. 22
( 34)


light of the consultation the Government announced that it favoured more
research and that it was not opposed to amending the 1981 Act, but that that
should not be done until there were speci¬c questions to be answered that could
not be su¬ciently investigated without altering the law.110
Studies of jury decision-making have mainly been based on the impressions
of judges, lawyers, or police o¬cers, or on simulations with ˜shadow™ or ˜mock™
One early study was the famous Chicago project based on the impressions of
judges conducted by Professors Harry Kalven, Jr. and Hans Zeisel of Chicago
University and published as The American Jury.111 The work was based on 3,576

Plotniko¬ and Woolfson research study.
The lengthy questionnaire addressed to jurors in the Crown Court study done for the
Runciman Royal Commission was not exempt from the provisions of the 1981 Act. The
questions asked were all carefully drafted and o¬cially approved on the basis that they did not
infringe the provisions of s. 8. Runciman, recommendation 1, p. 188.
Auld, Ch. 5, paras. 82“7, pp. 166“8.
Jury Research and Impropriety, DCA, consultation paper 04/2005, January 2005. For the text of
the consultation papers and the summary of responses see www.dca.gov.uk “ Publications “
Consultation Papers “ January 2005. For the writer™s actual response against the idea, contrary
to the view of the Runciman Royal Commission to which he signed up in 1993, see
www.lse.ac.uk “ Law Department “ Who™s Who “ academic sta¬ (emeritus).
For an account of the responses to the consultation paper and an assessment of the
Government™s decision see P. Ferguson QC, ˜Jury Research and Impropriety™, 155 New Law
Journal, 2 December 2005, p. 1840. (Little Brown & Co, 1966).
514 The jury

actual criminal trials and the replies to a questionnaire from the 555 trial judges
involved. (Jurymen could not be approached.) The results showed that judges
and juries agreed to acquit in 13 per cent of cases and agreed to convict in 62 per
cent of cases, yielding a total agreement rate of 75 per cent. In cases where judge
and jury disagreed, it was found that the jury was more lenient than the judge in
19 per cent and less lenient in 3 per cent. Just over half of the disagreements
which seemed explicable were caused by di¬erent approaches to the evidence.
Nearly one-third were due to jury reaction to the law and about one-tenth were
due to jury sentiments about the defendant himself. The authors summarised
their conclusions:

Harry Kalven, Jr. and Hans Zeisel, ˜The American Jury™, New Society, 25
August 1966, p. 290
It may be useful to put quite general and interrelated questions: why do judge
and jury ever disagree, and why do they not disagree more often?
The answer must turn on the intrinsic di¬erences between the two institu-
tions. The judge very often perceives the stimulus that moves the jury, but does
not yield to it. Indeed it is interesting how often the judge describes with sensi-
tivity a factor which he then excludes from his own considerations.
The better question is the second. Since the jury does at times recognise and
use its de facto freedom, why does it not deviate from the judge more often? Why
is it not more of a wildcat operation? In many ways our single most basic ¬nding
is that the jury, despite its autonomy, spins so close to the legal baseline.
The study does not answer directly, but it does lay the ground for three plau-
sible suggestions. As just noted, the o¬cial law has done pretty well in adjust-
ing to the equities, and there is therefore no great gap between the o¬cial values
and the popular. Again, the group nature of the jury decision will moderate and
brake eccentric views. Lastly, the jury is not simply a corner gang picked from
the street; it has been invested with a public task, brought under the in¬‚uence
of a judge, and put to work in solemn surroundings. Perhaps one reason why
the jury exercises its very real power so sparingly is because it is o¬cially told it
has none.
The jury thus represents a uniquely subtle distribution of o¬cial power; an
unusual arrangement of checks and balances. It represents also an impressive
way of building discretion, equity, and ¬‚exibility into a legal system. Not the
least of the advantages is that the jury, relieved of the burdens of creating prece-
dent, can bend the law without breaking it.
Whether or not one comes to admire the jury system as much as we have, it
must rank as a daring e¬ort in human arrangement to work out a solution to
the tensions between law and equity and anarchy.112

For an extended discussion of the book see Criminal Law Review, 1967, pp. 555“86, but
for doubts about the statistical methodology of the study see A.E. Bottoms and M. Walker,
˜The American Jury: A Critique™, 67 Journal of the American Statistical Association, 1972,
p. 773. For the authors™ rejoinder, see ibid, p. 779. For an assessment in 1991 see V. Hans
and N. Vidman, ˜The American Jury at Twenty-Five Years™, Law and Social Inquiry, 1991,
p. 323.
515 The quality of jury decision-making

The ¬rst English study, by the Oxford Penal Research Unit, was based primar-
ily on the views of barristers and the police. Its principal ¬nding was that most
acquittals were ˜attributable to a single cause “ the failure of the prosecution
(normally the police) to provide enough information, or to present it in court
in a way that would convince both judge and jury of the defendant™s guilt™.113
Very few verdicts were found to be perverse.
The writer™s study of acquittals at the Old Bailey and the Inner London
Crown Court was based on questionnaire interviews with the barristers for the
prosecution and the defence. It was striking that there was no great di¬erence
of view between prosecution and defence lawyers as to the likely reasons for the
acquittals. Again, there was little evidence of perverse verdicts.114
For the report of a series of experiments with ˜mock™ juries who listened to
tape recorded trials, see A.P. Sealy and W.R. Cornish, ˜Juries and their Verdicts™,
36 Modern Law Review, 1973, p. 496; and LSE Jury Project, ˜Juries and the Rules
of Evidence™, Criminal Law Review, 1973, p. 208.
A study based on thirty cases heard by ˜shadow™ juries conducted by the
Oxford Penal Research Unit showed the jury approaching its task very soberly.
The shadow juries listened to real cases and when the real jury withdrew to
consider their verdicts, so did the shadow jury. The authors summarised their

Sarah McCabe and Robert Purves, The Shadow Jury at Work, 1974,
pp. 60“3
Of course the ˜shadow™ jury discussions and verdicts were not comparable with
those of the real jury since the future of the defendant was not at risk, but the
fact that many of our volunteers felt like jurors encourages us to make certain
comparisons where real and ˜shadow™ jury verdicts agree . . .
Summary of results
3. The ˜shadow™ juries showed considerable determination in looking for evi-
dence upon which convictions could be based; when it seemed inadequate, they
were not prepared to allow their own ˜hunch™ that the defendant was involved in
some way in the o¬ence that was charged to stand in the way of an acquittal . . .
5. There was little evidence of perversity in the ¬nal decisions of these thirty
groups. One acquittal only showed that sympathy and impatience with the triv-
iality of the case so in¬‚uenced the ˜shadow™ jurors™ view of the evidence that they
refused to convict. One other unexpected acquittal seemed to be wholly due to
dissatisfaction with the evidence.
A less positive view of jury decisions emerged from a later piece of research by
Professors John Baldwin and Michael McConville.115 They selected a random
sample in Birmingham Crown Court of 500 defendants who pleaded not guilty.
In the event, 116 of these were acquitted by the judge before the case had run its

S. McCabe and R. Purves, The Jury at Work (Blackwell, 1972) p. 11.
M. Zander, ˜Are Too Many Professional Criminals Avoiding Conviction?™, 37 Modern Law
Review, 1974, p. 28. Jury Trials (Clarendon, 1979).
516 The jury

full course and another fourteen changed their plea to guilty during the case.
This left 370, of which 114 ended in acquittal. The researchers asked the trial
judge, the defence solicitor, the prosecuting solicitor, the police and the defen-
dant himself about these cases “ the ¬rst three groups by questionnaire and the
last two by interview. The response rate was very high (over 95 per cent for the
judges, the prosecuting solicitors and the police). The table below shows the
opinions of the di¬erent groups regarding the 114 acquittal cases.

Judge Defence Prosecuting Police
per cent solicitor solicitor per cent
per cent per cent

No strong view expressed 62 83 65 48
that the acquittal was
not justi¬ed
Some doubts about the 6 7 9 8
Serious doubts about the 32 10 26 44
Total 100 (114) 100 (114) 100 (114) 100 (114)

(Source: Jury Trials, Table 5, p. 46)

The acquittal was seen as doubtful or highly questionable by one respondent
in thirty instances (27 per cent of the 114), by two respondents in sixteen (14
per cent) and by three or more respondents in twenty-eight (25 per cent).116
There were forty-one cases in which both judge and one other respondent
found the acquittal doubtful.117
Convictions were less often found doubtful or highly questionable, but 8 per
cent were so regarded by one or more respondents (2 per cent by one respon-
dent, 3 per cent by two and 3 per cent by three or more respondents).118
The researchers concluded that in respect of a few acquittals it might be said
˜that the jury™s verdict was primarily conditioned by its sympathy for the defen-
dant or antipathy towards the victim™ and ˜some questionable convictions can
possibly be explained on the basis of sympathy with the victim or prejudice
against the defendant™, but in general ˜the performance of the jury did not
always appear to accord with the principle underlying the trial system in
England that it is better to acquit those who are probably guilty than to convict
any who are possibly innocent. On the contrary, the jury appeared on occasion
to be over ready to acquit those who were probably guilty and insu¬ciently pre-
pared to protect the possibly innocent™.119 There was nothing in the composi-
tion of the jury (age, sex or social class) that correlated with the decisions.

116 117 118 119
Ibid, Table 6, p. 47. Ibid, p. 54. Ibid, Table 9, p. 51. Ibid, p. 128.
517 The quality of jury decision-making

The study is signi¬cantly di¬erent from previous studies in suggesting a
considerable measure of disagreement between jury verdicts and those of the
other key actors.
By contrast the Crown Court Study was broadly very positive. As previously
noted, the study was based on the responses of jurors, prosecution and defence
barristers, judges and police o¬cers concerned in some 800 contested cases in
every Crown Court in England and Wales in a two-week period in February
Did the jury understand the evidence? Jurors were asked: ˜How di¬cult was it
for you to understand the evidence in this case?™ Over 90 per cent thought it ˜not
at all di¬cult™ (50 per cent) or ˜not very di¬cult™ (41 per cent).120 The same
question was asked in cases where there was scienti¬c evidence. The results were
very similar “ 56 per cent ˜not at all di¬cult™, 34 per cent ˜not very di¬cult™.121
Jurors were then asked: ˜Do you think the jury as a whole was able to under-
stand the evidence?™ The response broadly was Yes. Over 90 per cent thought
that all the jury understood the evidence (56 per cent) or that most understood
(41 per cent). The response from jury foremen was virtually identical.122 There
were 143 juries (17 per cent of the 821 in the study) in which one or more jurors
said ˜Only a few understood™ or ˜None of them understood™. 116 juries had one
such member, twenty had two such members, six had three and one had four.123
The prosecution and defence barristers were asked whether they thought the
jury had trouble understanding the evidence. 94 per cent of prosecution bar-
risters and 90 per cent of defence barristers thought they had no trouble.124 As
a result of an oversight this question was not put to the judges, but they were
asked whether the jury could understand the scienti¬c evidence in cases where
there had been some. In no fewer than 93 per cent of these cases the judges
thought all the scienti¬c evidence was understandable by the jury.125
The answers for the question ˜Could the jurors remember the evidence?™ were
much the same.126
Was the jury™s verdict surprising? Di¬erent participants in the trial were asked:
˜In your view, was the jury™s decision surprising in the light of the evidence?™ In
the great majority of cases the answer was No. The verdict was surprising in the
view of 27 per cent of the CPS, 25 per cent of the police, 18 per cent of the
defence solicitors, 15 per cent of the prosecution barristers and 14 per cent of
the judges and the defence barristers.127
The great majority of respondents in all the categories thought the verdict
was understandable in the light of the evidence. Those who thought it was
against the weight of the evidence but explicable gave a long list of explanations:
sympathy for the defendant, antipathy toward the complainant, case too trivial
or stale, misconduct by the police, concern over sentence and quality or lack of

120 121
M. Zander and P. Henderson, n. 69 above at section 8.2.1. Ibid at section 8.2.2.
122 123 124 125
Ibid at section 8.2.3. Ibid. Ibid, p. 177, Table 6.15. Ibid.
126 127
Ibid, p. 178, Table 6.16; p. 209, Table 8.9. Ibid, p. 163, Table 6.5.
518 The jury

quality of the respective counsel. Hardly any respondents thought the decision
was against the judge™s direction on law.
The prosecution and defence lawyers and the judges all agreed that 2“4 per
cent of jury decisions were inexplicable. The police thought that 8 per cent were
When the judges, the prosecution barristers and the police thought that the
verdict was against the weight of the evidence it was an acquittal in about 90 per
cent of instances. When defence barristers and defence solicitors thought the
verdict was against the weight of the evidence just under half were acquittals.128
On the basis of these ¬gures it appears that ˜problematic jury acquittals™ con-
stituted 31 per cent of all jury acquittals for prosecution barristers, 29 per cent
for the judges and 16 per cent for defence barristers.129 Jury acquittals are about
one-third of all acquittals in the Crown Court (see p. 538 below). On that basis,
problematic acquittals would be around one-tenth of all acquittals.
The Crown Court Study also showed that there were some (though far fewer)
problematic convictions. Judges and prosecution barristers thought that 2 per
cent of convictions were problematic, whereas defence barristers thought that
17 per cent were problematic.130
Research by Julie Vennard, then of the Home O¬ce Research and Planning
Unit, supports the view that juries decide rationally and on the basis of the evi-
Length of jury deliberations In most cases the jury was out for a very short
period “ in over half (52 per cent) for under two hours and in three-quarters
(77 per cent) for under four hours.132 There were eight cases (1 per cent) in
which the jury stayed together overnight. (At the time of the survey it was a rule
that once the jury had begun their deliberations, they were not allowed to
separate until they reached their verdict, known as ˜sequestration of the jury™.
Now, under the Criminal Justice and Public Order Act 1994, s. 43, it is in
the judge™s discretion whether he permits the jury to go home while they are
Not surprisingly, the length of jury deliberations was closely associated with
the length of the case. Thus where the case lasted under half a day, the jurors
reported being out for under two hours in 96 per cent of cases. When the case
lasted three to four days the jurors were back within two hours in only 15 per
cent of cases. When it lasted over two weeks, the jurors took more than four
hours in three-quarters of the cases.133 (The English record is probably still the
Maxwell trial in 1996 where the jury were out for seven days of deliberations
spread over ten days and eleven nights.)

128 129 130
Ibid, Table 6.9. Ibid, p. 170. Ibid, pp. 170“1.
J. Vennard, ˜The Outcome of Contested Trials™ in D. Moxon (ed.), Managing Criminal Justice
(1985) pp. 126“51; and Evidence and Outcome: a Comparison of Contested Trials in
Magistrates™ Courts and the Crown Court (Home O¬ce Research and Planning Unit, Research
Bulletin No. 20, 1986) p. 48.
132 133
M. Zander and P. Henderson, n. 69 above, Table 8.23, p. 225. Ibid.
519 Respective roles of judge and jury

For an unscienti¬c and distinctly jaundiced account of the experience of
serving on a jury, see the lecture of the late Professor Ely Devons of the London
School of Economics, ˜Serving as a Juryman in Britain™, 28 Modern Law Review,
1965, p. 561. See also articles in 140 New Law Journal, 14 September 1990,
pp. 1264“76; ˜Jury Service: A Personal Observation™, LAG Bulletin, 1979, p. 278;
and Trevor Grove, The Juryman™s Tale (Bloomsbury Publishing, 2000).

10. Respective roles of judge and jury
During the trial the jury is normally passive, simply listening to the case as it
develops. Sometimes the jury will ask a question by passing a note to the judge.
The judge will then decide whether, and if so how, the question should be
answered. Sometimes he will invite the views of the lawyers for both sides.
At each break they are usually warned by the judge not to discuss the case
among themselves or with anyone else until they reach their deliberations at
the end.
The judge™s role in the adversary system, as has been seen, is also largely
passive if the comparison is with that of the judge in the continental system, but
by comparison with the jury, the judge is quite actively involved. In particular,
he will have to rule on points of law as they arise, especially with regard to the
admissibility or otherwise of evidence. If this involves lengthy debate, the jury
will be asked to withdraw. The judge knows more about the case than the jury
in that he has access to the pre-trial papers. At the close of the prosecution™s case
he may have to deal with a submission that there is no case to answer (see
pp. 523“24 below).
For the view that there should be greater interaction between judge and jury
with regard to fact ¬nding see J. Jackson and S. Doran, ˜Judge and Jury: Towards
a New Division of Labour in Criminal Trials™, 60 Modern Law Review, 1997,
pp. 759“78.
The judge is supposed to sum up for the jury on both the law and the facts.134

Summing up the law
In McVey135 the Court of Appeal spelled out the minimum content of every
summing up: ˜it is trite to say that every summing up must contain at least a
direction to the jury as to the burden and standard of proof, and as to the ingre-
dients of the o¬ence or o¬ences which the jury are called upon to consider™. The
problem of what is meant by this dictum was considered in a lecture entitled
˜Summing Up the Law™ by the late Professor Edward Griew:136

For a valuable review of the empirical evidence see P. Darbyshire, ˜What can we Learn from
Published Jury Research? Findings for the Criminal Courts Review 2001™, Criminal Law
Review, 2001, pp. 970“9. See especially W. Young, ˜Summing-up to Juries in Criminal Cases “
What Jury Research says about Current Rules and Practice™, Criminal Law Review, 2003,
135 136
pp. 665“89. [1988] Crim LR 127. Criminal Law Review, 1989, pp. 768“80.
520 The jury

Directing the jury as to the burden of proof means telling them who has to prove
the case; it means telling them that the prosecution must prove the defendant™s
guilt, not the defendant his innocence. Directing them as to the standard of
proof means telling them that the case has to be proved beyond reasonable
doubt “ commonly expressed by saying that they may convict the defendant
only if they are sure of his guilt.137
In recent years the Court of Appeal has laid down ˜model™ or ˜specimen™ direc-
tions or standard forms of words in which directions on particular matters can
or ought to be given. A number are now embodied in a document issued to
all judges who sit in the Crown Court by the Judicial Studies Board with
the approval of the Lord Chief Justice. They are published on the Board™s
Website.138 The foreword warns: ˜They are an invaluable tool “ but must be a
servant not a master . . . They must be adapted to the needs of the individual
In his lecture, Professor Griew criticised the tendency of judges to give the
jury more law than it needed for the purpose of its decision139 and to use overly
technical and complex language. American research showed that a good many
judicial directions on law to juries were ˜totally incomprehensible to an alarm-
ing percentage of jurors™.140 No doubt similar research in this country would
yield similar results. ˜Our juries continue to be addressed in language relatively
rich in abstract and latinate words and in sentences that are often very long™.
In the Crown Court Study, 61 per cent of jurors said they found the judge™s
summing up ˜not at all di¬cult™ to understand and another 33 per cent found
it ˜not very di¬cult™ to understand. When asked if other members of the jury
found it di¬cult, a quarter were not sure but 65 per cent thought they did not.141
But saying that they understood does not mean that they did understand. Even
less does it show whether the jury followed the judge™s direction on the law.142
In ˜Summing up “ a judge™s perspective™, Madge J argued for greater use of
written directions to juries on the law.143 In his experience, giving the jury a
written copy of his instructions on the law had positive results: fewer requests

For an empirical study of what magistrates, ordinary citizens and professionals in the criminal
justice system understand by the admonition ˜only convict if you are sure of the defendant™s
guilt™ see M. Zander, ˜The Criminal Standard of Proof “ How Sure is Sure?™, 150 New Law
Journal, 20 October 2000, p. 1517.
www.jsboard.co.uk (Publications “ Bench books “ Specimen Directions).
Professor Griew suggested (at pp. 770“1) that in McVey, above, the Court of Appeal quashed a
conviction of a plainly guilty person because the judge™s direction on the ingredients of the
o¬ence was insu¬cient even though the missing words were unnecessary to the jury™s
Notably R.P. Charrow and V.R. Charrow, ˜Making Legal Language Understandable: A
Psycholinguistic Study of Jury Instructions™, Columbia Law Review, 1979, p. 1306; W.W.
Schwarzer, ˜Communicating with Juries: Problems and Remedies™, 69 California Law Review,
1981, p. 731. M. Zander and P. Henderson, n. 69 above, sections 8.6.2“3.
For references to other literature on whether jurors understand judges™ directions on the law
see Evidence in Criminal Proceedings: Previous Misconduct of a Defendant (Law Commission
consultation paper 141, 1996) pp. 127“8. Criminal Law Review, 2006, pp. 817“27.
521 Respective roles of judge and jury

from juries for further instructions, quicker jury decisions, more conviction
and fewer hung juries. He also argued for less judicial comment on the facts.

Summing up on the facts
The job of the judge in summing up the facts according to the Court of Appeal
is to ˜state matters impartially, clearly and logically™.144 His task therefore is to
remind them of the evidence and to marshal it in a convenient way which is fair
to both sides.145
To what extent can he go beyond this to comment on the evidence and
thereby seek to in¬‚uence the jury™s decision? There is no doubt that English
judges do this. (It is famously said that Sergeant Sullivan at the end of an Old
Bailey trial invited the judge to ask the jury whether they found for the defen-
dant or his Lordship.) In the notorious case of the ˜Birmingham Six™ whose con-
victions for involvement in IRA terrorist bombings were ultimately quashed,
the trial judge, Bridge J (as he then was), during a three-day summing up gave
innumerable indications that in his view the prosecution™s evidence was to be
preferred to that of the defence. Nor did he see anything wrong with leading the
jury to its conclusion. ˜I am of the opinion™, he told the jury, ˜that if a judge has
formed a clear view, it is much better to let the jury see that and say so and not
pretend to be a kind of Olympian detached observer™.
For an unusually strong summing up on the facts in a civil case see that of
Caul¬eld J in the libel action very unwisely brought in 1987 by Mr Je¬rey
Archer against the Daily Star, arising out of the allegation that he had visited a
prostitute.146 There was, the judge said, no accounting for the tastes of happily
married men and the fact that the jury would not expect Mr Archer, deputy
chairman of the Conservative Party, to visit a prostitute, did not mean that it
was not possible, but he asked the jury to consider whether it was probable. He
invited the jury to remember the evidence of Mrs Mary Archer. ˜Your vision of
her will probably never disappear. Has she elegance? Has she fragrance? Would
she have, apart from the strain of his trial, a radiance?™ Mr Archer, the judge
said, was a sportsman. ˜You may think he™s ¬t looking . . . Is he in need of cold,
unloving, rubber-insulated sex in a seedy hotel?™147
The danger that the judge will try to in¬‚uence the jury is the greater in cases
where the accused has previous convictions, because the judge knows of their exis-
tence from the outset. They are in his ¬le, allegedly so that he can steer counsel away

Berrada (1989) 91 Cr App Rep 131n. The trial judge had said that the defendant™s allegation
that police o¬cers had fabricated an interview was ˜really monstrous and wicked™ and ˜utterly
monstrous™. The court quashed the conviction.
It seems that in a short case in which the issues are simple it is not necessarily a fatal defect in
a summing up that the evidence has not been discussed: see Att¬eld (1961) 45 Cr App Rep
Four years later Lord Archer, as he had become, was imprisoned for perjury and perverting
the course of justice in the libel proceedings. The Times, 24 July 1987.
522 The jury

from questions which might otherwise lead to their becoming admissible. (This is
less signi¬cant now that previous convictions are much more widely admissible.)
In the United States the rule in most states is that the judge in a criminal trial
must express no opinion on the weight or credibility of the evidence of
witnesses or on the merits of either side.
Even if judges are prevented from commenting, they can of course still convey
to the jury their basic view through a mixture of in¬‚exion of the voice, ˜body lan-
guage™, timing and other signs which would not register in the o¬cial transcript.
(The Court of Appeal has, however, indicated that in extreme cases it would
allow evidence from those present in court as to ˜non-verbal communication™ by
the trial judge trying to persuade the jury.148) The only way to prevent such in¬‚u-
ence would be to prohibit the judge from summing up at all on the facts “ which
is the usual rule in the USA. There is little doubt that when the judge sums up
for a conviction the impression of impartial justice being done is diminished.149
In the Crown Court Study, the barristers, the CPS, the defence solicitor and
the judge were asked: ˜Did the summing up favour either side?™ In each category
of respondents the majority said No, but of those who said Yes, more in each
category thought it favoured the prosecution than the defence.150 Defendants
were asked about the fairness or otherwise of the judge during the trial and in
the summing up. The defendant thought the judge had been fair in the
summing up in 73 per cent of cases and unfair in 27 per cent.151
The Runciman Royal Commission thought it would be wrong to lay down a
rule as to how far the judge should sum up on the facts. Cases and circumstances
varied. Sometimes there would be no need for a summing up at all. The need to
be fair to both sides, the Commission said, required ˜that judges should be
wholly neutral in any comment that they make on the credibility of the evi-
dence™.152 It was ˜inappropriate for judges to intrude their own views of whether
or not a witness is to be believed™.153 Implementation of that recommendation
would presumably require some kind of ruling or Practice Statement by the
Lord Chief Justice. This has not happened.154
Lord Justice Auld took a radical approach to the judge™s summing up: ˜The
judge should no longer direct the jury on the law or sum up evidence in
the detail that he now does™.155 His basic recommendations were:156
• The judge should continue to remind the jury of the issues and, save in the
most simple cases, the evidence relevant to them, and should always give
the jury an adequate account of the defence, but he should do it in more
summary form than is now common.

Hircock [1969] 1 All ER 47.
See generally D. Wolchover, ˜Should Judges Sum Up on the Facts?™, Criminal Law Review,
1989, p. 781. M. Zander and P. Henderson, n. 69 above, Table 4.21, p. 130.
151 152 153
Ibid., p. 132. Ibid., p. 124, para. 23. Ibid.
For support for the Royal Commission™s view that it should happen see D. Osborne, ˜Breaking
New Ground™, Counsel, February 1998, pp. 16“17.
155 156
Auld, Ch. 11, para. 44, p. 533. Auld, pp. 537“8.
523 Respective roles of judge and jury

• The judge should devise and put to the jury a series of written factual ques-
tions, the answers to which could logically lead only to a verdict of guilty or
not guilty; the questions should correspond with those in the updated Case
and Issues Summary, supplemented as necessary in a separate written list pre-
pared for the purpose, and each question should be tailored to the law as the
judge knows it to be and to the issues and evidence in the case.
• The judge, where he considers it appropriate, should be permitted to require
a jury to answer publicly each of his questions and to declare a verdict in
accordance with those answers.
• So far as possible, the judge should not direct the jury on the law, save by
implication in the questions of fact that he puts to them for decision.
In Auld™s view, ˜simpli¬cation of the way in which judges direct and sum up to
juries was essential for the future well-being of our system of trial by judge and
jury™.157 The Court of Appeal bore ultimate responsibility for the elaborate and
complex structure now enshrined in the Judicial Studies Board™s specimen
directions. What was needed, he suggested, was ˜a fundamental and practical
review of the structure and necessary content of a summing up with a view to
shedding rather than incorporating the law and to framing simple factual ques-
tions that take it into account™. Perhaps, he thought, ˜a body drawn from the
judiciary and the Judicial Studies Board could be given a blank sheet of paper
and charged with the task™.
The Auld report was published in October 2001. Five years later, there was no
sign that such a body would be established or that Auld™s radical proposals
regarding the judge™s summing up would be adopted.

Directing an acquittal
A high proportion of acquittals are directed by the judge. As has been seen, there
are two forms of such acquittal: (1) where the prosecution enter no evidence at
all (called an ˜ordered acquittal™) and (2) after a submission by the defence at
some stage after the case has begun and usually at the end of the prosecution™s
case that there is no case to answer (a ˜directed acquittal™).
The withdrawal of a case from the jury poses a delicate problem. Can the
judge decline to put the case to the jury if he thinks that the prosecution™s case
is merely weak? The Court of Appeal considered this question in 1981 in R v.
Galbraith.158 Lord Lane, the Chief Justice, said there were two schools of
thought. One was that the judge should stop the case if in his view it would be
unsafe or unsatisfactory to convict.159 The other was that the judge should only
stop the case if there was no evidence on which a jury properly directed could

Ibid, para. 49, p. 535. Auld did not mention the ¬ndings of the Crown Court Study (sect.
8.6.2) suggesting that the jury may not in fact have as great di¬culties with the summing up
as some believe. [1981] 1 WLR 1039.
See for instance Mans¬eld (1977) 65 Cr App Rep 276.
524 The jury

properly convict. Before the Criminal Appeal Act 1966 the second test had been
applied, but under the 1966 Act the Court of Appeal was required to quash a
conviction where it found that under the circumstances it was unsafe or unsat-
isfactory (see p. 693 below). Since then a practice had grown up of asking the
trial judge to take a view as to whether conviction would be safe by submitting
that there was no case. This involved the judge invading the province of the jury.
It invited the trial judge to consider the weight and the reliability of the prose-
cution™s evidence “ precisely the issues that had to be considered by the jury.
In Galbraith Lord Lane answered the question in this way:
How then should the judge approach a submission of ˜no case™? (1) If there is no
evidence that the crime alleged has been committed by the defendant, there is
no di¬culty. The judge will of course stop the case. (2) The di¬culty arises
where there is some evidence but it is of a tenuous character, for example
because of inherent weakness or vagueness or because it is inconsistent with
other evidence. (a) Where the judge comes to the conclusion that the prosecu-
tion evidence, taken at the highest, is such that a jury properly directed could
not properly convict upon it, it is his duty, upon a submission being made, to
stop the case. (b) Where however the prosecution evidence is such that its
strength or weakness depends on the view to be taken of a witness™s reliability,
or other matters which are generally speaking within the province of the jury
and where on one possible view of the facts there is evidence upon which a jury
could properly come to the conclusion that the defendant is guilty, then the
judge should allow the matter to be tried by the jury. It follows that we think the
second of the two schools of thought is to be preferred.160
Even where the judge has rejected a submission of no case he may still direct the
jury to acquit if in light of the developing defence case he subsequently comes
to the conclusion that no reasonable jury properly directed could convict,161 but
such a power must be exercised very sparingly.
The Runciman Royal Commission Report recommended that Galbraith
should be reversed so that a judge could stop a case if he or she took the view
that the prosecution™s evidence was demonstrably unsafe or unsatisfactory or
too weak to be allowed to go to the jury.162 This recommendation has not been
acted upon.
Sometimes the judge, whilst not going so far as to direct the jury to acquit,
makes it very clear in his summing up that he thinks an acquittal is the right
result. He sums up strongly for an acquittal. There is nothing to prevent this even
if the judge goes beyond the proper limits. The matter is unlikely to become the
subject of comment from the Court of Appeal since the prosecution has no right

At p. 1042. For comment and discussion see R. Pattenden, ˜The Submission of No Case “
Some Recent Developments™, Criminal Law Review, 1982, p. 558; D. Wolchover, ˜Stopping the
Trial in Suspect Cases™, 132 New Law Journal, 1982, p. 527. For the e¬ect of this ruling in
magistrates™ courts see N. Yell, ˜Submissions of “No Case to Answer”™, Justice of the Peace,
1981, p. 406. Brown (Davina) [2001] Crim LR 675, CA.
Runciman, p. 59, para. 41.
525 Respective roles of judge and jury

of appeal against an acquittal (other than on a point of law taken by the Attorney
General, the outcome of which does not a¬ect the defendant “ see p. 667 below).
(For a classic instance of the judge ˜summing up for an acquittal™ see the
summing up of Mr Justice Cantley in the Jeremy Thorpe case.163)

Directing a conviction
Views have di¬ered as to whether it is ever legitimate for the judge to direct the
jury to convict. Lord Devlin thought it to be unconstitutional.164 There is no
doubt that the judge must leave to the jury any issue that has to be decided by
them. In Leer165 the Court of Appeal considered a direction to convict where the
accused had been charged with possessing an o¬ensive weapon after being
found with a ¬shing knife. The judge ruled that his answers to police question-
ing as to why he had the knife did not amount to a reasonable excuse and that
he therefore had no defence to put forward and he directed them to convict. The
Court of Appeal quashed the conviction because the judge should have left the
issue to the jury. It would have been surprising if the jury had decided to acquit
but such a decision on the evidence would not have been perverse.166
However, what if an acquittal would be perverse? Can the judge direct a con-
viction then? In DPP v. Stonehouse167 the House of Lords by a majority of three
to two held that he could not. Lord Salmon said there was a di¬erence between
directing the jury to acquit or to convict. If there was no evidence on which they
could reasonably convict, he should direct an acquittal. This rule had been
established a long time ago to protect the accused against being wrongly con-
victed. ˜But there is no converse rule . . . If the judge is satis¬ed that on the evi-
dence, the jury would not be justi¬ed in acquitting the accused and indeed that
it would be perverse of them to do so, he has no power to pre-empt the jury™s
verdict by directing them to convict. The jury alone has the right to decide that
the accused is guilty™.168 Lord Salmon did accept that it would be perfectly in
order for the judge to sum up to the jury ˜in such a way as to make it plain that
he considers the accused is guilty and should be convicted™.169 Lord Edmund-
Davies said there was an unfortunate tendency in the courts these days to with-
draw issues from the jury which were properly theirs. Whether this sprang from
distrust of the jury™s capacity ˜or from excessive zeal in seeking to simplify their
task, it needs careful watching™.170 The judge could give a strong lead to the jury
but he should not direct them to convict.
The view expressed in Stonehouse was con¬rmed by a unanimous House of
Lords in Wang, allowing an appeal from the Court of Appeal.171 The decision,
given in a single opinion, was a ringing endorsement of the dominant role of
the jury:
163 164
Daily Telegraph, 19 and 20 June 1979. Trial by Jury, 1966, p. 84 and Appendix II.
165 166
[1982] Crim LR 310. See also R v. Clemo [1973] RTR 176n.
167 168 169 170
[1978] AC 55. At 80. Ibid. At 88.
[2005] UKHL 9, [2005] 1 All ER 782 and see the commentary in [2005] Crim LR 646“8.
526 The jury

In England and Wales it has been possible to assume, in the light of experience and
with a large measure of con¬dence, that jurors will almost invariably approach
their important task with a degree of conscientiousness commensurate with what
is at stake and a ready willingness to do their best to follow the trial judge™s direc-
tions. If there were to be a signi¬cant problem, no doubt the role of the jury would
call for legislative scrutiny. As it is, however, the acquittals of such high pro¬le
defendants as Ponting,172 Randle and Pottle173 have been quite as much welcomed
as resented by the public, which over many centuries has adhered tenaciously to its
historic choice that decisions on the guilt of defendants charged with serious crime
should rest with a jury of lay people, randomly selected, and not with professional
judges . . . We would accordingly allow the appeal, quash the appellant™s convic-
tion and answer the certi¬ed question by saying that there are no circumstances in
which a judge is entitled to direct a jury to return a verdict of guilt [16, 18].174
In June 2006 the Court of Appeal put a gloss on Lord Bingham™s dictum in Wang
that there were ˜no circumstances in which a judge is entitled to direct a jury to
return a verdict of guilty™.175 Lord Justice Tuckey said this did not mean that in
every such case the conviction must be held to be unsafe.176 It depended on
whether the jury was given an opportunity to consider its verdict. In Kelleher177
the judge had directed the jury that ˜there can only be one verdict in this case
and that is one of guilty™. The jury retired and brought back a guilty verdict. The
Court of Appeal dismissed their appeals against conviction. In these two cases
the appellants raised no defence in law and in each case the judge had directed
the jury to convict which they did without retiring. The Court of Appeal
quashed the convictions as the question of the defendant™s guilt had been com-
pletely taken away from the jury. (One has to say that this distinction based on
whether the jury actually retires seems unconvincing.)
On the historic role of the jury see T.A. Green™s magisterial Verdict According
to Conscience: Perspectives on the English Criminal Trial, 1200“1800 (Chicago

[1985] Crim LR 318. Clive Ponting, a senior civil servant, was prosecuted under the O¬cial
Secrets Act for leaking to an MP information about the sinking of the Argentinian ship The
General Belgrano during the Falklands War. The judge thought Mr Ponting had no defence.
He therefore intended to direct the jury to convict but was dissuaded from doing so by
counsel for the prosecution who drew the judge™s attention to what he termed ˜recent
authorities™. The judge said that although all the elements of the o¬ence had been made out
and there was no defence, he told the jury that they were at liberty to bring in whatever verdict
they considered right. The jury acquitted “ presumably because they felt that Mr Ponting had
acted honourably and perhaps correctly (ed.).
[1991] 1 WLR 1087. The defendants were tried at the Old Bailey for helping the spy George
Blake to escape from prison twenty-¬ve years earlier. They were prosecuted after they wrote a
book about their exploit. The trial judge ruled that they had no defence to the charge. In his
closing speech to the jury Pat Pottle said: ˜We do not deny the things we are accused of doing.
Not only do we not deny it, we say it was the right thing to do™. The jury acquitted both men
(The Times, 5 July 1991) (ed.).
In Caley-Knowles and Jones [2006] EWCA Crim 1611, [2007] Crim LR 61, 4 October the
Court of Appeal said that Lord Bingham™s statement in Wang that there were ˜no
circumstances in which a judge is entitled to direct a jury to return a verdict of guilty™ meant
175 176
in every such case. At [17]. Caley-Knowles and Jones, n. 174 above.
[2003] EWCA Crim 3525.
527 Respective roles of judge and jury

University Press, 1985). Green shows that part of the role of the jury was to mit-
igate the rigour and harshness of the criminal law and its penalties by acquit-
ting guilty defendants, not just in the occasional case but on a massive scale.

Should the jury be prohibited from returning a perverse verdict?
In one of the most controversial passages in his report, Lord Justice Auld rec-
ommended ˜that the law should be declared, by statute if need be, that juries
have no right to acquit defendants in de¬ance of the law or in disregard of the
evidence, and that judges and advocates should conduct criminal cases accord-
ingly™.178 In his view, the ability of juries to acquit in de¬ance of the law and in
disregard of their oaths was ˜a blatant a¬ront to the legal process and the main
purpose of the criminal justice system “ the control of crime “ of which they are
so important a part™.179 The jury™s role was ˜to ¬nd the facts and, applying the
law to those facts, to determine guilt or no™.180 They were not there ˜to substi-
tute their view of the propriety of the law for that of Parliament or its enforce-
ment for that of the appointed Executive, still less on what may be irrational,
secret and unchallengeable grounds™.181
The writer criticised this proposal in his response to the Auld report:182
I regard this proposal as wholly unacceptable “ a serious misreading of the func-
tion of the jury. The right to return a perverse verdict in de¬ance of the law or the
evidence is an important safeguard against unjust laws, oppressive prosecutions or
harsh sentences. In former centuries juries notoriously de¬ed the law to save defen-
dants from the gallows. In modern times the power is used, sometimes to general
acclaim, sometimes to general annoyance, usually one imagines to some of each.
Auld quotes E.P. Thompson™s eloquent passage in describing the function of the
The English common law rests upon a bargain between the Law and the
People. The jury box is where people come into the court; the judge watches
them and the jury watches back. A jury is the place where the bargain is
struck. The jury attends in judgment, not only upon the accused, but also
upon the justice and humanity of the law . . . [Writing by Candlelight, 1980]
This exactly captures the position, which I would say is part of the unwritten
constitution of this country. Auld says that he regards the ability of juries to
acquit and to convict in de¬ance of the law and in disregard of their oaths, as a
˜blatant a¬ront to the legal process and the main purpose of the criminal justice
system “ the control of crime “ of which they are so important a part™. I believe
that this statement, perhaps the least attractive sentence in the whole report,
re¬‚ects deep distrust of the jury. It is based I believe on an authoritarian attitude
that disregards history and reveals a grievously misjudged sense of the proper
balance of the criminal justice system.

178 179 180 181
Auld, Ch. 5, para. 107, p. 176. Ibid. Ibid. Ibid.
Accessible on www.criminal-courts-review.org.uk “ Comments received, pp. 18“19.
528 The jury

In the Introduction to his Report183 Sir Robin quotes, with apparent approval,
from the concluding sentence in my Dissent to the Report of the Runciman
Royal Commission, ˜the integrity of the criminal justice system is a higher objec-
tive than the conviction of any individual™. But the concern for justice and for
the integrity of the system is too important to be entrusted solely to the judges.
The jury have a role in that regard too.
The Runciman Royal Commission dealt with this issue in a short paragraph “
which was not mentioned by Auld:
Although juries are under a solemn duty to return a verdict in accordance with
the evidence, they do from time to time perversely return a verdict contrary
to the evidence. Until there is research on jury deliberations it is impossible to
say con¬dently why this happens, but it is plausible to suppose that it is
because the jury has taken an unfavourable view of the prosecution or of the
law under which it is brought or the likely penalty. We do not, however, think
that these cases justify the introduction of a right of appeal against acquittal.184
I cannot imagine that on a constitutional matter of this importance any Govern-
ment would prefer the view of an individual judge, however distinguished, to the
unanimous contrary view of a recent Royal Commission. I believe that the present
system provides the right balance in telling the jury that they must decide the case
in light of the law and the evidence but allowing them to ignore either or both if
they believe that to be the right course. We have lived with that system for hun-
dreds of years. I believe that there is no acceptable reason to consider changing it.
The Government stated in its White Paper that it did not accept Lord Justice
Auld™s recommendation. (˜Nor do we intend to legislate to prevent juries from
returning verdicts regarded as perverse where the verdict ¬‚ies in the face of the
evidence, as has happened very occasionally™.185)

Asking the jury questions
The jury does not give reasons for its decisions. In a criminal case it simply says
˜guilty™ or ˜not guilty™. In a civil case it ¬nds for the plainti¬ or defendant and, if
for the plainti¬, it may have to determine the damages. But in a decision in
February 1999 the Court of Appeal created an exception to the general rule. It
held that where there was more than one possible basis for a verdict of guilty of
manslaughter ˜it might be convenient and desirable™ for the judge to invite the
jury to indicate the basis on which they returned that verdict. The purpose of
such an inquiry would be to assist the judge with regard to sentence. Lord Justice
Rose said that in summing up, the judge might hand to the jury written ques-
tions identifying the di¬erent possible verdicts as between murder and
manslaughter and also as to the reasons for manslaughter. Alternatively, after a
manslaughter verdict was returned, a judge might ask the jury what was the basis

183 184
Page 10, para. 8. Report, p. 177, para. 75.
White Paper, Justice for All, Cm. 5563, July 2001, para. 4.50.
529 Respective roles of judge and jury

of that verdict “ provided that he had warned the jury in his summing up of his
intention to ask that question. But there was no obligation on the jury to answer.
Nor was there any requirement of unanimity as to the reasons for the verdict.186
This proposal seems completely novel and highly problematic. So far as is
known, the suggestion has not been taken up.

Is the jury™s unreasoned verdict compatible with the European
Convention on Human Rights?
Article 6 of the ECHR states that one of the incidents of a fair trial is a public
pronouncement of a reasoned decision. If this applies to jury trial, the
Strasbourg Court could hold jury trial with its inscrutable verdict to be contrary
to the Convention. Addressing this question in his report, Lord Justice Auld
said: ˜For a number of reasons, I incline to the view of a number of eminent
British commentators187 that the Strasbourg Court, in taking account of the way
in which our system of jury trial works as a whole, would not consider our
juries™ unreasoned verdicts to breach Article 6™.188
He instanced the following reasons:
• The Strasbourg case law was not precise about the content of reasons required
to satisfy the fair trial test.
• The test was not exacting. As well as allowing for di¬erent national traditions,
the court had stressed that the general duty to give reasons did not require
detailed answers.
• Courts were not required by the Strasbourg case law to indicate the evidence
they accept and why.
• The Strasbourg Court had ruled that the publicly unreasoned decision of a
Danish jury was not contrary to the Convention.189
• In Condron v. United Kingdom190 the court had said: ˜The fact that the issue of
the applicant™s silence was left to a jury cannot of itself be incompatible with
the requirement of a fair trial™.191 This, Auld suggested, showed that the court
was prepared to accept the jury™s verdict as the ¬nal word in a judgment of
which the summing up furnished the overt reasoning process.
• In considering the fairness of the trial the Strasbourg Court looked at the trial
and the appeal process together. The Court of Appeal did have a limited capac-
ity to quash a conviction if it considers that it was contrary to the evidence.192

R v. Jones (Douglas) (1999) Times, 17 February.
He cited Professor John Spencer in his advice to the Auld Review and Harris, O™Boyle and
Warbrick, The Law of the European Convention on Human Rights (1995) p. 215.
Auld, Ch. 5, para. 92, p. 170.
Saric v. Denmark, application no. 31913/96, decision of 2 February 1999.
190 191
(2000) 31 EHRR 1. At [57].
However, Auld cited the court™s statement in Condron v. United Kingdom (above at [46]) that
jury verdicts in England ˜are not accompanied by reasons which are amenable to review on
530 The jury

• There was no general continental consensus as to what is meant by the rea-
soning (˜motivation™) of a judicial decision. In France, for instance, it could
mean no more than an indication of the legal principles applied by the court.
However, Lord Justice Auld did consider that ˜the time has come™ for the trial
judge to give the jury a series of written factual questions leading logically only
to a verdict of guilty or not guilty.193

11. Majority jury verdicts
In Scotland, since time immemorial, there has been a majority verdict based on
a bare majority of eight or more out of the ¬fteen who sit on a Scottish jury.
Historically in England, however, the jury™s decision had to be unanimous “
though the reality of unanimity must sometimes have been questionable. In
some cases dissenters would probably have ˜given in™ rather than have a hung
jury or just to bring the proceedings to a speedy conclusion.194
In 1967, the then Home Secretary, Mr Roy Jenkins, introduced proposals in
the Criminal Justice Bill to permit a majority verdict of not less than ten out of
twelve. The reason he gave was the spate of recent ˜jury nobbling™ cases “ though
the evidence for this was thin. The total proportion of jury disagreements
resulting in a retrial appeared to be about 4 per cent and few of these, presum-
ably, would have been due to any form of tampering with the jury.
The proposal provoked great controversy at the time, but in the interim it seems
to have become accepted.195 Lord Justice Auld dealt with this topic in nine lines.
His Review, he said, had ˜produced little support for change either in the levels of
the required majorities or for reversion to unanimity in all cases or for any form
of intermediate verdict, such as that of “not proven” in use in Scotland™.196
The proportion of guilty verdicts by majority is just over a ¬fth.197 There are
no equivalent o¬cial ¬gures for the proportion of acquittals by a majority, since
the jury is not permitted to reveal that an acquittal was by a majority, for fear
that it would be treated as a second-class acquittal.198 (In the Crown Court Study
it was possible to establish from the jury questionnaires that the proportion of
acquittals by a majority was exactly the same as convictions by a majority.199)
The court is not supposed to consider the possibility of a majority verdict
until at least two hours and ten minutes have elapsed.200 In a complex case the
judge will wait much longer than that.
Auld, Ch. 5, para. 97, p. 172.
For the particular danger of this happening on a Friday afternoon see P. Darbyshire, ˜Notes of
a Lawyer Juror™, 150 New Law Journal, 14 September 1990, p. 1264 at 6“7.
See, however, G. Maher, ˜Jury Verdicts and the Presumption of Innocence™, Legal Studies,
1983, p. 146 for a powerful argument that majority verdicts are inconsistent with the
requirement that proof of guilt be beyond a reasonable doubt.
Auld, Ch. 5, para. 75, p. 164.
See the annual Judicial Statistics, Table 6.11. In 2004 it was 23 per cent; in 2005 it was 21 per cent.
198 199
Criminal Justice Act 1967, s. 13(2). M. Zander and P. Henderson, n. 69 above, p. 162.
Practice Direction [2002] 3 All ER 936, para. 46.
531 Majority jury verdicts

In R v. Reynolds201 the Court of Appeal Criminal Division quashed a con-
viction for theft because the foreman of the jury stated that there was a major-
ity of ten in favour of conviction but he did not also state that there were two
members of the jury who disagreed. The court held that the provisions of
s. 17(3) of the 1967 Act were mandatory in stating that the court ˜shall not
accept [a majority verdict] unless the foreman of the jury has stated in open
court the number of jurors who respectively agreed to and dissented from the
verdict™. In R v. Pigg202 the House of Lords overruled Reynolds. The Law Lords
held unanimously that, although it was a mandatory requirement that the
number who agreed on conviction and the number who dissented must be
made known, the precise form of words used was not an essential part of that
requirement. It was enough if the words used by the foreman of the jury and
the clerk of the court made it clear to an ordinary person how the jury was
divided. If the foreman said that ten agreed to convict it could be inferred that
two dissented!
However, what if the jury is completely deadlocked? For many years the
judge was permitted in that situation to give what was known as a Walhein
direction, approved in the case of that name.203 In that case, the jury told the
judge that they were having di¬culty in reaching a unanimous verdict. (At that
date there was no such thing as a majority verdict.) The judge then directed
You are a body of twelve men. Each of you has taken an oath to return a true
verdict according to the evidence; but, of course, you have a duty not as indi-
viduals, but collectively. No one must be false to that oath; but in order to return
a collective verdict, the verdict of you all, there must necessarily be argument and
a certain amount of give and take and adjustment of views within the scope of
the oath you have taken; and it makes for great public inconvenience and
expense if jurors cannot agree owing to the unwillingness of one of their number
to listen to the arguments of the rest. Having said that, I can say no more.
This direction seemed to condone pressure on the dissenting minority to fall
into line. Since the introduction of majority verdicts in 1967 it has seemed inap-
propriate. In Watson204 the Court of Appeal approved a new direction to replace
that in Walhein:
Each of you has taken an oath to return a true verdict according to the evidence.
No one must be false to that oath, but you have a duty not only as individuals
but collectively. That is the strength of the jury system. Each of you takes into
the jury box with you your individual experience and wisdom. You do that by
giving your views and listening to the views of the others. There must necessar-
ily be discussion, argument and give and take within the scope of your oath.
That is the way in which agreement is reached. If, unhappily, [ten of] you cannot
reach agreement you must say so. It is a matter for the discretion of the Judge as

201 202 203
[1981] 3 All ER 849. [1983] 1 All ER 56. (1952) 36 Cr App Rep 167.
[1988] QB 690.
532 The jury

to whether he gives that direction at all and if so, at what stage of the trial. There
will usually be no need to do so.205

12. Retrials on jury disagreement
When the jury disagrees and cannot reach a verdict, the prosecution are enti-
tled to start afresh. It is a matter of discretion over which the court has no
control “ though occasionally the judge remonstrates with the prosecutor about
the desirability of pursuing a particular case. (It has been argued that the pros-
ecution should have to ask leave and that the court should take into account the
same factors regarded as relevant when the Court of Appeal considers whether
to order a retrial on quashing a conviction.206)
There are no regular statistics about the extent to which retrials occur as a
result of jury disagreements. In 1981, according to a Home O¬ce Research Unit
paper, there were some 370 retrials due to this cause “ about 1.5 per cent of the
25,000 or so contested cases in the Crown Court that year.207

13. Will the courts consider what happened in the jury room?
What happens if it is suggested that one or more jurors has behaved improperly?
If the matter is raised on appeal the ancient common law rule is clear.208 Appeal
courts will not receive evidence as to what transpired in the jury room. In Boston
v. Bagshaw & Sons209 all twelve members of the jury swore a¬davits that they had
given the opposite result of what they intended. The Court of Appeal refused to
change the decision. Giving the judgment of the court Lord Denning said:
To my mind it is settled as well as anything can be that it is not open to the court
to receive any such evidence as this. Once a jury has given their verdict, it is
accepted by the judge, and they have been discharged, they are not at liberty to
say they meant something di¬erent . . .
The reasons are twofold: ¬rst, to secure the ¬nality of decisions arrived at by
the jury; secondly, to protect the jury themselves and to prevent them being
exposed to pressure or inducement to explain or alter their views.
The rule is the same in criminal cases. In Thompson210 the Court of Criminal
Appeal refused to inquire into an allegation that the jury had been moving

See further M.J. Reville, ˜Directing the Hung Lamp of Freedom™, Law Society™s Gazette, 26
October 1988, p. 19; P. Robertshaw, ˜Exhorting Hung Juries™, Criminal Law Review, 1997,
p. 805. J. Hall, ˜Hung Juries and Retrials™, Archbold News, 27 June 2001, p. 6.
See S. Butler, Acquittal Rates (Home O¬ce Research and Planning Unit Paper No. 16, 1983)
p. 7. On the juror™s duty to reach a decision see R v. Schot and Barclay [1997] Crim LR 827. In
Robertshaw™s article, n. 205 above, it was stated that in 1991 there were ninety hung trials in
the whole country “ of which no fewer than seventy-nine (88 per cent) occurred in one court,
the Inner London Crown Court!
In 1785 in Vaise v. Delavel 1 TR 11, 99 ER 944 Lord Mans¬eld said that the court could not
receive an a¬davit from a juror as to the nature of the jury™s deliberations.
209 210
[1966] 1 WLR 1135n. [1962] 1 All ER 65.
533 Will the courts consider what happened in the jury room?

toward an acquittal until one member of the jury had produced a list of the
defendant™s previous convictions upon which they convicted. Lord Chief Justice
Parker gave the judgment of the court:
. . . There is absolutely no doubt that information as to the prisoner™s previous
convictions must be kept from a jury, and if what was said to have happened did
happen it would have been highly improper. This court is now asked to inquire
into the matter, and to adjourn the case in order to see whether the alleged state-
ment by the juryman can be supported by some statement or a¬davit made by
him. The court has come to the conclusion that it is perfectly idle to adjourn the
case for that purpose because the court is quite satis¬ed that they would have no
right at all to inquire what did occur in the jury room. It has for long been a rule
of practice, based on public policy, that the court should not inquire, by taking
evidence from jurymen, what did occur in either the jury box or the jury room.
The common law rule was endorsed in 2004 by a four to one decision of the
House of Lords in Connor and Mirza.211 In each case a letter had been written
to the trial court by a juror after majority verdicts of guilty. In one the letter
alleged that other jurors had been racially prejudiced against the defendant; in
the second the juror alleged that other jurors had rushed to ¬nish the case
against the two defendants with inappropriate speed. The Court of Appeal
applied the common law rule. It also held that the Contempt of Court Act 1981,
s. 8 prohibited complaint about or inquiry into such allegations. The House of
Lords, Lord Steyn dissenting, held that the principle of the con¬dentiality of a
jury™s deliberations underpinned the independence and impartiality of the jury
as a whole. The rule prohibiting inquiry applied even if it appeared that egre-
gious impropriety had occurred. This was because of the common law rule not
because of the Contempt of Court Act. Section 8 of the 1981 Act had no appli-
cation to the court that had responsibility for handling contempt.
The Law Lords conceded that there were some exceptions. The distinction is
not always easy to draw. In Ellis v. Deheer212 evidence was received that some of
the jurors had been unable to hear what the foreman said in giving the jury™s
verdict and that they were in disagreement with it. In Ras Behari Lal213 the Privy
Council upheld an inquiry held as to whether a juror had been able to under-
stand English. In Spencer214 the House of Lords quashed convictions of nurses
who had been found guilty of violence against patients at Rampton Hospital
when someone who had been removed from the jury because of possible bias
against the defendants had given lifts to and from the trial to three of the jurors.
In Young215 the Court of Appeal ordered a retrial after it emerged that during an
overnight adjournment in a hotel members of the jury had used a ouija board
to consult the deceased in a s©ance! (In Connor and Mirza the Law Lords said
that if the evidence was that the jurors had used a ouija board in the jury room

[2004] UKHL 2, [2004] 1 All ER 925, [2004] Crim LR 1041. The speeches take ¬fty-¬ve pages
212 213
in the law report. [1922] 2 KB 113. (1933) 50 TLR 1, PC.
214 215
[1986] 2 All ER 928. [1995] QB 324.
534 The jury

or that they had decided the case by drawing lots or tossing a coin, that evidence
could be admitted. There would in e¬ect have been no deliberations.) In K216
the Court of Appeal quashed a conviction when the day after the case was ¬n-
ished the jury baili¬ discovered in the jury room material that had been down-
loaded from the internet.
The House of Lords in Connor and Mirza said that if a juror had concerns
about the improper conduct of fellow jurors he should bring such concerns to
the attention of the trial judge.217 That, the Law Lords said, could extend to
other persons properly concerned, such as the jury baili¬ or the clerk of the
court. It would also be permissible to send a sealed letter to the defendant™s
lawyers or even to a Citizens™ Advice Bureau if it was to be passed on unopened
to the proper authorities. It was not permissible, however, to write as had hap-
pened in that case to the defendant™s mother.218
The judge then has a variety of options. He would normally discuss these
with counsel for both sides. He may warn the jury about the matter and exhort
them to behave properly. He has the option of discharging the jury. Whether he
can and should conduct an inquiry as to what happened in the jury room was
considered by the House of Lords in Smith.219 While the jury in that case were
considering their verdict the judge had received a letter from one of the jurors
alleging that some jurors had been badgering, coercing and even intimidating
other jurors into changing their verdict. The judge gave the jury a further direc-
tion which included exhortation not to be bullied or cajoled into a verdict with
which they did not agree. The Court of Appeal certi¬ed for the House of Lords
the question what inquiry could properly be made as to what had transpired in
the jury room. The ¬ve Law Lords agreed that the convictions should be
quashed on the ground that the judge™s direction to the jury had been inade-
quate. It had not mentioned the jury™s duty to follow the judge™s directions on
the law, to adhere to the evidence without speculation and to decide on the
verdict without bargaining or pressure.
As to whether the trial judge could inquire into what happened in the jury
room, Lord Carswell (with whom the four other judges agreed) said that
although there might be some matters into which the judge can and should
inquire “ he gave as an example an allegation that a juror had used a mobile tele-
phone “ ˜I do not think that it is necessary or desirable to attempt to draw up a

[2005] EWCA 346, [2005] Crim LR 574.
Practice Direction (Criminal Proceedings: Consolidation) [2002] 1 WLR 2870 was amended so
as to make this clear to jurors “ [2004] 1 WLR 665 (˜Judges should ensure that the jury is
alerted to bring any concerns about fellow jurors to the attention of the judge at the time and
not wait until the case is concluded™ and the judge should instruct the jury that they had ˜a
duty to bring to the judge™s attention promptly, any behaviour among jurors or by others
a¬ecting the jurors, that causes concern™).
A-G v. Scotcher [2005] UKHL 36, [2005] 3 All ER 1 in which sentence of the juror for
contempt of court was upheld. It was no defence to the charge that the juror™s purpose in
writing to the mother had been to prevent a miscarriage of justice.
[2005] UKHL 12, [2005] 2 All ER 29, [2005] Crim LR 476.
535 Will the courts consider what happened in the jury room?

precise de¬nition of the situations in which it would be legitimate for the judge
to question jurors™.220 In the case in question questioning jurors would not have
been appropriate. It would be unlikely to reveal the truth and could have made
the situation worse.
In Smith the letter to the judge identi¬ed the juror who wrote it. In the earlier
case of Robinson221 the judge received an anonymous letter about what was
going on in the jury room. In order to discover whether it came from a member
of the jury the judge drafted ¬rst one and then a second questionnaire. Both
were given to the jurors to complete in order to discover whether they had
written the anonymous letter. They all denied it and without further inquiry
(rightly or wrongly) the judge concluded that it had been written by someone
not on the jury. Commenting in Smith, Lord Carswell simply said: ˜I do not con-
sider that issuing a questionnaire of the type used in R v. Robinson would have
been appropriate in this case™.222 The question of what inquiries by the trial
judge are permissible remains open.
The Strasbourg Court has given decisions on these matters that pull in
slightly di¬erent directions. In Gregory v. United Kingdom G complained that he
had not had a fair trial due to racial discrimination. While the jury were con-
sidering their verdict, a note had been passed to the trial judge saying that the
jury were showing racial overtones and that one juror should be excused. The
judge warned the jury to put aside any prejudice and to decide the case on the
evidence. Rejecting the complaint, the court said it was accepted that ˜it was not
possible under English law for the trial judge to question the jurors about the
circumstances that gave rise to the note™. It also acknowledged ˜that the rule gov-
erning the secrecy of jury deliberations is a crucial and legitimate feature of
English trial law which serves to reinforce the jury™s role as the ultimate arbiter
of fact and to guarantee open and frank deliberations among jurors™.223
In Sander v. United Kingdom224 the Strasbourg Court held by four to three
that the defendant had not had a fair trial because the trial judge had not taken
su¬ciently robust action to deal with an allegation by one juror of racial prej-
udice by two other jurors. He had told the jurors to search their consciences,
overnight and to let the court know if they could not decide the case on the
evidence. Having received such assurance the next day, he allowed the trial to
proceed. The court held that he had failed to provide su¬cient guarantees
to exclude any objectively justi¬ed or legitimate doubts about the jury™s
In 2005, in its consultation paper regarding research in the jury room (p. 513
above), the Government raised the question whether any further steps should
be taken to deal with the problem of impropriety in the jury room. It indicated,
however, that it did not itself consider that any changes were needed, other than
220 221 222
At [20]. [2002] EWCA Crim 2489. At [21].
223 224
(1988) 25 EHRR 577 at [44]. (2001) 31 EHRR 1003.
For critical comment see M. Zander, ˜The Complaining Juror™, 150 New Law Journal, 19 May
2000, p. 723.
536 The jury

informing jurors about what could constitute impropriety and of how they
could draw it to the attention of the trial judge. In November 2005 the
Government stated that its decision was to allow the common law to develop on
a case by case basis rather than introduce statutory changes.226
For discussion of these issues see K. Quinn, ˜Jury Bias and the European
Convention on Human Rights: a Well-Kept Secret?™ Criminal Law Review, 2004,
pp. 998“1014; P.R. Ferguson, ˜The Criminal Jury in England and Scotland: the
Con¬dentiality Principle and the Investigation of Impropriety™, 10 International
Journal of Evidence and Proof, 2006, pp. 180“211.

14. Publication of the secrets of the jury room
Jurors are told that they must not reveal anything that occurs in the jury room
either during the trial or after it has ¬nished. But what is the position if the press
publish details of jury deliberations? The question came up for decision after
the sensational Jeremy Thorpe case227 when the New Statesman in 1979 pub-
lished an interview with a member of the jury in which he gave details of the
jury discussions. Proceedings for contempt were instituted by the Attorney
General but, surprisingly, the Attorney General lost.228 The Divisional Court
held that disclosure of the secrets of the jury room could be contempt but it
depended on the circumstances. It would be contempt if disclosure tended to
imperil the ¬nality of jury verdicts or to a¬ect adversely the attitude of future
jurors or the quality of their deliberations. In this case, the court found, there
were no special features which made publication a contempt. There had been
no payment of money to the juror. The article did not suggest that anything
improper had occurred. In fact it showed that the jury had approached their
task in a sensible and responsible manner. There was no suggestion that the
article could have interfered with the administration of justice.
The media™s victory was shortlived. As has been seen, the Contempt of Court
Act 1981, s. 8 made it contempt ˜to obtain, disclose or solicit any particulars of
statements made, opinions expressed, arguments advanced or votes cast by
members of a jury in the course of their deliberation in any legal proceedings™.
It makes no di¬erence whether the case is identi¬ed or whether any payment is
made for such disclosure. The clause was introduced against the advice of the
Government by Lords Hutchinson and Wigoder and was supported by the
Criminal Bar Association, the Senate of the Four Inns of Court and the Lord
Chief Justice. The Lord Chancellor, Lord Hailsham, declared it to be ˜far too
draconian™. However, he was defeated in the House of Lords and the
Government did not seek to have the clause overturned when the Bill returned
to the House of Commons.
See www.dca.gov.uk “ Publications “ Consultation Papers “ January 2005.
The leader of the Liberal Party was charged with conspiracy to murder his alleged homosexual
lover. He was acquitted after a trial lasting thirty-one days.
A-G v. New Statesman and Nation Publishing Co Ltd [1980] 1 All ER 644.
537 Does the jury acquit too many defendants?

It is noteworthy that when the issue had been put to the Criminal Law
Revision Committee for consideration in 1967 it did not think there was any
need for legislation. The Committee said juries were reminded of their duty to
maintain secrecy by a notice on the walls of the jury room and that there seemed
to be few breaches of this understanding: ˜We are of opinion that secrecy has
been well maintained and that such breaches or attempts to break it as have
become known so far have not established a mischief so extensive or serious that
it calls for legislation and punishment™.229 It accepted that it was not then a crim-
inal o¬ence to disclose what had happened in the jury room though in certain
circumstances it might amount to contempt of court. But it did not think the
problem was su¬ciently serious to warrant legislation. For one thing it did not
think it right to make punishable the inevitable minor disclosures as people
spoke to their families and friends after the case about the experience of being
jurymen. Such disclosures, the Committee said, though they should not be
encouraged, few would regard as deserving of punishment. Under the Contempt
of Court Act 1981 such disclosures could theoretically be the subject of pro-
ceedings for contempt “ though in practice this is unthinkable.
See further J. Jaconelli, ˜Some Thoughts on Jury Secrecy™, Legal Studies,
March 1990, p. 91.
In A“G v. Associated Newspapers Ltd230 the House of Lords rejected an appeal
by the owners of the Mail on Sunday which had been ¬ned £60,000 for contempt
in publishing views of jurors in the Blue Arrow fraud case. The information had
been obtained not from the jurors directly but from transcripts of paid inter-
views purportedly carried out by a researcher. The House of Lords held that it
made no di¬erence whether the publication of what had transpired in the jury
room came directly from jurors or indirectly from others.

15. Does the jury acquit too many defendants?
The suggestion that too many guilty defendants are acquitted was powerfully
urged in the 1970s by Sir Robert Mark when he was Commissioner of the
Metropolitan Police. The best-publicised occasion for the expression of these
views was his 1973 Dimbleby Lecture on BBC1:

Sir Robert Mark, ˜Minority Verdict™, BBC, 1973, pp. 8“14
What we do know about trials in higher courts doesn™t justify any complacency.
Indeed, there is one fact I can mention which should be enough in itself to
demand some kind of enquiry. This is the rate of acquittals. Of all the people in
England and Wales who plead not guilty and are tried by jury, about half are
acquitted . . .
Every acquittal is a case in which either a guilty man has been allowed to go
free or an innocent citizen has been put to the trouble and expense of defending

Criminal Law Revision Committee, Tenth Report, Secrecy of the Jury Room, 1968.
[1994] 1 All ER 556.
538 The jury

himself. There must be some rate of failure. We can™t always expect to convict
the guilty or never to prosecute the innocent, but in my opinion a failure rate of
one in two is far too high. I doubt whether it would be tolerated in many other
kinds of activity, so I think it™s something that certainly needs looking into. In
the absence of any reliable research no one can say with any certainty why the
acquittal rate is so high. A fairly high number of acquittals are undoubtedly by
direction of the judges, as soon as they™ve heard the prosecution case. Since 1967
cases are no longer sifted e¬ectively by a Magistrate, and the higher courts are
cluttered up by cases which in my opinion should never have got there at all.
My own view is that the proportion of those acquittals relating to those whom
experienced police o¬cers believe to be guilty is too high to be acceptable . . .
I wouldn™t deny that sometimes common sense and humanity produce an
acquittal which could not be justi¬ed in law, but this kind of case is much rarer
than you might suppose. Much more frequent are the cases in which the defects
and uncertainties in the system are ruthlessly exploited by the knowledgeable
criminal and by his advisers.
Sir Robert Mark™s strictures on the high ˜failure rate™ in English trials need some
further explication and comment.
The acquittal rate is based on contested cases, whereas the majority of defen-
dants in the Crown Court plead guilty. The proportion pleading guilty in
Crown Courts currently is around three-¬fths. (In 2004, of defendants tried in
the Crown Court, 58 per cent pleaded guilty to all charges. In 2005 the propor-
tion was 60 per cent.231 The proportion is lower than it was. For many years it
was around 70 per cent.)
The proportion of defendants pleading not guilty to all counts who are
acquitted is around two-thirds. (In both 2004 and 2005 it was exactly 66 per
cent.232) This is a high proportion “ higher even than the ¬gure of which Sir
Robert Mark complained.
However, only a minority of acquittals are actually by a jury. In 2005, 57 per
cent of all acquittals of defendants pleading not guilty to all charges were
ordered by the judge when the prosecution o¬ered no evidence at the start of
the case and another 12 per cent were directed by the judge at the close of the
prosecution™s case on the ground that there was insu¬cient evidence to put to
the jury.233 The jury were therefore only responsible for the remaining 31 per
cent of the acquittals. This represents 7 per cent of the total number tried in the
Crown Court.234
The fact that some 7 per cent of those tried in the Crown Court are acquit-
ted by the jury could be mainly attributable to the fact that the burden of proof
is a high one. Even assuming that all those acquitted were guilty (a wholly

Judicial Statistics 2004,Cmnd 6565, Table 6.8, p 90; ibid., 2005, Cmnd 6799, Table 6.8, p 90.
Ibid, Table 6.9.
Ibid, Table 6.10. For a study of these cases see J. Baldwin, ˜Understanding Judge Ordered and
Directed Acquittals in the Crown Court™, Criminal Law Review, 1997, pp. 536“55.
Ibid, calculated from Tables 6.8 and 6.10.
539 Does the jury acquit too many defendants?

impermissible and unrealistic assumption), a signi¬cant number would rightly
be acquitted simply because the prosecution failed to prove its case beyond a
reasonable doubt. If proof must be beyond a reasonable doubt (and no one has
suggested otherwise), it is inevitable that a considerable number of guilty defen-
dants will be acquitted because the evidence of their guilt cannot be produced.
Contrary to what Sir Robert Mark asserted, professional criminals do not
appear to do better in the criminal justice system than others. Indeed, if anything
the reverse. Taking defendants with a prior record, the evidence is that they have
a statistically lower chance of an acquittal than defendants with no prior record.235
Moreover, the worse the record, the worse the chances of an acquittal.236 In
Baldwin and McConville™s 1979 study they got from the police details not only of
prior convictions but also of prior acquittals and of suspected involvement in
criminal activity. From this they built up a pro¬le of each defendant on a scale of
criminal professionalism. For this exercise there were close to 5,000 defendants in
the sample “ 2,406 in Birmingham and 2,292 in London, a total of 4,698. Of these,
2,265 (48 per cent) were de¬ned by the police information as ˜low™ on the crimi-
nal professionalism score, 1,448 (31 per cent) as ˜medium™, 647 (14 per cent) as
˜high™ and 227 (5 per cent) as ˜very high™. When these scores were compared with
acquittals and especially the ˜questionable acquittals™, it was found that ˜only a
minuscule proportion of all cases end in the questionable acquittal of any defen-
dant who, on the measures used here, could be regarded as a professional crimi-
nal. Indeed, of those scoring highly on the professionalism scale in each city, no
more than one in eighty was said to have been questionably acquitted™.237
The only evidence that provides any support for Sir Robert Mark™s thesis was
that of John Mack, who contrasted the careers of the top criminals in his area
of research (from names supplied by the police) with that of two other cate-
gories of lesser criminals. He called his three groups the Main Group, the Lesser
Group and the Small Fry. On average the Small Fry were convicted on 85 per
cent of charges brought against them, the Lesser Group on 80 per cent and the
Main Group on 75 per cent.238 This does show that the Main Group were some-
what more successful in avoiding charges than the others, but the di¬erence can
hardly be said to be great and the police success rate in getting convictions in
three-quarters of the charges brought against the top villains seems, if anything,
remarkably high. Moreover, as Mack showed, when the serious criminals are
convicted they tend to get longer sentences. Mack compared the time not spent
in prison from the age of seventeen for his three groups. The Small Fry spent on
average 83 per cent of their time not in prison compared with 70 per cent for
the Lesser Group and 74 per cent for the Main Group.239

This is shown by a number of studies including even one conducted by the Metropolitan
Police “ see M. Zander, 37 Modern Law Review, 1974, p. 39, Table 3; McCabe and Purves, The
Jury at Work (Blackwell, 1972) p. 39, Table 4; Metropolitan Police, Law Society™s Gazette, 1
March 1973, Table 1. M. Zander, 37 Modern Law Review, 1974, p. 41.
237 238
Jury Trials (1979) pp. 110“12. J. Mack, 39 Modern Law Review, 1976, p. 255.
Ibid, p. 252.
540 The jury

Nor was Sir Robert™s attack on crooked lawyers supported by the small
amount of evidence on this issue. In Baldwin and McConville™s study of 370
contested jury trials in Birmingham they interviewed the police o¬cers in the
cases about the reasons for the acquittals. They reported: ˜There was not a single
serious allegation of any practice which could possibly be described as
corrupt™.240 In another study the same two authors looked at 2,000 cases heard
in seven London Crown Courts in the light of the ˜solicitors blacklist™ main-
tained by Scotland Yard. (They had been sent a copy anonymously.) The ¬rms
on the list appeared on behalf of 223 defendants in the sample. Of these, 50 per
cent pleaded guilty “ a proportion that was slightly higher than for the rest of
the sample. Of those who pleaded not guilty, the acquittal rate was 53 per cent,
which was not very di¬erent from that of 47 per cent of the rest of the
sample. Of the defendants identi¬ed to the researchers by the police as serious
professional criminals, only ten out of seventy-two had employed ¬rms on the
Finally, the minority of cases that are contested are likely, by de¬nition, to be
the doubtful ones in which one might expect a fairly high acquittal rate. This
common sense view is supported by the evidence which shows that many not
guilty pleas are based on a defence that the accused lacked the necessary knowl-
edge or intent (mens rea) to be guilty of the o¬ence. It is perhaps not surprising
that, in such cases particularly, the jury (or magistrates) will interpret con¬‚ict-


. 22
( 34)