5.96 In our consultation paper we did not propose any new avenues of appeal beyond
those which are currently in place, but we asked our consultees whether they
thought the question of evidentiary reliability should always be addressed before
the trial and whether there should be a further basis for an interlocutory appeal.91

5.97 Although some consultees favoured a procedure which would allow the judge’s
ruling on evidentiary reliability to be challenged on appeal before the trial, the
responses we received from the judiciary suggested that there should be no new
avenues of appeal. Mr Justice Treacy said the proliferation of rulings capable of
interlocutory appeal needs to be curbed because they distort the trial process and
over-burden the resources of the Court of Appeal, and there is in any event
already sufficient machinery in place to enable judges to make a pre-trial ruling
capable of interlocutory appeal in appropriate cases. A similar view was
expressed by the Council of HM Circuit Judges. The response we received from
the Rose Committee suggested that the senior judiciary are also opposed to any
new avenues of appeal.

5.98 We therefore do not recommend that there should be any avenue of interlocutory
appeal beyond those which currently exist.

CLOSING COMMENTS
5.99 In our consultation paper we said that the reforms we proposed would not
necessarily lead to a sea change in English criminal proceedings because (we
suggested) much expert evidence which is currently admitted would continue to
be admitted.92

5.100 A new reliability-based admissibility test would, however, put experts on notice
that they would be expected to provide sufficient material to enable the trial
judge, and indeed the other parties, to conclude that their opinion evidence is
sufficiently reliable to be admitted.

5.101 As we explained in Part 1, the increased level of scrutiny which comes with an
admissibility test focusing on the validity of the methodology and reasoning

90
The Court of Appeal (Criminal Division) often refers to “Wednesbury principles” or the
“Wednesbury test” when assessing the exercise of a judicial discretion as to the
admissibility or exclusion of evidence in a criminal trial (from Associated Provincial Picture
Houses Ltd v Wednesbury Corporation [1948] 1 KB 223, 229). This means that, so long
as all relevant factors were considered and all irrelevant factors disregarded, the trial
judge’s ruling will be upheld unless it was a decision no reasonable judge would have
reached. Sometimes the Court of Appeal will simply ask itself whether or not the judge’s
ruling was “plainly wrong” (or will use the “plainly wrong” approach alongside the traditional
Wednesbury test; see, for example, Hanson [2005] EWCA Crim 824, [2005] 1 WLR 3169
at [15]).
91
We set out the current framework in Appendix B to Consultation Paper No 190.
92
Consultation Paper No 190, paras 6.12 to 6.16.

84
underpinning an expert’s opinion evidence should encourage higher standards,
not only amongst expert witnesses themselves but also amongst scientists and
technicians working in forensic laboratories. Expert witnesses would need to
ensure that their opinion evidence, particularly the strength of any opinion they
wish to present, can be justified and will stand up to scrutiny if challenged.

5.102 If our proposed legislation leads to higher quality expert evidence being tendered
for admission, and as a result expert opinion evidence would only rarely be ruled
inadmissible, it would for that reason be a very successful reform project.

5.103 In this context it is worth mentioning that the Association of Forensic Science
Providers provided a very favourable response to our consultation paper because
our proposals were seen as directly complementing their own “Standards for
Expert Evidence” based on the four principles of “balance, logic, robustness and
transparency”.93 Similarly, the Forensic Science Regulator (whom we met) was
keen to support the main thrust of our proposals because our proposed
legislation would, he said, provide the missing link in the scheme of safeguards
he is now trying to introduce for many forensic scientists (including expert
scientific witnesses called by the prosecution).

5.104 We mention these comments because we too recognise the importance of


addressing the reliability of expert opinion evidence from different directions.
Greater quality control, sound organisational structures and proper accreditation
would go some way towards resolving some of the problems associated with
expert evidence in criminal proceedings.

5.105 There are, however, many areas of expertise which cannot realistically be
regulated or do not benefit from organisational structures, for example the work of
a specialist consultant in a field of medicine or an amateur lip-reader or a
podiatrist with a particular interest in forensic gait analysis. Furthermore, the
Forensic Science Regulator’s remit does not currently extend to forensic scientific
evidence tendered by the defence. There is therefore a powerful argument for
addressing the problems associated with expert evidence by the application of a
generally-applicable admissibility test along the lines we recommend. This
approach would:

(1) complement ongoing measures designed to ensure reliability in forensic


laboratories; and

(2) stimulate further measures to ensure reliability in other contexts.

5.106 Our admissibility test would ensure, first, that individuals claiming expertise are
properly screened for expertise and impartiality before giving evidence and,
secondly, that expert opinion evidence will be screened for evidentiary reliability
before being admitted if there is a plausible basis for doubting its reliability.

5.107 We should emphasise, however, that even if a qualified expert witness’s


proffered opinion evidence is ruled inadmissible on the ground that it is not
sufficiently supported by the available empirical research, it does not necessarily

93
The same factors were cited by the Forensic Science Service.

85
mean that the expert’s evidence in other respects would be inadmissible. The
expert might simply have to revise his or her original opinion in the light of the
opponent’s challenge and the judicial enquiry, and so present an opinion which is
less firm than the opinion he or she originally intended to present.

5.108 Accordingly, a prosecution expert’s evaluative opinion that a particular fact is


strong evidence of the accused’s guilt may be inadmissible on the ground that
that opinion is not warranted by the available data; but the same expert may be
permitted to give a different opinion, for example that the fact provides weak or
moderate support for the prosecution case.94

5.109 The key point is that, if our proposals are taken forward, the jury will only be
permitted to hear expert opinion evidence which can be properly substantiated,
and the stronger the expert’s opinion is, the greater will be the necessary degree
of substantiation.

5.110 Nor should it be assumed or thought that the difficulties experienced in the United
States, where expert opinion evidence is challenged on the ground of insufficient
reliability, are likely to arise in England and Wales under our proposed scheme.95
Our recommendations are positively different in a number of ways from the
approach used in jurisdictions which apply the “Daubert” test implied into rule 702
of the US Federal Rules of Evidence.96

5.111 First, our draft Bill sets out the fundamental criteria for determining the reliability
of expert opinion evidence, with reference to the underlying foundation material
and the strength of the opinion based on it,97 but it also supplements these
criteria by directing the judge to consider a number of examples and factors
which can affect the reliability of an expert’s opinion in a given case.98 Judges will
have the necessary guidance to determine evidentiary reliability.

5.112 Secondly, the Judicial Studies Board will provide trial and appeal judges with
practical training in how to assess the reliability of expert opinion evidence in

94
The Court of Appeal has accepted that in some contexts an expert may provide an
evaluative opinion based on a progressive scale, reflecting the likelihood of a match; see,
for example, Atkins [2009] EWCA Crim 1876, [2010] 1 Cr App R 8 (facial mapping) and T
[2010] EWCA Crim 2439 (footwear analysis). In T, however, the Court of Appeal took the
view (at [95]) that whilst a footwear analyst could in “appropriate cases use his [or her]
experience to express a more definitive evaluative opinion” than the mere observation that
a scene-of-crime mark “could have been made” by D’s shoe – eg where there was “an
unusual size or pattern” (at [74]) – no mathematical formula should be used to determine
that opinion because “there are far too many variables and uncertainties in the [underlying]
data” (at [85]). The court also took the view that in some cases a footwear analyst would
be able to go no further than provide an opinion that D’s shoe “could have made the mark”
(or “could not have made the mark”). With respect, however, we doubt whether an opinion
that D’s shoe “could have made the mark” would provide the jury with enough assistance
to justify the admission of the evidence; and we also query the court’s view (at [73]) that an
opinion of this sort would enable “a jury better to understand the true nature of the
evidence than the … phrase ‘moderate … support’”.
95
See fn 88 above.
96
Consultation Paper No 190, paras 4.41 to 4.49.
97
Clause 4(1).
98
Clause 4(2) and (3) and the factors in Part 1 of the Schedule.

86
practice, with reference to the test and factors in our draft Bill. Armed with the
guidance provided in our draft Bill, trial judges in England and Wales will be
properly equipped to address evidentiary reliability.99

5.113 Thirdly, the Court of Appeal will properly police the application of the test, by
approaching its assessment of the trial judge’s performance not on the basis that
it is the exercise of a broad discretion governed by “Wednesbury principles”, but
on the basis that the judge’s ruling is a judgment to be assessed according to
whether or not it is right. This will encourage a more critical approach to expert
opinion evidence at first instance, with reference to the relevant criteria in the
draft Bill, and it will ensure proper scrutiny on appeal.100

5.114 Fourthly, it will be seen in Part 6 that we are recommending a facility which would
allow the trial judge to call upon additional expertise to assist him or her in the
determination of evidentiary reliability in exceptionally complex cases. In addition,
in Part 7 we recommend a number of changes to the Criminal Procedure Rules
2010 which would ensure that the judge has all the relevant material he or she
needs to determine whether a party’s proffered expert opinion evidence is
sufficiently reliable to be admitted. These changes would also provide the parties
with the relevant information they need to make submissions to the judge on
evidentiary reliability, providing a further assurance that all relevant material will
be taken into consideration.

5.115 We close this explanation of our proposed reliability test by emphasising once
again the importance of training, both for lawyers and the judiciary, and the need
for a more proactive, enquiring approach to expert opinion evidence in criminal
proceedings. We fully endorse what the Criminal Cases Review Commission had
to say about this:

99
We do not therefore agree with the suggestion, predicated on data taken from the USA,
that some trial judges would not be able to understand and apply the evidentiary reliability
test in practice. The suggestion was recently made by J Hartshorne and J Miola in “Expert
evidence: difficulties and solutions in prosecutions for infant harm” (2010) 30 Legal Studies
279, 292 to 293. We note, however, that the authors could also see “the obvious benefits
that the introduction of the Commission’s proposed gate-keeping test would bring to
prosecutions for infant harm” and they favoured “a gate-keeping test that vests the court
with the responsibility for determining whether expert evidence should be admitted”
(p 291).
100
Compare J Hartshorne and J Miola, “Expert evidence: difficulties and solutions in
prosecutions for infant harm” (2010) 30 Legal Studies 279, 292, referring to the “significant
subjective factor” in r 702 of the US Federal Rules of Evidence and the concomitant risk in
the United States “of differing results depending on the idiosyncrasies or predisposition of
the trial judge”. (The authors justify their concern with reference to decisions made under
the Frye “general acceptance” test.)

87
Advocates who obtain, call and challenge expert evidence, and the
judges who preside over the cases in which such evidence is
deployed, must be encouraged to develop an approach of constant
scrutiny throughout the entire trial process in order to ensure that the
problems [associated with such evidence] can be identified and
addressed in any individual case … . The improved training of
solicitors, counsel and judges could by itself do much to reduce the
risk of miscarriages as a result of inaccurate or misleading expert
evidence.101

101
On training for lawyers, see fn 45 to para 1.43 above.

88
PART 6
COURT-APPOINTED EXPERTS

INTRODUCTION
6.1 One of the issues we considered in our consultation paper was whether a Crown
Court judge, for a trial on indictment, should be given a new statutory power to
appoint an independent expert to provide the judge with assistance and guidance
when addressing the question of evidentiary reliability.1 We took the view that an
independent expert of this sort would be able to provide the judge with valuable
help when determining the evidentiary reliability of especially complex scientific
(or purportedly scientific) evidence in advance of the trial.2

6.2 We suggested, but did not formally propose, that a Crown Court judge should
have this power; but we also suggested that it should be used only exceptionally,
to ensure that there would be no general lengthening of criminal proceedings or
increase in costs. Nevertheless, we also expressed the view that, in cases where
a judge made use of this power, time and other resources might be saved during
the trial because a party’s expert witness could be prevented from giving
unreliable opinion evidence.

6.3 We thought it would be justifiable to appoint an independent expert only if a


party’s proffered opinion evidence was complex evidence of a scientific nature
and it would not be reasonable to expect a judge to determine the question of
evidentiary reliability without assistance.3 We suggested that a court-appointed
expert could be selected from a list prepared or identified by the parties or in such
other manner as the court might direct.

6.4 We emphasised that the question whether a party’s expert opinion evidence was
sufficiently reliable to be admitted would in all cases be one of law for the judge to
determine. So, although the trial judge would treat with the greatest of respect the
views of the court-appointed expert, the final ruling on the question of reliability
and therefore admissibility would be for the judge alone.

6.5 In this report we have set out our recommendations that there should be a new
statutory admissibility test with an evidentiary reliability limb and a list of factors to
help judges apply it. We believe it is likely that the judiciary, when seeking to
apply the reliability test, will occasionally need additional expert assistance,
possibly in relation to fields such as psychology, psychiatry and statistical

1
By an “independent” expert we simply mean an expert witness called by the judge rather
than by a party. All expert witnesses have an overriding duty to give objective, unbiased
evidence, so in truth all expert witnesses are independent witnesses for the court. As
explained in paras 4.26 and 4.27 above, we have brought the overriding duty into cl 3(1)
and (2) of our draft Bill.
2
Consultation Paper No 190, para 6.67.
3
Consultation Paper No 190, paras 6.65 to 6.71. Because we originally proposed that there
should be two different bodies of guidelines to assist trial judges in their determination of
evidentiary reliability, we also suggested that a court-appointed expert might be able to
help the judge decide which guidelines should apply.

89
analysis. His Honour Judge Jeremy Roberts QC, a very experienced Old Bailey
judge, told us the following:

Although the need for a court-appointed expert will very rarely arise, it
seems clear that there may from time to time be cases in which the
judge cannot make the necessary decision without additional
information over and above that provided by the two sides. Where
that is the case, the interests of justice strongly favour a system by
which the judge can obtain that information from an independent
expert: otherwise there will be a real danger of an erroneous decision
being made … .

6.6 In its response to our consultation paper, the British Psychological Society
thought that it would be useful for judges to have the power to call upon an
independent expert in cases involving psychological evidence; and, similarly, the
Royal College of Psychiatrists suggested that the judge might require help from
“expert statisticians and [experts] in scientific methodology as applied to mental
health”.

6.7 Our original suggestion that Crown Court judges should be given a limited power
to appoint an independent expert, to assist in a pre-trial determination of
evidentiary reliability, would not have been a radical change in the law, but it
would have been a new power. We say this because the criminal courts already
have a common law power to call a witness of fact during a trial, if it is in the
interests of justice to do so,4 and the inherent flexibility of the common law would
presumably permit a Crown Court judge to call an expert witness to assist in the
determination of evidentiary reliability as a matter bearing on admissibility.

6.8 There is no case law to confirm this broad interpretation of the common law
power; and when we wrote our consultation paper we were unaware of any case
where the common law power had been used to call an expert witness. However,
we have since been told by the UK Register of Expert Witnesses that judges
have used their common law power to call expert witnesses during criminal trials,
albeit only very rarely.5

6.9 We are still unaware of any occasion where a criminal court has called an expert
to help in a pre-trial determination of evidentiary reliability, as a matter bearing on
admissibility, but this is hardly surprising. There are several reasons why we say
this.

6.10 First, given the current laissez-faire approach to the admissibility of expert
opinion evidence in criminal trials, there has until recently been very little
authority for the view that a trial judge should enquire into evidentiary reliability as
a matter bearing on admissibility.6

4
Roberts (1985) 80 Cr App R 89; R v Haringey Justices ex parte DPP [1996] QB 351.
5
Following a request we made to the UK Register, the Editor, Dr Chris Pamplin, kindly
emailed over 3,200 expert witnesses asking them whether they had ever been called by a
criminal court to provide expert evidence. Five experts replied that they had.
6
For recent developments, see paras 2.14 and 2.15 above.

90
6.11 Secondly, on account of the adversarial nature of criminal proceedings in
England and Wales, it is reasonable to assume that many trial judges may be
reluctant to enter the arena by using a common law power to call for additional
expert opinion evidence in the absence of an explicit authority permitting this,
whether the power would be exercised during the trial or before the jury is
empanelled.

6.12 Thirdly, trial judges are in any event unlikely to know which expert to appoint,
where there is a range of expertise on a matter, or how to go about finding a
suitably-qualified individual.

6.13 So, although the statutory power we suggested would not have changed the law
in any radical sense, it would have been a change, and in our view a positive
development. The courts would have had an explicit statutory power, replacing
the common law in this specific context, and this would have encouraged trial
judges to appoint independent experts to help them determine evidentiary
reliability when they really needed such help. But the statutory power to appoint
an independent expert would have been restricted in the way described above, to
ensure that it would be used only when necessary and so guard against
proceedings generally becoming longer or more expensive.

6.14 Before setting out the views of our consultees, and addressing the possible
problems associated with a new scheme for appointing independent experts, we
think it would be helpful if we first described how a court-appointed expert would
provide his or her evidence and how the interests of the parties would be
protected during a hearing – nearly always pre-trial – on evidentiary reliability.

THE PRE-TRIAL HEARING


6.15 Before any hearing, there would be disclosure of the various experts’ reports7 and
these would be sent to the court-appointed expert along with an explanation from
the judge as to the issue or issues to be addressed at the hearing and the court-
appointed expert’s role.

6.16 The court-appointed expert would also be provided with the information on which
the challenged expert opinion evidence is founded8 and the judge would direct
the court-appointed expert to provide the parties and the judge with a written
report, in advance of the hearing, setting out his or her preliminary view.9 Any
further evidence a party might wish to adduce at the hearing to counter the court-
appointed expert’s preliminary view would also need to be disclosed before the
hearing, with the court-appointed expert being given sufficient time to consider
the evidence and any objections to that preliminary view.

7
Under r 33.4 of the Criminal Procedure Rules 2010.
8
See paras 7.21 and 7.37 below.
9
Judge Jeremy Roberts QC told us that a report from the independent expert would be
desirable for two reasons. First, it would give the parties prior notice of any additional
material which ought to be brought to the attention of the judge and parties; and, secondly,
it would underline the fact that the independent expert, though appointed by the court, is a
witness like any other witness.

91
6.17 During the pre-trial hearing, which the court-appointed expert would attend, the
advocate for the party seeking to adduce the challenged opinion evidence would
make his or her opening submissions and call his or her expert witness or
witnesses to explain why the evidence is sufficiently reliable to be admitted, and
those witnesses would be subjected to cross-examination by the opposing party.
The judge would also be entitled to ask those witnesses questions at any stage,
in the normal way.

6.18 The opposing party’s advocate would then make submissions and call expert
witnesses to provide reasons why the challenged evidence is insufficiently
reliable to be admitted, and they too would face cross-examination and questions
from the judge.

6.19 The matters raised by the court-appointed expert in his or her written report
would no doubt form the basis of some of the questions asked by the advocates,
or by the judge, during examination and cross-examination of the parties’ own
expert witnesses.

6.20 The court-appointed expert would then provide his or her opinion evidence on the
question of evidentiary reliability. The principles of natural justice and the
accused’s right to a fair trial would demand that the court-appointed expert should
have to provide his or her opinion evidence from the witness box and that the
parties’ advocates should be able to ask him or her questions and make
submissions to the judge on the independent evidence provided.

6.21 We envisage that the procedure in relation to the court-appointed expert’s


evidence would be along the following lines. The judge would first ask the court-
appointed expert to confirm the contents of his or her written report, including the
summary of the expert’s qualifications and experience, and the judge might then
go on to ask supplementary questions in the light of the evidence already
presented by the parties’ witnesses. The parties’ advocates, acting on the advice
of their expert witnesses, would then put their own questions to the court-
appointed expert or raise further points for consideration. Again, it would be open
to the judge to ask questions during this part of the hearing to clarify or probe
points in the court-appointed expert’s evidence, although the judge would no
doubt be mindful of the need to ensure that the adversarial process is not unduly
disrupted by his or her interventions.

6.22 Once the court-appointed expert has provided his or her evidence, the parties’
expert witnesses could be recalled, if necessary, so that the judge could put
further points to them. The advocates would also have the opportunity to ask their
own questions about such points.

6.23 The parties’ advocates would then make their closing submissions and the judge
would give (or reserve) his or her reasoned ruling on admissibility. The judge
would determine the question of evidentiary reliability, and therefore admissibility,
in the light of all the evidence presented during the hearing, including the court-
appointed expert’s opinion evidence, and the submissions made by the parties.

92
6.24 The judge would rule on whether the proffered opinion evidence was sufficiently
reliable to be admitted. If the proffered opinion evidence was not admitted, it
would be open to the judge to rule on the opinion which could legitimately be
placed before the jury in the light of the evidence presented during the hearing.10

THE VIEW OF OUR CONSULTEES


6.25 There was considerable support for our view that a Crown Court judge should be
able to call upon an independent expert for exceptionally complex scientific
evidence. While it is fair to say that a significant number of our consultees
expressed some concern about the practical issues associated with selection,
cost and transparency, most consultees nevertheless believed that the
advantages of this reform measure would outweigh the perceived disadvantages.

6.26 Some of our consultees were extremely supportive of this measure. One
academic lawyer said it was “absolutely vital” and another (a statistician) thought
it was “an excellent proposal” because many trial judges may not be able to judge
scientific reliability without assistance from an independent expert. As noted
already, the British Psychological Society thought that it would be a useful power
for judges to have in cases involving psychological evidence, and the Royal
College of Psychiatrists suggested that the judge might require help from “expert
statisticians and those in scientific methodology as applied to mental health”.
Similarly, the Royal Statistical Society felt that we ought to be “more supportive of
a trial judge obtaining an independent expert to assist [him or her] when dealing
with new issues”; the United Kingdom Accreditation Service supported
independent experts for “complex scientific cases”, and we were told by Simon
Daniel, a Chartered Accountant in the Serious Fraud Office, that, “in the more
complex cases, involving complicated financial products or accounting
treatments”, it could be difficult for a judge to determine the question of
evidentiary reliability without a court-appointed expert.

6.27 One academic expert in forensic science11 even went so far as to suggest that
whenever forensic scientific evidence is fundamental to an issue in a criminal
case the judge should call upon an independent expert for guidance, the reason
being that an expert of this sort would not only be genuinely impartial but would
also be able to explain the alternative possible explanations of a phenomenon
(and likelihoods), particularly in the context of trace evidence. We believe such an
approach would be undesirable, given the potential for disruption and increased
costs it would bring, and bearing in mind the other reforms we recommend in this
report.12 We do, however, agree with the view of another consultee, Bruce
Houlder QC,13 that there could be “dangers for the criminal justice process” if
judges were unable to call upon independent assistance in some cases.

6.28 Bruce Houlder also made the very important point that the mere existence of the
power to appoint an independent expert, even if rarely used, would act as a
10
See paras 5.107 to 5.108 above.
11
Professor Pierre Margot (University of Lausanne).
12
See, in particular, paras 4.25 to 4.36 above (on impartiality) and paras 7.21(2)(c) and 7.22
to 7.25 below (on alternative explanations).
13
Director of Service Prosecutions.

93
deterrent against casual science, and might reduce costs in the long run as the
“market” in dubious expertise falls. He summarised his principled argument for a
new statutory power to appoint independent experts in the following terms:

(1) judges routinely choose between the evidence of one witness and
another in making civil judgments and in some criminal cases, so they
would be able to give the evidence of a court-appointed expert
appropriate weight;

(2) judges are already frequently required to make judgments about factual
circumstances, including matters of expertise outside their experience;14

(3) judges already have the power to exclude expert evidence without the
assistance of an independent expert, so there can be no objection to a
judge being assisted by an appropriate expert who has been cross-
examined by the parties;

(4) judges in criminal cases already have the power to call evidence of their
own motion, albeit a power they exercise only in cases of real need;

(5) any aberrant judgment would be open to appeal.

6.29 Importantly, the Rose Committee of the Senior Judiciary also agreed that it would
be useful for the trial judge to have a power to call upon an independent expert in
exceptional cases (subject to the limitations we set out in our consultation
paper).15 Mr Justice Treacy, responding separately, also supported the measure,
but stressed, in line with our own view, that it should always be for the judge to
determine whether assistance was needed and, if so, to make the ultimate
decision on admissibility.

6.30 A large number of other consultees also expressed support for our suggested
reform measure but, as mentioned above, raised practical concerns as a
potential obstacle. For example, the Crown Prosecution Service agreed with our
suggestion in principle, opining that the potential value of bringing in an
independent expert for areas of particular complexity could be significant, but
they were concerned as to how appointments would be made.16 Another body,
Forensic Access Ltd, welcomed our suggestion but argued that selection would
be both critical and problematic because independent experts would need to be
up-to-date and impartial, and they would need to have an understanding of the
forensic process.17 The Forensic Science Society felt that the independent expert
would need to “have the highest integrity with no affiliation to either prosecution
or defence”. In a similar vein, Northumbria University School of Law’s Centre for
14
On judicial fact-finding in Crown Court trials on indictment, see: R Pattenden, “Pre-verdict
judicial fact-finding in criminal trials with juries” (2009) 29 Oxford Journal of Legal Studies 1
and R Pattenden, “The proof rules of pre-verdict judicial fact-finding in criminal trials by
jury” (2009) 125 Law Quarterly Review 79.
15
Consultation Paper No 190, paras 6.68 to 6.71.
16
This problem also arises when the common law power is relied on, of course.
17
Another consultee, LGC Forensics, felt there would need to be agreement between the
parties as to the independence and expert status of the appointed expert and that this
might be difficult to achieve.

94
Criminal and Civil Evidence and Procedure felt that it would be difficult to identify
an appropriate individual for developing fields; and another consultee suggested
that it might be difficult to find an impartial expert for contentious medical
hypotheses.

6.31 The Criminal Bar Association (CBA) supported the idea that the trial judge
“should have a discretion to call upon the assistance of an independent expert …
to decide upon admissibility in exceptionally difficult cases”, but queried giving the
parties the right to agree on an appointment. (The RSPCA also suggested that
court-appointed experts should be chosen by the court without interference from
either party.) The CBA instead proposed that there should be an appointments
panel containing representatives of the Law Society and the Bar Council acting in
accordance with a set of agreed criteria, to ensure a measure of professional
agreement as to the suitability of potential appointees.18

6.32 We believe there is a great deal to be said for the CBA’s suggestion, or
something very much like it. An independent appointments panel of experienced
lawyers chaired by a Circuit Judge could liaise with professional organisations
such as those referred to in paragraph 6.26 above, consider possible candidates
against relevant criteria – knowledge, qualifications, experience, impartiality, no
appearance of partiality and no misconduct – and submit a shortlist of eminent
individuals from which the judge would be able to make his or her selection.19
This process would bring important measures of scrutiny, independence and
transparency to the selection process, mitigating or removing any possible
concerns there might otherwise be as to the suitability of the individual appointed
or the judge’s involvement in the proceedings. For example, a panel would not
shortlist an expert, no matter how eminent, if he or she had only ever provided
expert evidence for the prosecution and had consistently refused requests to
provide expert evidence for the defence. There would be an important issue of
apparent bias which would mean that his or her appointment as an independent
expert would be unacceptable.20

6.33 Additionally, the judge would not have to rely on the endeavours and agreement
of the parties. Although reliance on the parties would be considerably less
complex than the scheme suggested by the CBA, we now concede that it is
perhaps unrealistic to expect the parties to reach an agreement on a matter such
as this. The parties would be unlikely to co-operate on the suitability of any
individual or pool of individuals; and even if they were willing and able to come to
a joint position, the judge would not necessarily agree on the suitability of their
candidate.

18
Unlike the CBA, the Bar Law Reform Committee had serious reservations about the
suggested reform measure. The Committee expressed concern as to how the judge would
make his or her choice and the danger that he or she could be seen as less impartial by
becoming involved in evidential issues.
19
We appreciate, of course, that such a scheme would depend on the availability of such
experts.
20
For the less stringent requirements for experts called by the parties, see paras 4.33 to 4.36
above.

95
6.34 An independent appointments panel would be far better for the trial judge for the
reasons given above and also because, in the absence of any such panel, the
judge would largely be left to his or her own devices. We might add that if the
appointments panel were to comprise volunteers drawn from the legal profession,
as we envisage, this reform measure could be implemented relatively
inexpensively, with little recourse to public funds. We accept, however, that some
funding would be required to cover the basic administrative involvement of the
Ministry of Justice (for example, drafting correspondence and maintaining
records).

6.35 Returning to the views of our consultees, the minority who opposed the idea of
court-appointed experts principally cited practical objections. That is to say, they
were worried that, if there were a new statutory power to appoint an independent
expert witness, the benefits might be outweighed by the associated costs and
difficulties. The Council of HM Circuit Judges recognised the argument for
assistance for some types of case – for example, where the outcome of a case
might hinge upon the interpretation of statistical evidence – but expressed
concern about selection and impartiality and concluded that, on balance, these
practical difficulties would outweigh the benefits. The London Criminal Court
Solicitors’ Association also objected on a cost-benefit basis (although they also
feared that the appointed expert might decide the question of admissibility).21

6.36 The Academy of Experts expressed considerable unease about our suggestion
because of practical concerns about a loss of transparency, selection, the parties’
involvement (and right to object) and the possible cost. Questions were also
raised by other consultees as to how an independent expert would be appointed,
the issue of payment, the role of the parties (and whether they could object to a
selection), the difficulties associated with finding some suitably qualified
individuals and the selection criteria.22

6.37 It is worth pausing here to note that a scheme of the sort proposed by the
Criminal Bar Association, in tandem with the procedure outlined above, would
meet these concerns, save that, like any reform measure, there would inevitably
be some cost implications.

6.38 The Criminal Cases Review Commission objected to our suggested reform
measure on a different basis. They felt that the views of an independent expert
could lead to entrenchment of a preliminary view that a technique is reliable,
thereby stifling the need for further development; or the advice of such an expert
could lead to a view that a technique is unreliable, stifling further progress of that
technique for that reason. They suggested that, rather than calling on
independent help, the judge should be proactive by, for example, calling for
further information from the expert under examination or requesting evidence
from a further expert.

21
We do not believe that a court-appointed expert would usurp the trial judge’s role, as we
explained in Consultation Paper No 190, paras 6.68 to 6.69.
22
Attributes mentioned by consultees were impartiality, integrity, being a leader in the field
with up-to-date knowledge, the ability to exercise independent judgment and having both
forensic and academic experience.

96
6.39 We agree that the trial judge should be proactive; but we doubt whether the
power to call upon an independent (court-appointed) expert would stifle progress.
An expert of this sort would simply provide further information and therefore
assistance. It would in all cases be for the trial judge to determine admissibility
and, in appropriate cases, the judge’s ruling could be challenged on appeal with
further expert evidence being presented. Moreover, we believe it is highly unlikely
that the research endeavours of a particular scientific community would be stifled
by a judicial ruling on admissibility, at least in the long term. On the contrary, we
believe that a reasoned judicial ruling pointing out weaknesses in a particular
methodology could go some way towards stimulating appropriate research to
rectify the problem.

6.40 In any event, if we are to accept that a Crown Court judge should be able to call
for evidence from a further expert, it would make sense if the expert who is called
to provide such evidence is widely recognised in the field as having special
knowledge and has been independently screened to ensure impartiality and no
appearance of partiality. The Criminal Bar Association’s suggested panel, or
something like it, would undertake this screening function, acting in accordance
with a set of agreed criteria to ensure the suitability and acceptability of any court-
appointed expert.23 This would be a better approach in principle than the
alternative of calling for further expert evidence from one or more of the parties.

OUR REVISED APPROACH


6.41 Given the considerable support amongst our consultees for the reform measure
mooted in our consultation paper, the absence of any compelling objections
based on principle, and the likelihood that Crown Court judges will occasionally
need assistance when assessing evidentiary reliability against our new statutory
test, we now believe there should be a new statutory power to appoint an
independent expert in some cases.

6.42 We have moved on from the suggestion in our consultation paper, however,
because we also now believe, for reasons already given, that if assistance is to
be provided the judge requiring it should be provided with a structured basis for
finding and appointing an expert. This structured basis should incorporate not
only measures designed to ensure that the expert is properly screened,24 but also
rules to safeguard the parties’ rights and ensure transparency.

6.43 We must qualify these opening comments, however, with two important caveats.
The first is that, regardless of the principled arguments for a new scheme of this
sort, we should formally recommend it only if it would be effective in practice. The
second caveat is that, even if it would be practicable, we should recommend it
only if we believe it would be a cost-effective measure. We return to these issues
below.

23
As explained in para 6.32 above, the relevant requirements would be knowledge,
qualifications, experience, impartiality, no appearance of partiality and no misconduct. On
the last factor, the applicants would be considered in the same way as potential judges:
minor offences (such as minor road traffic violations) do not prevent a person from holding
a judicial office and they would not prevent an expert from being a court-appointed expert.
24
Above.

97
6.44 In principle, then, we support the idea that there should be an independent panel
(such as that proposed by the Criminal Bar Association) and we believe that
measures should be incorporated into the selection process to ensure
independence, transparency and the proper scrutiny of potential appointees. An
expert selected by the judge from a shortlist compiled by such a panel would be
able to provide disinterested, cogent assistance in a case where the judge is
required to determine the evidentiary reliability of particularly complex evidence.
Moreover, and just as important, the transparency and independence of the
selection process would ensure that the interests of the parties are properly
protected and would meet any concerns relating to the position of the judge.

6.45 We therefore believe there should be an independent, non-governmental panel of


experienced legal professionals (barristers and solicitors) which would undertake
the initial stage of the selection process in accordance with a set of agreed
criteria. That is to say, a quorate body of available panel members, chaired by an
experienced Circuit Judge, would convene when required.25 This quorate body of
panel members (“the panel”) would liaise with a relevant professional body (in
practice, a professional scientific or mathematical body) to create a shortlist of
eminent experts in the field, screened for their impartiality, special knowledge,
experience and good character.

6.46 An independent panel representing the views of prosecution and defence


lawyers, and chaired by a Circuit Judge, would ensure that the interests of all
parties would be protected during the initial stage of the selection process.

6.47 When the question of an appointment first arises, the trial judge would prepare a
draft note to be passed to the panel identifying the problem to be resolved and
setting out any specific points on which assistance was required. This draft would
first be passed to the parties for their comments, and would then (either in its
original form or as amended in the light of comments) go to the panel to help its

25
We envisage a pool of potential volunteers from the legal profession, from which a number
would be chosen to form an ad hoc panel at short notice, when required. There would
need to be agreement between the Law Society and Bar Council in liaison with the Ministry
of Justice to ensure that panel members are properly competent to sit as such and that
due regard is paid to equality issues when selecting volunteers and sitting as a panel. We
envisage that the Ministry of Justice’s administrative support for the panel would be based
in London, but equally we envisage that the professional membership of the panel would
be spread throughout England and Wales and that ad hoc appointment panels derived
from this pool would be able to meet outside London.

98
members identify suitable candidates.26 The panel would liaise with relevant
professional bodies and provide the judge with a shortlist of suitable individuals.27

6.48 The shortlist would include an appendix summarising the individuals’ relevant
attributes (including, ideally but not necessarily, an understanding of criminal
proceedings and the trial process). The question of selection from the shortlist
would principally be for the trial judge, but if the parties were willing to agree on a
particular individual on the list, the judge would probably agree with that choice
and appoint that individual. Ultimately, however, the final decision would lie with
the judge. It would always be open to the judge to override the parties’ wishes
and appoint a different individual if he or she concluded that there was a better
candidate. The individual selected would be a court-appointed, independent
expert whose fee would come from the courts’ central funds.

6.49 The selection criteria would be available to the parties as would be the
correspondence between the panel and relevant professional bodies and the
minutes recording the panel members’ deliberations. The panel would set out its
reasons for selecting the shortlisted experts and these too would be available to
the judge and the parties. It would make sense if the parties were first asked if
they might be able to agree to the appointment of a particular candidate from the
list. However, if agreement proved to be impossible, the judge would notify the
parties that he or she was provisionally minded to choose a particular individual
from the shortlist and invite submissions or alternative suggestions from the
parties; or the judge would invite the parties to make submissions for or against
the shortlisted experts more generally, without first suggesting a particular
individual. We envisage that preliminary matters of this sort could be resolved in
writing according to a set timetable.

6.50 Given the composition and independence of the panel, the transparency of the
process, and the judge’s role in deciding whom to select, a submission from
either party that a particular individual was unsuitable is unlikely to be well
founded. Nevertheless, a party (or indeed both parties) would be able to object to
a particular individual if there was a sound evidential basis for the objection and it
is something the panel overlooked or to which the panel attached insufficient
weight. For example, it might be that a shortlisted expert had never provided
evidence for the defence but had appeared in countless cases for the prosecution

26
Following a discussion with the parties’ advocates, the judge’s note would form the basis of
the instructions to the court-appointed expert. These instructions would explain the nature
of the party’s proffered evidence, the reason why there are doubts as to its reliability
warranting a pre-trial hearing, an explanation of the statutory reliability test and an
explanation of the court-appointed expert’s role and relevant procedure.
27
We would expect these individuals to maintain an up-to-date curriculum vitae setting out
their relevant qualifications, publications and academic and forensic experience. They
would also have to set out facts relevant to their character (material criminal convictions,
adverse disciplinary findings and the like). This information would be passed to the panel.
We would ordinarily expect the panel to compile a shortlist on the papers presented to it,
but it would be open to the panel to make further enquiries as appropriate. The panel
would be able to provide a shortlist of one or more individuals, depending on the field and
the availability (or unavailability) of suitable experts.

99
and the problem of apparent bias was not considered during the selection
process.28

6.51 A new statutory power incorporating such safeguards would be a substantial


improvement over the current legal position. The common law power – assuming
it extends to the present situation – provides no mechanism for ensuring
transparency or for ensuring that a court-appointed expert is sufficiently qualified
for the role. In addition, as Judge Jeremy Roberts QC pointed out to us when
referring to the common law power, many trial judges may be “worried about the
mechanics of calling the witness themselves, or of being accused of ‘entering the
arena’, or simply of ‘something going wrong’ and causing the trial to have to be
aborted or any conviction to be quashed as being unsafe”.

6.52 The process we favour in principle would be transparent; it would allow the
parties to be involved by agreeing to a shortlisted candidate or objecting to a
candidate; the parties would be allowed to question the independent expert
witness; and the judge would treat the court-appointed expert’s witness’s
evidence in the same way as any other witness’s evidence: the judge would give
this expert’s evidence as much weight as he or she thinks it deserves and would
be under no obligation to accept it. The parties would be able to make
submissions on the suitability of the shortlisted experts and on the appointed
expert’s evidence. The rights of the accused and the prosecution would be fully
protected.

6.53 In the light of our decision to abandon the dichotomy we originally proposed for
our statutory guidelines (as between scientific and non-scientific evidence), we
now accept that it would be undesirable and unnecessary to limit the field on
which an independent court-appointed expert should be able to provide
assistance. As a matter of practical reality, however, a Crown Court judge would
be unlikely to wish to seek independent assistance on matters which do not relate
to complex evidence of a scientific nature or involve complex statistical evidence,
as we intimated in paragraph 6.45 above.

6.54 We do not believe a scheme of the sort we favour would give rise to further
delays in criminal proceedings because the process of selection and appointment
would take place in parallel with the ongoing preliminary proceedings leading up
to the pre-trial hearing on evidentiary reliability. In addition, any risk of further
delays could be minimised if the panel’s administrative support established early
links with the various professional bodies, giving those bodies the opportunity to
compile a list of potential candidates who would be willing in principle to accept
an appointment.

6.55 Nevertheless, consistent with the suggestion we originally outlined in our


consultation paper, we still believe there should be a restriction on the power to
appoint an independent expert for an admissibility hearing, even more so if the

28
Northumbria University School of Law’s Centre for Criminal and Civil Evidence and
Procedure objected to the idea that the judge should be able to call upon an independent
adviser on the ground there are a number of fields in which the most reputable or highly-
regarded experts seem to work predominantly for the prosecution, with access to better
support in terms of training, and this might create the impression of unfairness. We agree
that the appearance of bias is an important consideration in this context.

100
appointee were to be selected from a shortlist compiled by an independent panel.
There is a pragmatic reason for this approach, but it is also based on principle.
We take the view that properly-trained Crown Court judges should ordinarily be
able to narrow the issues and determine reliability without the assistance of a
court-appointed expert, so the power to appoint should not be available for all
types of case.

6.56 The pragmatic reason is the desirability of minimising the likely costs and
inconvenience associated with the exercise of a statutory power of this sort. A
court-appointed expert would have to be paid for his or her report and attendance
in court, and a quorate body of independent panel members would need to be
convened to liaise with relevant professional bodies and draw up a shortlist of
suitable candidates.

6.57 Accordingly, the conclusion we have reached is that any new statutory power
which would allow a judge to seek the help of an independent expert should be
drawn very narrowly. The power to appoint should be available only if the
complexity and the likely importance of the disputed opinion evidence are such
that it would be in the interests of justice to call upon the assistance of an
independent expert. According to these criteria, the assistance of a court-
appointed expert would not be in the interests of justice in the vast majority of
criminal cases involving expert opinion evidence, so the power would be relied on
only very rarely.

6.58 This brings us to the first of the two caveats mentioned in paragraph 6.43 above.
Because our modified version of the Criminal Bar Association’s suggestion is
undeniably more complex than the alternatives (leaving the judge to his or her
own devices or calling upon the parties to agree on a compromise candidate) we
decided that we should formally recommend this new scheme only if we could be
confident that it would work in practice. To this end, we sought the advice of four
very experienced individuals, whom we now refer to as our “advisers”.29

THE EFFICACY OF OUR PROPOSED SCHEME


6.59 Three of our four advisers provided very positive responses on the proposals
outlined above, and their workability, save that one preferred a simpler system
which would depend on the parties reaching agreement on a suitable expert.30

6.60 Our fourth adviser set out a number of practical objections to our proposed
scheme.31 He was concerned that there would be time-consuming contests over
issues such as partiality and appointment and therefore opined that the potential
drawbacks associated with this scheme would outweigh the benefits. His
particular concern was the likely perception that, as a result of selection from an
approved list, the court-appointed expert would come with a judicial presumption
of correctness (that is, the judge would be slow to disagree with the court-
appointed expert’s view) and the court-appointed expert would therefore be

29
Anthony Edwards, a solicitor; Bruce Houlder QC, Director of Service Prosecutions; Edward
Rees QC; HH Judge Jeremy Roberts QC.
30
Anthony Edwards.
31
Edward Rees QC.

101
perceived to be a significant adversary by one of the parties and an ally by the
other.

6.61 In answer to this point, we believe the judiciary would recognise the desirability
and importance of not simply deferring to the view of a court-appointed expert,
particularly as the judge in the case would give a reasoned ruling on the question
of evidentiary reliability, explaining why he or she considers the party’s proffered
opinion evidence to be sufficiently or insufficiently reliable to be admitted. As we
explained in our consultation paper, the trial judge would treat with the greatest of
respect the views of the expert appointed to provide assistance, but the final
judgment on reliability would always remain a legal issue for the judge.32 The
judge would therefore give the court-appointed expert’s opinion the weight he or
she thinks it deserves, like any other item of evidence. The judge would not
simply defer to the view of the independent expert.

6.62 It should also be borne in mind that the court-appointed expert would simply be
appraising methodology and whether a party’s expert’s opinion as to an
affirmative proposition is logically in keeping with the research data (and so forth)
underpinning it, so the judge might allow the party’s expert to give an opinion, but
not the opinion originally proffered. Moreover, the reasons for the court-appointed
expert’s opinion would be articulated in his or her preliminary report, allowing the
parties to meet any criticisms or objections by the time of the hearing, and the
court-appointed expert would give oral evidence during the hearing and face
cross-examination (perhaps even robust cross-examination) on his or her
analysis of the situation. A party might therefore regard the court-appointed
expert to be an adversary, but this is likely to be the case only where that party’s
expert opinion evidence is flawed and the party appreciates (or ought reasonably
to appreciate) that this is the case.

6.63 We also envisage that in some cases an expert proffering a challenged opinion
will value the input of a court-appointed expert and offer a revised opinion in the
light of the court-appointed expert’s report, resulting in a shorter pre-trial hearing
on evidentiary reliability or obviating the need for any such hearing at all.

6.64 Our fourth adviser was also concerned by possible delays caused by challenges
to the appointments process. We accept that there might on occasion be a delay
during the pre-trial proceedings as the selection process runs its course, but we
do not accept that meaningful delays would be caused by such challenges. The
independent composition of the panel, the transparency of the selection process,
the fact that the panel would merely provide the judge with a shortlist of suitable
experts and the parties’ right to lodge objections as to the suitability of anyone on
the shortlist suggest there would be no significant problem in this respect. It
should also be remembered that the court-appointed expert would merely be
providing the trial judge with additional assistance on matters relating to expert
evidence. The judge would therefore treat the court-appointed expert as a
witness like any other expert witness, further undermining any objection to his or
her being appointed to give evidence.

32
Consultation Paper No 190, para 6.70.

102
6.65 Indeed, because of the very narrow remit of the appointments panel, the
safeguards built into the selection process (including transparency and the
parties’ right to raise objections), the appointed expert’s status as a witness like
any other witness, the obligation on the judge to give a reasoned ruling on
evidentiary reliability after an admissibility hearing, and, most importantly, the
availability of an appeal to the Court of Appeal, the Government will no doubt
wish to consider preventing challenges by way of judicial review in the High
Court, whether in relation to the selection process itself or the resulting shortlist.

6.66 A judicial ruling preventing the prosecution from adducing complex expert opinion
evidence, in the light of a court-appointed expert’s evidence, could allow the
prosecution to lodge an interlocutory appeal to the Court of Appeal, during which
the court-appointed expert’s opinion evidence (if accepted by the judge) would
again be scrutinised.33 Equally, a decision to exclude defence expert opinion
evidence in the light of a court-appointed expert’s opinion evidence could be
challenged on appeal if the accused is ultimately convicted.34

6.67 In the light of the responses we received on the original suggestion in our
consultation paper, and on the modified scheme placed before our four
experienced advisers, we are satisfied that a new statutory power permitting a
Crown Court judge to appoint an independent expert selected from a shortlist
provided by an independent selection panel is a power which would be desirable
in principle and effective in practice. We believe the selection process we favour
would be a substantial improvement over the current common law position.

6.68 This brings us to the second caveat mentioned in paragraph 6.43. The question
here is whether the benefits of the proposal would outweigh the potential for
additional costs, inconvenience and possibly also delays which a scheme of this
sort might engender. We take the view that a new appointments scheme should
be recommended only if the benefits would outweigh the costs.

COSTS, INCONVENIENCE AND POSSIBLE DELAYS


6.69 Given that court-appointed experts would be relied on only very rarely, we believe
the amount of money likely to be involved, in relation to the appointments process
and payment of fees from the courts’ central funds, would be low.35 Nevertheless,
in cases where a Crown Court judge decided to make use of the facility there

33
See in particular s 58 of the Criminal Justice Act 2003 which provides the prosecution with
a right to seek leave to appeal in respect of certain adverse rulings, whether the ruling was
made before or during the trial (s 58(13)). Note also the possibility of an interlocutory
appeal under s 35 of the Criminal Procedure and Investigations Act 1996 (in respect of
rulings made during a “preparatory hearing” at the start of a complex, serious or lengthy
case).
34
Note also the possibility of an interlocutory appeal under s 35 of the Criminal Procedure
and Investigations Act 1996 (above).
35
We do not think that competitive tendering would be necessary, but that the panel should
be permitted to fix a reasonable fee for the short-listed candidates on a case-by-case
basis, in line with guidance established by the Lord Chancellor. For present purposes,
however, we assume that a fixed daily or hourly rate could be set in line with the legal aid
rate for the parties’ experts.

103
could be increased costs, and possibly occasional delays, as a result of the
selection process.36

6.70 This does not mean that there would be delays, however. The selection process
would be undertaken alongside the conventional preliminary proceedings for
complex cases of this sort, where a Crown Court judge has ordered a pre-trial
hearing to address evidentiary reliability. Moreover, as we have already
intimated, the pre-trial hearing on admissibility could be shorter than otherwise, or
could even be rendered unnecessary, if the parties were to have advance
disclosure of an independent expert’s report or the trial judge were to have the
benefit of a court-appointed expert’s oral and written evidence. The pre-trial
hearing, and the judge’s enhanced knowledge and understanding gained from
having the help of a court-appointed expert, could also mean that the provision of
expert opinion evidence during the trial would be managed more efficiently or
limited, with concomitant savings in time and costs.

6.71 In short, we do not envisage any significant lengthening of proceedings in cases


of this sort or any significant additional costs being incurred by HM Courts
Service, the Crown Prosecution Service or the Legal Services Commission. Nor
do we believe there would be any significant additional costs for other
participants in the proceedings (such as HM Prison Service, the accused and
other witnesses). Additional costs and delays could result from the fact that a trial
judge has ordered a pre-trial admissibility hearing on reliability; but equally the
appointment of an independent expert could feasibly shorten the hearing or any
subsequent trial or both.

6.72 In any event, even if there might be additional delays in some cases because of
the selection and appointment process, these could be significantly shorter than
the delays which might be associated with the exercise by the trial judge of his or
her current common law power to find a suitable independent expert.

6.73 We accept, of course, that there would be some additional costs associated with
setting up and providing administrative support for an independent selection
panel. However, these costs should be low, given the limited support required
and the likely rarity of the process being applied. Accordingly, such costs could
probably be incorporated into an existing budget, at least in part. Indeed, given
that many public servants voluntarily take on additional tasks for the benefit of
their department, it is likely that the process could be run wholly or partly on the
back of volunteers in the Ministry of Justice or its associated offices, using
existing resources (meeting rooms, information technology support, filing and so
on). Certainly we have no doubt that the standing pool of panel members, from
whom the smaller ad hoc panels would be selected, could be drawn from
volunteers in the legal profession.37

36
It would take time to agree an ad hoc panel from the wider pool and there would need to
be an exchange of information between this panel and the relevant professional bodies.
Time would also need to be factored in for the involvement of the parties.
37
The fact that an ad hoc panel would be compiled from a larger standing group of
volunteers throughout England and Wales should mean that a small panel of this sort
could be convened at relatively short notice, wherever and whenever it is needed.

104
RECOMMENDATION
6.74 If trial judges are to be given a new reliability test for determining the admissibility
of expert evidence – in line with our recommendation in Part 5 – then it is likely
that Crown Court judges will occasionally need further expert assistance for some
of the myriad types of evidence tendered for admission in trials on indictment. If
this is accepted, then there are really only two alternatives for a trial judge faced
with determining the reliability of very complex evidence. The judge can be left to
cope as best he or she can by using the general common law power; or the judge
can be provided with a shortlist of eminent experts in the field who have been
independently screened for suitability, giving the judge access to the best expert
assistance available.

6.75 The second of these options is clearly desirable and preferable in principle. In
addition, if the new statutory power to appoint were to be limited so as to be
available only when really warranted, we believe a reform measure of this sort
could be achieved for little additional cost and with little if any adverse impact on
the length of criminal proceedings generally.

6.76 In the following paragraphs we therefore set out our recommendations for such a
power, including important limitations on its availability. However, on the basis
that there will be an additional start-up cost and the need for ongoing
administrative support for the panel, and because the Government might wish to
see whether the power is necessary – for example, by monitoring the extent to
which the common law power provides an adequate alternative – the relevant
clause in our draft Bill is free-standing and severable from the rest of the Bill.

6.77 The relevant clause – clause 9 – could therefore be brought into force some
years after the rest of the Bill; or it could be removed if the Government
concludes that the cost implications associated with an appointments panel
would outweigh the additional benefits the scheme would bring. If either approach
is adopted, the Court of Appeal might in due course wish to provide guidance on
the availability and use of the common law power in the context of admissibility
hearings on evidentiary reliability.

6.78 Subject to those caveats, we recommend that a Crown Court judge (for a
trial on indictment) should be provided with a statutory power to appoint an
independent expert to assist him or her when determining whether a party’s
proffered expert opinion evidence is sufficiently reliable to be admitted.38

6.79 We recommend that this power should permit a Crown Court judge to
appoint an independent expert only if he or she is satisfied that it would be
in the interests of justice to make an appointment, having regard to:

(a) the likely importance of the expert opinion evidence in the


context of the case as a whole;

(b) the complexity of that evidence, or the complexity of the


question of its reliability; and

38
Draft Bill, cl 9(1) and (2). This power would replace the common law power in this respect,
insofar as the common law power extends to the present situation (see cl 9(8)).

105
(c) any other relevant considerations.39

6.80 We further recommend that the judge should make his or her appointment
from a shortlist of individuals prepared by an independent panel of legal
practitioners, chaired by a Circuit Judge, reflecting the interests of both the
prosecution and the defence.40

6.81 An example of another relevant consideration for the purposes of paragraph


6.79(c) might be the fact that the party challenging the reliability of expert opinion
evidence has not called an expert witness to provide support for the challenge,
leading the judge to conclude that he or she needs the assistance of a court-
appointed expert. Alternatively, if the case is one where, exceptionally, the
question of reliability has to be addressed during the trial, the judge will no doubt
take into account considerations such as the likely disruption to the proceedings
and the delay and distress this would cause.

6.82 Our recommended statutory power would be available only in Crown Court cases
to be tried on indictment.41 There would be no power to appoint an independent
expert witness (to help determine whether a party’s expert opinion evidence is
sufficiently reliable to be admitted) in a magistrates’ court or in a Crown Court
hearing an appeal from summary proceedings.42

6.83 We have come to the conclusion that additional expertise will on occasion be
warranted for Crown Court trials on indictment, notwithstanding the additional
costs involved, because of the extremely complex expert evidence which can be
presented in such cases.

6.84 The argument for additional expertise in summary proceedings is considerably


weaker, and there was little support amongst our consultees for an extension of
this possible reform measure to such proceedings. Only the Crown Prosecution
Service suggested that magistrates’ courts might benefit from the appointment of
an independent expert in exceptional cases. The Justices’ Clerks’ Society did not
put forward any similar proposal.

39
Draft Bill, cl 9(2)(a) to (c).
40
Clause 9(3) to (6) of our draft Bill sets out the basic framework for the creation of
procedural rules to give effect to our recommendation and to remunerate court-appointed
experts.
41
Draft Bill, cl 9(1) and (8).
42
Above.

106
PART 7
PROCEDURAL MATTERS

INTRODUCTION
7.1 In Part 1 of this report we set out the reasons why we believe expert evidence is
a special type of evidence in criminal proceedings and why, therefore, such
evidence demands special treatment beyond the rules which apply to evidence
generally. Informed by our reasoning and conclusion in Part 1, we now turn to the
desirability of new procedural rules which would ensure that the new admissibility
requirements in our draft Bill would work effectively in practice.

7.2 The special nature of expert evidence is already recognised in primary and
secondary legislation. There are currently powers to create procedural rules on
the pre-trial disclosure and exclusion of expert evidence in criminal proceedings;1
and such rules, alongside other procedural rules, are now to be found in Part 33
of the Criminal Procedure Rules 2010:2

(1) Rule 33.3(1) sets out the matters which an expert’s report must contain if
tendered for admission in criminal proceedings.

(2) Rule 33.4(1) provides that if a party wishes to adduce expert evidence in
criminal proceedings, the expert evidence must be served (as an expert’s
report) on the other parties and on the court “as soon as practicable, and
in any event … with any application in support of which that party relies
on that evidence”.

(3) Rule 33.4(1) also provides that, following a request, another party must
be given a copy of (or a reasonable opportunity to inspect) the records of
any “examination, measurement, test or experiment on which the
expert’s findings and opinion are based, or that were carried out in the
course of reaching those findings and opinion, and anything on which
any such examination, measurement, test or experiment was carried
out”.

(4) Rule 33.4(2) provides that, if a party seeking to adduce expert evidence
does not comply with the requirements of rule 33.4(1), such evidence
cannot be adduced unless all the parties agree that it should be admitted
or the court gives leave for it to be admitted.

7.3 These obligations, and the potential sanction for non-compliance, apply to the
defence as well as to the prosecution, representing an important deviation from
the general principle that the accused is under no obligation to disclose his or her

1
Section 81 of the Police and Criminal Evidence Act 1984 and s 20(3) and (4) of the
Criminal Procedure and Investigations Act 1996 set out the powers to include in the
Criminal Procedure Rules provisions which (1) require a party to make pre-trial disclosure
of expert evidence which a party proposes to adduce and (2) prohibit the adduction of such
evidence if that party fails to make pre-trial disclosure as required.
2
Part 33 of the Rules is set out in full as Appendix B to this report.

107
evidence in advance of the trial.3 In addition, the possibility that the court will
prevent the accused from being able to adduce expert evidence under
rule 33.4(2) stands in stark contrast to the sanction of an adverse inference being
drawn if the accused fails to comply with his or her other disclosure obligations.4
Rule 33.4(2) is somewhat anomalous, therefore: it provides the criminal courts
with a discretion as to the admissibility of defence evidence which would
otherwise be admissible (if served).5

7.4 The special nature of expert evidence is also recognised at common law, as we
explained in Part 2. Importantly, the common law admissibility test developed to
provide a guarantee that expert evidence proffered for admission is sufficiently
helpful and reliable (in the round) to be taken into consideration by a jury in a
criminal trial. Its only major weakness, in our view, lies in the insufficient regard it
pays to the desirability of scrutinising the evidentiary reliability of expert opinion
evidence.

7.5 It also bears repeating that rule 33.2 of the Criminal Procedure Rules 2010
expressly provides that an expert “must help the court to achieve the overriding
objective [of the Rules] by giving objective, unbiased opinion on matters within his
[or her] expertise” and that this duty “overrides any obligation to the person from
whom he [or she] receives instructions or by whom he [or she] is paid”.6 Although
all witnesses in criminal proceedings are expected to provide impartial evidence,
it is only expert witnesses who are currently bound by an explicit obligation in
secondary legislation to be objective and impartial. For ease of exposition we
sometimes refer in this report to “defence experts” and “prosecution experts”, but
in truth there is no such thing as a defence or a prosecution expert. All expert
witnesses, whether called by the defence or by the prosecution, or indeed by the
court, are witnesses for the court with an overriding duty to provide objective,
impartial evidence. This overriding duty, a fundamental requirement rather than a
procedural issue, has been incorporated into clause 3 of our draft Bill.7

3
As a general rule the defence is under no obligation to disclose the evidence on which it
intends to rely, but there are exceptions. For example, if the accused is to be tried in the
Crown Court on indictment, he or she must make pre-trial disclosure of the particulars of
any defence of alibi (see fn 4 below). By contrast, the prosecution bears a heavy burden of
pre-trial disclosure, not only in respect of the evidence it intends to rely on at trial but also
in relation to any “unused” material which might reasonably undermine its case or support
the defence case.
4
For the accused’s disclosure obligations in the Criminal Procedure and Investigations Act
1996, see s 5 (obligation to provide a defence statement to the prosecution, for trials on
indictment); s 6A (what a defence statement must contain, including particulars of an alibi);
and s 6C (obligation to disclose details of defence witnesses). Section 11(5) allows an
adverse inference to be drawn from non-compliance.
5
Section 132(5) of the Criminal Justice Act 2003 provides a similar discretion for hearsay
evidence tendered for admission following non-compliance with the relevant rules in the
Criminal Procedure Rules 2010.
6
According to r 1.1(1), the overriding objective of the Rules is that “criminal cases be dealt
with justly”. Rule 1.1(2) sets out examples of what this requires.
7
Subsections (1) and (2). It should be noted that this duty was imported into the Rules from
the common law; see Harris [2005] EWCA Crim 1980, [2006] 1 Cr App R 5 at [271] and
Bowman [2006] EWCA Crim 417, [2006] 2 Cr App R 3 at [176].

108
7.6 It is the special nature of expert witnesses and the evidence they provide which
informs our recommendations in this Part. We have come to the conclusion that,
in addition to the existing powers in the Criminal Procedure Rules 2010, and the
new admissibility test we recommend for primary legislation, there should be a
number of other provisions. These additional provisions would further both the
overriding objective of the Rules and our own objective in making the
recommendations set out in Part 4 and Part 5, should those recommendations be
taken forward.

7.7 In particular, we believe a more stringent approach to pre-trial disclosure is


warranted for expert evidence so that matters bearing on the various limbs of our
proposed admissibility test can be properly investigated before the trial, whether
the party seeking to adduce the evidence is the prosecution or the defence.

7.8 Most of the recommendations in this Part are procedural, and for this reason
would best be effected, we believe, by the creation of additional rules within the
Criminal Procedure Rules 2010.

7.9 We do not therefore provide any clauses in our draft Bill for the recommendations
we set out below on procedure. Our draft Bill does, however, contain a clause –
clause 8 – which would extend the existing powers to make rules on matters
relating to expert witnesses.8

CLAUSE 8
7.10 Specifically, clause 8 of our draft Bill would permit the Criminal Procedure Rules
Committee to create rules requiring the parties to make pre-trial disclosure of
material information relating to the admissibility of their expert evidence, even if
such information, but for clause 8(2), would be protected by “litigation privilege”.9

7.11 Litigation privilege generally protects against the disclosure of:

x “communications between a professional legal adviser and his


client or any person representing his client or between such
an adviser or his client or any such representative and any
other person [such as an expert witness] made in connection
with or in contemplation of legal proceedings and for the
purposes of such proceedings”; and

8
The general power to make Criminal Procedure Rules has been conferred by s 69 of the
Courts Act 2003. As noted already, additional powers are provided by s 81 of the Police
and Criminal Evidence Act 1984 and s 20 of the Criminal Procedure and Investigations Act
1996.
9
Reform in this context would principally if not entirely relate to the defence because, as
noted in fn 3 above, the prosecution is already under an obligation to disclose the evidence
on which it proposes to rely and any “unused” information which might reasonably assist
the defence. See generally Part I of the Criminal Procedure and Investigations Act 1996,
Parts 21 and 22 of the Criminal Procedure Rules 2010 and the Attorney General’s
Guidelines on Disclosure,
www.attorneygeneral.gov.uk/Publications/Documents/disclosure.doc.pdf (last visited
3 February 2011).

109
x “items enclosed with or referred to in such communications
and made … in connection with or in contemplation of legal
proceedings and for the purposes of such proceedings, when
they are in possession of a person who is entitled to
possession of them.”10

7.12 We explain our recommendations on pre-trial disclosure below. It suffices here to


say that if a party – whether the defence or the prosecution – instructs an
individual to provide expert evidence, and that individual reveals material
information which suggests that he or she is not impartial (or credible) or is not
qualified to give expert evidence, or that his or her opinion evidence might not be
sufficiently reliable to be admitted, and that party persists in its desire to rely on
that individual’s evidence, we consider it to be right in principle that that party
should disclose the information to the other parties and to the court. The
alternative would be for the party to abandon that individual in favour of another
expert.11 We therefore believe a slight modification to the scope of litigation
privilege is justified in this context to ensure such pre-trial disclosure.12

7.13 We are confident that such a modification is justifiable in principle, desirable in


practice and compatible with the accused’s rights guaranteed under the
European Convention on Human Rights.13

7.14 Equally, we believe a sanction which would prevent the defence from relying on
expert evidence, where the defence has deliberately refused to comply with a

10
Section 10(1)(b) and (c) of the Police and Criminal Evidence Act 1984, reflecting the
common law. The general power in s 69 of the Courts Act 2003 does not allow the Criminal
Procedure Rules to displace this privilege; see R (Kelly) v Warley Magistrates' Court [2007]
EWHC 1836 (Admin), [2008] 1 WLR 2001. Nor are the powers in s 81 of the Police and
Criminal Evidence Act 1984 and s 20 of the Criminal Procedure and Investigations Act
1996 sufficiently wide to allow for the creation of disclosure obligations which would
displace this privilege. For an explanation and analysis of litigation privilege generally, see
C Passmore, Privilege (2nd ed 2006) pp 2, 40 to 48 and 143 to 265.
11
Section 6D of the Criminal Procedure and Investigations Act 1996 would place the defence
under an obligation to disclose the names and addresses of the individuals it has
instructed to provide expert evidence (to deter “expert shopping”), but this provision has
not yet been brought into force.
12
For the current procedural rules on pre-trial disclosure generally, see Part 22 of the
Criminal Procedure Rules 2010. The disclosure obligations specifically relating to expert
evidence are set out in Part 33 of the Rules.
13
The obligation to disclose such information would interfere with the relevant individuals’
right to respect for their confidential correspondence under Article 8(1), but this would
almost certainly be justified under Article 8(2). Importantly, information protected by “legal
advice privilege” (covering confidential communications between the accused and his or
her legal representatives) would continue to be fully protected against disclosure.
According to Michael Bowes QC, who commented on a draft of this report, our “proposed
limited incursion into litigation privilege (notably not into legal advice privilege) is both
reasonable and proportionate, in view of the special status held by expert witnesses”.

110
new disclosure obligation of the type just described, would be compatible with the
accused’s right to a fair trial under Article 6 of the Convention.14

7.15 However, in recognition of the desirability of interfering with litigation privilege


only to the minimum extent necessary to ensure that the court and parties are
properly equipped to address the admissibility of expert evidence, and given the
importance of not in any way inhibiting communications between the accused
and an expert or the way a defence representative instructs an expert, we have
framed clause 8 of our draft Bill as narrowly as we possibly can.

7.16 Clause 8(1)(a) and (b) provides the power to make rules which would require the
disclosure by a party of information relevant to the question whether expert
evidence which the party proposes to adduce is admissible (by virtue of
clause 1(1)(b) or (c) or (2)) or is worthy of belief; and clause 8(2) provides that
such information includes information otherwise protected by litigation privilege.15
However, clause 8(3) makes it clear that it would not be possible to make rules
which would remove the protection of litigation privilege from information
contained in a communication from the accused (or his or her representative) to
an expert. It is highly unlikely, of course, that any such communication would
contain information falling within the scope of clause 8(1), but by expressly
ensuring that procedural rules cannot extend to the privileged information in such
communications, the accused and his or her lawyers would not be inhibited in the
way they communicate with their experts in advance of the trial. It should also be
noted that any information currently protected by the accused’s legal advice
privilege (or privilege against self-incrimination) would continue to be protected
from disclosure.

AMENDING THE CRIMINAL PROCEDURE RULES


7.17 We have explained above that the parties and the trial judge should be in a
position to assess the individuals who are being called to provide expert evidence
and that they should also be properly equipped to scrutinise experts’ opinion
evidence for evidentiary reliability. Clause 8(1)(a) of our draft Bill refers to
“information relevant to the question whether expert evidence which a party
proposes to adduce is admissible” by virtue of clause 1(1)(b) (expertise) or clause
1(1)(c) (impartiality) or clause 1(2) (evidentiary reliability); and clause 8(1)(b)
refers to “information that might reasonably be thought capable of substantially
detracting” from an expert’s credibility.

7.18 Principally, we recommend amendments to Part 33 of the Criminal Procedure


Rules 2010 which would require experts to include additional matters in the

14
If the defence wilfully seeks to prevent the other parties and the court from having access
to material information about an expert or an expert’s evidence, a witness who has an
overriding duty to the court, it is difficult to see how the defence can argue that the
accused’s Article 6 rights have been infringed if, as a result, that expert’s evidence is
excluded. We also note that there is nothing in the case of Ensor [2009] EWCA Crim 2519,
[2010] 1 Cr App R 18 to suggest any violation of Article 6 if the defence is prevented from
relying on expert evidence through deliberate non-compliance with the disclosure
requirement in what is currently r 33.4 of the Criminal Procedure Rules 2010.
15
The disclosure rules would be set out in Part 22 or Part 33 of the Criminal Procedure
Rules. On information relating to an expert’s credibility, see paras 7.17 and 7.36 below.

111
written reports they disclose before the trial, and we address this issue under the
first of the following three sub-headings.

7.19 However, we also recommend amendments to the Rules which would place the
parties under a new obligation to disclose certain matters in advance of the trial,
and it is in this context that clause 8(2) of our draft Bill would bite.16 We discuss
this new disclosure obligation under the second sub-heading.

7.20 Under the third sub-heading we make recommendations which would build on the
judges’ current power in Part 33 of the Rules to direct experts to discuss the
expert issues in advance of the trial.

(1) Expert reports


7.21 We recommend that Part 33 of the Criminal Procedure Rules be amended to
include the following:

(1) a rule requiring an appendix to the expert’s report, setting out –

(a) sufficient information to show that the expertise17 and


impartiality requirements18 are satisfied;19 and

(b) a focused explanation of the reliability of the opinion


evidence with reference to the test and relevant examples
and factors in our draft Bill,20 concisely set out in a manner
which would be readily understood by a trial judge,21 along
with a summary of:

(i) other cases (if any) where the expert’s opinion


evidence has been ruled admissible or inadmissible
after due enquiry under the reliability test; and

(ii) other judicial rulings after due enquiry which the


expert is aware of (if any) on matters underlying the
expert’s opinion evidence;

16
Paragraph 7.16 above. The new obligation in the Criminal Procedure Rules to make pre-
trial disclosure would be supported by a sanction created under cl 8(1)(c); see para 7.38
below.
17
Draft Bill, cl 1(1)(b) and cl 2.
18
Draft Bill, cl 1(1)(c) and cl 3.
19
We include (a) here primarily for completeness, for these requirements are largely in the
Rules already. Rule 33.3(1)(a) requires details of qualifications, experience and
accreditation; and rule 33.3(1)(i) requires a statement as to the expert’s overriding duty.
We believe that an expert should also be able to state, as evidence of impartiality, that the
courts have not previously made an adverse ruling in relation to his or her evidence on the
ground of bias; but, if this is not the case, it should be sufficient if the expert simply has to
inform the party instructing him or her that his or her evidence has previously been excluded on
that basis. The party would then disclose that information separately; see para 7.37 below.
20
Draft Bill, cl 1(2), cl 4 and the Schedule.
21
By “focused” we mean that the expert would not need to provide the entire corpus of
knowledge on the area but only the direct foundation material for his or her opinion.

112
(2) a rule requiring an expert’s report to include –

(a) a statement explaining the extent to which the expert


witness’s opinion evidence is based on information falling
outside his or her own field of expertise and/or on the
opinions of other (named) experts;22

(b) a schedule identifying the foundation material underpinning


the expert witness’s inferences and conclusions;23 and

(c) a rule that where an expert witness is called by a party to


give a reasoned opinion on the likelihood of an item of
evidence under a proposition advanced by that party,24 the
expert’s report must also include, where feasible, a reasoned
opinion on the likelihood of the item of evidence under one
or more alternative propositions (including any proposition
advanced by the opposing party);25

(3) an extension of rule 33.4(2) so that, if a party seeking to adduce


expert evidence does not comply with the above requirements, the
evidence would be inadmissible unless all the parties agree that it
should be admitted or the court gives leave for it to be admitted.

7.22 The foregoing recommendations are largely self-explanatory and, we believe,


consistent with the overriding objective of the Criminal Procedure Rules. We do
however need to say something more about the proposed rule in
paragraph 7.21(2)(c) on the provision of alternative probabilities or likelihoods
(where feasible). We believe a rule of this sort is required for two reasons.

7.23 First, because all expert witnesses have an overriding duty to provide impartial
evidence, an expert should not provide an opinion on a particular probability or
likelihood favouring one party without explaining the probability or likelihood on
the alternative basis advanced by the opposing party (if it is feasible to provide
such an alternative). We would hope that the defence would disclose enough of
the accused’s case in his or her defence statement for a prosecution expert to be
able to provide an alternative probability or likelihood; but, as Professor Mike
Redmayne suggested to us, in the absence of an alternative defence proposition,
a prosecution expert should at least be explicit about what alternative hypotheses
he or she has considered and address the likelihood of the evidence occurring
under those hypotheses.
22
At present rule 33.3(1)(d) requires the expert to explain which of the facts stated in his or
her report are within his or her own knowledge.
23
At present rule 33.3(1)(b) to (c) requires details of any literature or other information relied
on and the “substance of all facts given to the expert which are material to the opinions
expressed”. As explained already, for the purposes of the reliability test, the expert would
not need to provide the entire corpus of knowledge on the area but only the direct
foundation material for his or her opinion.
24
That is, the likelihood that the item of evidence in question (say, a particular injury) would
occur under that proposition.
25
That is, the likelihood that the item of evidence in question (say, a particular injury) would
occur under that alternative proposition.

113
7.24 Secondly, a number of our consultees specifically argued for a requirement along
these lines so that any likelihood or probability an expert provides for the party
who called him or her could be seen in its proper context. The Association of
Forensic Science Practitioners, having cited their Standards for Expert Evidence
based upon the four principles of “balance, logic, robustness and transparency”,
proposed that forensic scientific experts should provide evidence that is balanced
in that they consider both the prosecution and the defence propositions; and they
suggested that the alternative propositions should be made apparent to the
expert to give him or her sufficient notice for proper consideration and
evaluation.26 Similarly, the Royal Statistical Society said:

In the evaluation of evidence at least two propositions need to be


considered. In the absence of a defence proposition, an alternative
could be provided by the evaluator such as “the opposite of the
prosecution proposition”. The role of probabilistic reasoning in the law
is to enhance the procedure for the evaluation of evidence under
each of two propositions, that of the prosecution and that of the
defence.27

7.25 Dr Phil Rose (of the Australian National University) argued that the alternative
probability should be given on two separate grounds: first, because it is a
necessary guarantee of impartiality; and, secondly, because, in relation to
“identification-of-the-source evidence”, the expert should be able to say how
probable the evidence is under both the prosecution and defence hypotheses (as
to how the evidence came to be where it was found), for if only one probability is
given under one hypothesis the opinion evidence is of “no use”.28

7.26 A related point made by the Bar Law Reform Committee was that an expert
should address opposing views in his or her report and add references to it in his
or her list of source material. Similarly, Adam Wilson (of Sheffield Hallam
University) argued, amongst other things, that experts should be trained to
identify matters which support or undermine their propositions.

26
An expert called by the defence should know the nature of the prosecution case in any
event, given the prosecution’s pre-trial disclosure obligations. An expert called by the
prosecution would no doubt receive from the prosecution the accused’s defence statement
(if available) and the expert reports on which the defence intends to rely.
27
The Society added the following comment: “A vital consideration in the assessment of
reliability is that the validation should go wider than just the technique – the court (and,
prior to that, the scientist) needs to assess whether there are sufficient data, knowledge
and understanding to assign robust likelihoods to the evidence, under the assumptions of
the [competing] propositions … presented.”
28
See also C Aitken and F Taroni, “Fundamentals of statistical evidence: a primer for legal
professionals” (2008) International Journal of Evidence and Proof 181, criticising Professor
Sir Roy Meadow’s statistic of one in 73 million (in Clark [2003] EWCA Crim 1020, [2003] 2
FCR 447, para 1.5 above) for two natural “cot death” incidents in a family on the ground
that, even if the figure was correct, the alternative likelihood of a double child murder in the
family should also have been given for a relative assessment, for in the absence of that
alternative “the figure has no significance or relevance”. A similar point was made by the
Forensic Science Service.

114
(2) Other pre-trial disclosure
7.27 A number of our consultees advocated greater pre-trial disclosure obligations (for
the defence) in relation to expert evidence.29 The Forensic Science Regulator
went further, suggesting that the defence should disclose reports prepared by
experts who are not called to testify so that prosecution witnesses would not be
ambushed by the use of the material in such reports. On this specific point, there
are several reasons why we believe that the reports prepared by defence experts
who are not called as witnesses should not be disclosed.

7.28 First, given the presumption of innocence, the concomitant obligation on the
prosecution to prove its case beyond reasonable doubt and the obligation on the
prosecution to demonstrate that the admissibility criteria for its expert evidence
are satisfied, we believe the onus should continue to be on the prosecution
expert witnesses to be prepared to justify their own opinion evidence (in terms of
admissibility, reliability and weight) without the assistance of the defence. Nor
should we overlook the fact that an undisclosed expert report is protected by
litigation privilege and may contain matters adversely affecting the accused’s
defence, potentially engaging his or her privilege against self-incrimination. The
justifications for compelling pre-trial disclosure of expert evidence the accused
wishes to rely on do not apply in relation to expert evidence the accused will not
be seeking to adduce.

7.29 Secondly, we believe the new admissibility test we are proposing, in tandem with
relevant organisational structures, will encourage greater care on the part of
prosecution expert witnesses to ensure that any opinion evidence they provide
will stand up to forensic scrutiny, thereby reducing the risk that they will be
ambushed by defence counsel armed with an undisclosed report.

7.30 Thirdly, we believe the suggestion that the defence should disclose such reports
would be unworkable. The defence could simply discuss the prosecution experts’
reports with a defence expert and obtain the material necessary for effective
cross-examination without commissioning a formal report. In some cases the
defence expert could support the defence advocate’s cross-examination of
prosecution experts by providing him or her with expert advice during the course
of the trial.

7.31 Fourthly, there is already legislation in place which could discourage the defence
from approaching experts with a view to obtaining reports which would never be
disclosed. If section 6D of the Criminal Procedure and Investigations Act 1996 is
brought into force, and supported by an appropriate sanction, the prosecution
would be given the names and addresses of experts the defence has instructed
but decided not to call. Because there is “no property in a witness”, the police
might even approach those individuals as possible experts for the prosecution.30

29
For the prosecution’s current disclosure obligations, see fn 9 above.
30
In practice, a conflict of interest would no doubt prevent many such experts from being
able to give evidence for the prosecution. Certainly such experts would not be able to
reveal or rely on matters relating to their previous instructions from the defence, because
litigation privilege would continue to protect such communications.

115
7.32 With regard to disclosure in other respects, Bruce Houlder QC31 argued that an
improved disclosure regime is “not only desirable but necessary”; the General
Medical Council felt that there should be complementary measures regarding
disclosure to assist in the effective screening of an opponent’s experts; the
Forensic Science Society suggested a pre-trial disclosure process which would
allow all parties to screen their opponents’ experts in respect of matters such as
qualifications, experience and extraneous conduct; and the RSPCA proposed
that expert witnesses should be required to make a declaration of any interest in
the proceedings.32

7.33 The Criminal Bar Association argued that there should be disclosure of names
and dates of previous cases in which expert witnesses have given evidence and
details of cases in which their evidence has been criticised. They also suggested
that there should be disclosure of the material forming the foundation from which
an expert’s conclusions are drawn (databases, photographs, recordings and the
like). Northumbria University School of Law’s Centre for Criminal and Civil
Evidence and Procedure suggested an obligation on all experts to include in their
reports adverse findings concerning their competence or credibility.

7.34 As we have already explained, we agree that there should be an enhanced


disclosure regime for expert witnesses and their evidence, regardless of which
party wishes to adduce the evidence.33 The special nature of expert opinion
evidence and all expert witnesses’ overriding duty to the court to provide impartial
evidence militate against the validity of any principled objection to a requirement
to disclose information relevant to the reliability (in the round) of expert evidence.
Part 33 of the Criminal Procedure Rules 2010 already sets out disclosure
obligations for expert evidence which apply to the defence and prosecution alike,
and we have already recommended an extension of rule 33.3 with regard to what
an expert’s report should contain.

7.35 Further disclosure obligations are also warranted by the new admissibility
requirements we have recommended in this report. If our recommendations are
taken forward, the criminal courts will be explicitly directed by primary legislation
to consider not only whether expert evidence is needed34 but also whether the
witnesses called to provide expert evidence are impartial and qualified.35 The
parties and the courts will also need to address the reliability of the experts’
proffered opinion evidence and, in cases where there is a pre-trial hearing on the
question, the court will need to inquire into and rule on evidentiary reliability.36 To
ensure that these statutory measures would operate effectively in practice, we
believe the parties – and, more to the point, trial judges and magistrates – should
have access to the information which could reasonably be said to have a bearing
on these admissibility requirements.
31
Director of Service Prosecutions.
32
That is, any connection with any of the parties or any commercial or scientific advantage
there may be to the expert in proffering his or her views.
33
See paras 7.6 and 7.7 above.
34
Draft Bill, cl 1(1)(a).
35
Draft Bill, cl 1(1)(b) and (c).
36
Draft Bill, cl 1(2) with cl 4; and see also cl 6.

116
7.36 Furthermore, given the special status of expert witnesses, we also believe that
the defence, like the prosecution, should be expected to disclose matters
adversely affecting its expert witnesses’ credibility beyond evidence of bias.37 All
witnesses enjoy the right to have their private life respected under Article 8(1) of
the European Convention on Human Rights, so in principle only evidence which
is likely to carry substantial probative value in relation to the question of an
expert’s credibility should need to be disclosed.38 A recent conviction for an
offence involving untruthfulness would need to be disclosed, for example, as
would the fact that the expert’s work has been roundly criticised for a good
reason by a judge in other proceedings. However, there should be no obligation
on the defence to disclose the fact that an expert has committed a peccadillo or
that an expert has been on the receiving end of patently ill-founded criticism in
previous proceedings (although we appreciate the difficulties which may arise in
the latter case, given that different individuals may have different views on such
matters).

7.37 We therefore recommend that the Criminal Procedure Rules should require
pre-trial disclosure by the parties of the following matters to the other
parties and to the court:

(1) information relevant to the application of the expertise and


impartiality tests;

(2) if requested, information relevant to the application of the reliability


test (including, in particular, the evidence underpinning the expert’s
opinion);39 and

(3) information which could substantially undermine the credibility of


the experts being relied on.40

7.38 We also recommend, in line with the current position under rule 33.4(2) of
the Criminal Procedure Rules 2010, that a party’s failure to comply with

37
It should be noted that, whilst evidence of an expert’s “bad character” may be admitted in
criminal proceedings under s 100 of the Criminal Justice Act 2003 and relevant evidence
not amounting to bad character may be admitted at common law, if such evidence is
tendered to undermine the credibility of an expert it will not be admitted unless it would
undermine his or her credibility in a substantial way. Although we believe that matters
adversely affecting credibility should be disclosed to the other parties and to the court, we
do not believe it is necessary or desirable to include such matters in the experts’ reports
(see fn 19 above).
38
See, in the context of prosecution witnesses in Scotland: HM Advocate v Murtagh [2009]
UKPC 36, [2009] SCCR 790 (disclosure of information which could have a material
adverse bearing on a witness’s credibility).
39
Databases, photographs, the relevant opinions of other experts, assumptions, statistical
models and so on. Again, the expert should not have to provide the entire corpus of
knowledge on the area but only the direct foundation material for his or her opinion. As
noted above, r 33.4 of the Criminal Procedure Rules 2010 already provides for the
inspection of experimental records (and the like) which the expert has relied upon and the
things which have been measured or tested, if another party requires such inspection.
40
Clause 8(1)(b) of our draft Bill would permit a rule requiring disclosure limited to
information “that might reasonably be thought capable of substantially detracting from” an
expert’s credibility. We use the word “substantially” rather than “materially” for consistency
with the admissibility test in s 100(1)(b) of the Criminal Justice Act 2003.

117
such disclosure requirements should render that party’s expert evidence
inadmissible, unless the judge gives leave (or all the parties agree that the
evidence should be admitted).41

7.39 The parties would need to explain these disclosure obligations to their experts
and that the experts bear a duty to provide them with the information referred to
in paragraph 7.37. Indeed, the Criminal Procedure Rules Committee might wish
to include in the Rules a specific provision requiring a party’s legal representative
to set out in his or her instructions to an expert a summary of the legal obligations
borne by the party and the expert (under the Rules and our draft Bill) and, in
particular, the expert’s obligation to provide the legal representative with the
information required by the disclosure rules. The expert’s report could then
include a statement confirming that the expert is aware of his or her obligations
and has satisfied them.

7.40 To minimise unnecessary disclosure, and therefore costs, the obligation in


paragraph 7.37(2) should arise only if a request for disclosure is expressly made
by another party. This is in line with what is currently rule 33.4(1)(c) of the
Criminal Procedure Rules 2010.

7.41 There would of course need to be sufficient time during the course of pre-trial
proceedings (following disclosure of expert reports) for investigations to be
conducted, if necessary, into matters which might affect admissibility.

7.42 We would expect the courts and the parties to exercise their powers and
discharge their duties under Part 3 of the Criminal Procedure Rules (“Case
Management”) to ensure that the disclosure requirements for evidentiary
reliability are met before any pre-trial hearing on the question. Needless to say,
there should be flexibility in the time limits to ensure that the opposing party has
sufficient time to address complex or voluminous material. That is to say, the
greater the volume or difficulty of the material disclosed, the more time should be
available to the other party to address the material and conduct any necessary
investigations.

(3) Developing rule 33.6 of the Criminal Procedure Rules


7.43 Rule 33.6 of the Criminal Procedure Rules currently empowers the court to direct
that the parties’ experts meet and discuss the “expert issues” and prepare a joint
statement for the court explaining the matters on which they agree and the areas
of disagreement. One of our judicial consultees spoke most highly of this power
and its practical utility in helping to narrow the issues,42 but a number of
consultees suggested that there should be an extension of this power. The
proposal was that the judge should be able to direct the parties’ legal
41
Draft Bill, cl 8(1)(c).
42
Mr Justice (Sir Peter) Gross (now Lord Justice Gross). The Academy of Experts also
suggested that the power to direct experts to meet should be more widely used. In the
context of appeals against conviction, see Henderson and others [2010] EWCA Crim 1269,
[2010] 2 Cr App R 24 at [5]: “The Vice-President [Hughes LJ] conducted a detailed case
management hearing providing timetables and giving directions as to how the evidence
was to be prepared. Importantly, meetings were held between the experts so as to identify
clearly those issues upon which agreement had been reached and those issues which
remained a matter of debate … .”

118
representatives and experts to attend a pre-trial hearing chaired by the judge to
discuss the expert issues. This would be more formal than the type of meeting
held pursuant to rule 33.6; but it would be less formal, and therefore less
adversarial, than the pre-trial hearing we considered in Part 5, at which the judge
would need to rule on the question of evidentiary reliability.43

7.44 The UK Register of Expert Witnesses, who held their own internal consultation on
the proposals in our consultation paper, told us that their respondents strongly
believed that in cases where the complexity of the expert evidence warrants it,
and/or the principal basis of the prosecution case is its expert evidence, the court
should be able to call a pre-trial hearing at which the judge, lawyers and experts
would come together to appraise and probe the expert evidence in context and
be given time for reflection on the expert issues. They told us that the careful
analysis of complex evidence takes time and often requires the opportunity for
quiet and considered reflection which “simply cannot be done at trial”.44

7.45 A similar argument was provided by Dr Robert Moles (of Network Knowledge), an
author who has written several books on miscarriages of justice. He suggested
that there should be a pre-trial investigation of this sort before the jury is sworn.
This, he said, would focus on the adequacy of the scientific principles involved
and whether they are capable of producing reliable conclusions.

7.46 We agree that a pre-trial meeting of experts chaired by the judge (in the presence
of the parties’ representatives) could be beneficial for the experts or the court or
both, at least in cases where expert opinion evidence is central to the prosecution
case and a meeting under rule 33.6 has already revealed a significant dispute
between the parties’ experts.45 In this sort of situation the judge and experts
might be assisted, and the conclusion of the trial expedited, by a hearing chaired
by the trial judge, with time provided after the hearing for the experts to reflect on
what was discussed, possibly with another pre-trial discussion of the expert
evidence under rule 33.6 (and therefore a revised joint statement of the experts’
respective positions under rule 33.6(2)(b)).46

7.47 In a case where the prosecution evidence is probed and scrutinised during a
hearing of this sort it may be that the prosecution experts would be willing to
withdraw from a previous opinion without the need for a more formal hearing to
address evidentiary reliability, which may lead to savings in time and money. A
hearing of this sort would no doubt also assist the judge in his or her
understanding of the expert issues and the areas of disagreement, which would
facilitate the more effective management of the trial.

43
Paragraphs 5.46 to 5.50 above.
44
The need for time to reflect was also made by an individual respondent on the on-line
forum.
45
The presence of legal representatives would not be obligatory, but they would almost
certainly wish to be present.
46
The venue would depend on the number of experts and legal representatives. In many
cases the discussion could take place in the judge’s chambers. In other cases it may need
to be in a courtroom.

119
7.48 It may be that there is already sufficient scope within Part 3 of the 2010 Rules
(“Case Management”) to allow a judge to direct that there be a hearing of the sort
suggested by our consultees. We note in particular that rule 3.2(3) provides that
the court “must actively manage the case by giving any direction appropriate to
the needs of that case as early as possible” and that rule 3.3 provides that each
party must actively assist the court in fulfilling its duty under rule 3.2. In addition,
rule 3.5 provides that, in fulfilling its duty under rule 3.2, “the court may give any
direction and take any step actively to manage a case unless that direction or
step would be inconsistent with legislation”; and rule 3.5(6) provides that if a party
fails to comply with a rule or direction “the court may … exercise its powers to
make a costs order; and impose such other sanction as may be appropriate”.

7.49 Nevertheless, if there is scope within Part 3 of the Rules, we are unaware of the
power being exercised to direct the experts to attend a hearing of this sort. The
explicit focus on pre-hearing discussions in rule 33.6 in tandem with the call for
reform from the UK Register of Expert Witnesses would seem to suggest that
such hearings either do not happen or, if they do, they happen only very rarely;
but this is hardly surprising. We appreciate that many judges may be unwilling to
exercise a power unless it is expressly set out and precisely particularised in the
Criminal Procedure Rules.

7.50 Since there is probably already sufficient scope in the Rules to direct the parties
and experts to attend such a hearing, and that such hearings may in some cases
be useful, we believe that Part 33 of the Criminal Procedure Rules should
expressly empower the criminal courts to direct the parties’ experts to attend a
pre-trial discussion chaired by the trial judge.47 As with the power in rule 33.6 to
direct the experts to attend a pre-trial discussion, there would be no obligation on
the judge to use this power, but it would be available for use in appropriate cases.

7.51 The expert witnesses’ overriding duty to the court justifies a power which would
compel expert witnesses to participate in such a discussion, and a rule similar to
rule 33.6(4) of the Criminal Procedure Rules (leave required to admit an expert’s
evidence if he or she fails to comply with a direction to attend) would ensure
compliance. It is true that one or more of the parties might not wish to have their
expert evidence scrutinised during a discussion of this sort, but they are unlikely
to refuse to co-operate if their experts have been directed to attend and there is
the possibility that their expert evidence will be excluded if they do not co-
operate.48

7.52 We therefore recommend that:

(1) Part 33 of the Criminal Procedure Rules be amended to make


explicit provision for a judge-led meeting of the parties’ legal
representatives and experts if there is a dispute on the expert
issues and the judge believes that such a meeting would be
beneficial in resolving or reducing the dispute; and

47
In line with what we say in paras 5.37 to 5.41 above, we believe that a hearing of this sort
in a magistrates’ court should be chaired by a District Judge (Magistrates’ Court) who
would also be the trial judge.
48
No privilege would attach to the communications in such a hearing.

120
(2) this power be supported by a provision similar to that now set out
in rule 33.6(4) of the Rules.

DIRECTING THE JURY


7.53 Our proposed evidentiary reliability test should ensure that expert opinion
evidence of doubtful reliability is properly screened before it can be heard by the
jury. If the evidence is insufficiently reliable to be admitted, the jury will not hear it.
Nevertheless, once the evidence is admitted the jury will still need to determine
for itself whether the evidence is reliable and how much weight it should be given.

7.54 One of our consultees suggested that the trial judge should warn the jury about
relying on disputed expert opinion evidence adduced by the prosecution, if the
expert evidence relates to a developing field of medical science.49 The Court of
Appeal has recently intimated that a warning of this sort may be appropriate for
some such cases.50

7.55 The Criminal Bar Association proposed that the Judicial Studies Board should
provide enhanced specimen directions for prosecution expert evidence, to
provide an “additional safeguard by way of explanation and warnings” setting out
the “limits and any potential for error” because there are factors which impede an
effective and critical examination of much expert evidence, such as “a
misunderstanding of the limits of ‘science’”.51

7.56 We agree that warnings and explanations of this sort (“cautionary warnings”)
would be desirable in some trials, certainly in many of the trials where the
prosecution case depends fundamentally on disputed expert opinion evidence,
such that the evidence is central to its case and does not merely provide support
for a cogent body of other evidence.

7.57 However, we also believe that this is an area where it would be best not to be too
prescriptive. In our view the judge should always consider whether the jury
should be given a cautionary warning about the prosecution’s expert opinion
evidence, if that evidence is of substantial importance to the prosecution case.
And if a cautionary warning is thought to be appropriate in such a case, the judge
should provide one tailored to the type of evidence, the strength of the disputed
opinion and the facts of the case. But we do not believe the judge should be
49
M John Batt (Solicitor), proposing that the judge should give examples of hypotheses
which have been discredited.
50
See Holdsworth [2008] EWCA Crim 971 at [57]: “As knowledge increases, today’s
orthodoxy may be tomorrow’s outdated learning. Special caution is also needed where
expert opinion evidence is not just relied upon as additional material to support a
prosecution but is fundamental to it.” In addition, the Court of appeal has suggested
cautionary warnings for certain types of expert evidence; see, for example: Flynn [2008]
EWCA Crim 970, [2008] 2 Cr App R 20 at [64] (expert evidence on voice recognition),
Luttrell [2004] EWCA Crim 1344, [2004] 2 Cr App R 31 at [44] (expert lip-reading evidence)
and Atkins [2009] EWCA Crim 1876, [2010] 1 Cr App R 8 at [23] (expert facial-mapping
evidence).
51
For the current position, see the Judicial Studies Board’s Crown Court Bench Book,
Directing the Jury, March 2010, p 157, on cases where there are “serious and respectable
disagreements between experts as to the conclusions which can be drawn from post
mortem findings”, and p 153 on “the limitations of expert evidence at the boundaries of
medical knowledge”.

121
under a duty to provide the jury with a cautionary warning in all such cases. In
some cases a warning would be unnecessary and potentially confusing.

7.58 Although we appreciate that there are some special categories of prosecution
evidence in criminal proceedings in respect of which the judge must explain the
special need for caution – for example, where the prosecution case depends
wholly or substantially on disputed eye-witness identification evidence52 or wholly
or substantially on a confession made by a mentally-handicapped accused53 –
the trend in recent years, certainly in relation to potentially unreliable witnesses of
fact, has been to give greater latitude to trial judges as to when a cautionary
warning should be given and, if so, the terms of the warning.54 To take a recent
example, in Stone55 the Court of Appeal refused to accept that the judge must
always give a cautionary warning if the prosecution has relied on a confession
purportedly made by the accused to another inmate while in prison, but the court
did accept that in such cases the judge should always consider whether a warning
should be given.

7.59 Eye-witness identifications, voice identifications and confessions made by the


mentally-ill all carry an inherent risk of unreliability, but this risk is not the same
with expert opinion evidence because there are so many fields of expertise and
so many individuals who are competent to provide such evidence. Undoubtedly
some fields of expert evidence and some expert opinions derived from those
fields are more likely to be unreliable than others, but it would be impracticable
and undesirable to create a rule requiring a warning for some fields but not
others. We believe that whether a warning should be given, and how any such
warning should be framed, should be left to the judge to determine in the light of
all the circumstances of the case.

7.60 What is desirable and important is for the judge to consider the issue. If the judge
does this, it is very likely that he or she will provide the jury with an appropriate
warning in an appropriate case.

7.61 We therefore recommend that the Criminal Procedure Rules should provide
that, for trials on indictment (before a judge and jury), if the judge
determines at the end of the trial that the prosecution case depends wholly
or substantially on disputed expert opinion evidence, the judge should:

(1) consider whether to provide the jury with a cautionary warning in


relation to that evidence; and

(2) if a cautionary warning is thought to be appropriate, provide the


jury with an appropriate warning tailored to the facts of the case.

52
Turnbull [1976] QB 224. A similar approach is developing for voice-identification evidence;
see, for example, Roberts [2000] Criminal Law Review 183 and Erskine [2001] EWCA Crim
2513, [2001] All ER (D) 23 (Nov).
53
Section 77(1) of the Police and Criminal Evidence Act 1984.
54
See, for example: Makanjuola [1995] 1 WLR 1348, Muncaster [1999] Criminal Law Review
409, Causley [1999] Criminal Law Review 572, Mountford [1999] Criminal Law Review 575
and Whitehouse [2001] EWCA Crim 1531.
55
[2005] EWCA Crim 105, [2005] Criminal Law Review 569.

122
7.62 The second limb would provide the Crown Court judge with a broad discretion as
to the nature and extent of any cautionary warning he or she feels the jury should
be given, in line with the view we express above.

7.63 It should also be noted that this test would not prevent the judge from considering
whether to give a warning in other cases, where the prosecution’s expert opinion
evidence is not so important.

123
PART 8
THE NEW TEST IN PRACTICE

INTRODUCTION
8.1 In this report, we have set out our conclusion that there should be a new
framework of statutory provisions governing the admissibility of expert evidence
in criminal proceedings. Most importantly, we have recommended that for expert
opinion evidence this new framework should include an evidentiary reliability limb
relating to matters such as methodology, peer review and the expert’s reasoning.
We will now attempt to demonstrate how our recommendations would work in
practice, with reference to:

(1) the flawed prosecution evidence we described in Part 2 of our


consultation paper;1 and

(2) a hypothetical case where we examine the application of our test to


defence evidence.

8.2 The application of the new test to the cases we discussed in our consultation
paper must, however, be read with important caveats relating to the reasons for
our recommendations, the benefits of hindsight and the desirability of a new,
critical approach to expert evidence (supported by appropriate training for legal
practitioners and the judiciary).

8.3 First, as we have explained in Part 1 and Part 5, our recommendations, if


implemented, would not only establish a proper framework in criminal
proceedings for screening expert evidence at the admissibility stage; they should
also encourage higher standards amongst expert witnesses, and the specialists
on whom they rely, resulting in expert evidence of greater reliability being
tendered for admission. It follows that if our proposed legislation had been in
force at the relevant times – that is, at the time of the trials when the flawed
prosecution evidence was adduced – it is highly unlikely that the evidence of
doubtful reliability we criticised in our consultation paper would even have been
proffered for admission, and it would not have been necessary to apply our
proposed test. The outcome would have been the same, however: the unreliable
evidence would not have been admitted and could not have been relied on by the
juries in those trials.

8.4 Secondly, we fully appreciate how easy it can be to criticise a decision on


admissibility after the event, with the benefit of hindsight, and we have borne this
in mind. Nevertheless, it is also apparent from a review of past cases that there
have been failings on the part of some individuals, so this is an apt time to stress
that the effectiveness of our reforms, as applied to unreliable evidence which is
tendered for admission, depends on legal practitioners and trial judges having an
understanding of the factors bearing on evidentiary reliability and on their being
willing to adopt a more critical, enquiring approach to expert evidence.

1
Consultation Paper No 190, paras 2.14 to 2.24, summarised again in paras 1.4 to 1.7
above.

124
8.5 The party opposing the admissibility of an expert’s opinion evidence under the
reliability limb of our proposed test would have to formulate a sound argument to
explain why there is a need to investigate evidentiary reliability.2 The trial judge
would, however, have ultimate control and would therefore have the power to
require an investigation in appropriate cases even if a challenge has not been
made.3

8.6 We cannot therefore provide an absolute guarantee that our reforms would have
prevented the unreliable prosecution evidence being admitted in the cases we
referred to in our consultation paper. But what we can say with some confidence
is that:

(1) the existence of our statutory test means it is highly unlikely that the
experts in those cases would have wished to give the expert opinion
evidence in question;

(2) if the experts had endeavoured to give the evidence in question, the
existence of our test would probably have led the prosecution to
conclude that it should not be tendered for admission;

(3) if the unreliable evidence had been tendered for admission, our test
would probably have led the experts, legal practitioners and judges to
scrutinise it for reliability more effectively in advance of the trial, before a
ruling on admissibility; and

(4) given the foregoing, it is almost certain that the unreliable evidence would
not have been placed before the jury.

8.7 The success of our proposed framework as an effective barrier to the admission
of unreliable expert opinion evidence in future cases (assuming such evidence is
tendered for admission) will in large measure depend on lawyers and judges
adopting a more critical approach to expert evidence, with a new culture of
engaged enquiry. For this reason, it bears repeating that judges and criminal
lawyers will need to undergo training on factors bearing on evidentiary reliability,
particularly, but not exclusively, on the factors to be borne in mind when
assessing the validity of evidence of a scientific nature.4

8.8 This would not be an unduly burdensome or expensive obligation. It should be


possible to incorporate training, with reference to the examples and factors set
out in our draft Bill, into existing programmes for the judiciary5 and for

2
Paragraphs 5.43 to 5.56 above; see cl 6(2) of our draft Bill.
3
See cl 6(3) of our draft Bill.
4
Consultation Paper No 190, paras 1.15(3) and 6.72; and see para 1.43 above.
5
Responsibility for the training of the judiciary rests with the Lord Chief Justice and this is
exercised through the Judicial Studies Board.

125
practitioners.6 It is therefore likely that the associated costs would be quite
modest.7 In this context it is worth repeating a comment we set out in our
consultation paper with reference to evidence of a scientific nature:8 “judges do
not need to be trained to become scientists, they [merely] need to be trained to
be critical consumers of the science that comes before them.”9

THE CASES IN OUR CONSULTATION PAPER


8.9 With our caveats established, we now turn to the expert opinion evidence heard
by the juries in Dallagher,10 Clark,11 Cannings12 and Harris and others,13 the
cases we considered in our consultation paper.14

Ear-print identification evidence – Dallagher


8.10 D’s conviction for murder in Dallagher15 was based almost entirely on prosecution
expert opinion evidence relating to the comparison of an ear-print made by D with
a latent ear-print found on a window at the scene of the crime.16 One of the
prosecution experts opined at D’s trial that he was “absolutely convinced” that D
had left the latent print, an opinion we suggested was insufficiently reliable to be
considered by the jury.17

8.11 At the time of D’s trial there was an insufficient body of research data to support a
hypothesis (or assumption) that every human ear leaves a unique print and that
the identity of an offender could confidently be determined solely on the basis of
an ear-print comparison. Moreover, the expert’s opinion relied heavily on

6
Solicitors and barristers are required to undertake a certain number of continuing
professional development (“CPD”) hours training per year in order to maintain their
practising certificates. Any cost would be borne by the practitioners (or their employers)
who choose to undertake training to assist their work in this regard. We hope that guidance
on assessing the reliability of expert evidence will in due course become an important
feature in the training of newly-qualified barristers and solicitors. Indeed, as we suggested
in Part 1 (fn 45), the CPD requirements for practising solicitors and barristers who
undertake work in criminal law should be amended to require attendance at approved
lectures covering statistics and scientific methodology (in the context of criminal
proceedings).
7
For the likely cost implications, see Appendix C.
8
Consultation Paper No 190, para 6.73.
9
SI Gatowski and others, “Asking the Gatekeepers: A National Survey of Judges on Judging
Expert Evidence in a Post-Daubert World” (2001) 25 Law and Human Behavior 433, 455.
10
[2002] EWCA Crim 1903, [2003] 1 Cr App R 12 (on ear-print evidence).
11
Clark (Sally) (No 2) [2003] EWCA Crim 1020, [2003] 2 FCR 447 (on the statistical evidence
given by an expert paediatrician).
12
[2004] EWCA Crim 1, [2004] 1 WLR 2607 (on the hypothesis that two or more unexplained
infant deaths in the same family meant that murder had been committed).
13
[2005] EWCA Crim 1980, [2006] 1 Cr App R 5 (on the hypothesis that a non-accidental
injury could always be inferred from the presence of a triad of intra-cranial injuries).
14
Consultation Paper No 190, paras 2.14 to 2.24.
15
[2002] EWCA Crim 1903, [2003] 1 Cr App R 12.
16
DNA evidence taken from the latent print later established that it had been left by someone
other than D.
17
Consultation Paper No 190, paras 2.14 and 2.15.

126
subjective factors (an experience-based skill in making visual comparisons)
rather than on objectively verifiable measuring techniques.18

8.12 Under our proposed test, the prosecution would have had to prove that the
witness claiming expertise was skilled in the comparison of ear-prints and
therefore qualified to provide expert evidence in a criminal trial. If the defence had
then made submissions on the poor data and doubtful hypothesis underpinning
the expert’s proffered opinion evidence, or the judge had raised the matter
independently, there would have been an enquiry into the reliability of the
opinion. The judge may have been able to conclude without a hearing that the
expert’s opinion (that D could be identified with absolute certainty from ear-prints
alone) was insufficiently reliable to be admitted. Alternatively, there would have
been a pre-trial hearing on the issue, which no doubt would have led to the same
conclusion. The expert would not have been permitted to give an opinion that he
was “absolutely convinced” that D had left the latent print at the scene of the
murder. He might, however, have been able to give a weaker opinion on
similarities between the latent print and D’s print (assuming the jury had required
the assistance of an expert in this respect).

Statistical evidence on SIDS – Clark (Sally) (No 2)


8.13 C’s convictions for the murder of her two infant sons were quashed (in Clark
(Sally) (No 2))19 primarily because of a prosecution expert’s failure to disclose
test results. In our consultation paper, however, we focused specifically on the
unreliable statistical evidence given by a professor of paediatrics and child
health.20

8.14 The expert opined that there was only a one in 73 million chance of two natural
cot deaths (sudden infant death syndrome or “SIDS”) in the same family. In
reaching this figure, the expert relied on a draft Confidential Enquiry into
Stillbirths and Deaths in Infancy (CESDI) report.21 From this information, the
possibility of one SIDS death in a family such as the Clark family – a middle-class
family of non-smoking parents, with at least one income and where the mother
was at least 26 years old – was one in 8,543.22 The expert simply squared this
improbability to reach his opinion that the likelihood of two infant deaths in the
same family would be one in 73 million.23 His opinion was therefore based on a
18
According to the Court of Appeal in this case (at [9]), the prosecution expert “had simply
become interested in ear print identification and read what was available on the topic. He
had built up a portfolio of about 600 photographs and 300 ear prints and from his
experience and what he had read he was satisfied that no two ear prints are alike in every
particular.” D’s case on appeal (at [11]) was that there was “no empirical research, and no
peer review to support the conclusion that robust decisions can be founded on
comparisons which in turn are critically dependent on the examiner’s judgment in
circumstances where there are no criteria for testing that judgment”.
19
[2003] EWCA Crim 1020, [2003] 2 FCR 447.
20
Clark’s earlier appeal was unsuccessful; see Clark (Sally) (No 1) (2000) 1999/07495/Y3,
[2000] All ER (D) 1219, cited on www.bailii.org as [2000] EWCA Crim 54.
21
See Clark (Sally) (No 1) [2000] EWCA Crim 54 at [102].
22
Clark (Sally) (No 1) [2000] EWCA Crim 54 at [118]; Clark (Sally) (No 2) [2003] EWCA
Crim 1020, [2003] 2 FCR 447 at [96].
23
Clark (Sally) (No 1) [2000] EWCA Crim 54 at [118]; Clark (Sally) (No 2) [2003] EWCA
Crim 1020, [2003] 2 FCR 447 at [98].

127
hypothesis (or assumption) that genetic or other environmental factors do not
affect the likelihood of SIDS. The Court of Appeal said that it was “unfortunate
that the trial did not feature any consideration as to whether the statistical
evidence should be admitted in evidence”.24

8.15 Under our proposed reforms, the trial judge would have ruled on the scope of the
paediatrician’s competence to give expert evidence and would have monitored
his evidence to ensure that he did not drift into other areas.25 Insofar as the
paediatrician was adjudged to be competent to provide some basic statistical
evidence to supplement his opinion as an expert on child health, he would have
been conscious of the limitations on his entitlement to opine as an expert. Indeed,
before giving his expert testimony, he would have been reminded of his duty to
give impartial opinion evidence only on matters falling within his areas of
expertise.26 The paediatrician would not have been asked questions in the
witness box on matters beyond his competence; and if he was inadvertently
asked such a question while giving his expert evidence, the judge would have
intervened to prevent an impermissible opinion being given. It follows that, insofar
as the expert may have had some knowledge of statistical analysis, he would
probably have been prevented from giving an opinion on the statistical likelihood
of multiple SIDS deaths.

8.16 With regard to the reliability of the statistical evidence – insofar as the expert
paediatrician was competent to provide it and would have wished to proffer it for
admission (given the existence of the statutory test), and assuming that the figure
would have been disclosed before the trial in his written report – the defence or
court would presumably have raised the matter as a preliminary issue in the pre-
trial proceedings and the judge would no doubt have directed that the parties and
their experts attend a pre-trial hearing to assess the reliability of the figure of one
in 73 million.

8.17 The reliability of the hypothesis (or assumption) underlying the figure of one in 73
million would then have been examined against our proposed statutory test,
examples and factors. The expert would have been required to demonstrate the
evidentiary reliability (the scientific validity) of his hypothesis and the chain of
reasoning leading to his opinion, with reference to properly conducted scientific
research and an explanation of the limitations in the research findings and the
margins of uncertainty associated with them.

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.27 It is to be noted that when C appealed against her convictions
the Court of Appeal accepted that there was evidence to suggest that the figure
of one in 73 million “grossly” misrepresented the chance of two sudden deaths

24
[2003] EWCA Crim 1020, [2003] 2 FCR 447 at [173].
25
Paragraph 4.46(2) above.
26
Paragraph 4.46(1) above.
27
It is to be noted that the report the paediatrician relied on acknowledged the possibility of
other familial factors increasing the risk of a SIDS death; see Clark (Sally) (No 2) [2003]
EWCA Crim 1020, [2003] 2 FCR 447 at [101].

128
within the same family from unexplained but natural causes28 and said that, if the
question of the statistical evidence had been fully argued on appeal, it would
probably have provided a distinct basis upon which to quash C’s convictions.29

8.19 In the absence of sufficiently cogent research findings supporting the underlying
hypothesis, the expert paediatrician would almost certainly not have been
permitted to give his figure of one in 73 million at C’s trial for murder.

8.20 Moreover, even if it had been possible for the prosecution to call a competent
statistician to provide a reliable figure as to the probability of two SIDS death in
one family, couched with appropriate qualifications to reflect the uncertainties and
gaps in the scientific knowledge on SIDS and the dangers associated with a
retrospective approach to probabilities, under our recommendations that expert
would have been expected to try to formulate a counterbalancing probability
reflecting the defence case. That is to say, he or she would have been expected
to try to come to a figure reflecting the unlikelihood that the accused would have
murdered her two children (if such a calculation were feasible).30

Inferring murder from unexplained infant deaths – Cannings


8.21 In our consultation paper we explained that the Court of Appeal quashed C’s
convictions for the murder of her two infant sons in Cannings31 on the ground that
the mere fact of two or more unexplained infant deaths in the same family could
not be allowed to lead inexorably to the conclusion that murder had been
committed.32 The Court of Appeal rejected the dogmatic view held by a number of
paediatricians that murder could confidently be inferred from two or more
unexplained deaths.33 Fresh evidence suggested that multiple cot (SIDS) deaths
in the same family could have an underlying genetic cause; and a report relating
to the largest follow-up study of cot-death families concluded that “the occurrence
of a second unexpected infant death within a family is … usually from natural
causes”.34

8.22 Under our proposed reforms, the defence would presumably have challenged the
evidentiary reliability of the paediatricians’ hypothesis during the pre-trial
proceedings on the ground that it was insufficiently supported by data generated
by sound empirical research. Given the prosecution’s dependence on the
hypothesis and opinion evidence founded on it, the judge would no doubt have

28
[2003] EWCA Crim 1020, [2003] 2 FCR 447 at [178].
29
[2003] EWCA Crim 1020, [2003] 2 FCR 447 at [178] to [180]. The Court of Appeal pointed
out (at [176]) that if the paediatrician’s figure of one in 73 million was correct the purpose of
the Care of Next Infant scheme, designed to provide guidance and to monitor the
possibility of a second SIDS death in the same family, was “wasted effort” because the risk
could “effectively be discounted”. See also A Coghlan, “Infant deaths: Justice for the
innocents” (2005) 187 New Scientist (issue 2510), 6, citing The Lancet (2004) vol 365, 29
referring to a study which found that a second similar death in the same family was nine
times as likely to be natural as inflicted.
30
Paragraph 7.21(2)(c) and paras 7.22 to 7.25 above.
31
[2004] EWCA Crim 1, [2004] 1 WLR 2607.
32
Consultation Paper No 190, para 2.20.
33
[2004] EWCA Crim 1, [2004] 1 WLR 2607 at [18] to [20].
34
[2004] EWCA Crim 1, [2004] 1 WLR 2607 at [141]. See also fn 29 above.

129
accepted that there was an issue of evidentiary reliability to be addressed and a
pre-trial hearing would have been arranged, during which the extent and quality
of the underlying research data and the associated margins of uncertainty would
have been considered along with other relevant factors such as peer review,
differences of opinion in the expert community and research findings supporting
an alternative hypothesis (such as a genetic contribution).35

8.23 If our admissibility test had been applied, the trial judge would have permitted the
prosecution experts to give opinion evidence at trial only to the extent justified by
the available data and ongoing uncertainties.36 If the judge had ruled that the
experts could not give an opinion that the deaths were almost certainly non-
accidental, and if the judge concluded that there was no independent evidence to
suggest non-accidental deaths, the case would not have proceeded to trial.37

Inferring a non-accidental cause from intra-cranial injuries – Harris


8.24 Until the appeal in Harris and others38 the prosecution could secure a murder (or
serious assault) conviction on the basis of expert opinion evidence that a non-
accidental head injury to an infant child could confidently, in effect always, be
inferred from the presence of a particular triad of intra-cranial injuries39 (and that,
accordingly, the accused’s exculpatory explanation could be disregarded as
untrue). So, a person charged with murdering a child could be convicted solely on
the basis of evidence of shaken baby syndrome (“SBS”).40

8.25 However, as we explained in our consultation paper,41 the evidence base for the
hypothesis of SBS when the appellants in Harris and others were tried has been
described as an inverted pyramid “with a very small database (most of it poor-
quality original research, retrospective in nature, and without appropriate control

35
In Cannings [2004] EWCA Crim 1, [2004] 1 WLR 2607 at [138] the Court of Appeal
acknowledged that there was a “substantial body of research” which suggested that infant
deaths “can and do occur naturally, even when they are unexplained”. As noted already,
the court referred to a study (the CONI study) which found that a second SIDS death in a
family is usually from natural causes. The court went on to accept (at [142]) that even three
unexpected and apparently unexplained infant deaths in the same family could have
natural causes.
36
We do of course appreciate the difficulties of experimental research in this area because of
the rarity of multiple infant deaths and the inability to conduct experiments to test a
hypothesis.
37
In Cannings [2004] EWCA Crim 1, [2004] 1 WLR 2607 at [178] to [179], the Court of
Appeal took the view that, until all known causes of death had been excluded, the cause
should remain unknown and, accordingly, parents should not be prosecuted if experts
disagreed over causation (that is, a reasonable body of experts opined that the deaths
were natural) and there was no other evidence to suggest murder.
38
[2005] EWCA Crim 1980, [2006] 1 Cr App R 5.
39
Acute encephalopathy (a disorder of the brain), subdural haemorrhage (bleeding around
the brain) and retinal haemorrhage (bleeding in the retinas).
40
See Editorial, British Medical Journal 29 July 2010 (issue 2771): “For 40 years,
mainstream medical experts who give evidence in court have largely agreed that shaken
baby syndrome can be unambiguously diagnosed by a triad of symptoms at post-mortem
… . Murder convictions are often secured on the basis of these alone, even in the absence
of other signs of abuse … .”
41
Consultation Paper No 190, fn 31 to para 2.24.

130
groups) spreading to a broad body of somewhat divergent opinions”.42 The same
paper concluded that there “was inadequate scientific evidence to come to a firm
conclusion on most aspects of causation, diagnosis, treatment, or any other
matters pertaining to SBS”.43

8.26 If our proposed admissibility test had been in force at the time when the
prosecution was seeking to rely on the triad of intra-cranial injuries as compelling
evidence of a non-accidental head injury – and assuming for the sake of
argument that our test would not have deterred the prosecution’s medical experts
from wishing to give an opinion that the presence of the triad standing alone
justified a diagnosis of non-accidental head injury – the defence would
presumably have challenged the reliability of the opinion evidence (or the judge
would have raised the matter of his or her own motion) and the question of
evidentiary reliability would have been addressed at a pre-trial hearing.

8.27 At the hearing, the hypothesis that the triad of intra-cranial injuries was sufficient
evidence to justify a conviction for murder or a serious assault would have been
critically appraised. The judge would have scrutinised the prosecution experts’
opinion evidence for reliability, with reference to the nature and extent of the
empirical research underpinning the hypothesis of SBS.

8.28 The prosecution experts would have had to show that their hypothesis was
supported by sufficient observational data and/or simulations. There would need
to have been properly conducted research showing a sound correlation between
the intra-cranial injuries and a non-accidental cause (from independent evidence)
and demonstrating the absence of such injuries in cases where there have been
accidents (such as choking or small falls) or congenital conditions.44 Given the
strength and importance of the opinion evidence proffered for admission, the
observational data consistent with the hypothesis (and the absence of
observational data inconsistent with it) would need to have been considerable to
justify the proffered evidence being admitted.

8.29 In short, the judge would have permitted a prosecution expert to give opinion
evidence only to the extent that any opinion could be justified by the research
data available.45 The expert opinion evidence in support of the prosecution
assertion of a non-accidental injury would have been modified or weakened to
reflect the uncertainties associated with the hypothesis of SBS, the quality and
42
M Donohoe, “Evidence-based Medicine and Shaken Baby Syndrome” (2003) 24 American
Journal of Forensic Medicine and Pathology 239, 241. See also D Tuerkheimer, “The Next
Innocence Project: Shaken Baby Syndrome and the Criminal Courts” (2009) 87
Washington University Law Review 1, 12 to 14 and 17 to 18.
43
M Donohoe, “Evidence-based Medicine and Shaken Baby Syndrome” (2003) 24 American
Journal of Forensic Medicine and Pathology 239, 241.
44
See M Donohoe, “Evidence-based Medicine and Shaken Baby Syndrome” (2003) 24
American Journal of Forensic Medicine and Pathology 239: “Genuine hypothesis testing
requires use of appropriate research methodologies, including collection of relevant control
data, and suitable statistical analysis. The interpretation of individual study findings may be
constrained by factors such as whether the cohort examined was adequately
representative of the patient population in general. Replication across studies and in
independent research centres is a key factor in the reliability of evidence.”
45
As the Court of Appeal recently acknowledged in Henderson [2010] EWCA Crim 1269,
[2010] 2 Cr App R 24 at [6]: “the strength of a proposition in medicine depends upon the
strength of the medical evidence on which it is based”.

131
extent of the research underpinning it, evidence suggesting that the triad could
result from another cause generally and any evidence in the instant case
suggesting an accidental or congenital cause.

8.30 Given the limited research data supporting the hypothesis, and evidence that the
triad could have some other cause, the judge would not have allowed prosecution
experts to give opinion evidence that, standing alone, the triad of injuries was
certain proof of non-accidental trauma; and the judge would not have permitted
the prosecution to seek a conviction solely on the basis that the infant exhibited
the triad of intra-cranial injuries associated with SBS.46

A HYPOTHETICAL CASE
8.31 Under this heading we apply our proposed framework to a hypothetical case of
alleged murder, relating to an infant child who died unexpectedly while alone with
his mother (D). In this case, however, we focus on defence evidence, developing
a point we made in our consultation paper47 and drawing on the facts of a recent
case in civil proceedings.48

8.32 The deceased child was found to have the triad of intra-cranial injuries associated
with violent shaking, including subdural and retinal haemorrhages (bleeding
around the brain and in the retinas) but no external injuries. The prosecution will
allege that D violently assaulted the child and will seek to rely on expert opinion
evidence relating to the triad in tandem with extraneous circumstantial evidence
of D’s guilt (evidence that D assaulted the child on previous occasions).

8.33 D’s defence is that she did not injure the child and was asleep when he died. In
support of her defence, D will seek to call an expert witness to opine that the
cause of the child’s subdural and retinal haemorrhages could have been severe

46
However, a conviction would have been possible on the basis of the triad of intra-cranial
injuries in association with other sufficiently cogent circumstantial evidence of the
accused’s guilt (such as separate injuries consistent with abuse). In Harris and others
[2005] EWCA Crim 1980, [2006] 1 Cr App R 5 the Court of Appeal noted evidence which
suggested that the triad of injuries could be caused, albeit only rarely, by a minor fall or
non-violent handling. The court therefore held that, without more, the mere presence of the
triad could not automatically or necessarily lead to a diagnosis of non-accidental head
injury, but the court also maintained that the triad was a “strong pointer” to non-accidental
head injury. See also Henderson [2010] EWCA Crim 1269, [2010] 2 Cr App R 24 at [6]: the
triad is “strong prima facie evidence of shaking”. Of course the cogency of the triad as
strong prima facie evidence of shaking would be profoundly weakened by circumstantial
evidence suggesting some other (innocent) cause; see Butler [2010] EWCA Crim 1269,
[2010] 2 Cr App R 24 at [84] to [118]. Note also the Crown Prosecution Service’s updated
guidance for prosecutors, explaining that it is unlikely a charge of murder, attempted
murder or assault will be justified if the only evidence against the accused is the triad of
injuries; see CPS, Non-Accidental Head Injury Cases (NAHI, formerly referred to as
Shaken Baby Syndrome [SBS]) – Prosecution Approach,
http://www.cps.gov.uk/legal/l_to_o/non_accidental_head_injury_cases/ (last visited 26
January 2011).
47
Consultation Paper No 190, para 6.56.
48
A Local Authority v S [2009] EWHC 2115 (Fam), [2010] 1 FLR 1560.

132
hypoxia (oxygen deficiency in the tissues) resulting from choking.49 The defence
expert proposes to rely on the “Geddes III” hypothesis50 and a recent research
paper co-authored by a Dr C which the expert says supports it. The prosecution
will counter that the hypothesis is invalid and that the research paper does not
support it, relying on a recent judgment of the Family Division of the High Court,
A Local Authority v S,51 where Dr C’s paper was criticised. The prosecution will
therefore submit that there should be a hearing on evidentiary reliability.

8.34 Given the evidence on Geddes III before the Court of Appeal in Harris and
others,52 and the court’s critical comments on the hypothesis in that case,53 the
trial judge in our hypothetical case would no doubt decide that the question of
evidentiary reliability must be addressed at a pre-trial hearing. The court would
focus on the scientific validity of Geddes III (that hypoxia could have caused the
haemorrhages) and the research paper the defence expert wishes to rely on in
support of that hypothesis.

8.35 The case of A Local Authority v S was an application made by a local authority
for a care order54 in respect of a three-year-old child, S. The proceedings arose
out of the death of S’s sibling, Z, who died at the age of 13 weeks. The local
authority brought proceedings alleging that Z died as the result of a shake or
impact injury inflicted by the mother, as evidenced by the triad of intracranial
injuries;55 and, while the majority of experts in the case considered the most likely
cause of Z’s death to be a non-accidental injury, two experts, Dr C56 and Dr S,57

49
According to this hypothesis, the lack of oxygen causes the brain to swell and this swelling
increases the pressure on the brain causing haemorrhages. This was a central feature of
the “unified hypothesis” known as “Geddes III”, a summary of which can be found in A
Local Authority v S [2009] EWHC 2115 (Fam), [2010] 1 FLR 1560 at [32]. In Consultation
Paper No 190, fn 64 to para 6.56, we explained that Geddes III is regarded as invalid, even
by its proponent, and was originally published merely to stimulate debate. See Harris and
others [2005] EWCA Crim 1980, [2006] 1 Cr App R 5 at [57], [58] and [66] to [69]. One
criticism of the hypothesis is that, if it were correct, there would be far more children
showing the triad, but this is not the case.
50
Above.
51
[2009] EWHC 2115 (Fam), [2010] 1 FLR 1560.
52
[2005] EWCA Crim 1980, [2006] 1 WLR 2607 at [58] and [68].
53
[2005] EWCA Crim 1980, [2006] 1 WLR 2607 at [69]: “In our judgment … [Geddes III] can
no longer be regarded as a credible or alternative cause of the triad of injuries.” See also A
Local Authority v S [2009] EWHC 2115 (Fam), [2010] 1 FLR 1560 at [207] to [210] referring
to a research paper published in 2007 following a study of 82 foetuses, infants and
toddlers, which showed no causal link between hypoxia and subdural haemorrhage,
leading the authors to conclude that Geddes III “can no longer be regarded as a credible
alternative cause of the triad of injuries”. (It is to be noted, however, that the High Court
judge criticised the inclusion of foetuses in the researchers’ cohort.)
54
Under s 31 of the Children Act 1989.
55
A Local Authority v S [2009] EWHC 2115 (Fam), [2010] 1 FLR 1560 at [29].
56
A consultant paediatric histopathologist.
57
A consultant neuropathologist.

133
subscribing to Geddes III, opined that hypoxia could have been the cause of Z’s
haemorrhages.58

8.36 In A Local Authority v S the court noted that Drs C and S’s support for Geddes III
was controversial and contrary to “the mainstream of current thinking and the
analysis of the Court of Appeal in R v. Harris”.59 In the same paragraph of the
court’s judgment, mention was made of the fact that there were only about three
or four experts in the country (including Drs C and S) who subscribe to
Geddes III. The court explained that Drs C and S regarded themselves as having
built on Dr Geddes’ original work; but the court went on to consider the legitimacy
of their belief in Geddes III, referring to their use of research material, their
deference to experts in another field, their tendency to opine on matters beyond
their expertise and the question of factual accuracy.60

8.37 Dr C had co-authored a paper in 2009 in which the authors concluded that
subdural haemorrhage was not an unusual finding where children had died from
a non-traumatic cause and that this was “confirmatory of Geddes III”.61 However,
having heard expert evidence which was highly critical of the way the research
had been conducted and the conclusions drawn from the data, the court noted
that Dr C could not explain why clinicians, forensic pathologists and
neurosurgeons are not finding subdural haemorrhages in cases of pure
hypoxia;62 and the court went on to hold that, as evidence, the paper provided
nothing beyond confirmation of what was already known, namely that a number
of small babies have subdural haemorrhages following the traumatic process of
birth.63 The court noted that Dr C and her co-author had focused exclusively on
foetuses and babies who had lived no longer than 19 days, even though the
preponderant medical view is that a cohort of foetuses cannot provide useful
data, on account of differences between foetuses and live children, and because
subdural haemorrhages are not unusual in babies up to 19 days old. The court
went on to conclude that Dr C and her co-author had been selective in the way
they had chosen children for their study and that this “must inevitably undermine
the value of the study”.64

8.38 The court then listed other concerns regarding the paper and Dr C’s evidence: (1)
it was disingenuous of Dr C to suggest that her research provided confirmation of
Geddes III; (2) the results were entirely at odds with other research on a larger
58
A Local Authority v S [2009] EWHC 2115 (Fam), [2010] 1 FLR 1560 at [43]. All the experts
in the case accepted that a non-accidental head injury was a possible cause of Z’s death,
but Drs C and S considered this to be an unlikely explanation in the absence of any
independent evidence of trauma (see [44] and [58]). Dr C was of the view that the triad can
be used as a diagnostic tool to prove non-accidental injury only if there are also injuries
associated with trauma such as grip marks (bruises) or fractures (or eye-witness
testimony); Dr S refused to acknowledge that the triad was any evidence of a non-
accidental injury in the absence of independent evidence of serious trauma (see [63] and
[198]).
59
[2009] EWHC 2115 (Fam), [2010] 1 FLR 1560 at [199].
60
[2009] EWHC 2115 (Fam), [2010] 1 FLR 1560 at [204] and [205].
61
[2009] EWHC 2115 (Fam), [2010] 1 FLR 1560 at [209] and [214].
62
[2009] EWHC 2115 (Fam), [2010] 1 FLR 1560 at [219].
63
[2009] EWHC 2115 (Fam), [2010] 1 FLR 1560 at [223].
64
[2009] EWHC 2115 (Fam), [2010] 1 FLR 1560 at [224] and [225].

134
cohort where there had been no selection; (3) Dr C’s findings were at odds with
the experience of all the clinicians who gave evidence; (4) Dr C’s conclusions
were reached without examining the babies for retinal haemorrhages; (5) the
paper was predicated on the soundness of Geddes III; and (6) Dr C had claimed
in her evidence that there were research papers confirming a link between
hypoxia and retinal haemorrhages, contrary to the evidence of the ophthalmic
expertise before the court.65

8.39 The court also criticised Dr S’s evidence because she had relied heavily, but
selectively, upon other research material and, in respect of one paper, had made
an unfounded assumption.66 The court commented that expert witnesses must
display professionalism and rigour, meaning “not only drawing the court’s
attention to research that is contrary to their view, but [also being] rigorous in the
use they make of research papers.”67 The court added that Dr S should at the
very least have drawn attention to the fact that the cases she relied on involved a
child who had been shaken and one who was regarded as the victim of a non-
accidental injury (resulting in the child being taken into care);68 and the court felt
compelled to conclude that her use of the research material had been
“disingenuous”.69

8.40 Returning now to our hypothetical criminal case, if the reliability of Geddes III and
Dr C’s paper as purported support for the hypothesis were to be examined at a
pre-trial hearing, it is possible, likely even, that the judge would conclude that the
defence expert’s opinion evidence should be ruled inadmissible. For the defence
expert to be able to opine that D’s child might plausibly have died as a result of
haemorrhaging caused by choking, generating a reasonable doubt as to D’s guilt,
it would be necessary to show that Geddes III was supported by sound, properly
conducted scientific research. The evidence presented in A Local Authority v S
suggests there is no such empirical support. On the contrary, what research there
is, and the widespread absence of the triad in cases where there has been no
trauma, critically undermine the validity of the hypothesis.

8.41 As we explained earlier in this report,70 a minority opinion (even a lone voice)
could be admissible for the defence in criminal proceedings if our test is taken
forward into legislation. But to be admissible any such opinion would need to be
based on the rigorous application of sound scientific principles. Hypotheses
unsupported by appropriate research, and undermined by what research has
been undertaken, are hardly likely to provide a sound basis for expert opinion
evidence in criminal proceedings.

65
[2009] EWHC 2115 (Fam), [2010] 1 FLR 1560 at [225].
66
[2009] EWHC 2115 (Fam), [2010] 1 FLR 1560 at [226] to [235] and [247].
67
[2009] EWHC 2115 (Fam), [2010] 1 FLR 1560 at [247].
68
[2009] EWHC 2115 (Fam), [2010] 1 FLR 1560 at [247].
69
[2009] EWHC 2115 (Fam), [2010] 1 FLR 1560 at [248].
70
Paragraph 3.104.

135
8.42 Finally, it should always be borne in mind that an expert may not even be
permitted to provide evidence at all if he or she has become so wedded to a
hypothesis that he or she has lost the essential quality of impartiality.71

71
Doctors C and S were also criticised in A Local Authority v S for opining beyond their areas
of expertise, and Dr S was criticised for making an unwarranted assumption to plug a gap
in their hypothesis, thereby demonstrating a lack of scientific rigour. Further criticisms were
made in relation to factual inaccuracies in their evidence and for being experts who have
“developed a scientific prejudice”. See [2009] EWHC 2115 (Fam), [2010] 1 FLR 1560
at [249] to [285].

136
PART 9
SUMMARY OF RECOMMENDATIONS

A NEW RELIABILITY TEST


9.1 We recommend that there should be a statutory admissibility test which would
provide that an expert’s opinion evidence is admissible in criminal proceedings
only if it is sufficiently reliable to be admitted (“the reliability test”).

Paragraph 3.36

9.2 We recommend a rule which would provide, for the reliability test, that if there is
any doubt on the matter expert evidence presented as evidence of fact should be
treated as expert opinion evidence.

Paragraph 3.39

9.3 We recommend that trial judges should be provided with a single list of generic
factors to help them apply the reliability test and that these factors should be set
out in the primary legislation containing the test.

Paragraph 3.62

9.4 We recommend that the trial judge should be directed to take into consideration
the factors which are relevant to the expert opinion evidence under consideration
and any other factors he or she considers to be relevant.

Paragraph 3.63

9.5 We recommend that:

(1) criminal courts should have a limited power to disapply the reliability test
so that it does not have to be applied routinely and unnecessarily;

(2) but, equally, the power to disapply must not be such that the reliability
test becomes only a nominal barrier to the adduction of unreliable expert
opinion evidence.

Paragraph 3.77

9.6 We recommend for our proposed reliability test that, where the test is applied, the
party wishing to adduce the expert opinion evidence should bear the burden of
demonstrating that it is sufficiently reliable to be admitted.

Paragraph 3.88

9.7 We recommend that there should be a single framework in primary legislation


governing the admissibility of all expert evidence in criminal proceedings.

Paragraph 3.136

137
CODIFICATION OF THE COMMON LAW
9.8 We recommend that primary legislation should provide that expert evidence is
admissible in criminal proceedings only if:

(1) the court is likely to require the help of an expert witness; and

(2) it is proved on the balance of probabilities that the individual claiming


expertise is qualified to give such evidence.

9.9 We also recommend that this legislation should provide that expert evidence is
inadmissible if there is a significant risk that the expert has not complied with, or
will not comply with, his or her duty to provide objective and unbiased evidence,
unless the court is nevertheless satisfied that it is in the interests of justice to
admit the evidence.

Paragraphs 4.8 and 4.9

9.10 We recommend that the Criminal Procedure Rules be amended to include the
following additional requirements:

(1) before giving oral evidence, an expert witness should be referred to his
or her overriding duty to give expert evidence which is

(a) objective and unbiased, and

(b) within his or her area (or areas) of expertise;

(2) the trial judge or magistrates’ court should rule on the expert witness’s
area (or areas) of expertise before he or she gives evidence and monitor
the position to ensure that he or she does not give expert evidence on
other matters.

Paragraphs 4.46

THE RELIABILITY TEST


9.11 We recommend for criminal proceedings:

(1) a statutory provision in primary legislation which would provide that


expert opinion evidence is admissible only if it is sufficiently reliable to be
admitted;

(2) a provision which would provide our core test that expert opinion
evidence is sufficiently reliable to be admitted 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;

(3) a provision which would set out the following key (higher-order) examples
of reasons why an expert’s opinion evidence is not sufficiently reliable to
be admitted:

138
(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 stand up to
scrutiny;

(b) the opinion is based on an unjustifiable assumption;

(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;

(e) the opinion relies on an inference or conclusion which has not


been properly reached.

(4) a provision which would direct the trial judge to consider, where relevant,
more specific (lower-order) factors in a Schedule to the Act and to any
unspecified matters which appear to be relevant.

Paragraph 5.17

9.12 We recommend that a trial judge who has to determine whether an expert’s
opinion evidence is sufficiently reliable to be admitted should be directed to have
regard to:

(1) the following factors (insofar as they appear to be relevant):

(a) the extent and quality of the data on which the expert’s opinion is
based, and the validity of the methods by which they were
obtained;

(b) if the expert’s opinion relies on an inference from any findings,


whether the opinion properly explains how safe or unsafe the
inference is (whether by reference to statistical significance or in
other appropriate terms);

(c) if the expert’s opinion relies on the results of the use of any
method (for instance, a test, measurement or survey), whether
the opinion takes proper account of matters, such as the degree
of precision or margin of uncertainty, affecting the accuracy or
reliability of those results;

(d) the extent to which any material upon which the expert’s opinion
is based has been reviewed by others with relevant expertise (for
instance, in peer-reviewed publications), and the views of those
others on that material;

(e) the extent to which the expert’s opinion is based on material


falling outside the expert’s own field of expertise;

139
(f) the completeness of the information which was available to the
expert, and whether the expert took account of all relevant
information in arriving at the opinion (including information as to
the context of any facts to which the opinion relates);

(g) whether there is a range of expert opinion on the matter in


question; and, if there is, where in the range the expert’s opinion
lies and whether the expert’s preference for the opinion proffered
has been properly explained;

(h) whether the expert’s methods followed established practice in the


field; and, if they did not, whether the reason for the divergence
has been properly explained;

(2) approved factors, if any, for assessing the reliability of the particular type
of expert evidence in question (insofar as they appear to be relevant);
and

(3) any other factors which appear to be relevant.

Paragraph 5.35

9.13 We recommend the following for criminal proceedings:

(1) there should be a presumption that expert opinion evidence tendered for
admission is sufficiently reliable to be admitted, but this presumption
would not apply if:

(a) it appears to the court, following a reasoned challenge, that the


evidence might not be sufficiently reliable to be admitted, or

(b) the court independently rules that the presumption should not
apply;

(2) if the presumption no longer applies, the court should direct that there be
a hearing to resolve the question of evidentiary reliability, unless the
question can be properly resolved without a hearing; and

(3) for Crown Court jury trials, the reliability hearing should ordinarily take
place before the jury is sworn, but, exceptionally, it should be possible to
hold a hearing during the trial in the absence of the jury.

Paragraph 5.56

9.14 We recommend that, if challenged on appeal, the trial judge’s ruling under the
reliability test should be approached by the appellate court as the exercise of a
legal judgment rather than the exercise of a judicial discretion.

Paragraph 5.94

140
COURT-APPOINTED EXPERTS
9.15 We recommend that a Crown Court judge (for a trial on indictment) should be
provided with a statutory power to appoint an independent expert to assist him or
her when determining whether a party’s proffered expert opinion evidence is
sufficiently reliable to be admitted.

9.16 We recommend that this power should permit a Crown Court judge to appoint an
independent expert only if he or she is satisfied that it would be in the interests of
justice to make an appointment, having regard to:

(a) the likely importance of the expert opinion evidence in the context
of the case as a whole;

(b) the complexity of that evidence, or the complexity of the question


of its reliability; and

(c) any other relevant considerations.

9.17 We recommend that the judge should make his or her appointment from a
shortlist of individuals prepared by an independent panel of legal practitioners,
chaired by a Circuit Judge, reflecting the interests of both the prosecution and the
defence.

Paragraphs 6.78 to 6.80

PROCEDURAL MATTERS

Expert reports
9.18 We recommend that Part 33 of the Criminal Procedure Rules be amended to
include the following:

(1) a rule requiring an appendix to the expert’s report, setting out –

(a) sufficient information to show that the expertise and impartiality


requirements are satisfied; and

(b) a focused explanation of the reliability of the opinion evidence


with reference to the test and relevant examples and factors in
our draft Bill, concisely set out in a manner which would be
readily understood by a trial judge, along with a summary of:

(i) other cases (if any) where the expert’s opinion evidence
has been ruled admissible or inadmissible after due
enquiry under the reliability test; and

(ii) other judicial rulings after due enquiry which the expert is
aware of (if any) on matters underlying the expert’s opinion
evidence;

(2) a rule requiring an expert’s report to include –

141
(a) a statement explaining the extent to which the expert witness’s
opinion evidence is based on information falling outside his or her
own field of expertise and/or on the opinions of other (named)
experts;

(b) a schedule identifying the foundation material underpinning the


expert witness’s inferences and conclusions; and

(c) a rule that where an expert witness is called by a party to give a


reasoned opinion on the likelihood of an item of evidence under a
proposition advanced by that party, the expert’s report must also
include, where feasible, a reasoned opinion on the likelihood of
the item of evidence under one or more alternative propositions
(including any proposition advanced by the opposing party);

(3) an extension of rule 33.4(2) of the Criminal Procedure Rules so that, if a


party seeking to adduce expert evidence does not comply with the above
requirements, the evidence would be inadmissible unless all the parties
agree that it should be admitted or the court gives leave for it to be
admitted.

Paragraph 7.21

Other pre-trial disclosure requirements


9.19 We recommend that the Criminal Procedure Rules should require pre-trial
disclosure by the parties of the following matters to the other parties and to the
court:

(1) information relevant to the application of the expertise and impartiality


tests;

(2) if requested, information relevant to the application of the reliability test


(including, in particular, the evidence underpinning the expert’s opinion);
and

(3) information which could substantially undermine the credibility of the


experts being relied on.

9.20 We also recommend, in line with the current position under rule 33.4(2) of the
Criminal Procedure Rules 2010, that a party’s failure to comply with the
requirements of sub-paragraph (1) or with a request for disclosure under sub-
paragraph (2), should render that party’s expert evidence inadmissible, unless
the judge gives leave (or all the parties agree that the evidence should be
admitted).

Paragraph 7.37 and 7.38

142
Developing rule 33.6 of the Criminal Procedure Rules
9.21 We recommend that:

(1) Part 33 of the Criminal Procedure Rules be amended to make explicit


provision for a judge-led meeting of the parties’ legal representatives and
experts if there is a dispute on the expert issues and the judge believes
that such a meeting would be beneficial in resolving or reducing the
dispute; and

(2) this power be supported by a provision similar to that now set out in
rule 33.6(4) of the Rules.

Paragraphs 7.52

Directing the jury


9.22 We recommend that the Criminal Procedure Rules should provide that, for trials
on indictment (before a judge and jury), if the judge determines at the end of the
trial that the prosecution case depends wholly or substantially on disputed expert
opinion evidence, the judge should:

(1) consider whether to provide the jury with a cautionary warning in relation
to that evidence; and

(2) if a cautionary warning is thought to be appropriate, provide the jury with


an appropriate warning tailored to the facts of the case.

Paragraph 7.61

(Signed) JAMES MUNBY, Chairman


ELIZABETH COOKE
DAVID HERTZELL
DAVID ORMEROD
FRANCES PATTERSON

MARK ORMEROD, Chief Executive


21 February 2011

143
APPENDIX A

Criminal Evidence (Experts) Bill

CONTENTS

Admissibility
1 Basic rules
2 “Qualified to do so”
3 Impartiality
4 Reliability: meaning
5 Reliability: reconsideration on appeal

Procedural matters
6 Reliability: procedural matters
7 Reliability: procedure in magistrates’ courts
8 Disclosure

Court-appointed experts
9 Court-appointed experts

Final provisions
10 Interpretation
11 Short title, commencement, application and extent

Schedule — Reliability: factors


Part 1 — Generic factors
Part 2 — Factors for specific fields
Part 3 — Amendment of Part 1 factors

144
Admissibility of expert evidence in criminal
proceedings: the basic framework
EXPERT EVIDENCE PROFFERED
FOR ADMISSION

Does the expert evidence provide


information which is likely to be outside a
Expert
judge or jury’s experience and knowledge
No evidence
such that it would give them help they need inadmissible
in arriving at their conclusions?
(Cl 1(1)(a))

Yes

Has the party proffering the expert evidence Expert


proved that the person who is to give it is No evidence
qualified to do so? (Cl 1(1)(b) and cl 2) inadmissible

Yes

Is there a significant risk that the expert has


Expert
failed to fulfil, or that the expert will fail to Is the evidence opinion evidence?
No No evidence of fact
fulfil, his or her duty to give objective and (Cl 1(2)) admissible
unbiased evidence? (Cl 1(1)(3) and cl 3(1))

Yes Yes
Yes

Is it nevertheless in
Expert the interests of justice
evidence No that the evidence
inadmissible should be admitted?
(Cl 3(3)) Is there an issue
of evidentiary Expert
reliability to be No opinion evidence
determined? admissible
(Cl 6(2) and (3))

Yes

Evidentiary reliability is usually determined pre-


trial or, if during the trial, in the absence of the
jury. (Cl 6(4)) Has the party proffering the
expert opinion evidence shown Expert
No opinion
that it is sufficiently reliable to evidence
be admitted? (Cl 1(2), cl 4 and inadmissible
An independent expert may be appointed by the
the Schedule)
court to help the judge assess the reliability of
the expert opinion evidence. (Cl 9)
Yes

Expert
opinion evidence
admissible

145
Criminal Evidence (Experts) Bill 1

DRAFT

OF A

BILL
TO

Make provision about expert evidence in criminal proceedings.

B by the Queen’s most Excellent Majesty, by and with the advice and
E IT ENACTED
consent of the Lords Spiritual and Temporal, and Commons, in this present
Parliament assembled, and by the authority of the same, as follows:—

Admissibility

1 Basic rules
(1) Expert evidence is admissible in criminal proceedings only if—
(a) the court is satisfied that it would provide information which is likely
to be outside a judge or jury’s experience and knowledge, and which 5
would give them help they need in arriving at their conclusions,
(b) the person who gives it is qualified to do so (see section 2), and
(c) the evidence is not made inadmissible as a result of section 3
(impartiality).
(2) In addition, expert opinion evidence is admissible in criminal proceedings only 10
if it is sufficiently reliable to be admitted (see section 4).
(3) If there is a doubt about whether an expert’s evidence is evidence of fact or is
opinion evidence, it is to be taken to be opinion evidence.

2 “Qualified to do so”
(1) For the purposes of section 1(1)(b), a person may be qualified to give expert 15
evidence by virtue of study, training, experience or any other appropriate
means.
(2) The court must be satisfied on the balance of probabilities that the person is so
qualified.

146
EXPLANATORY NOTES
A.1 This draft Bill would make provision about expert evidence in criminal proceedings, but
only in England and Wales.1

Clause 1
A.2 Clause 1(1) sets out the basic admissibility rules for expert evidence in criminal
proceedings, whether the evidence is expert opinion evidence or expert evidence of fact:2
x Paragraph (a) in effect restates the common law “Turner test”.3 The reference to
“a judge or jury’s experience and knowledge” is a reference to a notional judge or
jury, so the test would be applied in magistrates’ courts, in Crown Court trials (or
appeals) without a jury and in Crown Court trials with a jury.

x Paragraph (b) provides that expert evidence can be given only by an individual
who is an expert, in accordance with clause 2.

x Paragraph (c) provides that expert evidence cannot be admitted if rendered


inadmissible by clause 3 (which sets out the impartiality requirement).

A.3 Clause 1(2) provides, in addition, that expert opinion evidence is admissible in criminal
proceedings only if it is sufficiently reliable to be admitted (in accordance with clauses 4
and 6 and the Schedule to the Bill).
A.4 Clause 1(3) provides (for the purposes of clause 1(2)) that expert evidence presented as
evidence of fact is nevertheless to be regarded by the court as expert opinion evidence if
there is any doubt on the matter.

Clause 2
A.5 Clause 2 addresses the question of expertise, replacing the common law requirement of
(expert) competence.4 An individual must be “qualified” before he or she is able to provide
expert evidence in criminal proceedings (clause 1(1)(b)).
A.6 Clause 2(1) lists key examples of the ways in which an individual claiming expertise can
satisfy the requirement in clause 1(1)(b).
A.7 Clause 2(2) provides that the standard of proof to be applied when assessing whether or
not an individual is qualified to be an expert is the balance of probabilities (that is, more
likely than not). Although this subsection does not expressly provide that the burden of
proof lies with the party wishing to adduce the individual’s evidence, it is implicit.

1
Clause 11(6).
2
Clause 10.
3
From Turner [1975] QB 834; see paras 2.3 to 2.5 of this report.
4
The word “competence” is already used in the Youth Justice and Criminal Evidence Act 1999 to
mean witness competence generally, so it is not used in this Bill.

1
2 Criminal Evidence (Experts) Bill

3 Impartiality
(1) An expert has a duty to the court to give objective and unbiased expert
evidence for the purpose of criminal proceedings.
(2) That duty overrides any obligation to the person from whom the expert
receives instructions or by whom the expert is paid. 5
(3) If it appears to the court that there is a significant risk that the expert will not
comply (or has not complied) with that duty in connection with the
proceedings, the expert evidence is not admissible unless the court is satisfied
that it is in the interests of justice that it should be admitted.
(4) The fact that the expert has an association (for example, an employment 10
relationship) which could make a reasonable observer think that the expert
might not comply with that duty does not in itself demonstrate a significant
risk.
(5) Criminal Procedure Rules may make further provision in connection with that
duty. 15

4 Reliability: meaning
(1) Expert opinion evidence is sufficiently reliable to be admitted 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. 20
(2) Any of the following, in particular, could provide a reason for determining that
expert opinion evidence is 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 stand up to scrutiny; 25
(b) the opinion is based on an unjustifiable assumption;
(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; 30
(e) the opinion relies on an inference or conclusion which has not been
properly reached.
(3) When assessing the reliability of expert opinion evidence, the court must have
regard to—
(a) such of the generic factors set out in Part 1 of the Schedule as appear to 35
the court to be relevant;
(b) if any factors have been specified in an order made under Part 2 of the
Schedule in relation to a particular field, such of those factors as appear
to the court to be relevant;
(c) anything else which appears to the court to be relevant. 40

5 Reliability: reconsideration on appeal


(1) This section applies if the court hearing an appeal (“the appellate court”) must
determine, or thinks it appropriate to determine, whether or not a ruling to
which subsection (2) applies involved an error.

148
EXPLANATORY NOTES

Clause 3
A.8 Subsections (1) and (2) set out the overriding duty to the court all experts have to give
objective and unbiased evidence, in accordance with the position at common law5 and as
currently required by rule 33.2 of the Criminal Procedure Rules 2010.6
A.9 Clause 3(3) provides that where there is a significant risk that an expert will not comply,
or has not complied, with his or her duty to the court in relation to the instant proceedings,
the expert’s evidence is inadmissible unless the court rules that it is in the interests of
justice to admit it. An example of an exceptional situation where an expert’s evidence
might be admitted, despite a significant risk of bias, could be where the risk is relatively
low (but significant), the expert’s evidence would materially support the accused’s
defence if believed, evidentiary reliability is not in dispute and there is a dearth of
alternative expert evidence for the accused to draw upon.
A.10 Clause 3(4) provides that the mere appearance of bias or partiality on account of the fact
of an association (such as an employment relationship) does not in itself demonstrate a
significant risk of the type described in clause 3(3). So, although there may be an
association which might “make a reasonable observer think” that the expert might not
comply with his or her overriding duty, this fact alone does not give rise to a “significant
risk that the expert will not comply (or has not complied) with that duty in connection with
the [instant] proceedings”. The rule of inadmissibility in clause 3(3) would arise only if the
particular factual nature of the association is such that it appears to the court that there is
a real, significant risk of non-compliance with that duty.
A.11 Clause 3(5) provides the Criminal Procedure Rules Committee with the power to create
further rules relating to the expert’s overriding duty.

Clause 4 (and the Schedule)


A.12 Clause 4(1) sets out the basic reliability test for expert opinion evidence (see clause 1(2)).
A.13 Clause 4(2) provides five key reasons for ruling that a party’s expert opinion evidence is
insufficiently reliable to be admitted. The list in this subsection is not exhaustive.
A.14 Clause 4(3) directs the court to consider relevant factors in Part 1 of the Schedule (and
relevant “specified” factors for a particular field, if any) when determining whether or not a
party’s expert opinion evidence is sufficiently reliable to be admitted. The court is also
directed to have regard to anything else which appears to be relevant. Part 1 of the
Schedule lists eight generic factors which have a bearing on the evidentiary reliability of
expert opinion evidence, supplementing the higher-level reasons listed in clause 4(2).

Clause 5
A.15 Clause 5 sets out the rules for the situation where a magistrates’ court or Crown Court
ruling on the admissibility of expert opinion evidence under clause 1(2) (with reference to
clause 4) is addressed by an appellate court.7 The rules apply whether the appellate court
has to determine the issue or simply decides in its discretion to consider the issue.8

5
Harris [2005] EWCA Crim 1980, [2006] 1 Cr App R 5 at [271]; Bowman [2006] EWCA Crim 417,
[2006] 2 Cr App R 3 at [176].
6
Rule 33.2(1)–(2) would be removed from the Rules if this draft Bill is enacted, but r 33.2(3)
would remain (albeit amended to encompass expert evidence of fact as well as expert opinion
evidence). Part 33 of the 2010 Rules is set out as Appendix B to this report.
7
See cl 5(2).
8
See cl 5(1).

1
Criminal Evidence (Experts) Bill 3

(2) This subsection applies to a ruling made by a magistrates’ court or the Crown
Court (“the original court”) as to whether expert opinion evidence was, or was
not, sufficiently reliable to be admitted in particular criminal proceedings.
(3) For the purposes of subsection (2) it does not matter whether the ruling was
express, or was implied by the original court’s admission of the evidence. 5
(4) The appellate court is to make its own determination of whether or not the
evidence was (or, as the case may be, is) sufficiently reliable to be admitted.
(5) So far as that determination is inconsistent with the ruling made by the original
court, the appellate court is to be taken to have determined that the ruling is
wrong in law (and may exercise its powers to dispose of the appeal 10
accordingly).
(6) In this section—
“appeal” means an appeal or application to the High Court, the criminal
division of the Court of Appeal or the Supreme Court, other than an
application for judicial review, and 15
references to a ruling made by the Crown Court include a ruling made by
a judge of the Crown Court.

Procedural matters

6 Reliability: procedural matters


(1) This section applies if a party to criminal proceedings proposes to adduce 20
expert opinion evidence.
(2) If a representation is made to the court that the evidence is not sufficiently
reliable to be admitted, and it appears to the court that it might not be, it is for
the party proposing to adduce the evidence to show that it is.
(3) As a condition of allowing the party to adduce the evidence, the court may of 25
its own motion require the party to show that it is sufficiently reliable to be
admitted.
(4) In the Crown Court, in a trial on indictment with a jury, if the question whether
or not expert evidence is sufficiently reliable to be admitted has not been
determined before the jury is sworn, it is to be determined in the absence of the 30
jury.

7 Reliability: procedure in magistrates’ courts


(1) A magistrates’ court which determines the question whether or not expert
opinion evidence is sufficiently reliable to be admitted in criminal proceedings
(“the question”) must be composed of a District Judge (Magistrates’ Courts)— 35
(a) when it determines the question, and
(b) during any subsequent part of the proceedings.
(2) If the question arises in circumstances in which the court could hold a pre-trial
hearing on it, the court must do so unless it appears to the court that it would
not be in the interests of justice to make a ruling on the question at such a 40
hearing.
(3) Criminal Procedure Rules may make further provision about the composition
of a magistrates’ court which determines the question, including—

150
EXPLANATORY NOTES
A.16 Clause 5(4) provides that, when a ruling on the evidentiary reliability of expert opinion
evidence is addressed on appeal,9 the question is not whether the ruling was a
reasonable one for the court to have made but whether or not the evidence was (or is)10
sufficiently reliable to be admitted. So, importantly, the ruling will not be addressed on
appeal as one made pursuant to the exercise of a judicial discretion.
A.17 Clause 5(5) provides that if the appellate court’s conclusion is different from the ruling, the
ruling is wrong in law. The appellate court will then act accordingly to rectify the error.

Clause 6
A.18 Clause 6(2) provides that it is for the party seeking to adduce expert opinion evidence to
show that it is sufficiently reliable to be admitted, but only if it appears to the court,
following a party’s representation, that the evidence might not be sufficiently reliable to be
admitted. For the purposes of this subsection, there is a weak presumption of sufficient
reliability (in respect of admissibility) which stands until a party presents a credible
argument for setting it aside. It will be set aside if it appears to the court that the opinion
evidence might not be sufficiently reliable to be admitted.
11
A.19 No such presumption applies in subsection (3). This provision allows the court to require
the party proffering the expert opinion evidence to show that it is sufficiently reliable to be
admitted even though the requirements of subsection (2) have not been met. The courts
are unlikely to exercise this power very often, but they will do so if there is an appearance
of unreliability which has not been raised by a party or perhaps if there is a morass of
confusing information which needs to be properly marshalled.
A.20 If the party proffering expert opinion evidence is required to show that it is sufficiently
reliable to be admitted, there is no requirement that this be proved to any particular
standard of proof; but that party must demonstrate to the court’s satisfaction that the
12
opinion is “soundly based” and that the strength of the opinion is warranted having
13
regard to its basis. Evidentiary reliability will ordinarily be addressed at a pre-trial
hearing (“before the jury is sworn”),14 but if the issue needs to be addressed during the
trial, the jury must not be present.15

Clause 7
A.21 Clause 7 sets out the procedure for magistrates’ courts. Subsections (1) to (3) provide
that where it appears to the court that proffered expert opinion evidence might not be
sufficiently reliable to be admitted, then (subject to exceptions in the Criminal Procedure
Rules) the question will be addressed by a District Judge (Magistrates’ Courts), usually at
a pre-trial hearing, and that judge will then try the case.16

9
An appeal to the High Court, Court of Appeal (Criminal Division) or the Supreme Court; see
cl 5(6).
10
For interlocutory appeals.
11
That is, cl 6(3) allows the court to set aside the presumption in clause 6(2).
12
Clause 4(1)(a).
13
Clause 4(1)(b).
14
Clause 6(4).
15
Clause 6(4).
16
For the meaning of pre-trial hearing, see cl 7(6), referring to s 8A of the Magistrates’ Courts Act
1980. Clause 7(2) refers to the “interests of justice” test in s 8A(3)(c) of the 1980 Act.

15
4 Criminal Evidence (Experts) Bill

(a) exceptions in relation to subsection (1),


(b) provision specifying circumstances in which a trial which has begun
otherwise than before a District Judge (Magistrates’ Courts) may be
terminated, and a new trial started before a District Judge (Magistrates’
Courts). 5
(4) Provision made by virtue of subsection (3) may confer a discretion on a court.
(5) Section 121(6) of the Magistrates’ Courts Act 1980 is not to be taken to prevent
the making under subsection (3)(a) of provision for the court to be composed
of a District Judge (Magistrates’ Courts) during part, but not the whole, of the
proceedings before it. 10
(6) In this section—
“pre-trial hearing” has the same meaning as in section 8A of the
Magistrates’ Courts Act 1980;
“trial” includes a trial of the facts following a plea of guilty.

8 Disclosure 15
(1) Criminal Procedure Rules may include—
(a) provision for the disclosure by a party to criminal proceedings of
information relevant to the question whether expert evidence which
the party proposes to adduce in the proceedings is admissible by virtue
of section 1(1)(b) or (c) or (2), 20
(b) provision for the disclosure by a party to criminal proceedings of
information that might reasonably be thought capable of substantially
detracting from the credibility of an expert on whom the party
proposes to rely,
(c) provision prohibiting (except in such circumstances, if any, as are 25
specified in the Rules) a party who fails to comply with a requirement
imposed by virtue of paragraph (a) or (b) from adducing the expert
evidence in question in the proceedings.
(2) The information in relation to which provision may be made by virtue of
subsection (1)(a) and (b) includes information that would otherwise be 30
privileged from disclosure on the ground that it is contained in a
communication made in connection with or in contemplation of legal
proceedings and for the purposes of such proceedings.
(3) But no provision may be made by virtue of subsection (2) in relation to
information contained in a communication from a defendant, or a person 35
acting on a defendant’s behalf, to an expert.

Court-appointed experts

9 Court-appointed experts
(1) Subsections (2) to (7) apply where the Crown Court has to determine whether
expert opinion evidence which a party proposes to adduce in a trial on 40
indictment is sufficiently reliable to be admitted.
(2) The court may appoint another expert to help it determine that question if
satisfied that it would be in the interests of justice to do so, having regard to—

152
EXPLANATORY NOTES
A.22 Subsections (3) to (5) of clause 7 set out new powers for the Criminal Procedure Rules
Committee to create procedural rules, including exceptions to the rule in subsection (1).
These exceptions could in certain circumstances permit:
x a bench of magistrates (rather than a District Judge) to determine evidentiary
reliability;17

x a bench of magistrates to try the case once a District Judge has determined
evidentiary reliability;

x a bench of magistrates to terminate a trial and rule that it be restarted before a


District Judge (who would then determine evidentiary reliability and try the case in
accordance with subsection (1)).18

A.23 The reference in subsection (6) to “a trial of the facts following a plea of guilty” refers to a
Newton hearing. A Newton hearing is a hearing to determine the facts where there is a
dispute following a guilty plea (so that an appropriate sentence can be passed).

Clause 8
A.24 Clause 8(1)(a) and (b) provides the Criminal Procedure Rules Committee with the power
to create rules requiring the pre-trial disclosure of information relevant to the admissibility
of expert evidence (under clause 1(1)(b), clause 1(1)(c) or clause 1(2)) and information
which could have a substantial adverse impact on an expert’s credibility.
A.25 By virtue of clause 8(2), this power would permit the creation of rules which would require
the disclosure of such information in communications currently protected from disclosure
by “litigation privilege” (but not if the communications are from the party or the party’s
19
agent to the expert or if the information is protected by another head of privilege). A
legal representative would therefore have to disclose information which had been
communicated to him or her by an expert witness – for example, evidence that the expert
is biased or untruthful or facts suggesting that the expert’s opinion is unreliable – even if it
was communicated in confidence.
A.26 Clause 8(1)(c) would allow the Criminal Procedure Rules Committee to create a rule
prohibiting the adduction of expert evidence by a party who has failed to comply with
disclosure rules made under this clause.

Clause 9
A.27 This clause would allow a Crown Court judge for a trial on indictment to call upon a
further expert witness (a court-appointed expert) in a hearing convened to determine the
evidentiary reliability of a party’s proffered expert opinion evidence.
A.28 Clause 9(2) provides the “interests of justice” test which would determine whether the
case is one which warrants having a court-appointed expert.

17
The magistrates would be guided by their legal adviser.
18
Clause 7(3)(b).
19
Clause 8(3).

15
Criminal Evidence (Experts) Bill 5

(a) the likely importance of the evidence in the context of the case as a
whole,
(b) the complexity of the evidence or of the question of its reliability, and
(c) any other relevant considerations.
(3) The other expert must be a person nominated for the purpose of the particular 5
proceedings—
(a) by a selection panel established by the Lord Chancellor, and
(b) in accordance with any procedure specified by the Lord Chancellor.
(4) The selection panel must if practicable nominate a number of persons, and if it
does the court may appoint any of them. 10
(5) The remuneration to be paid to an expert appointed under this section is to be
determined by the Lord Chancellor, and is to be paid out of central funds.
(6) The nomination and appointment of experts under this section, and the
procedure to be followed in relation to their evidence, are subject to any further
provision in Criminal Procedure Rules. 15
(7) The duty in section 3(1) applies to experts appointed under this section, but
otherwise sections 1 to 8 do not apply in relation to their evidence.
(8) The Crown Court has no power apart from this section to appoint an expert to
help it determine the question mentioned in subsection (1); and a magistrates’
court has no such power in relation to criminal proceedings before it. 20

Final provisions

10 Interpretation
In this Act—
“criminal proceedings” means criminal proceedings in relation to which
the strict rules of evidence apply; 25
references to an expert are to a person who gives, or is to give, expert
evidence;
references to expert evidence, however expressed, include all such
evidence, in any form and however given.

11 Short title, commencement, application and extent 30


(1) This Act may be cited as the Criminal Evidence (Experts) Act 2011.
(2) Section 10 and this section come into force on the day on which this Act is
passed, but otherwise this Act comes into force on such day as the Lord
Chancellor may by order made by statutory instrument appoint.
(3) An order made under subsection (2) may appoint different days for different 35
purposes.
(4) Nothing in this Act affects—
(a) any power of the court to exclude expert evidence at its discretion
(whether by preventing questions from being put or otherwise), or
(b) any rule of law (except so far as inconsistent with the provisions of this 40
Act), or any other enactment, so far as either relates to the admissibility
of expert evidence.

154
EXPLANATORY NOTES
A.29 Subsections (3) to (6) of clause 9 set out rules, or the power to create procedural rules,
relating to the appointment process and payment of fees. In particular, subsections (3)
and (4) provide that a selection panel must be convened and that the panel must (if
practicable) nominate a shortlist of experts from which list the trial judge would make his
or her appointment.
A.30 Clause 9(7) provides that, save for one exception, clauses 1 to 8 of the Bill do not apply
to the evidence of court-appointed experts, the reason being that the selection panel will
scrutinise any such individual for his or her expertise and the court-appointed expert’s
opinion is not one which (in the present context) is being tendered for admission before a
jury or equivalent fact-finding magistrates’ court.
A.31 The one exception mentioned in the previous paragraph is that a court-appointed expert
is under the same duty as any other expert to provide evidence which is objective and
unbiased (see clause 3(1)).
A.32 Clause 9(8) provides that there is no other power beyond the new rule in this clause to
appoint an expert to assist in the determination of evidentiary reliability (under the
admissibility test in clauses 1(2) and 4), whether in the Crown Court or in a magistrates’
court. The subsection expressly provides that a magistrates’ court has no power at all to
appoint an expert in this context.

Clause 10
A.33 Clause 10 explains some expressions used in the Bill:
x in line with the evidence provisions in the Criminal Justice Act 2003, the term
“criminal proceedings” is limited to criminal proceedings to which the strict rules of
the law of evidence apply (including criminal trials and Newton hearings);20

x an expert is a person who is to give (or gives) expert evidence in criminal


proceedings;

x references to expert evidence and expert opinion evidence “include all such
evidence, in any form and however given”, so “expert evidence” includes expert
evidence of fact or opinion (or both), and any reference to expert evidence or
expert opinion evidence includes such evidence given orally in court or in a written
report.

Clause 11
A.34 Save for subsection (4), this clause is self-explanatory.
A.35 Subsection (4) provides that the provisions in the Bill do not affect other existing statutory
or common law rules relating to the admissibility of expert evidence (such as the rules
governing the admissibility of hearsay evidence). The subsection also provides that the
judicial discretions criminal courts currently have at common law or under statute to
exclude evidence continue to operate in relation to expert evidence.

20
See Bradley [2005] EWCA Crim 20, [2005] 1 Cr App R 24 at [29] and [36]. For Newton
hearings, see para A.23 above.

1
6 Criminal Evidence (Experts) Bill

(5) This Act does not affect criminal proceedings in which any person has entered
a plea before sections 1 to 8 have come fully into force.
(6) This Act extends to England and Wales only.

156
Criminal Evidence (Experts) Bill 7
Schedule — Reliability: factors

SCHEDULE Section 4(3)

RELIABILITY: FACTORS

PART 1

GENERIC FACTORS

1 The factors referred to in section 4(3)(a) are as follows. 5


(a) The extent and quality of the data on which the opinion is based, and
the validity of the methods by which they were obtained.
(b) If the opinion relies on an inference from any findings, whether the
opinion properly explains how safe or unsafe the inference is
(whether by reference to statistical significance or in other 10
appropriate terms).
(c) If the opinion relies on the results of the use of any method (for
instance, a test, measurement or survey), whether the opinion takes
proper account of matters, such as the degree of precision or margin
of uncertainty, affecting the accuracy or reliability of those results. 15
(d) The extent to which any material upon which the opinion is based
has been reviewed by others with relevant expertise (for instance, in
peer-reviewed publications), and the views of those others on that
material.
(e) The extent to which the opinion is based on material falling outside 20
the expert’s own field of expertise.
(f) The completeness of the information which was available to the
expert, and whether the expert took account of all relevant
information in arriving at the opinion (including information as to
the context of any facts to which the opinion relates). 25
(g) Whether there is a range of expert opinion on the matter in question;
and, if there is, where in the range the opinion lies and whether the
expert’s preference for the opinion proffered has been properly
explained.
(h) Whether the expert’s methods followed established practice in the 30
field; and, if they did not, whether the reason for the divergence has
been properly explained.
2 These factors are not arranged in any hierarchical order.

PART 2

FACTORS FOR SPECIFIC FIELDS 35

3 The Lord Chancellor may by order made by statutory instrument provide


for other factors relevant to specific fields of expertise.
4 An order made under paragraph 3 must—
(a) state the field to which the factors are relevant, and
(b) set out the factors in question. 40

157
8 Criminal Evidence (Experts) Bill
Schedule — Reliability: factors
Part 2 — Factors for specific fields

5 A statutory instrument containing an order made under paragraph 3 is


subject to annulment in pursuance of a resolution of either House of
Parliament.

PART 3

AMENDMENT OF PART 1 FACTORS 5

6 The Lord Chancellor may by order made by statutory instrument amend


paragraph 1 of this Schedule so as to add, omit or amend any factor.
7 A statutory instrument containing an order under paragraph 6 may not be
made unless a draft of the instrument has been laid before and approved by
a resolution of each House of Parliament. 10

158
The Criminal Procedure Rules 2010 Part 33 as in force on 5 April 2010

APPENDIX B
PART 33 OF THE CRIMINAL PROCEDURE
RULES 2010

PART 33
EXPERT EVIDENCE

Contents of this Part

Reference to expert rule 33.1


Expert’s duty to the court rule 33.2
Content of expert’s report rule 33.3
Service of expert evidence rule 33.4
Expert to be informed of service of report rule 33.5
Pre-hearing discussion of expert evidence rule 33.6
Court’s power to direct that evidence is to be given by
a single joint expert rule 33.7
Instructions to a single joint expert rule 33.8
Court’s power to vary requirements under this Part rule 33.9
[Note. For the use of an expert report as evidence, see section 30 of the Criminal
Justice Act 1988(a).]

Reference to expert
33.1. A reference to an ‘expert’ in this Part is a reference to a person who is required
to give or prepare expert evidence for the purpose of criminal proceedings, including
evidence required to determine fitness to plead or for the purpose of sentencing.
[Note. Expert medical evidence may be required to determine fitness to plead under
section 4 of the Criminal Procedure (Insanity) Act 1964(b). It may be required also
under section 11 of the Powers of Criminal Courts (Sentencing) Act 2000(c), under
Part III of the Mental Health Act 1983(d) or under Part 12 of the Criminal Justice Act
2003(e). Those Acts contain requirements about the qualification of medical experts.]

Expert’s duty to the court


33.2.—(1) An expert must help the court to achieve the overriding objective by
giving objective, unbiased opinion on matters within his expertise.

(a) 1988 c. 33; section 30(4A) was inserted by section 47 of, and paragraph 32 of Schedule 1 to, the Criminal Procedure and
Investigations Act 1996 (c. 25) and is repealed by section 41 to, and paragraph 60(1) and (6) of Schedule 3 and Schedule 37
to, the Criminal Justice Act 2003 (c. 44), with effect from a date to be appointed.
(b) 1964 c. 84; section 4 was substituted, together with section 4A, for section 4 as originally enacted, by section 2 of the
Criminal Procedure (Insanity and Unfitness to Plead) Act 1991 (c. 25), and amended by section 22 of the Domestic
Violence, Crime and Victims Act 2004 (c. 28).
(c) 2000 c. 6.
(d) 1983 c. 20.
(e) 2003 c. 44.

159
The Criminal Procedure Rules 2010 Part 33 as in force on 5 April 2010
(2) This duty overrides any obligation to the person from whom he receives
instructions or by whom he is paid.
(3) This duty includes an obligation to inform all parties and the court if the expert’s
opinion changes from that contained in a report served as evidence or given in a
statement.

Content of expert’s report


33.3.—(1) An expert’s report must—
(a) give details of the expert’s qualifications, relevant experience and
accreditation;
(b) give details of any literature or other information which the expert has relied
on in making the report;
(c) contain a statement setting out the substance of all facts given to the expert
which are material to the opinions expressed in the report, or upon which
those opinions are based;
(d) make clear which of the facts stated in the report are within the expert’s own
knowledge;
(e) say who carried out any examination, measurement, test or experiment which
the expert has used for the report and—
(i) give the qualifications, relevant experience and accreditation of that
person,
(ii) say whether or not the examination, measurement, test or experiment was
carried out under the expert’s supervision, and
(iii) summarise the findings on which the expert relies;
(f) where there is a range of opinion on the matters dealt with in the report—
(i) summarise the range of opinion, and
(ii) give reasons for his own opinion;
(g) if the expert is not able to give his opinion without qualification, state the
qualification;
(h) contain a summary of the conclusions reached;
(i) contain a statement that the expert understands his duty to the court, and has
complied and will continue to comply with that duty; and
(j) contain the same declaration of truth as a witness statement.
(2) Only sub-paragraphs (i) and (j) of rule 33.3(1) apply to a summary by an expert
of his conclusions served in advance of that expert’s report.
[Note. Part 27 contains rules about witness statements. Declarations of truth in
witness statements are required by section 9 of the Criminal Justice Act 1967(a) and

(a) 1967 c. 80; section 9 was amended by section 56 of and paragraph 49 of Schedule 8 to, the Courts Act 1971 (c. 23), section
69 of the Criminal Procedure and Investigations Act 1996 (c. 25), section 168 of, and paragraph 6 of Schedule 9 to, the
Criminal Justice and Public Order Act 1994 (c. 33) and regulation 9 of, and paragraph 4 of Schedule 5 to S.I. 2001/1090. It
is amended by section 72 of, and paragraph 55 of Schedule 5 to, the Children and Young Persons Act 1969 (c. 54), section
65, and paragraph 1 of Schedule 4 to, the Courts Act 2003 (c. 39) and sections 41 and 332 of, and paragraph 43 of Schedule
3 and Part 4 of Schedule 37 to, the Criminal Justice Act 2003 (c. 44), with effect from a date to be appointed.

160
The Criminal Procedure Rules 2010 Part 33 as in force on 5 April 2010
section 5B of the Magistrates’ Courts Act 1980(a). A party who accepts another
party’s expert’s conclusions may admit them as facts under section 10 of the Criminal
Justice Act 1967(b). Evidence of examinations etc. on which an expert relies may be
admissible under section 127 of the Criminal Justice Act 2003(c).]

Service of expert evidence


33.4.—(1) A party who wants to introduce expert evidence must—
(a) serve it on—
(i) the court officer, and
(ii) each other party;
(b) serve it—
(i) as soon as practicable, and in any event
(ii) with any application in support of which that party relies on that evidence;
and
(c) if another party so requires, give that party a copy of, or a reasonable
opportunity to inspect—
(i) a record of any examination, measurement, test or experiment on which
the expert’s findings and opinion are based, or that were carried out in the
course of reaching those findings and opinion, and
(ii) anything on which any such examination, measurement, test or
experiment was carried out.
(2) A party may not introduce expert evidence if that party has not complied with
this rule, unless–
(a) every other party agrees; or
(b) the court gives permission.
[Note. Under section 81 of the Police and Criminal Evidence Act 1984(d), and under
section 20(3) of the Criminal Procedure and Investigations Act 1996(e), rules may—
(a) require the disclosure of expert evidence before it is introduced as part of a
party’s case; and
(b) prohibit its introduction without the court’s permission, if it was not disclosed
as required.]

Expert to be informed of service of report


33.5. A party who serves on another party or on the court a report by an expert must,
at once, inform that expert of that fact.

(a) 1980 c. 43; section 5B was inserted by section 47 of, and paragraph 3 of Schedule 1 to, the Criminal Procedure and
Investigations Act 1996 (c. 25), and is amended by section 72(3) of, and paragraph 55 of Schedule 5 to, the Children and
Young Persons Act 1969 (c. 54), with effect from a date to be appointed. It is repealed by sections 41 and 332 of, and
paragraph 51(1) and (3) of Schedule 3 and Schedule 37 to, the Criminal Justice Act 2003 (c. 44), with effect from a date to
be appointed.
(b) 1967 c. 80.
(c) 2003 c. 44; section 127 was amended by article 3 of, and paragraphs 45 and 50 of the Schedule to, S.I. 2004/2035.
(d) 1984 c. 60; section 81 was amended by section 109(1) of, and paragraph 286 of Schedule 8 to, the Courts Act 2003 (c. 39).
(e) 1996 c. 25; section 20(3) was amended by section 109(1) of, and paragraph 378 of Schedule 8 to, the Courts Act 2003
(c. 39).

161
The Criminal Procedure Rules 2010 Part 33 as in force on 5 April 2010
Pre-hearing discussion of expert evidence
33.6.—(1) This rule applies where more than one party wants to introduce expert
evidence.
(2) The court may direct the experts to—
(a) discuss the expert issues in the proceedings; and
(b) prepare a statement for the court of the matters on which they agree and
disagree, giving their reasons.
(3) Except for that statement, the content of that discussion must not be referred to
without the court’s permission.
(4) A party may not introduce expert evidence without the court’s permission if the
expert has not complied with a direction under this rule.
[Note. At a pre-trial hearing, a court may make binding rulings about the
admissibility of evidence and about questions of law under section 7 of the Criminal
Justice Act 1987(a); sections 31 and 40 of the Criminal Procedure and Investigations
Act 1996(b); and section 45 of the Courts Act 2003(c).]

Court’s power to direct that evidence is to be given by a single joint expert


33.7.—(1) Where more than one defendant wants to introduce expert evidence on an
issue at trial, the court may direct that the evidence on that issue is to be given by one
expert only.
(2) Where the co-defendants cannot agree who should be the expert, the court
may—
(a) select the expert from a list prepared or identified by them; or
(b) direct that the expert be selected in another way.

Instructions to a single joint expert


33.8.—(1) Where the court gives a direction under rule 33.7 for a single joint expert
to be used, each of the co-defendants may give instructions to the expert.
(2) When a co-defendant gives instructions to the expert he must, at the same time,
send a copy of the instructions to the other co-defendant(s).
(3) The court may give directions about—
(a) the payment of the expert’s fees and expenses; and
(b) any examination, measurement, test or experiment which the expert wishes to
carry out.
(4) The court may, before an expert is instructed, limit the amount that can be paid
by way of fees and expenses to the expert.

(a) 1987 c. 38; section 7 was amended by section 168(1) of, and paragraph 30 of Schedule 9 to, the Criminal Justice and Public
Order Act 1994 (c. 33) and section 80 of, and paragraph 2 of Schedule 3 and Schedule 5 to, the Criminal Procedure and
Investigations Act 1996 (c. 25). It has been further amended by sections 45 and 310 of, and paragraphs 52 and 53 of
Schedule 36 to, the Criminal Justice Act 2003 (c. 44), for certain purposes, with effect from 24 July 2006, and for remaining
purposes from a date to be appointed).
(b) 1996 c. 25; section 31 was amended by sections 310, 331 and 332 of, and paragraphs 20, 36, 65 and 67 of Schedule 36 and
Schedule 37 to, the Criminal Justice Act 2003 (c. 44).
(c) 2003 c. 39.

162
The Criminal Procedure Rules 2010 Part 33 as in force on 5 April 2010
(5) Unless the court otherwise directs, the instructing co-defendants are jointly and
severally liable for the payment of the expert’s fees and expenses.

Court’s power to vary requirements under this Part


33.9.—(1) The court may—
(a) extend (even after it has expired) a time limit under this Part;
(b) allow the introduction of expert evidence which omits a detail required by this
Part.
(2) A party who wants an extension of time must—
(a) apply when serving the expert evidence for which it is required; and
(b) explain the delay.

163
APPENDIX C Impact Assessment (IA)
The admissibility of expert evidence in IA No: LAWCOM0002
criminal proceedings in England and Wales Date: 22 March 2010
Lead department or agency: Stage: Final
The Law Commission
Source of intervention: Domestic
Other departments or agencies:
Ministry of Justice Type of measure: Primary legislation
Contact for enquiries on this project:
Raymond Emson: 020 3334 0272

Summary: Intervention and Options


What is the problem under consideration? Why is government intervention necessary?
The admissibility of expert evidence in criminal trials is governed by an unduly liberal, common law
“relevance and reliability” test; and judges have little, if any, guidance on how they should determine
reliability. Expert evidence can therefore be admitted without sufficient enquiry into its reliability, which
means that juries may rely on unreliable expert evidence in reaching their verdicts. There have been a
number of wrongful convictions involving unreliable expert evidence in recent years, suggesting a real,
ongoing problem. Government intervention is required to replace the common law approach with a more
robust admissibility test, and to provide judges with the guidance they need in order to apply it.
What are the policy objectives and the intended effects?
The policy objectives for our proposals are:
x to provide a more robust admissibility test, so that only reliable expert evidence is admitted;
x to provide judges with uniform criteria against which to assess reliability; and
x to encourage parties to tender only reliable expert evidence for admission.
The effects will be:
x to improve the reliability of expert evidence used in criminal proceedings; and
x to avoid wrongful convictions and acquittals based on unreliable expert evidence.

What policy options have been considered? Please justify preferred option (further details in Evidence Base)
Option 0: Do nothing.
Option 1: Judicial assessment of evidentiary reliability (including both 1A and 1B). Codification of the
uncontroversial common law admissibility requirements for expert evidence, a new statutory reliability
test to govern the admissibility of expert opinion evidence and guidelines to assist the judge when
determining the evidentiary reliability of expert opinion evidence.
1A: Amendments to the rules on pre-trial disclosure and the introduction of a judge-led meeting of
experts (with parties).
1B: New judicial power to appoint a further expert witness to assist in the determination of evidentiary
reliability and the introduction of a selection system to guarantee the suitability of such experts; but
the power to be used only exceptionally, if it is in the interests of justice.
Option 1 is the preferred option because it offers the best solution to the problems and was broadly
supported by consultees.
Option 2: Judicial assessment of evidentiary reliability (including 1A, but excluding 1B).
When will the policy be reviewed to establish its impact and the extent to which It will not be reviewed
the policy objectives have been achieved?

Are there arrangements in place that will allow a systematic collection of Not applicable
monitoring information for future policy review?

Chair’s Sign-off For final proposal stage Impact Assessments:


I have read the Impact Assessment and I am satisfied that (a) it represents a fair and reasonable
view of the expected costs, benefits and impact of the policy, and (b) the benefits justify the costs.

Signed by the responsible Chair:........................................................................... Date:........................................

164 URN 10/899 Ver. 1.0 04/10


Summary: Analysis and Evidence Policy Option 1
Description: Judicial assessment of evidentiary reliability (including 1A and 1B).
Price Base PV Base Time Period Net Benefit (Present Value (PV)) (£m)
Year 09/10 Year 2011 Years 10 Low: +£16.641 High: -£45.379 Best Estimate: +£3.565

COSTS (£m) Total Transition Average Annual Total Cost


(Constant Price) Years (excl. Transition) (Constant Price) (Present Value)
Low £0.473 £0.000 £0.473
High £1.177 5 £9.547 £80.222
1
Best Estimate £0.766 £2.582 £22.151
Description and scale of key monetised costs by ‘main affected groups’
Criminal justice system – Transitional costs: JSB (training of legal professionals): £5,000 in year 0;
CPS/LSC/HMCS (increased appeals over 5 years): £65,000 annually; experts (experts’ appendices) (1A):
£468,000 in year 0. Ongoing costs: Experts/Police/CPS/LSC/HMCS (annual increase in pre-trial hearings
under the new test): £1,607,097; (increase in judge-led pre-trial meetings) (1A): £1,607,097; MoJ/HMCS
(costs of panel and fees for witnesses) (1B): £13,209.
Other key non-monetised costs by ‘main affected groups’ None

BENEFITS (£m) Total Transition Average Annual Total Benefit


(Constant Price) Years (excl. Transition) (Constant Price) (Present Value)
Low £0 £2.060 £17.114
High £0 0 £4.205 £34.843
Best Estimate £0 £3.101 £25.716
Description and scale of key monetised benefits by ‘main affected groups’
Criminal justice system – Ongoing benefits: Individuals/HMCS/LSC/CPS/Police (reduction in trials): £1,472,800;
(shorter trials if less evidence is tendered): £775,108; (savings in experts’ fees if fewer reports are
commissioned): £19,917; (reduction in appeals): £50,000; (reduction in trial time from judge-led meetings) (1A):
£542,867; (reduction in pre-trial discussions) (1A): £234,989; (shorter trials from court appointed expert) (1B):
£20,798.
Other key non-monetised benefits by ‘main affected groups’
Criminal justice system: improved justice with fewer wrongful convictions, fewer wrongful acquittals; enhanced
public confidence; increased clarity, consistency and uniformity of law; benefits to the lives and wellbeing of
persons who would otherwise be wrongly convicted (and their families).
Key assumptions/sensitivities/risks Discount rate (%) 3.5%
Key assumptions: Appeals are heard by the Court of Appeal and each costs £25,000. 95% of expert
evidence tendered in the Crown Court and 98% tendered in the magistrates’ courts will pass the new
test. An item of expert evidence adds three hours to the trial in the Crown Court and one hour in a
magistrates’ court. Experts are paid £156 per hour. A wrongful conviction costs at least £123,548.
Sensitivities: Between 0 and 5 (2 best) additional appeals in years 1 to 5. Between 1 and 3 (2 best) fewer
appeals annually. Between 0% (low) and 40% (high) of expert evidence which would not pass the test
will still be tendered for admission. Pre-trial hearings will take 0.5 to 3 days (1 best) in the Crown Court
and 1 hour to 1 day (0.5 days best) in a magistrates’ court. Judge-led meetings of experts will take 0.5 to
1 day (1 best). Judge-led meetings of experts will be used in 0% to 2.5% (2% best) of cases where
expert evidence is tendered. The power under 1B to appoint an independent expert will be used
between 0 and 10 (5 best) times a year.
Risks: Possible increase in appeals under the new test, increasing costs. The test could reveal deficiencies
in documentation, and corrections to meet its requirements would result in higher costs for businesses.
Judge-led, pre-trial hearings under 1A might be used more often and the power exercised under 1B
might also be used more often. The independent panel selecting court-appointed experts might not work
on a voluntary basis, resulting in higher costs.

Direct impact of business(Equivalent Annual £m) In scope of OOIO Measure qualifies as:
Costs: £0.468 Benefits: £0 Net: -£0.468 In In

1
Figures reflect the present value of transitional costs, not the annual transitional cost.
165
Enforcement, Implementation and Wider Impacts
What is the geographic coverage of the policy/option? England and Wales
From what date will the policy be implemented? Unknown
Which organisation(s) will enforce the policy? The Judiciary of England and
Wales and lay magistrates.
What is the annual change in enforcement cost (£m)? £0
Does enforcement comply with Hampton principles? Yes
Does implementation go beyond minimum EU requirements? Yes
What is the CO2 equivalent change in greenhouse gas emissions? Traded: Non-traded:
(Million tonnes CO2 equivalent) 0 0
Does the proposal have an impact on competition? No
What proportion (%) of Total PV costs/benefits is directly attributable to Costs: Benefits:
primary legislation, if applicable? 100% 100%
Annual cost (£m) per organisation Micro < 20 Small Medium Large
(excl. Transition) (Constant Price) £0 £0 £0 £0 £0
Are any of these organisations exempt? No No No No No

Specific Impact Tests: Checklist


Set out in the table below where information on any SITs undertaken as part of the analysis of the policy
options can be found in the evidence base. For guidance on how to complete each test, double-click on
the link for the guidance provided by the relevant department.
Please note this checklist is not intended to list each and every statutory consideration that departments
should take into account when deciding which policy option to follow. It is the responsibility of
departments to make sure that their duties are complied with.
Does your policy option/proposal have an impact on…? Impact Page ref
within IA
Statutory equality duties2 No 203
Statutory Equality Duties Impact Test guidance
Economic impacts
Competition Competition Assessment Impact Test guidance No 203
Small firms Small Firms Impact Test guidance Yes 203
Environmental impacts
Greenhouse gas assessment Greenhouse Gas Assessment Impact Test guidance No 204
Wider environmental issues Wider Environmental Issues Impact Test guidance No 204
Social impacts
Health and well-being Health and Well-being Impact Test guidance Yes 204
Human rights Human Rights Impact Test guidance Yes Throughout
Justice system Justice Impact Test guidance Yes Throughout
Rural proofing Rural Proofing Impact Test guidance No 204
Sustainable development No 204
Sustainable Development Impact Test guidance

2
Race, disability and gender impact assessments are statutory requirements for relevant policies. Equality
statutory requirements will be expanded 2011, once the relevant provisions of the Equality Act 2010 come into
force.
166
Summary: Analysis and Evidence Policy Option 2
Description: Judicial assessment of evidentiary reliability (including 1A but excluding 1B)
Price Base PV Base Time Period Net Benefit (Present Value (PV)) (£m)
Year 09/10 Year 2011 Years 10 Low: +£16.641 High:-£45.867 Best Estimate: +£3.504

COSTS (£m) Total Transition Average Annual Total Cost


(Constant Price) Years (excl. Transition) (Constant Price) (Present Value)
Low £0.473 £0 £0.473
High £1.177 5 £9.501 £79.841
3
Best Estimate £0.766 £2.569 £22.040
Description and scale of key monetised costs by ‘main affected groups’
Criminal justice system – Transitional costs: JSB (training of legal professionals): £5,000 in year 0;
CPS/LSC/HMCS (increased appeals over 5 years): £65,000 annually; experts (experts’ appendices) (1A):
£468,000 in year 0.
Ongoing costs: Experts/CPS/Police/LSC/HMCS (annual increase in pre-trial hearings under the new test):
£962,122; (increase in judge-led pre-trial meetings) (1A): £1,607,097.

Other key non-monetised costs by ‘main affected groups’ None


BENEFITS (£m) Total Transition Average Annual Total Benefit
(Constant Price) Years (excl. Transition) (Constant Price) (Present Value)
Low £0 £2.060 £17.114
High £0 0 £4.050 £33.974
Best Estimate £0 £3.079 £25.544
Description and scale of key monetised benefits by ‘main affected groups’
Criminal justice system – Ongoing benefits: Individuals/HMCS/LSC/CPS/Police (reduction in trials): £1,472,800;
(shorter trials if less evidence is tendered): £775,108; (savings in experts’ fees if fewer reports are
commissioned): £19,917; (reduction in appeals): £50,000; (reduction in trial time from judge-led meetings) (1A):
£542,867; (reduction in pre-trial discussions) (1A): £234,989.

Other key non-monetised benefits by ‘main affected groups’


Criminal justice system: improved justice with fewer wrongful convictions and acquittals and enhanced public
confidence; increased clarity, consistency and uniformity of law; benefits to the lives and wellbeing of persons
who would otherwise be wrongly convicted (and their families).
Key assumptions/sensitivities/risks Discount rate (%) 3.5%
Key assumptions: Appeals are heard by the Court of Appeal and each costs £25,000. 95% of expert
evidence tendered in the Crown Court and 98% tendered in the magistrates’ courts will pass the new
test. An item of expert evidence adds three hours to the trial in the Crown Court and one hour in a
magistrates’ court. Experts are paid £156 per hour. A wrongful conviction costs at least £123,548.
Sensitivities: 0 and 5 (2 best) additional appeals in years 1 to 5. Between 1 and 3 (2 best) fewer appeals.
Between 0% (low) and 40% (high) of expert evidence which would not pass the test will still be tendered
for admission. Pre-trial hearings will take 0.5 to 3 days (1 best) in the Crown Court and 1 hour to 1 day
(0.5 days best) in a magistrates’ court. Judge-led meetings of experts will take 0.5 to 1 day (1 best).
Judge-led meetings of experts will be used in 0% to 2.5% (2% best) of cases where expert evidence is
tendered.
Risks: Possible increase in appeals under the new test, increasing costs. The test could reveal deficiencies
in documentation, and corrections to meet its requirements would result in higher costs for businesses.
Judge-led, pre-trial hearings under 1A might be used more often.

Direct impact of business(Equivalent Annual £m) In scope of OIOO: Measure qualifies as:
Costs: £0.468 Benefits: £0 Net: -£0.468 Yes In

3
Figures reflect the present value of transitional costs, not the annual transitional cost.
167
Enforcement, Implementation and Wider Impacts
What is the geographic coverage of the policy/option? England and Wales
From what date will the policy be implemented? Unknown
Which organisation(s) will enforce the policy? The Judiciary of England and
Wales and lay magistrates.
What is the annual change in enforcement cost (£m)? £0
Does enforcement comply with Hampton principles? Yes
Does implementation go beyond minimum EU requirements? Yes
What is the CO2 equivalent change in greenhouse gas emissions? Traded: Non-traded:
(Million tonnes CO2 equivalent) 0 0
Does the proposal have an impact on competition? No
What proportion (%) of Total PV costs/benefits is directly attributable to Costs: Benefits:
primary legislation, if applicable? 100% 100%
Annual cost (£m) per organisation Micro < 20 Small Medium Large
(excl. Transition) (Constant Price) £0 £0 £0 £0 £0
Are any of these organisations exempt? No No No No No

Specific Impact Tests: Checklist


Set out in the table below where information on any SITs undertaken as part of the analysis of the policy
options can be found in the evidence base. For guidance on how to complete each test, double-click on
the link for the guidance provided by the relevant department.
Please note this checklist is not intended to list each and every statutory consideration that departments
should take into account when deciding which policy option to follow. It is the responsibility of
departments to make sure that their duties are complied with.
Does your policy option/proposal have an impact on…? Impact Page ref
within IA
Statutory equality duties4 No 203
Statutory Equality Duties Impact Test guidance
Economic impacts
Competition Competition Assessment Impact Test guidance No 203
Small firms Small Firms Impact Test guidance Yes 203
Environmental impacts
Greenhouse gas assessment Greenhouse Gas Assessment Impact Test guidance No 204
Wider environmental issues Wider Environmental Issues Impact Test guidance No 204
Social impacts
Health and well-being Health and Well-being Impact Test guidance Yes 204
Human rights Human Rights Impact Test guidance Yes Throughout
Justice system Justice Impact Test guidance Yes Throughout
Rural proofing Rural Proofing Impact Test guidance No 204
Sustainable development No 204
Sustainable Development Impact Test guidance

4
Race, disability and gender impact assessments are statutory requirements for relevant policies. Equality
statutory requirements will be expanded 2011, once the relevant provisions of the Equality Act 2010 come into
force.
168
Evidence Base (for summary sheets) – Notes
Use this space to set out the relevant references, evidence, analysis and detailed narrative from which you
have generated your policy options or proposal. Please fill in References section.

References
Include the links to relevant legislation and publications, such as public impact assessment of earlier
stages (e.g. Consultation, Final).

No. Legislation or publication


Expert Evidence in Criminal Proceedings in England and Wales (2011) Law
1
Commission No 325
The Admissibility of Expert Evidence in Criminal Proceedings in England and
2
Wales (2009) Law Commission Consultation Paper No 190.

Evidence Base
Ensure that the information in this section provides clear evidence of the information provided in the
summary pages of this form (recommended maximum of 30 pages). Complete the Annual profile of
monetised costs and benefits (transition and recurring) below over the life of the preferred policy (use
the spreadsheet attached if the period is longer than 10 years).
The spreadsheet also contains an emission changes table that you will need to fill in if your measure has
an impact on greenhouse gas emissions.

Annual profile of monetised costs and benefits* - (£m) constant prices

Y0 Y1 Y2 Y3 Y4 Y5 Y6 Y7 Y8 Y9
Transition costs £0.473 £0.065 £0.065 £0.065 £0.065 £0.065 £0 £0 £0 £0
Annual recurring cost £0 £2.582 £2.570 £2.570 £2.570 £2.570 £2.570 £2.570 £2.570 £2.570

Total annual costs £0.473 £2.647 £2.635 £2.635 £2.635 £2.635 £2.570 £2.570 £2.570 £2.570

Transition benefits £0 £0 £0 £0 £0 £0 £0 £0 £0 £0
Annual recurring benefits £0 £3.101 £3.082 £3.086 £3.090 £3.094 £3.094 £3.094 £3.094 £3.094

Total annual benefits £0 £3.101 £3.082 £3.086 £3.090 £3.094 £3.094 £3.094 £3.094 £3.094

* For non-monetised benefits please see summary pages and main evidence base section

Microsoft Office
Excel Worksheet

169
Evidence Base
1. Introduction
Background to the problem
Expert evidence in criminal trials can have a persuasive effect, particularly in cases where the field
of expertise is difficult to understand. This is not necessarily problematic, if the evidence is reliable;
but clearly the admission of unreliable expert evidence is likely to be harmful. The real possibility of
jury deference to expert opinion evidence means that the admission of unreliable expert evidence
is likely to distort the jury’s understanding of the facts, adversely affect its deliberations and result
in erroneous conclusions, as evidenced by a number of wrongful convictions in recent years. In the
past eight years, there have been at least 11 wrongful convictions caused by (or involving)
unreliable prosecution expert evidence, suggesting a real, ongoing problem. The most well-known
cases are Dallagher,5 Clark (Sally),6 Cannings7 and Harris and others.8
Such cases demonstrate that unreliable expert evidence can be admitted too readily. A further
problem is that the trial process does not provide sufficient safeguards which would prevent
miscarriages of justice in cases where unreliable evidence is admitted. In particular, cross-
examination may not be an effective tool for bringing out weaknesses in the foundation material
underpinning an expert’s opinion evidence.

Problem under consideration


The current law on the admissibility of expert opinion evidence in criminal proceedings is
unsatisfactory. Although there are the four common law admissibility requirements (assistance,
expertise, impartiality and evidentiary reliability), a laissez-faire attitude exists in relation to the
question of evidentiary reliability. The absence of effective scrutiny before expert opinion evidence
is placed before juries in criminal trials means that unreliable evidence can be admitted too freely.
This problem is compounded by the absence of a clear test or guidelines which would help trial
judges assess evidentiary reliability.
Insufficient judicial scrutiny and the possibility that juries may base their verdicts on unreliable
expert evidence means there is a “pressing danger” of wrongful convictions.9 There may also be
wrongful acquittals if the accused is allowed to adduce unreliable expert opinion evidence with a
view to undermining a credible prosecution case. A further problem is that an otherwise reputable
expert may stray outside his or her field of expertise and put forward, unchallenged, an unreliable
hypothesis, meaning that the jury will ultimately reach a verdict based in part on flawed evidence.
This can also lead to miscarriages of justice.
In the absence of reform through government intervention, the pressing danger of wrongful
convictions, and the risk of wrongful acquittals, will remain. Reform in the form of a new approach
to the admissibility of expert evidence would address these problems and safeguard public
confidence in the criminal justice system.

Rationale for intervention


The reason why expert evidence is admitted in criminal trials is to help jurors (or the other fact-
finding individuals) come to a correct decision on the facts of the case before them. Given the risk
of jury deference and the increasing complexity of much expert evidence, a strong case for reform
can be made on the principled ground that only expert opinion evidence which has been properly
screened for reliability should be considered by a jury in a criminal trial. Although there is a

5
[2002] EWCA Crim 1903, [2003] 1 Cr App R 12.
6
[2003] EWCA Crim 1020, [2003] 2 FCR 447.
7
[2004] EWCA Crim 1, [2004] 1 WLR 2607.
8
[2005] EWCA Crim 1980, [2006] 1 Cr App R 5.
9
D Ormerod and A Roberts, “Expert evidence: where now? What next?” (2006) 5 Archbold News 5.

170
common law reliability test for expert evidence, certainly for evidence of a scientific nature,10
judges have been given little if any guidance on how it should be applied in practice, and the
common law test is in any event far from robust.11
The case for reform can also be made on the ground that the judiciary should have the tools and
guidance they need to do what they are already duty-bound to do at common law, at least for
evidence of a scientific nature.
The conventional economic approach to government intervention, to resolve a problem, is based
on efficiency or equity arguments. The Government may consider intervening if there are strong
enough failures in the way markets operate (for example, monopolies overcharging consumers) or
if there are strong enough failures in existing interventions (for example, waste generated by
misdirected rules). In both cases the proposed intervention should avoid creating a further set of
disproportionate costs and distortions. The Government may also intervene for equity (fairness)
and redistributional reasons (for example, to reallocate goods and services to more needy groups
in society).
In economic terms, there is currently a failure in the “market” for expert evidence because judges
and juries will often have insufficient information to address its reliability. Expert evidence could
therefore be said to be a “credence good”.12 Unreliable expert evidence may be relied on because
the jury is not in a position to determine whether or not the evidence is reliable. As explained
above, this could lead to wrongful convictions and wrongful acquittals.

Policy objectives
Our policy objectives are to ensure:
x that where unreliable expert opinion evidence is tendered for admission in criminal
proceedings it is not admitted, which would be achieved by providing judges with a new,
robust reliability test and the guidance required to apply it;
x that the common law admissibility requirements in relation to assistance, expertise and
impartiality are clarified, publicised and properly enforced, which would be achieved
through codification;
x that judges and parties in criminal proceedings have access to the information they need to
determine whether the admissibility requirements for expert evidence are satisfied in a
given case, which would be achieved by enhanced disclosure requirements;
x that judges have the tools to be able properly to manage expert evidence in advance of
trials, which would be achieved by the foregoing and by a new power permitting judge-led
hearings with the experts (and parties) to narrow the disputed expert issues;
x that judges have access to the best available help when applying the reliability test to
extremely complex expert evidence, which would be achieved by providing judges with a
new power to call upon a further expert witness who has been independently screened for
impartiality and expertise;
x that, as a result of the above:
o only reliable expert evidence is tendered for admission in criminal proceedings;
o only reliable expert evidence is admitted in criminal proceedings;
o clarity, certainty and consistency are brought to the law;
o the risk of wrongful convictions and wrongful acquittals is reduced; and

10
Reed [2009] EWCA Crim 2698, [2010] 1 Cr App R 23 at [111]: “expert evidence of a scientific nature is not
admissible where the scientific basis on which it is advanced is insufficiently reliable for it to be put before
the jury”.
11
Reed [2009] EWCA Crim 2698, [2010] 1 Cr App R 23 at [111]: there is “no enhanced test of admissibility”,
even for evidence of a scientific nature.
12
A “credence good” is a term used in economics for something whose utility impact is difficult or impossible
for the consumer to ascertain, even after consumption.

171
o public confidence in the criminal justice system is strengthened.
As intimated above, any reform measure designed to address the problems associated with expert
evidence in criminal proceedings must be a proportionate response to the problem. We have borne
this in mind when formulating our recommendations to achieve our policy objectives.

Scale and scope


There is information on the use and procurement of expert evidence generally, but it is very difficult
to quantify the scale of the problem associated with unreliable expert evidence in criminal
proceedings for a number of reasons. First, there is no central collection of data on the number of
cases in which expert evidence is tendered or admitted, or on the number of cases in which an
expert opinion has wrongly been allowed to go before a criminal court. Secondly, and more to the
point, it is impossible to ascertain the number of cases where a wrongful conviction has occurred
because of unreliable expert evidence. This is in part due to the adversarial nature of our legal
system and in part due to the secrecy which surrounds the deliberations of juries in criminal trials.
In short, there is no way of knowing whether or not, or to what extent, a jury has relied on expert
evidence to reach its verdict; so, in a case where unreliable expert evidence was placed before the
jury, it is not possible to know whether the same verdict would have been reached if that evidence
had not been admitted. The available, relevant information is presented below.

Structure of the industry


The UK Register of Expert Witnesses lists over 2,500 experts. Many experts provide evidence in
criminal courts in addition to their other work. A recent bi-annual survey of experts suggests that
about 85% of experts work on their own and the remaining 15% work as part of a group. Even the
larger forensic science companies rarely employ more than about 50 people.

Key stakeholders
Key stakeholders in the present context include: the judiciary, magistrates, HM Court Service, the
Crown Prosecution Service (and other prosecuting bodies), the police, expert evidence
organisations and the professional bodies representing solicitors and barristers.

Forensic Science Service (FSS)


The Forensic Science Service accounts for most of the forensic science market in England and
Wales. It has been estimated that for the year 2009/10 the FSS was involved in approximately
84,500 cases and 1,300 court attendances. In 2005 the FSS was transformed from a Government
agency to a company wholly owned by the Government.13

Police
Individual forces and providers, previously using bilateral contracts, are now encouraged to sign up
to the National Forensic Framework Agreement, managed and supported by the National Policing
Improvement Agency (NPIA). Police forces contract for packages of work by experts. Evidence
relating to preparation for a trial is usually paid for by the police, whereas evidence which relates to
the trial process is usually paid for by the prosecution. In addition to procuring expert evidence, the
police employ their own fingerprint experts.

Crown Prosecution Service (CPS)


The CPS has procurement arrangements with LGC Forensics, FSS and Document Evidence Ltd.
These arrangements stipulate fees for expert witnesses but occasionally the CPS negotiates a
price in advance with an expert. Between April 2009 and 2010, of 110,000 Crown Court finalised
defendant cases14 and 873,000 magistrates’ courts finalised defendant cases,15 there were 5,420

13
The Government has, however, recently announced that the FSS is to be wound up by 2012.
14
Finalised defendant case figures refer to prosecutions completed within a specified time period. They
include cases proceeding to trial or guilty plea in the Crown Court, together with cases discontinued or
dropped after the defendant has been committed or sent for trial.

172
invoices to the CPS for expert witnesses. The total cost of expert witnesses and forensic services
was £6,296,305. The average invoice was for £1,176.66. The invoices do not give a breakdown of
time spent between report writing and court attendance. Nor do they give the hours spent working.
It has been estimated that, on average, 1.7 prosecution expert witnesses give evidence in a trial
involving expert witnesses. In half of such trials, however, only one expert witness is used by the
prosecution. The CPS’s approved rates for expert witnesses appear in Table 1, below. The CPS
also occasionally commissions experts in specific areas. Fees are negotiated independently. In
addition, the CPS has negotiated rates with its three main providers.

Table 1: CPS fees for expert witnesses (September 2008 rates)


Category of expert and work Minimum Maximum

Consultant medical practitioner, psychiatrist, pathologist

Preparation (per hour) £70 £100


Attendance at court (full day) £346 £500
Fire expert (assessor), explosives expert
Preparation (per hour) £50 £75
Attendance at court (full day) £255 £365

Forensic Scientist, surveyor, accountant, medical


practitioner, meteorologist, architect, engineer, document
examiner, veterinary surgeon

Preparation (per hour) £47 £100


Attendance at court (full day) £226 £490
Fingerprint expert
Preparation (per hour) £32 £52
Attendance at court (full day) £153 £256

Source: Crown Prosecution Service, Expert Witnesses – Scale of Guidance (September 2008)

Central Funds
The Legal Services Commission (LSC) and the courts’ central funds share responsibility for paying
for defence experts. We do address funding in this project, but a brief discussion may provide
useful background information.
Where an expert is used in a way which assists the court, such as in the preparation of a
psychiatric report, the expert’s fee will usually be paid by the courts’ central funds. Payment to a
witness attending to give expert evidence (and for associated preparatory work) is made from
central funds via the courts. Part V of the Costs In Criminal Cases (General) Regulations 1986
provides that expenses properly incurred by an expert witness attending court to give evidence in
criminal proceedings will be allowed out of central funds (unless the court directs that the expenses
are not to be paid out of central funds). Paragraph 5.49 of the LSC's Standard Crime Contract
Specification stipulates that payment to a witness attending court to give evidence in criminal
proceedings cannot be claimed from the LSC unless there is a direction from the court that the
witness expenses may not be claimed from central funds and they are not recoverable from any
other source.

15
This figure comprises all defendants whose case was completed in a magistrates' court during the period,
including cases tried, guilty pleas, cases discontinued, and cases which could not proceed. Cases
committed or sent for trial in the Crown Court are not included.

173
The scale of payments for expert witnesses is set out in the Guide to Allowances under Part V of
the Costs in Criminal Cases (General) Regulations 1986. This scale follows Table 1, above. We
have been informed that, since this rates have not been increased since 2003, in practice experts
are sometimes paid at a higher rate.
We do not have any more information on the average rates paid out of central funds, or on the
number of experts paid, so we have used the information from the LSC as a proxy for all expert
witnesses called by the defence.

Legal Services Commission (LSC)


As explained above, the LSC does not pay all the fees for experts appearing on behalf of the
defendant. If during preparation a defendant applies to the LSC for prior authority to use an expert
and the authority is granted, then the LSC will cover those fees. This is often the case.
Information on expert witnesses is available from HM Court Service’s witness monitoring surveys
(information on all witnesses involved in a four-week period for a given year). In one of its surveys,
witnesses may be counted more than once if they are requested to attend on more than one day
and, as there are no unique witness identifiers in the dataset, there is no way of knowing how
many times one witness has been counted. Estimates based on the survey data for the number of
expert witnesses appearing in 2009 are presented below, in Table 2. For context, in 2009
approximately 94,600 cases committed and sent for trial were closed in the Crown Court.

Table 2: estimated number of expert witnesses appearing in court in 2009


Witnesses
Total number Witnesses Witnesses
who attended
of witnesses expected but who gave
but did not
expected did not attend evidence
give evidence
Crown Court
Defence 1,413 0 299 1,115
0% 21% 79%
Prosecution 7,266 319 1,652 5,295
4% 23% 73%
Magistrates Court
Defence 1,322 119 559 644
9% 42% 49%
Prosecution 3,526 441 1,593 1,492
13% 45% 42%
Total 13,527 879 4103 8546
Source: HMCS Witness Monitoring Survey, 2009
We have assumed that when experts were expected at court and did not attend, or did attend but
did not give evidence, it was because their evidence was not challenged at trial or because there
was a late guilty plea. In calculating the impact of our proposals we use the numbers of witnesses
who attended and gave evidence because this evidence is more likely to be controversial and,
perhaps, unreliable.
Note that there is a difference between the number of witnesses who gave evidence for the
prosecution and the number of invoices received by the CPS. This is probably due to the fact that
the police pay for certain evidence used by the prosecution.
In cases where expert witnesses were used, an average of 1.2 expert witnesses gave evidence for
the defence. Applying that to the estimated attendance of 1,759 defence expert witnesses, it is
further estimated that expert witnesses for the defence appeared in 1,466 cases in 2009. As with
prosecution witnesses, in half of all trial cases involving expert witnesses for the defence only one

174
expert witness was used.
The Ministry of Justice (MoJ) conducted a file review in 2010 which looked at a sample of 4,566
expert witness invoices collected by one regional office between 26 April and 23 July. From this
review it is estimated that the average hourly rate the LSC pays an expert is £96.41 and the
average total cost for an expert in a criminal case is £1,155.34.
The Bond Solon and Legal Services Commission Expert Witness Survey reports the average costs
for experts, for writing reports and attending court. Experts are also divided between legal aid and
non-legal aid cases, although the average cost is the same.

Table 3: average cost of experts’ time for court appearances and report writing (2009/10 prices)
Writing reports Attending court Average
Legal Aid £142 £168 £156
Non-Legal Aid £155 £157 £156
Average £148 £162 £156
Source: Bond Solon and Legal Services Commission Expert Witness Survey16

Wrongful convictions and wrongful acquittals


A wrongful conviction can have a significant impact on the lives of the convicted person and his or
her family, as well as on the victim and the victim’s family. These impacts can relate to matters
such as liberty, an individual’s health and mental wellbeing, family life and financial wellbeing. If no
crime has been committed, but an individual has been wrongfully convicted, there is an additional
cost to the criminal justice system and the prison service. The state may have to pay compensation
where an individual has been wrongfully convicted. Where an individual has been wrongly
convicted for a crime committed by another individual, or there has been a wrongful acquittal, it
may be the case that the perpetrator is free to offend again, exposing the public to increased risk
and the financial consequences of further crime. Equally, a wrongful acquittal can have a
significant impact on the victim. It may also potentially increase the risk of reoffending. Importantly,
wrongful convictions and acquittals can also have an adverse impact on society and society’s
perception of the efficacy of the criminal justice system.
It is difficult to know how many wrongful convictions and wrongful acquittals there have been as a
result of unreliable expert evidence going before juries. There were at least two successful appeals
in 2002, one in 2003, two in 2004, two in 2005, one in 2007, two in 2008 and one in 2010, so in
recent years there have certainly been wrongful convictions caused by, or at least involving,
unreliable prosecution expert evidence. Consider the following two examples:
1. In Clark (Sally),17 part of the prosecution case was opinion evidence provided by an expert
paediatrician. That expert, who was not a statistician, had formulated his opinion on the
assumption that there were no genetic or environmental factors affecting the likelihood of
cot death. He gave evidence that there was only a one in 73 million chance that two natural
cot deaths would occur in the same family. The Court of Appeal took the view that the
figure grossly misrepresented the chance of two sudden deaths within a family from natural
causes.
2. Until the judgment in Harris and others,18 the prosecution had been allowed to rely on a
hypothesis that a non-accidental head injury to a young child could confidently be inferred
from nothing more than the presence of a particular triad of intra-cranial injuries. The

16
Bond Solon and Legal Services Commission Expert Witness Survey 2008,
www.legalservices.gov.uk/docs/about_us_main/2008surveyanalysis.pdf (last visited 9 February 2011), p 3.
See also, JS Publications, Expert Witnesses Survey (2007), www.jspubs.com/Surveys/feesurveys.cfm
(last visited 1 February 2011).
17
[2003] EWCA Crim 1020, [2003] 2 FCR 447.
18
[2005] EWCA Crim 1980, [2006] 1 Cr App R 5.

175
prosecution had in effect been able to rely on nothing more than expert opinion evidence
based on the triad to secure convictions for very serious offences, including murder.19 This
was the case even though the diagnosis of a violent assault was predicated on empirical
research which has been criticised as comprising only a small, poor-quality database.20
Despite the limitations in assessing the scope and scale of the problem, this is an area of law that
has been highlighted as being in need of urgent reform and our recommendations are aimed at
ensuring, as far as possible, that only reliable expert evidence will be put before a jury. Our
recommendations would minimise the number of wrongful convictions and wrongful acquittals and
maintain or enhance public confidence in the criminal justice system.

Consultation
Our Consultation Paper (CP), The Admissibility of Expert Evidence in Criminal Proceedings in
England and Wales, was published in April 2009 and the consultation process took place between
April and July. During that period, individuals could also post their views on an online forum. In our
CP we asked for comments on whether the benefits of our provisional proposals, if implemented,
would outweigh the financial costs in the medium to long term and whether or not the potential
benefits would outweigh the potential costs when compared with the alternative proposals set out
the CP (including the cost of doing nothing).
Responses to the consultation were received from a wide range of consultees including judges,
academics, experts, legal practitioners, expert organisations and prosecuting bodies. There was
broad support for our core proposal to codify the uncontroversial common law requirements and to
introduce a more robust reliability-based admissibility test with guidelines to help the judiciary apply
it (Option 1). The broad consensus was that the potential benefits of our proposals would outweigh
the potential costs.
There was also considerable support for the proposal that a judge should be able to call upon an
independent expert in cases involving exceptionally complex scientific evidence (now, in a revised
form, 1B), although a number of consultees expressed some concern about practicalities as we
explain below. In the light of these concerns, further advice was sought from four experienced
individuals as to how this proposal might work in practice (two barristers, a solicitor and a judge).
Three of these consultees provided very positive responses.
There was little, if any, support for the other options set out in the CP (exclusionary discretion
without guidance, exclusionary discretion with guidance and consensus amongst experts
(deference)). Accordingly, those options have not been discussed in our final report. Nor are they
addressed any further in this impact assessment.
In the final stages of the project, we conducted further consultations with experts, academics,
judges and practitioners to gauge opinion on our draft recommendations and on the reliability test
set out in our draft Criminal Evidence (Experts) Bill. These further consultations also elicited broad
support for what we proposed to recommend.

Description of options considered


Two options for reform, in addition to the “do nothing” case, have been considered.21

Option 0: do nothing
This is the “base case” against which the other options are compared; it demonstrates the costs of

19
See Editorial, British Medical Journal 29 July 2010 (issue 2771): “For 40 years, mainstream medical
experts who give evidence in court have largely agreed that shaken baby syndrome can be unambiguously
diagnosed by a triad of symptoms at post-mortem … . Murder convictions are often secured on the basis
of these alone, even in the absence of other signs of abuse … .”
20
See M Donohoe, “Evidence-based Medicine and Shaken Baby Syndrome” (2003) 24 American Journal of
Forensic Medicine and Pathology 239, 241. See also D Tuerkheimer, “The Next Innocence Project:
Shaken Baby Syndrome and the Criminal Courts” (2009) 87 Washington University Law Review 1, 12 to
14 and 17 to 18.
21
The options in this impact assessment are numbered differently from those originally described in the CP.

176
non-intervention. It would bring no change to the current common law position or the current
disclosure requirements and procedural provisions.

Common law admissibility test


At common law, four requirements need to be satisfied before expert evidence can be admitted in
criminal proceedings:
1. Assistance. Expert evidence is admissible only if it would provide the court with information
which is likely to be outside a judge or jury’s knowledge and experience, such that it would
give the court the help it needs in forming its conclusions. This issue is determined by the
judge or magistrates at or in advance of the trial.
2. Expertise. An individual can give evidence as an expert only if he or she is sufficiently
qualified to do so, on account of knowledge, experience or training in the relevant field. A
witness wishing to give expert evidence must give details in his or her report of the
“qualifications, relevant experience and accreditation” relied on to satisfy this requirement.22
The threshold for demonstrating expertise is quite low, however,23 although logically there
must (we suggest) be a minimum requirement of proof of expertise on the balance of
probabilities. It would seem to be the case in practice that, so long as the information set out
in the report suggests that the witness has relevant expertise on the back of formal
qualifications or relevant experience, he or she will ordinarily be regarded as an expert and
allowed to give expert evidence in a criminal trial.
3. Impartiality. There is authority for the view that, at common law, an expert can provide expert
evidence only if he or she is able to provide impartial, objective evidence on the matters
within his or her field of expertise.24 There is also a common law duty of impartiality which
has been repeated in rule 33.2 of the Criminal Procedure Rules 2010. The mere appearance
of bias is an insufficient reason for ruling that an expert’s evidence is inadmissible in criminal
proceedings; and there is little evidence in the case law to suggest that experts are often
prevented from giving evidence on the ground that they are biased.
4. Reliability. The expert’s opinion evidence must pass a threshold of acceptable reliability,
certainly for evidence of a scientific nature,25 but this requirement of threshold reliability
merely requires that the field of expertise is “sufficiently well-established to pass the ordinary
tests of relevance and reliability”.26
The issue of evidentiary reliability in criminal proceedings prior to a trial on indictment may be
addressed at a plea and case management hearing (PCMH), a preparatory hearing (for complex,
serious or lengthy cases) or at some other pre-trial hearing following disclosure of expert reports. If
the trial has already started, a “voir dire” (trial-within-the-trial) will be held in the absence of the jury.
It is possible for magistrates' courts to consider the admissibility of expert evidence during the
summary trial itself or at a pre-trial hearing.

Procedure
Part 33 of the Criminal Procedure Rules 2010 sets out requirements on the content of any expert
report a party intends to rely on as evidence and also the parties’ obligation to disclose and serve
their reports in advance of the trial. Part 33.6 permits the trial judge to direct the parties’ experts to
meet and discuss the “expert issues” and part 33.7 permits the court to direct that co-defendants
call a joint expert.27

22
Criminal Procedure Rules 2010, r 33.3(1)(a).
23
R (Doughty) v Ely Magistrates’ Court [2008] EWHC 522 (Admin) at [24].
24
Field v Leeds City Council [2000] 1 EGLR 54; Toth v Jarman [2006] EWCA Civ 1028, [2006] 4 All ER 1276.
25
Reed [2009] EWCA Crim 2698, [2010] 1 Cr App R 23; Weller [2010] EWCA Crim 1085; Henderson and
others [2010] EWCA Crim 1269, [2010] 2 Cr App R 24.
26
Dallagher [2002] EWCA Crim 1903, [2003] 1 Cr App R 12; Luttrell [2004] EWCA Crim 1344, [2004] 2 Cr
App R 31.
27
We refer to a meeting of experts under part 33.6 as a pre-hearing discussion.

177
Court-appointed experts
There is a common law power for judges in criminal cases to call a witness of fact during a trial, if
this is in the interests of justice.28 The power is used only sparingly, but it is probably flexible
enough to allow a judge to call an expert witness to assist him or her in the determination of
evidentiary reliability as a matter bearing on admissibility (although we are unaware of any case
where this has happened). This common law power provides no mechanism for ensuring that a
court-appointed expert is fit for the role, or for ensuring transparency or for safeguarding the
parties’ interests.

Option 1: judicial assessment of evidentiary reliability


1: Statutory admissibility test
Our central recommendation is that there should be a robust reliability test in primary legislation
(an Act of Parliament) to replace the current common law “relevance and reliability test”. Our new
test would provide that an expert’s opinion evidence is admissible only if sufficiently reliable to be
admitted, meaning that the opinion must be soundly based and its strength must be warranted
having regard to the grounds on which it is based. Our draft Bill sets out a number of guiding
examples and factors designed to help the judge determine whether or not an expert’s opinion
evidence is sufficiently reliable to be admitted. However, the trial judge would ordinarily apply the
new test and investigate reliability only if it appeared to him or her that the expert’s opinion
evidence might not be sufficiently reliable to be admitted. If the reliability test is applied, the judge
would ordinarily call a pre-trial hearing to resolve the matter.
The statutory test would make it clear to the expert, and the party wishing to call the expert, what is
required for the expert’s evidence to be admitted, which would encourage better practice in the
preparation of expert evidence, dissuading any existing practice whereby unreliable evidence is
tendered for admission in the expectation that it will be admitted (under the laissez-faire common
law test).
We also recommend that the uncontroversial common law admissibility requirements should be
codified in the same Act of Parliament. The assistance requirement would be codified without
change; the expertise (“competence”) requirement would be codified with an explicit obligation on
the party calling the ‘expert’ to prove his or her expertise on the balance of probabilities; and the
impartiality requirement would be codified with reference to the duty in rule 33.2 of the Criminal
Procedure Rules 2010, but with an “interests of justice” inclusionary discretion. It would then be
clear to all concerned that expert witnesses must demonstrate their expertise (on the balance of
probabilities) and that their evidence would probably be excluded if there is a real risk that they are
not impartial.
Since we recommend no significant changes to the law on assistance, expertise and impartiality,
we anticipate that codification will be largely cost-neutral.
Under our recommendations, all four requirements in the statutory test would need to be satisfied,
otherwise the evidence would be inadmissible. If the judge is satisfied that the evidence is
sufficiently reliable to be admitted, the trial would commence and the expert evidence would be
adduced in the usual way. The expert witness would be challenged in cross-examination and no
doubt contradicted by the adduction of contrary expert evidence.
As well as discouraging the proffering of expert opinion evidence of doubtful reliability, our
recommendations would deter dubious challenges to admissibility. It would ordinarily be for a party
challenging the admissibility of an expert’s opinion evidence to provide a coherent argument that
the evidence is insufficiently reliable to be admitted. But where a challenge is well-founded, and it
appears to the court that an expert’s opinion evidence might not be sufficiently reliable to be
admitted, the party seeking to rely on the evidence would have to show that it is sufficiently reliable
to be admitted. If the party proffering the evidence cannot show that it is sufficiently reliable to be
admitted it will be inadmissible. Unreliable expert evidence would be kept from the jury, reducing
the possibility of an erroneous verdict.

28
Roberts (1984) 80 Cr App R 89; R v Haringey Justices ex parte DPP [1996] QB 351.

178
1A: changes to the procedural regime
The current disclosure requirements for expert evidence are set out in Part 33 of the Criminal
Procedure Rules 2010.
Our recommendations also include new pre-trial disclosure requirements designed to ensure that:
1. the admissibility requirements in our draft Bill would work in practice (by allowing matters
bearing on the admissibility of proffered expert evidence to be properly aired and investigated
before the trial);
2. the parties would be properly equipped to challenge admissible expert evidence placed
before the jury;
3. the parties would not proffer unreliable expert opinion evidence or call unreliable witnesses to
give expert evidence;
4. there would be effective pre-trial case management.
We recommend enhanced pre-trial disclosure requirements (for the defence as well as the
prosecution) and rules which would allow the judge to chair a meeting of the experts to reduce the
issues and the potential for conflict during the trial. An expert witness would also have to
summarise in an appendix to his or her report the reasons why his or her evidence is admissible,
with reference to the new statutory requirements.

1B: court-appointed expert witness


When considering the reliability of very complex expert opinion evidence under our reliability test, a
judge might in exceptional cases require the assistance of an additional expert witness. There
exists a common law power for a judge in criminal proceedings to call a witness during a trial and
our recommendation builds on this by providing for the selection and appointment of an expert
witness to provide evidence in a hearing to address the evidentiary reliability of a party’s expert
opinion evidence. We recommend an independent, non-governmental selection panel which would
liaise with relevant professional bodies to compile a list of possible experts from which the trial
judge would make his or her selection. The appointments system would be transparent and would
ensure that any court-appointed expert is properly screened for expertise and impartiality.
However, the power would be available only for trials on indictment (in the Crown Court) and only if
the interests of justice warrant an appointment by virtue of the complexity of the opinion evidence,
the likely importance of the evidence in the trial and any other relevant circumstances.

Option 2: judicial assessment of evidentiary reliability (including 1A but excluding 1B)


Option 2 includes all the proposals in Options 1 and 1A, but excludes our proposals for court-
appointed experts as outlined in 1B.
Although we are recommending Option 1 (including 1A and 1B), we recognise that concerns have
been expressed about the practicalities of selection and the possible costs associated with setting
up a panel with appropriate administrative support. 1B is therefore a stand-alone option; and the
relevant clause in our draft Bill (clause 9) is also free-standing and severable from the rest of our
reform package. Our draft Bill in all other respects and our recommended changes to Part 33 of the
Criminal Procedure Rules could therefore be taken forward immediately, but clause 9 could be
brought into force at some later date, if the Government determines that there is a real need for
this measure for cases involving very complex expert opinion evidence.

Summary
Under our preferred option (Option 1), the current procedural framework and rights of appeal would
continue to operate and our proposals, if taken forward, would fit into this existing framework.29 A
recent case illustrates current practice, the judiciary’s support for the reforms we are
recommending and why government intervention is necessary. In Reed,30 the Court of Appeal
held, in line with the central proposal in our consultation paper, that expert opinion evidence “of a

29
See Appendix B of CP 190.
30
[2009] EWCA Crim 2689, [2010] 1 Cr App R 23.

179
scientific nature” should not be admitted if “the scientific basis on which it is advanced is
insufficiently reliable for it to be put before the jury”.31
However, the court’s judgment also highlights the ongoing rationale for statutory intervention. The
court reaffirmed the traditional laissez-faire approach to admissibility which has developed at
common law,32 and provided no guidance which would help trial judges determine whether or not
expert opinion evidence “of a scientific nature” has a sufficiently reliable scientific basis (or indeed
whether any other expert opinion evidence is sufficiently reliable to be admitted).

2. Cost benefit analysis


This impact assessment identifies both monetised and non-monetised impacts of intervention, with
the aim of understanding the overall impact on society and the wider environment. The costs and
benefits of each option are measured against the “do nothing” option. Impact assessments place a
strong emphasis on valuing the costs and benefits in monetary terms (including estimating the
value of goods and services that are not traded). However there are important aspects that cannot
sensibly be monetised. These might include impacts on equity and fairness, either positive or
negative, or enhanced (or diminished) public confidence.
The impact assessment process requires that we make an assessment of the quantifiable costs
and benefits even when there is insufficient material on which to base those calculations. Where
possible we have spoken to practitioners to inform our view of the number of cases likely to be
affected by aspects of the policy and have used this as the basis for our calculations. Where it has
not been possible to obtain a rough indication of numbers in this way we have had to make a
realistic estimate. In such cases we have taken a conservative approach and have tended to use
figures that we considered likely to under-estimate benefits and over-estimate costs.
In the absence of sufficient data we have used ranges of estimates in our calculations. Some of the
assumptions apply in both the cost and benefit calculations. Since the net present value (NPV) has
been calculated by subtracting the low costs from the low benefits, and the high costs from the
high benefits, it is important that the same assumptions underlying the calculation of the low costs
apply to the calculations of the low benefits, etc. This has resulted in some values in the lower
benefits column being greater than those in the high benefits column. In addition our high NPVs
are negative, as they are calculated using the high estimates of costs, which are not offset by high
benefits. Our low NPVs are positive.
When calculating the NPVs for the impact assessment we have used a time frame of ten years,
with the current year (2011) being year 0.33 With the exception of the increased cost of appeals we
have assumed that the transitional costs and benefits occur in year 0, and ongoing costs and
benefits accrue in years 1 to 10. In the case of the cost of appeals – which we have identified as a
transition cost lasting over a five year period – we have discounted the values accordingly. A
discount rate of 3.5% has been used in all cases in accordance with HM Treasury guidance.
Unless stated, all figures are in 2009/10 prices, and have been uprated using the GDP deflator.
A summary analysis and evidence sheet is available for our preferred Option 1 (with a separate
cost/benefit assessment of 1A and 1B) and also for Option 2 (Option 1 including 1A, but excluding
1B).

Option 0: do nothing
Option 0 is the base case against which our other options are measured. Because the do-nothing
option is compared against itself, its costs and benefits are of course zero, as is its NPV. While
there would not be any additional costs, current costs incurred would continue to be incurred.
These are discussed below to provide context for the assessment of the other options.

31
[2009] EWCA Crim 2689, [2010] 1 Cr App R 23 at [111].
32
The court held at [111] that there is “no enhanced test of admissibility” for expert evidence of a scientific
nature.
33
The net present value is the discounted stream of benefits less the discounted stream of costs. The
present value of an annual cost is the discounted stream of that cost.

180
Costs
There would continue to be no accepted means by which trial judges are able to assess the
evidentiary reliability of expert evidence. This means expert evidence would not be adequately
assessed before it is presented to the jury in many criminal trials.
The risk would be that unreliable expert evidence would continue to contribute to wrongful
convictions and wrongful acquittals. We cannot estimate how many miscarriages of justice are
caused by unreliable expert evidence, but we can say that a wrongful conviction involves
significant costs to individuals, government and society in general, and has been estimated (below)
to cost at least £123,548. The costs, which have not all been monetised, include:
x loss of liberty for the individual (most obviously if the individual receives a custodial
sentence);
x loss in earnings and potential earnings (with a concomitant impact on the welfare system);
x loss of an individual’s home and possessions (with a concomitant impact on the welfare
system);
x adverse impact on the private life of the individual as well as on that of his or her family;
x adverse health impact and wellbeing of the individual (with a possible impact on the NHS);
and
x stigma and possibly ostracism.
Wrongful convictions adversely affect the criminal justice system and can seriously erode public
confidence in the system. They use up time and resources, and yet the Court of Appeal is already
overstretched. In the year 2009/10, 3,346 applications before the Court of Appeal (Criminal
Division) were outstanding with an average waiting time of just over 10.1 months for appeals
against conviction and 5 months for appeals on sentencing.34
If there is a wrongful conviction but an offence was committed, the real perpetrator is not punished
and society does not benefit from the incapacitation of the offender. The same problems arise in
cases where there has been a wrongful acquittal.
Benefits
The only benefit is the avoidance of the cost of reform.

Option 1: judicial assessment of evidentiary reliability (including 1A and 1B)


1: Statutory admissibility test
Costs
Transitional costs
1. Training
Training is required for two reasons: to inform judges about the new law and procedure; and to
guide judges in the practical application of the reliability test. The training and education of
practitioners and the judiciary was recommended by the House of Commons’ Science and
Technology Committee, which also recommended that the judiciary receive an annual update on
scientific developments relevant to their work.35
Responsibility for training the judiciary rests with the Lord Chief Justice and is exercised through
the Judicial Studies Board (JSB), an independent body chaired by Lady Justice Hallett. The JSB
recognises the benefits which would come from ensuring that judges have the relevant training to
be able to apply the reliability test we recommend.

34
HMCS, The Court of Appeal Criminal Division Review of the Legal Year 2009 to 2010 (2010),
http://www.judiciary.gov.uk/media/media-releases/2010/jco-news-release-coa-crim-div-review-legal-yr (last
visited 1 March 2011), p 1.
35
House of Commons’ Science and Technology Committee, Forensic Science on Trial (2004–2005) HC 96-
1, p 78.

181
Under our recommendations, training would need to be provided to District Judges (Magistrates’
Courts)36, Crown Court judges and appeal court judges (for criminal appeals). In England and
Wales in 2010, there were 143 District Judges (Magistrates’ Courts), 1,233 recorders, 680 Circuit
Judges, 72 Queen’s Bench Judges and 37 Lord Justices of Appeal.37
The JSB has indicated that one possible way in which Circuit Judges and recorders could be
trained is at the annual Circuit Criminal Seminars, which are provided for all judges of the Crown
Court. This would incur no additional costs, unless an outside speaker were to be invited to attend
(generating a cost of about £500 per session for 19 sessions a year). Similarly, training could be
provided to District Judges (Magistrates’ Courts) during their annual training events, and if this
were done, the only additional cost would be for the hire of any speakers, as outlined above.
Training High Court judges would be dependent on the direction of the senior judiciary. If they
perceived a need for training, and High Court judges were able to be released for one day, the cost
for 74 Queen’s Bench Division judges would be around £3,700, based on a price of £50 per
individual, excluding any additional speaker costs (if required). Taking speaker costs into account,
we have estimated the cost to be £5,000. There is potential scope for some High Court judges to
be included in the Circuit Criminal Seminars (but only if judicial release time is agreed with the
senior judiciary).
The costs associated with training legal professionals would be borne by the practitioners (or their
employers) who choose to undertake training to assist their work in this regard. It is unlikely that
this training would add significant cost or time to the training already required by the Solicitors
Regulation Authority and the Bar Standards Board in order for barristers and solicitors to maintain
their practising certificates (although one solicitor told us that courses could be expensive if they
involved the attendance of experts). In any case, the CPS informed us that training would be cost-
beneficial for prosecutors in the long term. We would expect defence practitioners to be of the
same view.
It should be noted that practitioners and the judiciary should receive training on the determination
of evidentiary reliability in any event, certainly in relation to evidence “of a scientific nature”, given
the Court of Appeal’s recent judgment in Reed.38
We should stress, however, that no guarantees relating to judicial training on evidentiary reliability
can be provided. Any final decision must depend on competing priorities and available resources.
The low and best estimates for the total training cost are £5,000. However, to take account of the
low risk of an outside speaker being invited to attend Circuit Criminal Seminars, we also include a
high estimate of £9,500 (£500 x 19 sessions).

2. Increase in appeals
During the first five years there could be a temporary increase in the number of appeals as
practitioners and judges come to terms with the new reliability test. This additional cost would be
mitigated, however, by a concomitant improvement in standards. That is to say, in the medium to
long term the quality of the expert evidence tendered for admission in criminal trials should
improve.
Although this project addresses criminal trials heard in magistrates’ courts, most serious crimes are
heard in the Crown Court, and appeals against conviction in the Crown Court are heard by the
Court of Appeal (Criminal Division). In this impact assessment we have assumed that any
additional appeals will be heard in the Court of Appeal. There is no current data on the average
cost of an appeal to the Court of Appeal. We do, however, have the following data:
x The estimated cost of a day’s sitting for the Court of Appeal (Criminal Division) in 2009/10 is
£16,635.
36
Given that the number of challenges in the magistrates’ courts is likely to be low we consider that it would
be disproportionate to train lay magistrates in addition to district judges.
37
The Judiciary of England and Wales, Judicial Statistics 2010 (2010),
http://www.judiciary.gov.uk/publications-and-reports/statistics/judges/judicial-statistics (last visited
1 February 2011).
38
[2009] EWCA Crim 2689, [2010] 1 Cr App R 23.

182
x A simple model of the average cost to the criminal justice system of an appeal against a
conviction or sentence imposed by the Crown Court is £20,821 (in 2009/10 prices).39
x If leave is refused on the papers, the court will not sit and the cost will be far lower. However,
an application for leave to appeal potentially increases the workload for those who handle the
leave applications – the judges and staff of the Crown Court – and for those who handle the
appeals against the refusal of leave, namely the judges and staff of the Court of Appeal
(Criminal Division), even if leave is refused. We have estimated that an application for leave
to appeal costs £3,000.
The figure of £20,821 for the cost of an appeal includes legal aid costs and costs to the CPS. It
does not include any private costs to the defendant and so the figure might be an underestimate.
To account for this we have used the estimate of £25,000 for the cost of an appeal in the Court of
Appeal.
We have estimated between 0 and 10 additional applications for leave to appeal per year, and that
an appeal will be heard in 0 to 5 of these cases. The best estimate is 5 applications for leave to
appeal and 2 for an appeal hearing. Relevant calculations are provided in Table 4 below. To
calculate the present value we have assumed that the additional appeals would start in year 1 and
end in year 5.

Table 4: annual total cost of additional appeals and present value over 5 years
Low Best High
Applications for leave to appeal 0 5 10
Cost of applications £0 £15,000 £30,000
Number of appeals heard 0 2 5
Cost of appeals £0 £50,000 £125,000
Annual cost £0 £65,000 £155,000
Present value £0 £293,478 £699,833

The best estimate of the cost of additional appeals over the five-year period (annual total cost and
present value) is £65,000 per year and £293,478 in total.

3. Legislative costs
Excluded from the cost estimates are the additional legal costs associated with the creation of a
new Act of Parliament and rules of criminal procedure. By introducing statutory reform there would
be the initial cost of enacting primary legislation. This cost should be quite low, however, because
a draft Bill accompanies our recommendations.
The Criminal Procedure Rules Committee has started a revision of all of the Criminal Procedure
Rules, which should be finished in 2015. As they are reviewing the rules in any event, no additional
work would be required to incorporate our recommendations, so there would be no additional cost
in this respect.

Ongoing costs
1. Court costs
Where there is a legitimate doubt as to the reliability of an expert’s opinion evidence, the trial judge
would have the power to convene a pre-trial hearing to assess the issue. Our recommendations
would put experts on notice that they might be required to provide sufficient material to

39
R Harries, Cost of Criminal Justice (Home Office Research, Development and Statistics Directorate
Research Findings No 103, 1999). This research has excluded some costs, such as compensation.

183
demonstrate the reliability of their opinion evidence. Because of this, there would be fewer
instances of unreliable expert evidence being tendered for admission in criminal proceedings.
Nevertheless, if a judge were to convene a pre-trial hearing to assess the reliability of expert
opinion evidence, the cost of the trial (in terms of court time, judicial time and the time of legal
representatives) could increase.
The Bond Solon and Legal Services Commission Expert Witness Survey (above) estimated that for
2009 there were a total of 6,410 experts who gave evidence in the Crown Court and 2,136 in
magistrates’ courts.
In order to calculate the number of additional pre-trial hearings (Table 5) we have used the
following assumptions:
x 5% of all expert opinion evidence currently tendered for admission in the Crown Court, and
2% in magistrates’ courts, might not pass the new test. The difference comes from the
different types of expert evidence proffered for admission in the Crown Court and
magistrates’ courts.
x Of this, between 0% (low estimate), 20% (best estimate) and 40% (high estimate) would still
be tendered if the new reliability test was in place.
x Of the evidence not tendered or found inadmissible, 50% would be replaced by weaker
opinion evidence given by the same expert.
x Of the unreliable evidence tendered for admission, we assume that 100% would be
challenged (by a party or the by the judge applying the test of his or her own motion).
x There would be challenges to reliable evidence (including the replacement weaker opinion
evidence). We have estimated that an additional 0% to 2% (best estimate 1%) of all reliable
expert opinion evidence would be challenged.
x In 65% to 90% (best estimate 80%) of these cases a pre-trial hearing would be necessary. In
the remaining cases the judge would be able to decide the question of admissibility without a
hearing.
x We have made our calculations using the number of witnesses who attended court and gave
evidence. We assume that the evidence of those who were not expected at court, or who did
not attend and give evidence, was likely to have been accepted and was therefore more
likely to be reliable.

Table 5: additional pre-trial hearings


Magistrates’ Courts Crown Court
Low Best High Low Best High
Total expert evidence tendered 2,136 6,410
% of evidence unreliable 2% 5%
Pieces of unreliable evidence 43 321
% of unreliable evidence tendered 0% 20% 40% 0% 20% 40%
Unreliable evidence challenged 0 9 17 0 64 128
% of reliable evidence challenged 0% 1% 2% 0% 1% 2%
Reliable evidence challenged 0 21 42 0 62 125
Total challenges 0 30 59 0 127 253
% of challenges requiring hearings 65% 80% 90% 65% 80% 90%
Total hearings 0 24 53 0 101 228

184
In order to calculate the cost of a pre-trial hearing (Table 6) we have used the following
assumptions:
x On average a pre-trial hearing on reliability would take between half a day and three days
(best estimate one day) in the Crown Court and between one hour and one day (best
estimate half a day) in a magistrates’ court.
x A Crown Court sitting occupies a whole day (4.45 hours on average) and costs £4,454 on
average. A session in a magistrates’ court takes only half a day (2.5 hours) and costs £2,005
on average.40 Consequently, there is one Crown Court sitting per day and two sessions per
day in a magistrates’ court. A full day in a magistrates’ court is estimated to cost £4,010.
x We have assumed that each side will employ an expert in a pre-trial hearing. We have used
the Bond Solon and Legal Services Commission Expert Witness Survey figures throughout
and assumed that the average hourly fee of an expert witness is £156.41 Per side, the expert
evidence fee for half a day is £390 (£156 x 2.5). The estimate for a full day is £780 (£156 x
5), and for three days is £2,340 (£156 x 15).
x We have assumed throughout that for every hour an expert witness presents evidence in
court, they spend one hour preparing. This is also costed at £156 per hour.
The Advocate Graduated Fee Funding Summary lists fees claimable from the LSC. We make
the following assumptions:
x The fees payable for these additional hearings would be identical to those under the hearings
relating to the admissibility of evidence generally.
x The costs to the CPS would be identical to that of the LSC.
x As a best estimate in the Crown Court 50% of cases would be presented by a junior
advocate alone (£143 half day, £263 full day) and 50% would be presented by a junior led by
a Queen’s Counsel (£430 half day, £812 full day). We have assumed that in the magistrates’
courts all cases will be presented by a junior advocate working alone.
x LSC fees in the paragraph above will be reduced in April 2011 and April 2012 by 4.5%.

Table 6: costs of a pre-trial hearing


Magistrates’ Courts Crown Court
Low Best High Low Best High
Time 1 hour 0.5 days 1 day 0.5 days 1 day 3 days
Court costs £802 £2,005 £4,010 £2,227 £4,454 £13,362
Expert fees £312 £780 £1,560 £780 £1,560 £4,680
Expert preparation work £312 £780 £1,560 £780 £1,560 £4,680
LSC & CPS costs 2011 £109 £273 £502 £547 £1,026 £3,078
Total costs 2011 £1,535 £3,838 £7,632 £4,334 £8,600 £25,800
LSC & CPS costs 2012 £104 £261 £480 £523 £980 £2,939
Total costs 2012 £1,530 £3,826 £7,610 £4,310 £8,554 £25,661

40
Assumed by HMCS to be a total of 5 working hours over two sessions per day.
41
Bond Solon and Legal Services Commission Expert Witness Survey 2008,
www.legalservices.gov.uk/docs/about_us_main/2008surveyanalysis.pdf (last visited 9 February 2011), p 3.
See also, JS Publications, Expert Witnesses Survey (2007), www.jspubs.com/Surveys/feesurveys.cfm
(last visited 1 February 2011).

185
The calculations in Table 5 and Table 6 have been combined in Table 7 to provide the total costs
of the change.

Table 7: total costs to criminal justice sector of additional pre-trial hearings


Low Best High
Costs in 2011 £0 £962,122 £6,287,072
Costs in 2012 and beyond £0 £957,155 £6,254,302
Present value of costs £0 £7,965,082 £52,046,221

We have assumed that 100% of unreliable evidence tendered will be challenged. In further
calculations we have assumed that unreliable expert evidence which is challenged will not be
admitted. Reliable expert evidence which is challenged will pass the new test.

2. Prison costs
If our recommendations were to reduce the number of wrongful acquittals, there would be an
increase in the demand for prison spaces. We know that the estimated average annual cost for a
prison space in 2009/10 is £44,703. However we do not know how many wrongful acquittals may
be prevented, and we have no reliable data which we could use to estimate a figure. For this
reason we have not quantified this cost. In any event, it should be borne in mind that this cost
would be offset by the benefit which would come from fewer wrongful convictions.

Cost summary
The costs have financial implications for HM Court Service, the Crown Prosecution Service and
other prosecuting authorities, individual defendants and the LSC. The Criminal Bar Association told
us that they envisaged higher defence costs because of the need for more pre-trial court
attendances by experts, meetings between solicitors and experts and further research. The CBA
also expressed the view that the LSC would have to adjust its funding arrangements so that the
defence would not be prejudiced. Professor Paul Roberts also felt that public funding would need
to be made available to enable the prosecution and defence advocates to provide the court with
the assistance needed to make a properly informed ruling. For their part, the LSC expressed
concern that our proposals could have a significant cost impact on the legal aid budget because of
the possible need for more information to be provided in expert reports and the possible increase
in time spent in court proceedings (and therefore an increase in lawyers’ costs and fees paid to
expert witnesses).

Benefits
Transitional benefits
We do not foresee any transitional benefits of Option 1.

Ongoing benefits
A clearer admissibility test would bring clarity, consistency and uniformity to the admissibility of
expert evidence in criminal trials. Hearings would be conducted in a more structured, efficient and
cost-effective manner. An important benefit of the provisions in our draft Bill is that they would
provide guidance for judges, helping them determine evidentiary reliability for expert opinion
evidence (something they are now required to do at common law for evidence of a scientific
nature).42 At present, judges have no clear test or guidance to help them assess the reliability of
expert opinion evidence, notwithstanding the importance of such evidence and the clear dangers
associated with the admission of unreliable expert opinion evidence. The guidance in our draft Bill
would be particularly useful for cases involving scientific expert evidence.
The prominent benefits of Option 1 are detailed below.

42
Reed [2009] EWCA Crim 2698, [2010] 1 Cr App R 23.

186
1. Reduction in trials as a result of the exclusion of unreliable expert evidence
We estimate that as a result of our recommendations, 5% of expert evidence currently tendered for
admission might be ruled insufficiently reliable to be admissible under the new test and guidance in
the Crown Court, and 98% in magistrates’ courts. Some of this evidence would be ruled
inadmissible by the judge following a pre-trial hearing but, as practitioners become familiar with the
new test, we anticipate that increasingly both the prosecution and the defence would avoid
tendering unreliable expert evidence for admission at all.
In Table 8 below have assumed the following:
x 95% of expert opinion evidence tendered for admission would pass the new test in the
Crown Court, and 98% in magistrates’ courts;
x Between 0% and 40%, best estimate 20%, of the expert opinion evidence that might not
pass the test would still be tendered for admission. It would all be challenged before the main
trial proceeded and found inadmissible.
x Of the evidence not tendered or found inadmissible, 50% would be replaced by weaker
opinion evidence given by the same expert.
x We therefore estimate a net reduction of expert evidence tendered for admission after the
pre-trial stage of 2.5% in the Crown Court and 1% in magistrates’ courts.
Table 8 below is derived from the prosecution and defence statistics in Table 2 (above). In our
calculations, we only include witnesses who attended court and gave evidence. This is because
their evidence was more likely to be controversial and, for that reason, unreliable.

Table 8: annual reductions of expert evidence tendered and admitted after the pre-trial stage
Magistrates'
Crown Court
Courts
Estimated expert evidence tendered now 2136 6410
Evidence which might not pass the test (5%) 43 321
Net reduction in expert evidence 1% 2.5%
Reduction in expert evidence tendered 21 160

In a number of the cases where expert opinion evidence is found not to satisfy the new reliability
test, the party’s case would be fatally undermined. If the party is the prosecution, and its case
depends fundamentally or critically on the expert opinion evidence – such that without it there
would no longer be a realistic prospect of a conviction – no evidence would be offered against the
defendant in accordance with the Code for Crown Prosecutors. If a vital piece of defence expert
evidence were to be considered unreliable under the new test, the defendant might still legitimately
put the prosecution to proof (that is, see if the prosecution is able to prove its case) in which case
the trial would proceed, at least to the end of the prosecution evidence. Alternatively the defendant
might decide to plead guilty given the reduction in sentence available for an early guilty plea.
Because there is a significant cost in instructing an expert in a criminal trial, it is fair to assume that
an expert’s opinion evidence is of vital importance in many of the trials in which such evidence is
used. In estimating how many trials would fail to proceed as a result of the early identification of
unreliable expert evidence, we have worked on the basis that between 40% and 60% (best
estimate 50%) of expert opinion evidence found to be unreliable under the new test would be so
important to the party seeking to adduce it that the trial would not proceed.
We have calculated the savings from a reduction in trials by multiplying the cost per hour in court
(Table 9) by the average trial length. We have used the MoJ’s estimate of 12.94 hours as the best

187
estimate of the average hearing time in the Crown Court when a not guilty plea is entered. This is
likely to be an under-estimate since trials involving expert evidence tend to be more complex than
average. In the event that the defendant chose to plead guilty as a result of expert evidence being
unreliable under the new test these savings in time would be offset slightly by the time to deal with
a guilty plea. We have estimated that the average hearing length in a magistrates’ court is half a
day, or 2.5 hours. Note that the trial lengths do not include the length of a pre-trial hearing, which is
costed separately.
We have assumed that an expert will appear on each side, each appearing for 3 hours in the
Crown Court and 1 hour in a magistrates’ court. We have assumed that for each hour in court an
expert requires one hour’s preparation. We have assumed that the experts are not present for
each other’s presentations.

Table 9: costs per trial in magistrates’ courts and the Crown Court
Magistrates'
Crown Court
courts
Length of trial (hours) 2.5 12.94
Court Costs per hour £802 £1,000
Hours of evidence presentation 2 6
Expert: hour in court £156 £156
Expert: hour of preparation £156 £156
LSC & CPS costs per hour 2011 £109 £246
Costs per trial 2011 £2,902 £17,994
LSC & CPS costs per hour 2012 and beyond £104 £235
Costs per trial 2012 and beyond £2,890 £17,851

The savings from a reduction in trials are calculated in the tables below.

Table 10A: savings from trials avoided in magistrates’ courts


Low Best High
Reduction in expert evidence tendered 21 21 21
% of cases where trial does not proceed 40% 50% 60%
Cases where trial does not proceed 9 11 13
Savings from trials avoided in 2011 £24,796 £30,995 £37,194
Savings from trials avoided in 2012 £24,691 £30,863 £37,036
Present value of savings £205,445 £256,806 £308,167

188
Table 10B: savings from trials avoided in the Crown Court
Low Best High
Reduction in expert evidence tendered 160 160 160
% of cases where trial does not proceed 40% 50% 60%
Cases where trial does not proceed 64 80 96
Savings from trials avoided in 2011 £1,153,445 £1,441,806 £1,730,167
Savings from trials avoided in 2012 £1,144,265 £1,430,331 £1,716,397
Present value of savings £9,525,268 £11,906,585 £14,287,901

The range of annual savings in 2011 is £1,178,240 (low), £1,472,800 (best) and £1,767,361 (high).
The range of the total present value of savings is £9,730,713 (low), £12,163,391 (best) and
£14,596,069 (high).

2. Reduction in trial time


If the trial goes ahead in the absence of an expert opinion which fails the new admissibility test, or
a different (weaker) opinion is admitted instead, the shorter trial would lead to savings. We have
estimated that each expert opinion which is not tendered will save three hours of court time in the
Crown Court and one hour in a magistrates’ court.44 In the case of the replacement weaker expert
evidence no new report would be commissioned. Because the opinion is weaker the presentation
of that opinion, in terms of examination in chief and cross examination, would take half as long as
the original opinion.
The savings per hour in court are set out in Table 11 below. We have assumed that experts would
be paid for their reports and pre-trial attendance, but there would be a saving on expert fees as
they would not attend the trial. We have assumed that the opposing party’s expert would not have
been present when an expert gave evidence at trial, and so there would be no saving in respect of
their fees.

Table 11: savings in per hour in magistrates’ courts and Crown Court
Magistrates'
Crown Court
Courts
Best Best
Court Costs per hour £802 £1,000
Expert: hour in court £156 £156
Expert: hour of preparation £156 £156
LSC & CPS costs per hour 2011 £109 £246
Costs per hour 2011 £1,223 £1,558
LSC & CPS costs per hour 2012 and beyond £104 £235
Costs per hour 2012 and beyond £1,218 £1,547

The estimated annual savings are set out in Tables 12A and 12B below, based on the magistrates’
courts / Crown Court breakdown in Table 2 (above).
44
This figure is conservative and reflects the fact that it can take anything from a few minutes to adduce
undisputed expert opinion evidence to more than a week for complex expert opinion evidence (eg in a
murder trial).

189
Table 12A: savings in trial time in magistrates’ courts
Low Best High
Items tendered 2,136
Net reduction in evidence tendered 1%
Items failing the test 21

Percentage of trials still proceeding 60% 50% 40%


Items not tendered 13 11 9
Hours saved per item 1
Hours saved as less tendered 13 11 9

Number of replacement opinions 21


Hours saved per item 0.5
Hours saved with weaker opinions 11 11 11

Total hours saved 23 21 19


Total annual savings 2011 £28,742 £26,129 £23,516
Total annual savings 2012 £28,626 £26,024 £23,421
Present value of savings £238,183 £216,530 £194,877

Table 12B: savings in trial time in the Crown Court


Low Best High
Items tendered 6,410
Net reduction in evidence tendered 2.5%
Items failing the test 160

Percentage of trials still proceeding 60% 50% 40%


Items not tendered 96 80 64
Hours saved per item 3
Hours saved as less tendered 288 240 192

Number of replacement opinions 160


Hours saved per item 1.5
Hours saved with weaker opinions 240 240 240

Total hours saved 529 481 433


Total annual savings 2011 £823,877 £748,979 £674,081
Total annual savings 2012 £818,025 £743,659 £669,293
Present value of savings £6,808,842 £6,189,857 £5,570,871

190
The low estimate of total annual savings is £852,619, the best estimate is £775,108 and the high
estimate is £697,597. The low estimate of the present value of total savings is £7,047,025, the best
estimate is £6,406,387 and the high estimate is £5,765,748.

3. Saving in experts’ fees from reduction in evidence commissioned


We anticipate that the new admissibility test under Option 1 would encourage a cultural shift over
time such that the parties would seek to rely on expert opinion evidence of questionable reliability
less often. We estimate a net reduction of expert evidence adduced, after some strong evidence is
replaced with weaker opinion, of between 2.5% in the Crown Court and 1% in magistrates’ courts
(compared with current levels). Initially the reduction would come from judicial findings of
unreliability at pre-trial hearings and by parties coming to their own assessment of reliability in the
light of their experts’ reports, but with time this cultural shift would lead to a reduction in requests
for certain types of expert evidence. We anticipate that there would be a reduction in respect of
requests for expert opinion evidence of a type which has previously been held to be unreliable, and
this reduction would bring concomitant savings. We estimate that in year 1 there would be a
reduction of 10%, 20% in year 2, 30% in year 3, 40% in year 4, and 50% in year 5 and beyond.
We have assumed that it typically takes six hours to prepare and adduce an item of expert
evidence in the Crown Court and two hours in a magistrates’ court. By deducting the cost of
preparing and adducing evidence from the average invoiced total typically paid by LSC, we can
estimate the cost of preparing an expert’s report. We have been conservative in calculating
savings, and have deducted six hours, not two, to calculate the savings per report in the
magistrates’ courts.
The average hourly rate that the LSC pays is £96.41, and the average total cost of an expert in a
criminal case is £1,155.34. To be cautious, and consistent with our earlier approach, we will use an
average hourly rate of £156. We have used the average value of the invoice from the LSC to
calculate the cost of a report, because it is possible that the average invoice to the CPS
(£1,176.66) does not reflect the complete cost of the expert witness. This is because the police
may be responsible for paying part of the expert’s fees. The saving from each report no longer
commissioned would be £1,155.34 – (£156 x 6) = £219.34.

Table 13: annual in reductions of expert evidence reports


Total
Estimated expert reports commissioned now (magistrates' courts) 2,136
Evidence which might not pass the test (2%) 43

Estimated expert reports commissioned now (Crown Court) 6,410


Evidence which might not pass the test (5%) 321

Total evidence which might not pass 363


Evidence converted to weaker opinion (50%) 182
Evidence liable for challenge 182
Savings per report £219
Savings year 1 £3,983
Savings year 2 £7,967
Savings year 3 £11,950
Savings year 4 £15,934
Annual Savings year 5 and beyond £19,917
Net present value of savings £128,436

191
4. Reduction in the number of appeals on the basis of unreliable expert opinion evidence
The admission of unreliable evidence at the trial stage may result in costs in the form of appeals.
Screening expert opinion evidence for reliability before it is admitted in criminal proceedings could
result in a decrease in the number of appeals, thereby reducing court costs.
Appeals against conviction in the Crown Court are heard by the Court of Appeal (Criminal
Division). The Court of Appeal is already overstretched. In the year 2009/10, 3,346 applications
before the court were outstanding, with an average waiting time of 10.1 months for appeals against
conviction and about 5 months for appeals on sentencing.45 An appeal against conviction based on
unreliable expert evidence would cause further delays for other court users.
The appeal process itself has cost implications for the appellant, whether he or she is privately
funded or publicly funded, and also for the court service and prosecution authorities. As discussed
above, we have estimated the costs of an appeal to be £25,000.
We do not know how many appeals will be prevented if our recommendations were to be taken
forward, so we have used a range of scenarios. The low estimate is that the reforms will save one
appeal annually; the best estimate is that two would be saved and the high estimate is that three
would be saved. The savings are set out below:
x Low estimate: £25,000 annually and £207,915 over 10 years.
x Best estimate: £50,000 annually and £415,830 over 10 years.
x High estimate: £75,000 annually and £623,745 over 10 years.
Note that these savings are offset against the initial increase in appeals. For the sake of clarity we
have separated the costings of the long-term decrease in appeals and the initial increase in
appeals.

5. Increased confidence in the criminal justice system


A reduction in the number of wrongful convictions and fewer appeals would bring a concomitant
increase in public confidence in the criminal justice system. It is impossible to quantify this crucial
benefit.

1A: Changes to the procedural regime


Costs
Transitional costs
1. Appendix to expert reports
The addition of an appendix to an expert’s report on the admissibility criteria would be a transitional
cost. The underlying basis of an expert’s opinion evidence (and the evidence of his or her
expertise) is unlikely to be case-specific, so it should be possible for the expert to use the same
material in his or her appendix in any subsequent case. The Expert Witness Institute told us that
expert witnesses would initially spend extra time preparing reports, but the guiding factors in our
draft Bill would provide a framework for experts to follow. Accordingly, the Institute did not think our
“proposed changes would increase costs in any significant way since an expert witness’s particular
responses to the criteria are unlikely to change markedly from case to case”.
If experts are already conforming to best practice, we anticipate that it will not take them much
more than one hour each to compile the appendix of the sort we recommend. The average cost
per hour for an expert’s time is £156, and so, assuming 3000 experts, the total transitional cost is
£468,000. We have used this as a low, best, and high estimate.
With regard to the Criminal Procedure Rules, Jonathan Solly from the MoJ Rules Committee
informed us there will be a complete revision of the Rules to be completed in 2015. No additional
work should be required to incorporate our recommended changes into the Rules, so there would
45
HMCS, The Court of Appeal Criminal Division Review of the Legal Year 2009 to 2010 (2010),
http://www.judiciary.gov.uk/media/media-releases/2010/jco-news-release-coa-crim-div-review-legal-yr (last
visited 1 March 2011), p 1.

192
be no additional costs in this respect.

2. Increase in challenges
It is possible that there would be an increase in the number of challenges to expert evidence
tendered for admission, which could result in longer proceedings in some cases. However, we
believe that after an initial increase, challenges would soon drop back close to present levels.

Ongoing costs
1. Judge-led pre-trial meetings with experts
There is already a power in the Criminal Procedure Rules 2010 to direct experts to meet, discuss
expert issues and prepare a joint statement on areas of agreement and disagreement. Our
recommendation would complement this. A judge would be able to order a meeting of experts,
chaired by the judge. We estimate that such a judge-led meeting would occur in 0% to 5% (best
estimate 2.5%) of cases where expert evidence is tendered for admission, and in half of those
cases they would replace a pre-hearing discussion under the existing rules (see below).
Using the figures from the Bond Solon and Legal Services Commission Expert Witness Survey
(Table 2), in 2009 there were 6,410 experts who gave evidence in the Crown Court and 2,136 who
gave evidence in magistrates’ courts.
A judge-led meeting of experts is estimated to take between half a day and one day (best estimate
one day). The cost inputs are nearly identical to those for pre-trial adversarial hearings: each party
would bring an expert and the CPS and LSC’s costs would be the same. Since the judge-led
meetings would require fewer overheads, we estimate that court costs would be about 10% lower
than the costs associated with a pre-trial reliability hearing.
Our calculations are set out in Tables 14A, B and C below.

Table 14A: annual total costs of judge-led meetings


Low Best High
Annual costs 2011 £0 £1,607,097 £3,214,194
Annual costs 2012 £0 £1,599,354 £3,198,708
Present value £0 £13,308,678 £26,617,356

Table 14B: annual total costs of judge-led meetings in magistrates’ courts


Low Best High
Estimated expert evidence tendered 2136
Judge-led meetings (0%, 2.5%, 5%) 0 53 107
Length of meeting (days) 0.5 1 1
Cost per meeting 2011 £3,454 £6,869 £6,869
Cost per meeting 2012 £3,443 £6,849 £6,849
Annual costs 2011 £0 £366,810 £733,620
Annual costs 2012 £0 £365,723 £731,447
Present value £0 £3,042,627 £6,085,253

193
Table 14C: annual total costs of judge-led meetings in Crown Courts
Low Best High
Estimated expert evidence tendered 6410
Judge-led meetings (0%, 2.5%, 5%) 0 160 321
Length of meeting (days) 0.5 1 1
Cost per meeting 2011 £3,901 £7,740 £7,740
Cost per meeting 2012 £3,879 £7,698 £7,698
Annual costs 2011 £0 £1,240,287 £2,480,575
Annual costs 2012 £0 £1,233,631 £2,467,261
Present value £0 £10,266,051 £20,532,102

Benefits
Transitional benefits
We do not anticipate any transitional benefits.

Ongoing benefits
1. Overall reduction in trial time
Our recommendations would encourage further pre-trial disclosure of matters relevant to the issue
of evidentiary reliability. Such disclosure would ensure that the parties and the trial judge are
properly equipped to scrutinise expert evidence for reliability before the trial.
Due to the narrowing of issues preceding the trial in judge-led meetings, we have estimated that
the time taken to adduce expert opinion evidence would be reduced by around one third. Assuming
that it takes on average three hours to adduce an expert’s opinion evidence in the Crown Court
and one hour in a magistrates’ court, this would translate to an hour per opinion in the Crown court
and 20 minutes in a magistrates’ court. Assuming that judge-led meetings would tend to be used
where there are experts for both the prosecution and the defence, this would be a saving of two
hours / 40 minutes. We also assumed an equal saving in experts’ preparation time.
The savings calculations are presented in Table 15, below. The savings per hour in court are
shown in Table 11.

Table 15: savings from shorter trials (Crown and magistrates’ courts)
Low Best High
Estimated expert evidence tendered 8546
Judge-led meetings (0%, 2.5%, 5%) 0 214 427
Trial hours saved 0 356 712
Savings 2011 £0 £542,867 £1,085,735
Savings 2012 £0 £539,145 £1,078,290
Present value £0 £4,487,454 £8,974,909

2. Reduction in pre-hearing discussions under rule 33.6 of the Criminal Procedure Rules 2010
Currently it is possible to hold pre-hearing discussions between experts under rule 33.6 of the 2010
Rules. The intended outcome of these discussions is a statement for the court explaining the
matters on which the experts agree and the areas of disagreement. A pre-hearing discussion can

194
be ordered by the judge when more than one party tenders expert evidence. The discussion might
reveal areas of disagreement and result in a judge-led meeting under our recommendations. When
it is apparent that there are areas of significant disagreement, we anticipate that the judge will
order a judge-led meeting instead of a pre-hearing discussion.
We have assumed that 50% of the new judge-led meetings will replace a pre-hearing discussion.
In such cases there would be a saving of the cost of a pre-hearing discussion to be offset against
the cost of the judge-led meeting.
On the advice of the CPS we have assumed that a pre-hearing discussion will involve experts and
legal representatives, but no court fees will be incurred. We have assumed that they will meet to
discuss for 2 hours, and spend one hour writing up the statement for the court together. Experts
will also prepare for 3 hours each. To be conservative we have assumed that, although Queen’s
Counsel may attend judge-led meetings and pre-trial hearings, legal representatives at pre-hearing
discussions will typically be junior advocates. The savings are set out in the tables below.

Table 16: costs per discussion


Hours of experts' time 12
Costs of expert per hour £156
Length of discussion 3
LSC & CPS costs per hour 2011 £109
Costs per discussion 2011 £2,200
LSC & CPS costs per hour 2012 and beyond £104
Costs per discussion 2012 and beyond £2,185

Table 17: savings from reduced pre-trial discussions under 1A


Low Best High
Judge-led meetings 0 214 427
Reduction in pre-hearing discussions 0 107 214
Cost of discussion 2011 £2,200 £2,200 £2,200
Costs per discussion 2012 £2,185 £2,185 £2,185
Annual savings 2011 £0 £234,989 £469,978
Annual savings 2012 £0 £233,413 £466,827
Present value 0 £1,942,729 £3,885,458

3. Enhanced pre-trial scrutiny of expert evidence


This would strengthen the operation of the recommendations made under Option 1, contributing to
the increased likelihood that unreliable expert evidence would be tendered for admission less
often. There would be fewer wrongful convictions and acquittals; there would be fewer appeals;
and there would be greater public confidence in the criminal justice system.

1B: Court-appointed expert witness


Costs
Transitional costs
Initial costs involved in setting up the panel are assumed to be negligible as we anticipate the legal

195
professionals involved would donate their time on a pro bono basis.
The initial costs associated with setting up the scheme could include recruiting panel members,
drafting procedures and assembling preliminary lists of potential court-appointed experts for
various fields. While this could be done pro bono, some administrative support would be required.

Ongoing costs
1. Administrative costs
Under 1B the judge would have the power to appoint an independent expert in cases where a
party’s expert opinion evidence is exceptionally complex. We recommend that the selection of
court-appointed experts would be undertaken by an independent panel, which would be self-
governing. A small ad hoc panel composed of voluntary members would meet when required to
identify an appropriate expert. The administrative support for the panel would be handled by the
MoJ.
Whenever a judge called upon the independent panel to fulfil its task, MoJ staff would be required
to arrange a meeting room, liaise with the panel and take minutes during meetings. The staff might
also have to draft correspondence and maintain records. It might be possible to incorporate much
of the cost into existing budgets, using this department’s existing resources (meeting rooms,
information technology support, filing and so on). If so, this would reduce costs further.
Notwithstanding the scope for limiting costs through voluntary arrangements and by utilising
existing resources, we have proceeded on the assumption that there would be administrative
costs.
We assume the panel would be called upon to appoint an independent expert 0 (low estimate) to
10 (high estimate) times a year (best estimate 5 times). We also assume that 8 hours of
administrative staff time would be used whenever the panel is convened:
x Correspondence: 5 hours
ƒ Between judge and administrator: 1.5 hours
ƒ Between administrator and panel: 2 hours
ƒ Between administrator and 2 experts (includes substitute expert): 1 hour
ƒ Administrator to sign-off on completed tasks: 0.5 hours
x Minutes of panel meeting: 3 hours
The median gross hourly pay of people working in administrative occupations in 2009 was £9.61,
so a meeting of the panel would incur an administrative staff cost of £77. Total annual
administrative staff costs would range between £0 and £769 (best estimate £384). The present
value ranges between £0 (low estimate) and £6,394 (high estimate), the best estimate being
£3,197.

2. Fees for panel members


If panel members were to act in a voluntary capacity, fees would be £0. This is a best and low
estimate.
If fees are payable, we work on the basis that there would be five members of any ad hoc panel.
The chairperson would be a Circuit Judge and the remaining four members would be experienced
legal practitioners. The MoJ assigns an annual salary of £128,296 to a Circuit Judge as of 1 April
2010. Taking a simple average, this is approximately £10,691 per month and £535 per day
(assuming a 20-day month).
High-earning legal practitioners with at least 10 years’ experience earn on average £80,000.46
However, this figure masks significant variations in earnings and is likely to be rather conservative.
The average earnings of panel members would very much depend on the composition of the

46
Workgateways, Legal Jobs UK, http://www.workgateways.com/job-legal.html#salary (last visited 25
February 2011).

196
particular panel.
If the £80,000 annual income average is used, the monthly earnings, based on a simple average,
would be £6,666, equivalent to £333 per day (assuming a 20-day working month).
If the panel meets for one day, ten times per year, this equates to: 10 x (£535 + [£333 x 4]) =
£18,670 total annual cost. This would be a present value of £155,271 over 10 years. This is our
high estimate.

3. Expenses for the panel


The members of the panel would act in a voluntary capacity but would receive reasonable out-of-
pocket expenses. If we were to assume average out-of-pocket expenses per member at £30 per
day, the total amount would be £150 per day (for five members). If the panel were to meet 0 to 10
times per year, for a maximum of one day each time, the total cost would range between £0 and
£1,500 per year. The low estimate is that members are based in London and no meetings take
place (£0).
As the meetings take place in London, the costs to attend the meeting will increase if panel
members need to travel into London. If we assume three members from London and two from
outside London, and we increase the average out-of-pocket expenses by 50% to £45 per day, the
total would be £225 per day for all five members. If the panel were to meet for one day, 0 to 10
times per year, the total cost would range from £0 to £2,250 per year. The best estimate is a panel
comprising both London- and non-London-based members with five meetings per year. This gives
an annual total cost of £1,125 and present value of £9,356 over 10 years.
If the panel consists of all non-London-based members and we assume an increase of 100% from
the London-based average to £60 per day, then between 0 and 10 meetings per year would cost
between £0 and £3,000 (£24,950 best estimate), which we have used as our high estimate.

Table 18: panel expenses


Average daily Cost of 0 to Cost of 5 Present value
Panel Members (5)
expense 10 meetings meetings of 5 meetings
All London based £30 £0 – £1,500 £750 £6,237
3 London/2 Non-London £45 £0 – £2,250 £1,125 £9,356
All Non-London £60 £0 – £3,000 £1,500 £12,475

4. Fees for court-appointed expert witnesses


Where the assistance of an independent court-appointed expert is required, the fee for this
individual would come from the courts’ central funds.
Using for present purposes the standard LSC rates for court attendance, the average per hour
attendance and preparation fee is £156. We assume that an independent expert would be required
to appear at court for one day for a pre-trial hearing (5 hours) and prepare for 10 hours, which
would result in a cost of £2,340 in expert fees every time an independent expert was used. We
have assumed that an independent expert would be used 0 to 10 times per year (best estimate 5).
The estimated annual cost would be between £0 and £23,400 (best estimate £11,700). The
present value over a 10-year period ranges between £0 and £194,609, best estimate £97,304
The total annual cost of a panel and the present value over 10 years is summarised in Table 20.

197
Table 19: total cost of a panel including witness fees
Low Best High
Costs
(No meetings) (5 meetings) (10 meetings)
Administrative £0 £384 £769
Panel members’ fees £0 £0 £18,670
Panel expenses £0 £1,125 £3,000
Experts’ fees £0 £11,700 £23,400
Total cost £0 £13,209 £45,639
Present value £0 £109,854 £381,225

5. Delays to the hearing


By providing Crown Court judges with a statutory power to call an independent expert witness, we
do not envisage any significant lengthening of proceedings or any significant additional costs being
incurred by HM Courts Service, the Crown Prosecution Service or the LSC (Ministry of Justice).
The parties will wish to scrutinise the evidence provided by the court-appointed expert and possibly
make representations on it following consultation with their experts, but in cases of this sort
additional costs and delays are likely to result from the fact that a trial judge has ordered a pre-trial
admissibility hearing on reliability.
That is to say, the infrequency with which this power would be used and the fact that the hearing
would nearly always be conducted in advance of the trial, before the jury is empanelled, means
that the selection process would be undertaken as part of the usual pre-trial proceedings, with no
adverse impact on the trial itself. Indeed, the appointment of an independent expert could feasibly
shorten the hearing or any subsequent trial or both.
Nevertheless, we accept that where the judge decides to appoint an independent expert, there
could occasionally be a delay in the overall proceedings. This is because the ad hoc panel would
have to draw up a list of potential experts, having liaised with relevant professional bodies. Delays
could be minimised, however, if the panel’s administrative support established early links with the
various professional bodies, giving those bodies the opportunity to compile a list of potential
candidates in advance of any request for assistance.
But even if additional delays are occasionally generated by the selection and appointment process,
these could be significantly shorter than the delays which might occur if the trial judge were to use
his or her common law power to find a suitable expert.

Benefits
Transitional benefits
We do not believe there would be any transitional benefits associated with 1B.

Ongoing benefits
An independent court-appointed expert could help judges assess the evidentiary reliability of very
complex (almost certainly scientific or mathematical) expert evidence. This reform measure would
help judges discharge their duty to investigate evidentiary reliability, thereby enhancing the benefits
of Option 1. Bruce Houlder QC foresaw “dangers for the criminal justice process in judges not
receiving such help.” He suggested that the existence of the power would also act as a deterrent
against casual science and might reduce costs in the long run by reducing the amount of unreliable
expert opinion evidence being tendered for admission in criminal proceedings.

1. Lower court costs


We have estimated that the presence of court-appointed independent experts would save between
10% and 30% (best estimate 20%) of the time it takes to adduce expert evidence. Since the court-

198
appointed expert would be used only in cases involving complex expert evidence we have applied
these percentage savings to evidence taking between 2 and 5 days to adduce, best estimate 3
days. A day is the Crown Court is 4.45 hours. The hourly cost of a trial in the Crown Court has
been taken from Table 11 above.

Table 20: savings from shorter trials (Crown Court)


Low Best High
Number of cases 0 5 10
Days of expert evidence 2 3 5
Hours per case 0 13.35 22.25
Total hours 0 66.75 225
Time savings 10% 20% 30%
Cost per hour 2011 £1,558 £1,558 £1,558
Cost per hour 2012 £1,547 £1,547 £1,547
Annual savings 2011 £0 £20,798 £105,161
Annual savings 2012 £0 £20,651 £104,414
Present value £0 £171,887 £869,091

Net impact of Option 1


The net impact of Option 1 is presented in Table 21 below.

Table 21: net impact of Option 1


Low estimate Best estimate High estimate
Transitional costs £473,000 £766,478 £1,177,333
Ongoing costs (annual) £0 £2,582,428 £9,547,106
Present value of costs £473,000 £22,150,815 £80,222,134
Transitional benefits £0 £0 £0
Ongoing benefits (annual) £2,059,843 £3,100,547 £4,204,815
Present value of benefits £17,114,089 £25,716,113 £34,843,455
Net present value £16,641,089 £3,565,298 -£45,378,678

In addition to the monetised costs and benefits associated with our recommendation, this policy
would carry significant non-monetised benefits. There have been well-publicised cases where
unreliable expert evidence has been wrongly admitted in high profile criminal trials. In such cases
the human costs for those convicted, and for their families, have the potential significantly to
undermine public confidence in the criminal justice system.

Option 2: judicial assessment of evidential reliability (including 1A, excluding 1B)


The costs and benefits of this option are identical to the sum of the cost and benefits for option 1,
excluding 1B.
The net impact of Option 2 is presented in Table 22, below.

199
Table 22: net impact of Option 2
Low estimate Best estimate High estimate
Transitional costs £473,000 £766,478 £1,177,333
Ongoing costs (annual) £0 £2,569,219 £9,501,267
Present value of costs £473,000 £22,040,238 £79,840,910
Transitional benefits £0 £0 £0
Ongoing benefits (annual) £2,059,843 £3,079,748 £4,099,654
Present value of benefits £17,114,089 £25,544,227 £33,974,365
Net present value £16,641,089 £3,503,988 -£45,866,545

Summary of options
This discussion is informed by the comparative table of estimated net present values for the
options.

Table 23: comparative net present values


Low estimate Best estimate High estimate
Option 0 £0 £0 £0
Option 1 £16,641,089 £3,565,298 -£45,378,678
Option 2 £16,641,089 £3,503,988 -£45,866,545

Option 1, complemented by 1A and 1B, is the preferred option because it would offer the best
solution to the problems associated with expert evidence and it would result in the highest NPV.
We believe that Option 1 would bring significant and important benefits.
We must emphasise that the inability to monetise the full extent of the benefits means that the full
value of our proposals could be under-estimated in this assessment.
Nevertheless, we believe the benefits of this option would outweigh the anticipated costs. We also
support implementation of 1A and 1B. The implementation of these reform measures would bring
positive benefits to the criminal justice system and would complement Option 1.
At one level our proposed reforms would not place any additional financial burden on the experts
and other individuals involved in criminal litigation. It is already open to a party to challenge the
admissibility of an expert’s opinion evidence on the ground of unreliability and for a pre-trial hearing
to be convened to address the issue. Moreover, parties and their experts should not be seeking to
adduce unreliable opinion evidence for admission, and should already be prepared to show why
their evidence ought to be relied on by juries and magistrates in criminal proceedings.
Furthermore, Part 33 of the Criminal Procedure Rules 2010 already sets out certain obligations as
to what must be set out in an expert witness’s written report, so the proposals introduced through
1A would not result in any significant change. Nor would our proposals require any fundamental
changes to the law of criminal procedure: the current procedural framework and rights of appeal
would continue to operate and the proposals would largely fit into this existing framework. In
addition, our recommendations would make it easier for trial judges to do what they already have
to do in appropriate cases. It is worth noting the view of Bruce Houlder QC (Director of Service
Prosecutions) that “the law has for some time been moving naturally towards [the changes we
recommend] and the impact will not be as great as feared”.
Our recommendations are in line with the development of the common law in the past year or so,

200
as exemplified most starkly by the judgment of the Court of Appeal in Reed.47 This development
reflects the attitude of the judiciary to the problems associated with expert opinion evidence and
further strengthens the value and desirability of our recommendations and draft Bill. We would be
providing the courts with the guidance they need and abolishing a common law admissibility test
which is generally considered to be flawed. The Court of Appeal fully recognises the desirability of
an investigation into evidentiary reliability in appropriate cases, certainly for evidence of a scientific
nature.
Moreover, we are confident that the potential benefits and savings generated by our proposals
would outweigh the initial and ongoing financial costs and, in line with our policy objective of
proportionality, an admissibility hearing would be held only if there was a legitimate doubt about the
basis of an expert’s opinion evidence. Over time, with widespread awareness of the rules and how
they are to be applied, the costs we have outlined would fall.
The difference between Option 1 and Option 2 comes down to whether or not the proposal for
court-appointed experts is introduced. Our recommendation that there should be a new power to
call upon an independent witness would provide a transparent, independent selection process,
offering important guarantees as to the expertise and impartiality of the court-appointed expert.
Our recommendations, if adopted, would ensure that convictions and acquittals are always
founded on sound expert opinion evidence. The expert proffering the evidence would have to
demonstrate the soundness of any underlying hypothesis and methodology and the soundness of
his or her reasoning. The courts would focus on the strength of the expert’s opinion and whether or
not it is warranted by the foundation material.

Risks, assumptions and sensitivities


Key assumptions
For the purposes of this cost benefit analysis we have made several assumptions.
1. There are 3,000 expert witnesses currently working in the UK.
2. For every wrongful conviction the imprisoned individual would appeal. Most expert evidence
is submitted in Crown Court cases, so we have assumed that all appeals would be heard by
the Court of Appeal (Criminal Division) and that each would cost £25,000.
3. Experts who wrote reports but did not attend court or did not give evidence were excluded
from our calculations as we have assumed that their evidence is likely to be uncontroversial
and so less likely to be unreliable. In 2009 there were an estimated 4,103 experts who
attended court but did not give evidence, compared with 8,546 who did, so some benefits
and costs could be underestimated.
4. 95% of expert evidence currently tendered for admission in the Crown Court would pass the
new test, as would 98% of the expert evidence currently tendered for admission in
magistrates’ courts.
5. 100% of unreliable evidence tendered would be challenged and would be ruled inadmissible.
6. Experts are paid at an average rate of £156 per hour, or £780 per day (5 hours). It is
assumed that experts take an average of 3 hours to adduce evidence in the Crown Court and
1 hour in the magistrates’ courts. In most cases we have assumed that preparation time
required is equal to the length of time that the expert will be giving evidence.
7. A wrongful conviction costs at least £123,548. The average annual cost per prison space is
£44,703, and an appeal against a wrongful conviction is heard after 10.1 months in prison.
An imprisoned individual earns the median national wage, which we have used as a proxy for
their loss of income while in prison, and it takes one year to return to work after release.

Sensitivities
In order to reduce the risk that our assumptions are incorrect we have used sensitivity ranges.

47
[2009] EWCA Crim 2689, [2010] 1 Cr App R 23.

201
1. We have estimated between 0 and 10 additional applications for leave to appeal, and the
appeal heard in 0 to 5 of these cases in each year over a five year period. The best estimate
is 5 applications for leave to appeal each year, and the appeal heard for two of these. An
increase in the number of appeals will decrease the NPV.
2. In the long term the number of appeals per year will reduce by between 1 and 3, best
estimate 2.
3. 0% to 40% of expert evidence which might not pass the test will still be tendered.
4. Normal pre-trial hearings in the Crown Court take between 0.5 and 3 days (best estimate 1
day), and one hour to one day in the magistrates’ courts (best estimate half a day). The new
judge-led meetings of experts take between 0.5 and 1 day (best estimate 1 day). Longer pre-
trial and judge-led meetings will lead to a lower NPV.
5. In 40% to 60% of cases where expert evidence is unreliable, the trial will not proceed either
because it is crucial to the prosecution case or because it is crucial to the defence case and
the defendant decides to plead guilty rather than put the prosecution to proof.
6. Judge-led meetings of experts will be used in 0% to 5% (best estimate 2.5%) of cases where
expert evidence is tendered. The more judge-led meetings of experts there are, the lower the
NPV.
7. We have assumed that the power under 1B to appoint independent experts will be used 0 to
10 times a year, best estimate 5 times. The more times the power is used, when needed, the
higher the NPV.

Risks
Risks to be considered include the risk that our assumptions are incorrect. In addition there is a
risk that we have under-estimated the potential increase in costs associated with our
recommended changes. We believe, however, that the risk of under-estimation is low because we
have used conservative figures and ranges in our costing. We set out particular risks below:
1. We have used the information from the LSC as a proxy for all expert witnesses presenting
evidence for the defence. The average costs and numbers paid from the courts’ central funds
might be very different and hence we might have under or over estimated our costs and
benefits.
2. There might be a higher than estimated increase in appeals under the new test, or the
increase could continue for longer than we have estimated. The appeal itself could take
longer than average.
3. We have miscalculated the impact of our proposals on the amount of expert evidence
tendered for admission in criminal cases, the number of times the reliability test would be
applied, and how often a pre-trial hearing would be conducted. If we have underestimated
these, then the costs of our proposals might be somewhat higher than we have anticipated.
4. If experts are not currently employing best practice regarding disclosure they could take more
than one hour each to compile an appendix demonstrating reliability. This could result in
higher costs for businesses. There may be more than 3,000 experts compiling such
appendices.
5. We have underestimated the number of times the new judge-led pre-trial hearing under 1A
will be used.
6. We have assumed that the panel selecting the independent experts will be prepared to work
on a voluntary basis and can meet and receive administrative support from the MoJ. If this is
not the case the start-up and ongoing costs of the panel could be significantly higher.
7. There is the low risk that the power under 1B would be used when not absolutely necessary,
and that costs would increase (but with concomitant benefits). However, we believe the
explicit limits on the power would prevent this from happening.

202
One in, one out
The new procedural requirements in 1A arguably impose some costs on experts, but they clarify
what experts are already expected to do. We have estimated that if experts are already following
disclosure requirements the total additional cost on experts would be £468,000. The total additional
costs for businesses and sole traders offering expert evidence would amount to £468,000.

3. Specific impact tests


An impact assessment must consider the specific impacts of a policy option upon various groups
within society. These specific tests are carried out below and refer to the implementation of
Option 1.

Statutory equality duties: gender, disability and race.


If the recommendation for a court-appointed expert (1B) is taken forward, convening an ad hoc
appointments panel might have some impact on individuals who have a role as a carer (evening or
weekend meetings might be necessary, to utilise available resources). However, because the
judicial discretion to call an expert witness would be used only very rarely, we do not think the
impact of this proposal would be significant. Moreover, the impact of the proposal could be
minimised. For example, where a meeting of the panel is convened, members could be given
alternative dates, times and venues to accommodate, as far as possible, those with caring
responsibilities.
After considering the equality impact assessment initial screening questions, we believe the
recommendations set out in Option 1 will have no significant impact in terms of gender, disability,
or race.

Competition assessment
According to Office of Fair Trading guidance,48 the competition assessment must consider whether
in any affected market, the proposal would directly or indirectly limit the number or range of
suppliers, reduce the supplier incentives to compete vigorously, or limit the ability of suppliers to
compete.
The Option 1 recommendations will affect the market for expert evidence. This has to be
understood in the context of the commercialisation of expert evidence in England and Wales. Our
recommendations will set minimum quality standards by which experts would be expected to work.
The overall effect on competition between suppliers will be negligible because the reform
measures will have a marked impact on only a small number of cases, on the assumption that
most expert evidence tendered for admission in criminal cases is currently reliable. Moreover, any
change would apply to all criminal cases involving expert evidence and so all experts would be
subject to the same standards.

Small firms impact test


The size of businesses is determined by the number of employees. Some forensic companies are
small or very small businesses; they employ fewer than 50 people. At least 75% of expert
witnesses do not work in a larger expert witness grouping. Implementing Option 1 would therefore
have an impact on small businesses. The policy would entail increased costs to experts because
they would have to undertake the necessary additional preparatory work required to be able to
demonstrate their expertise, to adhere to the pre-trial disclosure requirements and to be prepared
to demonstrate the evidentiary reliability of their opinion evidence. Experts would, however, be paid
for this work. Moreover, we anticipate the additional cost involved would decrease over time as
experts would be able to reuse prepared material in subsequent cases.

48
Office of Fair Trading, Completing competition assessments in Impact Assessments: Guideline for policy
makers (OFT876) (August 2007), htpp://www.oft.gov.uk/shared_oft/reports/comp_policy/oft876.pdf (last
visited 9 February 2011).

203
The impact identified here must be placed in context. The proposed statutory test does not require
experts to do anything they are not already required to do under the common law. The test would
codify the existing requirements and provide factors to help judges to assess the evidentiary
reliability of expert evidence.
The recommendations under 1A to amend the pre-trial requirements of experts would support the
statutory admissibility test and ensure, as far as possible, that only reliable expert evidence is
tendered for admission in criminal proceedings. We do not anticipate that implementing 1A would
have any significant adverse impact on small businesses. There are already disclosure
requirements experts must adhere to. The recommended reform measures would merely add to
these existing requirements and so there would be no substantial adverse impact.

Carbon assessment
We do not foresee any impact on carbon emissions.

Other environment
We do not foresee any impact on the environment.

Health impact assessment


As outlined above, we expect that our proposals under Option 1 would have a beneficial impact on
health by reducing the likelihood of wrongful convictions and acquittals. Wrongful convictions and
acquittals can have a detrimental impact on the mental and physical health of many people.

Human rights
If implemented, the proposed reforms set out in Option 1 would minimise the risk of unreliable
expert evidence being responsible for wrongful convictions and acquittals. This has the obvious
benefit of ensuring a fairer criminal justice system and promoting the protection of human rights
under the European Convention on Human Rights and the Human Rights Act 1998.
Several Convention articles relate to fairness in criminal justice systems and, above all, require any
convictions and punishments to be lawfully imposed. Most obviously, article 6 (the right to a fair
trial) is involved, but other rights that can also be considered, including article 3 (protection against
inhuman or degrading treatment or punishment), article 5 (the right to liberty and security of
person) and article 8 (respect for private and family life).
Our recommended reforms would comply with the objectives of promoting and protecting human
rights under the Human Rights Act 1998.

Justice impact test


The impact on the justice system of our proposals is considered throughout this impact
assessment.

Rural proofing
We do not foresee any differential impact on rural areas.

Sustainable development
We do not foresee any impact on sustainable development.

204
Annexes
Annex 1 should be used to set out the Post Implementation Review Plan as detailed below. Further
annexes may be added to provide further information about non-monetary costs and benefits from
Specific Impact Tests, if relevant to an overall understanding of policy options.

Annex 1: Post Implementation Review (PIR) Plan


A PIR should be undertaken, usually three to five years after implementation of the policy, but
exceptionally a longer period may be more appropriate. A PIR should examine the extent to which
the implemented regulations have achieved their objectives, assess their costs and benefits and
identify whether they are having any unintended consequences. Please set out the PIR Plan as
detailed below. If there is no plan to do a PIR please provide reasons below.
Basis of the review: [The basis of the review could be statutory (forming part of the legislation), it could be to review
existing policy or there could be a political commitment to review];
N/A

Review objective: [Is it intended as a proportionate check that regulation is operating as expected to tackle the problem
of concern?; or as a wider exploration of the policy approach taken?; or as a link from policy objective to outcome?]
N/A

Review approach and rationale: [e.g. describe here the review approach (in-depth evaluation, scope review of
monitoring data, scan of stakeholder views, etc.) and the rationale that made choosing such an approach]
N/A

Baseline: [The current (baseline) position against which the change introduced by the legislation can be measured]
N/A

Success criteria: [Criteria showing achievement of the policy objectives as set out in the final impact assessment; criteria
for modifying or replacing the policy if it does not achieve its objectives]
N/A

Monitoring information arrangements: [Provide further details of the planned/existing arrangements in place that
will allow a systematic collection systematic collection of monitoring information for future policy review]
If Option 2 is implemented then the application of the test should be monitored to see whether,
at some later date, 1B should be brought into force.
Currently there is no central collection of data on the number of cases in which expert evidence is
tendered or admitted, or on the number of cases in which an expert opinion has wrongly been
allowed to go before a criminal court. Increased data collection in this area would inform any PIR.
Reasons for not planning a PIR:
The Law Commission does not implement policy and does not therefore review policy
implementation.

205
APPENDIX D
ACKNOWLEDGMENTS

CONSULTEES TO CONSULTATION PAPER 190


The Academy of Experts

Lord Justice Aikens

The Association of Forensic Science Providers

WE Bache (Solicitor)

M John Batt (Consultant, Batt Broadbent Solicitors)

Berolena (Jacqui Cooper) (online forum)

Better Trials Unit, Ministry of Justice

Bodriche (online forum)

The British Association for Shooting and Conservation

British Medical Association

The British Psychological Society

British Standards Institution

Andrew Campbell-Tiech QC (Dyers Chambers)

Centre for Criminal and Civil Evidence and Procedure, School of Law,
Northumbria University

Dr Tim Clayton (Barrister and expert witness employed by the Forensic Science
Service Ltd)

The Criminal Bar Association

Criminal Cases Review Commission

Crown Prosecution Service

Michael Curry (online forum)

Professor Tim David (Professor of Child Health and Paediatrics, University of


Manchester)

Frank De Silva

Devon County Council’s Trading Standards Service

Dr Déirdre Dwyer (Faculty of Law, University of Oxford; Barrister of Lincoln’s Inn)

206
Dr Gary Edmond (Associate Professor and Director, Expertise, Evidence & Law
Program, Faculty of Law, the University of New South Wales)

Anthony Edwards (Solicitor, TV Edwards LLP)

Richard Emery (4Keys International)

The Hon Theodore R Essex (Administrative Law Judge, United States


International Trade Commission)

The Expert Witness Institute

Roy Everett (online forum)

Dr Ian Webber Evett (Scientist and member, Forensic Science Service)

Forensic Access Ltd

The Forensic Institute

Forensic Science Regulator (Andrew Rennison)

Forensic Science Service Ltd

Forensic Science Society

General Medical Council

His Honour Judge Andrew Gilbart QC (Honorary Recorder of Manchester)

Dr Cedric Gilson (Visiting Fellow, Department of Advanced Legal Studies, School


of Law, University of Westminster)

Mr Justice Peter Gross

Professor David Hand (Professor of Statistics, Imperial College, London)

John Hemming MP

Bruce Houlder QC (Director of Service Prosecutions)

Laura Hoyano (Fellow & Tutor in Law, Wadham College, Oxford)

Michael Innis (Scientist)

Justices’ Clerks’ Society

Law Reform Committee of the Bar Council

The Law Society of England and Wales

Legal Services Commission

The London Criminal Courts Solicitors’ Association

207
LGC Forensics

Dr David Lucy (Department of Mathematics & Statistics, Lancaster University)

Campbell Malone (Solicitor, Stephensons Solicitors LLP)

Professor Pierre Margot (Professor of Forensic Science in the School of Criminal


Justice; Vice-Dean of the Faculty of Law and Criminal Justice, University of
Lausanne, Switzerland)

The Medical Defence Union

Penny Mellor (Campaigner, Dare to Care)

Dr Bob Moles (Networked Knowledge)

Dr Geoffrey Morrison (Research Associate, School of Language Studies,


Australian National University; Visiting Fellow, School of Electrical Engineering
and Telecommunications, University of New South Wales)

Associate Professor William O’Brian Jr (University of Warwick)

Old Bailey Judges

Dr Malcolm Park (Centre for Spatial Data Infrastructures & Land Administration,
University of Melbourne)

Simon Phillips (Barrister, Park Court Chambers)

Police Superintendents’ Association of England and Wales, Crime and Criminal


Justice Business Area

Professionals Against Child Abuse

Gary Pugh (Director of Forensic Services, Metropolitan Police)

David Ranson (Deputy Director, Victorian Institute of Forensic Medicine; Hon


Clinical Associate Professor, Department of Forensic Medicine, Monash
University, Australia)

Steve Redhead (Partner, R&M Chartered Accountants – Forensic Experts)

Professor Mike Redmayne (London School of Economics)

The Rose Committee of the Senior Judiciary

Dr Keith JB Rix (Consultant Forensic Psychiatrist)

Associate Professor Andrew Roberts (University of Warwick)

Professor Paul Roberts (Professor of Criminal Jurisprudence, University of


Nottingham)

Dr Phil Rose (Associate Professor in Phonetics and Chinese Linguistics; Forensic


Phonetics Consultant, Australian National University)

208
The Royal College of Paediatrics and Child Health

The Royal College of Psychiatrists

Royal College of Veterinary Surgeons

The Royal Society for the Prevention of Cruelty to Animals (RSPCA)

The Royal Statistical Society (Prepared by Professor CGG Aitken, Chairman,


Statistics and Law Working Group)

Oriola Sallavaci (Middlesex University)

Alec Samuels (Barrister, formerly Reader in Law, University of Southampton)

Society of Expert Witnesses

Skills for Justice

Geoffrey Thursfield

The Honourable Mr Justice Treacy

UK Accreditation Service

UK Forensic Speech Science Community

UK Register of Expert Witnesses (Dr Chris Pamplin)

Dr Glyn Walters (Consultant Chemical Pathologist, retired)

Tony Ward (Reader in Law, University of Hull) (online post)

Dr Matthew Weait (Reader in Socio-Legal Studies, Birkbeck College, University


of London)

Susan Weston

Adam Wilson (Senior Lecturer in Law, Sheffield Hallam University)

Dr Ian Wilson

Professor Wesley Vernon

One further consultee who wishes to remain anonymous.

MEMBERS OF THE WORKING GROUP


Sir Louis Blom-Cooper QC (Doughty Street Chambers)

Graham Cooke (Kings Bench Chambers)

John Cooper QC (25 Bedford Row)

209
Professor Nigel Eastman (Consultant Forensic Psychiatrist; Professor of Law and
Ethics in Psychiatry, St Georges University of London)

Dr Chris Pamplin (Editor of the UK Register of Expert Witnesses)

Professor Mike Redmayne (London School of Economics)

Andrew Rennison (Forensic Science Regulator)

Professor Paul Roberts (University of Nottingham)

CONSULTEES TO ADDITIONAL CONSULTATATIONS


The Academy of Experts

Attorney General’s Office

Better Trials Unit, Ministry of Justice

Michael Bowes QC (Outer Temple)

British Medical Association

The British Psychological Society

Andrew Campbell-Tiech QC (Dyers Chambers)

Crown Prosecution Service

Simon Daniel (Serious Fraud Office)

Professor Tim David (Professor of Child Health and Paediatrics, University of


Manchester)

Professor Nigel Eastman (Consultant Forensic Psychiatrist; Professor of Law and


Ethics in Psychiatry, St Georges University of London)

Anthony Edwards (TV Edwards LLP)

Richard Emery (4Keys International)

The Expert Witness Institute**

Dr Ian Webber Evett (Scientist and member, Forensic Science Service)

Forensic Access Ltd

The Forensic Institute

Forensic Science Regulator (Andrew Rennison)

General Medical Council

Anthony Heaton-Armstrong (Barrister, 9-12 Bell Yard)

210
Max Hill QC (18 Red Lion Court)

Bruce Houlder QC (Director of Service Prosecutions)

Judicial Studies Board

Paul Keleher QC (25 Bedford Row)

Legal Committee of HM Council of District Judges (Magistrates’ Courts)

Dr David Lucy (Department of Mathematics & Statistics, Lancaster University)

The Medical Defence Union

Gary Pugh (Director of Forensic Services, Metropolitan Police)

Edward Rees QC (Doughty Street Chambers)

Professor Mike Redmayne (London School of Economics)

Dr Keith JB Rix (Consultant Forensic Psychiatrist)

His Honour Judge Jeremy Roberts QC

The Royal College of Paediatrics and Child Health

The Royal College of Psychiatrists

Royal College of Veterinary Surgeons

The Royal Society for the Prevention of Cruelty to Animals (RSPCA)

The Royal Statistical Society (Prepared by Prof CGG Aitken, Chairman, Statistics
and Law Working Group)

David Spens QC (Garden Court Chambers)

Michael Topolski QC (Tooks Chambers)

UK Accreditation Service

UK Forensic Speech Science Community

UK Register of Expert Witnesses (Dr Chris Pamplin)

** additional response to an earlier draft provided by Penny Cooper and James


Badenoch QC. Professor Cooper also provided further comments from
academics at the City Law School.

211
Published by TSO (The Stationery Office) and available from:

Online
www.tsoshop.co.uk

Mail, Telephone Fax & E-Mail


TSO
PO Box 29, Norwich, NR3 1GN
Telephone orders/General enquiries: 0870 600 5522
Order through the Parliamentary Hotline Lo-Call 0845 7 023474
Fax orders: 0870 600 5533
E-mail: customer.services@tso.co.uk
Textphone: 0870 240 3701

The Parliamentary Bookshop


12 Bridge Street, Parliament Square,
London SW1A 2JX
Telephone orders/General enquiries: 020 7219 3890
Fax orders: 020 7219 3866
Email: bookshop@parliament.uk
Internet http://www.bookshop.parliament.uk

TSO@Blackwell and other Accredited Agents

Customer can also order publications from


TSO Ireland
16 Arthur Street, Belfast BT1 4GD
Telephone orders/General enquiries: 028 9023 8451
Fax orders: 028 9023 5401

You might also like