R v Gough (Robert) [1993] UKHL 1 (20 May 1993)

Regina

Gough (Appellant)
(On Appeal from the Court of Appeal (Criminal Division))

JUDGMENT

Die Jovis 20° Maii 1993

Upon Report from the Appellate Committee to whom was
referred the Cause Regina against Gough, That the Committee had
heard Counsel as well on Wednesday the 27th as on Thursday the
28th days of January last upon the Petition and Appeal of Robert
Brian Gough, praying that the matter of the Order set forth in
the Schedule thereto, namely an Order of the Court of Appeal
(Criminal Division) of the 2nd day of June 1992, might be
reviewed before Her Majesty the Queen in Her Court of Parliament
and that the said Order might be reversed, varied or altered or
that the Petitioner might have such other relief in the premises
as to Her Majesty the Queen in Her Court of Parliament might seem
meet; as upon the case of the Director of Public Prosecutions (on
behalf of Her Majesty) lodged in answer to the said Appeal; and
due consideration had this day of what was offered on either side
in this Cause:

It is Ordered and Adjudged, by the Lords Spiritual and
Temporal in the Court of Parliament of Her Majesty the Queen
assembled, That the said Order of the Court of Appeal (Criminal
Division) of the 2nd day of June 1992 complained of in the said
Appeal be, and the same is hereby, Affirmed and that the said
Petition and Appeal be, and the same is hereby, dismissed this
House: And it is further Ordered. That the certified guestion
be answered in accordance with the principles set out in the
speech of the Lord Goff of Chieveley.

Cler: Parliamentor:

Judgment: 20 May 1993

HOUSE OF LORDS

REGINA

GOUGH
(APPELLANT)

(ON APPEAL FROM THE COURT OF APPEAL)
(CRIMINAL DIVISION)

Lord Goff of Chieveley
Lord Ackner
Lord Mustill
Lord Slynn
Lord Woolf

LORD GOFF OF CHIEVELEY

On 25 April 1991, at Liverpool Crown Court, the appellant Robert
Brian Gough was convicted on an indictment containing a single count of
conspiracy to rob, and was sentenced to a term of 15 years imprisonment.

The indictment was based upon the commission of eight robberies in
Liverpool between 13 April 1989 and 6 March 1990. The first seven
robberies bore features of striking similarity. In all seven cases the premises
concerned were a betting shop; the robbery was committed by two masked
men, either at the beginning or at the end of the day; the men were armed,
one with a shotgun and the other with a knife; and the modus operandi was
similar. The prosecution contended that the first seven robberies had been
committed by the same two men, the appellant and his brother David Stephen
Gough. There was however insufficient evidence to link this brother with the
eighth robbery, and the evidence against him on the other seven was weak.
In the result, at the committal proceedings the prosecution applied for David
Stephen Gough to be discharged on the ground that there was insufficient
evidence against him; and at the trial the appellant was indicted on a single
count that between the relevant dates he conspired with David Stephen Gough
to commit the robberies.

– 1 –

On appeal, the appellant claimed that the learned judge should on his
own motion have required the prosecution to proceed on an indictment
containing eight substantive counts of robbery and not on the conspiracy
count. That submission was rejected by the Court of Appeal. There was
however another ground of appeal, which is the subject of the present appeal
to your Lordships’ House. This was that, by reason of the presence on the
jury of a lady who was David Stephen Gough’s next door neighbour, there
was a serious irregularity in the conduct of the trial and for that reason the
conviction of the appellant should be quashed. That submission was also
dismissed by the Court of Appeal, and the appellant now appeals to your
Lordships’ House from that part of the decision of the Court of Appeal, with
the leave of your Lordships’ House.

It was not until after the trial that it emerged that a member of the jury
was David Stephen Gough’s next door neighbour. In opening and in the
indictment, he was referred to as David Gough; but in closing speeches he
was referred to as David Stephen Gough. The defence case was based on the
premise that David Stephen Gough was one of the robbers. He had a record
of previous convictions, as had the appellant. During the trial, photographs
of both brothers had been produced to the jury, and retained by them.
Furthermore the vehicle alleged to have been used in the eighth robbery was
owned by Elaine Gough, the wife of David Stephen Gough, and her statement
including her address was read to the jury. The car must have been parked
outside the juror’s house for a number of months, and at the time at least of
the eighth robbery.

After sentence was passed, David Stephen Gough, who was then
present in court for the first time, started shouting; and it was at this point
that the juror, Mrs Smith, recognised him. He in his turn informed the
defence that one member of the jury was his next door neighbour. This was
drawn to the attention of the judge, but he rightly decided that he was by then
functus officio. However the juror was later interviewed by the police, and
subsequently swore an affidavit. The effect of the affidavit was summarised
by the Court of Appeal as follows:

1. 1. When she began her service on the jury she did not recognise
    the name ‘Gough’ as she knew her neighbour as ‘Steve’. Similarly she
    knew David’s wife as Elaine during the two years that they had been
    her next door neighbours.
  2. The name David Gough was mentioned on a number of
    occasions during the course of the trial.
  3. She had no recollection of ever seeing the appellant before the
    trial; and she had no idea that he was the brother of her next door
    neighbour.

-2-

1. 1. On April 24, 1991 during the trial, prosecution counsel read
    out a statement which contained the address, 3 Buckley Way (Mrs
    Smith lives at No. 2) and concerned the Capri motor car. She
    wondered whether Steve was David Gough but thought it could not be
    him as he was called Steve. She was confused.
  2. The photographs of the appellant and David Gough respectively
    were shown to the jury during the trial of the appellant. They were
    police photographs colloquially known as ‘mug shots’. Mrs Smith did
    not recognise David.

6. The fact that David Gough was her neighbour did not influence
her thinking as a juror and she did not mention the matter to her
fellow members of the jury.

The affidavit was and remains unchallenged.

It was on these facts that the question arose whether the courts should
conclude that, by reason of the presence of Mrs Smith on the jury, there was
such a possibility of bias on her part against the appellant that his conviction
should be quashed. As I have already recorded, that question was answered
by the Court of Appeal in the negative. The Court of Appeal however
identified in the cases two strands of authority, revealing that differing criteria
have been applied in the past when considering the question of bias. The two
tests have, as will appear, themselves been variously described. The Court
of Appeal identified them as being (1) whether there was a real danger of bias
on the part of the person concerned, or (2) whether a reasonable person
might reasonably suspect bias on his part. In the end, the court concluded
that the former test was to be applied in cases concerned with jurors, and the
latter in those concerned with magistrates or other inferior tribunals. The
court therefore applied the real danger test in the present case and, on that
basis, held that the appeal must fail, as indeed had been accepted by counsel
for the appellant.

In considering the subject of the present appeal, Your Lordships have
been faced with a series of authorities which are not only large in number, but
bewildering in their effect. It is only too clear how great a difficulty courts
of first instance, and indeed Divisional Courts and the Court of Appeal, must
face in cases which come before them; and there is a compelling need for
your Lordships’ House to subject the authorities to examination and analysis
in the hope of being able to extract from them some readily understandable
and easily applicable principles, thus obviating the necessity of conducting on
each occasion a trawl through authorities which are by no means easy to
reconcile. It is on that exercise that I now propose to embark.

A layman might well wonder why the function of a court in cases such
as these should not simply be to conduct an inquiry into the question whether
the tribunal was in fact biased. After all it is alleged that, for example, a

– 3 –

justice or a juryman was biased, i.e. that he was motivated by a desire
unfairly to favour one side or to disfavour the other. Why does the court not
simply decide whether that was in fact the case? The answer, as always, is
that it is more complicated than that. First of all, there are difficulties about
exploring the actual state of mind of a justice or juryman. In the case of both,
such an inquiry has been thought to be undesirable; and in the case of the
juryman in particular, there has long been an inhibition against, so to speak,
entering the jury room and finding out what any particular juryman actually
thought at the time of decision. But there is also the simple fact that bias is
such an insidious thing that, even though a person may in good faith believe
that he was acting impartially, his mind may unconsciously be affected by bias
– a point stressed by Devlin L.J. in Reg. v. Barnsley Licensing Justices, Ex
parte Barnsley and District Licensed Victuallers’ Association [1960] 2 Q.B.
167, 187. In any event, there is an overriding public interest that there should
be confidence in the integrity of the administration of justice, which is always
associated with the statement of Lord Hewart C.J. in Rex. v. Sussex Justices,
Ex parte McCarthy [1924] 1 KB 256, 259, that it is “of fundamental
importance that justice should not only be done, but should manifestly and
undoubtedly be seen to be done”. I shall return to that case in a moment, for
one of my tasks is to place the actual decision in that case in its proper
context. At all events, the approach of the law has been (save on the very
rare occasion where actual bias is proved) to look at the relevant
circumstances and to consider whether there is such a degree of possibility of
bias that the decision in question should not be allowed to stand.

My initial reaction to the conclusion of the Court of Appeal in the
present case was one of surprise that it should be necessary to draw a
distinction between cases concerned with justices and those concerned with
jurymen, and to conclude that different criteria fell to be applied in
investigating allegations of bias in the two categories of case. Evidently, the
Court of Appeal was itself unhappy in having to reach this conclusion, which
it felt bound to reach on the authorities. Of course, there are some
distinctions between the two groups of cases. For example, in the case of
jurymen there is the inhibition, to which I have already referred, against
investigating the state of mind of a juryman when reaching his decision in the
privacy of the jury room. There is also the fact that the possibility of bias
may come to light in the course of a jury trial – for example, a juryman may
have unwisely indulged in conversation with a witness, or previous convictions
of the accused may have accidentally been revealed to the jury. Situations
such as these have to be dealt with by the judge when they arise; and he may
be able to deal with the situation on the spot, for example by issuing a
warning to the jury, or by discharging the particular juryman involved. And,
if a verdict is challenged before the Court of Appeal on the ground of bias,
the ultimate principles to be applied are to be found in section 2 of the
Criminal Appeal Act 1968. But, even taking these matters into account, I am
left with the feeling that there should be no reason, in principle, why the test
of bias should be different in the two groups of cases – those concerned with
justices and those concerned with juries. I shall however, as a matter of

– 4 –

convenience, submit the authorities concerning these two categories of case
to separate consideration, before reaching any final conclusion on this point.

The argument before the Appellate Committee was presented on the
basis that there were two rival, alternative tests for bias to be found in the
authorities, and that the result in the present case depended on the choice
made by your Lordships’ House between them. The first test, favoured by
Mr Hytner for the appellant, was whether a reasonable and fair minded person
sitting in the court and knowing all the relevant facts would have had a
reasonable suspicion that a fair trial by the defendant was not possible. The
second test, favoured by Mr Leveson for the Crown, was whether there was
a real likelihood of bias. I shall for convenience refer to these two tests
respectively as the reasonable suspicion test, and the real likelihood test. It
was recognised by Mr Hytner before the Appellate Committee, as before the
Court of Appeal, that if the real likelihood test is to be preferred, the appeal
must fail.

In fact, examination of the authorities reveals that selection of the
appropriate test does not simply involve a choice between the two tests
formulated by counsel in the present case. Thus, when the appropriate test
in cases concerned with juries fell to be considered by your Lordships’ House
in Reg. v. Spencer [1987] AC 128, a variant of the real likelihood test, viz.
whether there was a real danger of bias, was adopted, as it was by the Court
of Appeal in the present case. There are also to be found in the authorities
variants of the reasonable suspicion test; and sometimes the two tests seems
to have been combined. At the heart of the present inquiry lies the need to
identify the precise nature of these tests, and to consider what, if any, are the
differences between them. For that purpose, I propose to consider first the
cases concerned with justices and other inferior tribunals, where the principal
problems appear to have arisen; and then to turn to the cases concerned with
juries, of which Reg. v. Spencer is of great importance.

Before I do so, however, I wish to draw attention to the fact that there
are certain cases in which it has been considered that the circumstances are
such that they must inevitably shake public confidence in the integrity of the
administration of justice if the decision is to be allowed to stand. Such cases
attract the full force of Lord Hewart’s requirement that justice must not only
be done but must manifestly be seen to be done. These cases arise where a
person sitting in a judicial capacity has a pecuniary interest in the outcome of
the proceedings. In such a case, as Blackburn J. said in Reg. v. Rand (1866)
L.R. 1 Q.B. 230, 232:

“. . . any direct pecuniary interest, however small, in the subject of
inquiry, does disqualify a person from acting as a judge in the matter.”

The principle is expressed in the maxim that nobody may be judge in his own
cause (nemo judex in sua causa). Perhaps the most famous case in which the
principle was applied is Dimes v. Grand Junction Canal (1853) 3 H.L.C. 759,

– 5 –

in which decrees affirmed by Lord Cottenham L.C. in favour of a canal
company in which he was a substantial shareholder were set aside by this
House, which then proceeded to consider the matter on its merits, and in fact
itself affirmed the decrees. Lord Campbell said (at p. 793):

“No one can suppose that Lord Cottenham could be, in the remotest
degree, influenced by the interest that he had in this concern; but, my
Lords, it is of the last importance that the maxim that no man is to be
a judge in his own cause should be held sacred.”

In such a case, therefore, not only is it irrelevant that there was in fact no bias
on the part of the tribunal, but there is no question of investigating, from an
objective point of view, whether there was any real likelihood of bias, or any
reasonable suspicion of bias, on the facts of the particular case. The nature
of the interest is such that public confidence in the administration justice
requires that the decision should not stand.

I turn next to the broader question of bias on the part of a member of
the relevant tribunal. Here it is necessary first to put on one side the very
rare case where actual bias is shown to exist. Of course, if actual bias is
proved, that is an end of the case; the person concerned must be disqualified.
But it is not necessary that actual bias should be proved; and in practice the
enquiry is directed to the question whether there was such a degree of
possibility of bias on the part of the tribunal that the court will not allow the
decision to stand. Such a question may arise in a wide variety of
circumstances. These include, but are by no means limited to, cases in which
a member of the tribunal has an interest in the outcome of the proceedings,
which falls short of a direct pecuniary interest. Such interests may vary
widely in their nature, in their effect, and in their relevance to the subject
matter of the proceedings; and there is no rule, as there is in the case of a
pecuniary interest, that the possession of such an interest automatically
disqualifies the member of the tribunal from sitting. Each case falls to be
considered on its own facts.

I turn first to the authorities concerned with justices, with whom I
bracket members of other inferior tribunals. Of the authorities cited to the
Appellate Committee in the course of argument, the first in point of time was
Reg. v. Rand (1866) L.R. 1 Q.B. 230, to which I have already referred, in
which Blackburn J. stated the law in terms of the real likelihood test. He
referred (at p. 232) to cases in which there was “a real likelihood that the
judge would, for kindred or any other cause, have a bias in favour of one of
the parties” in which event “it would be very wrong in him to act”. That test
was later approved by three members of the Appellate Committee of this
House in Frome United Breweries Co. Ltd. v. Bath Justices [1926] A.C. 586
(a case concerned with licensing justices): see p. 591 per Viscount Cave
L.C.; p. 607 per Lord Atkinson (citing Rex. v. Sunderland Justices [1901]
2 K.B. 357); and p. 610 per Lord Sumner (quoting from the dissenting
judgment of Atkin L.J., sub nom. Rex v. Bath Compensation Authority [1925]

– 6 –

1 K.B. 685, 712). Furthermore Lord Shaw of Dunfermline agreed with
Viscount Cave L.C.; and, although the other member of the Appellate
Committee, Lord Carson, spoke simply of “a likelihood of bias” (see p. 617),
there is no reason to suppose that he intended any different test.

At this stage, however, I must turn to the well known case of Rex. v.
Sussex Justices, Ex parte McCarthy [1924] 1 KB 256. There the applicant
came before magistrates charged with the offence of dangerous driving, which
had involved a collision between his vehicle and another vehicle. The
solicitor acting as magistrates’ clerk on this occasion was also acting as
solicitor for the other driver in civil proceedings against the applicant arising
out of the collision. At the conclusion of the evidence before the magistrates,
the acting clerk retired with them in case his help should be needed on a point
of law; but in fact the magistrates did not consult him, and he himself
abstained from referring to the case. The magistrates convicted the applicant,
but his conviction was quashed by a Divisional Court. This is of course the
case in which Lord Hewart C.J. let fall his much-quoted dictum, to which I
have already referred. I think it helpful, however, to quote from his judgment
in extenso (see pp. 258-9):

“It is said, and, no doubt, truly, that when that gentleman retired in
the usual way with the justices, taking with him the notes of the
evidence in case the justice might desire to consult him, the justices
came to a conclusion without consulting him, and that he scrupulously
abstained from referring to the case in any way. But while that is so,
a long line of cases shows that it is not merely of some importance but
is of fundamental importance that justice should not only be done, but
should manifestly and undoubtedly be seen to be done. The question
therefore is not whether in this case the deputy clerk made any
observation or offered any criticism which he might not properly have
made or offered; the question is whether he was so related to the case
in its civil aspect as to be unfit to act as clerk to the justices in the
criminal matter. The answer to that question depends not upon what
actually was done but upon what might appear to be done. Nothing is
to be done which creates even a suspicion that there has been an
improper interference with the course of justice. Speaking for myself,
I accept the statements contained in the justices’ affidavit, but they
show very clearly that the deputy clerk was connected with the case in
a capacity which made it right that he should scrupulously abstain from
referring to the matter in any way, although he retired with the
justices; in other words, his one position was such that he could not,
if he had been required to do so, discharge the duties which his other
position involved. His twofold position was a manifest contradiction.
In those circumstances I am satisfied that this conviction must be
quashed, . . .”

The case was therefore concerned with the possibility that the acting
magistrates’ clerk, who plainly had such an interest in the outcome of the civil

– 7 –

proceedings that he might well be biased against the applicant in the
proceedings before the magistrates, might influence the decision of the
magistrates adversely to the applicant. Lord Hewart C.J. clearly thought that
the acting magistrates’ clerk’s involvement in the civil proceedings was such
that he should never have participated in the hearing before the magistrates,
and went so far as to indicate that “even a suspicion that there had been an
improper interference with the course of justice” is enough to vitiate the
proceedings, an observation which has been invoked as the origin of the
reasonable suspicion test. Indeed, following the Sussex Justices case, there
developed a tendency for courts to invoke a test requiring no more than a
suspicion of bias.

However in a later case, also concerned with alleged bias on the part
of a magistrates’ clerk, Reg. v. Camborne Justices, Ex parte Pearce [1955]
1 Q.B. 41, a Divisional Court, having received the assistance of the Solicitor-
General as amicus curiae, approached the question on the basis that a real
likelihood of bias must be established. In that case, the applicant was
convicted of an offence under the Food and Drugs Act 1938. The information
alleging the offence had been laid by a sampling officer, for the Cornwall
County Council. The magistrates’ clerk, who in the course of the hearing was
invited into the magistrates’ private room in order to advise them, was a
member of the County Council (though not of the relevant committee of the
Council, the Public Health and Housing Committee). For this reason, the
applicant alleged that a reasonable suspicion of bias might arise, and that his
conviction should be quashed. The court dismissed the application, holding
that in the circumstances there was no real likelihood of bias on the part of the
magistrates’ clerk. Moreover the court was at pains to reject any suggestion
that mere suspicion of bias was sufficient; and, while endorsing and fully
maintaining the integrity of the principle reasserted by Lord Hewart C.J. in
the Sussex Justices case, nevertheless deplored the principle “being urged as
a warrant for quashing convictions or invalidating orders upon quite
unsubstantial grounds and, indeed, in some cases, upon the flimsiest pretext
of bias” (see pp. 51-52, per curiam).

In the Sussex Justices case it must have been plain that there was a real
likelihood of bias on the part of the acting magistrates’ clerk; and the court
went on to hold that, despite the fact that there had been no discussion about
the case between the magistrates and the clerk, nevertheless the decision of the
magistrates must be quashed, because nothing may be done which creates even
a suspicion that there has been a wrongful interference with the course of
justice. It appears that this decision was later used to suggest that a mere
suspicion of bias on the part of a person involved in the process of
adjudication is enough to require that the decision should be quashed. That
approach was rejected in the Camborne Justices case, in which it was held
that, since there was no real likelihood of bias on the part of the magistrates’
clerk, there was no ground for quashing the magistrates’ decision. The cases
can therefore be distinguished on the facts. But the question remains
whether, in a case involving a magistrates’ clerk, it is enough to show that

-8-

there was a real likelihood of bias on the part of the clerk, or whether it must
also be shown that, by reason of his participating in the decision-making
process, there was a real likelihood that “he would impose his influence on the
justices or give them wrong legal advice” (see [1955] 1 Q.B. 41, 46, per Sir
Reginald Manningham-Buller Q.C., S.G., arguendo as amicus curiae). In
my opinion, the latter view is to be preferred. Of course, nowadays a
magistrates’ clerk will not withdraw with the justices, but will only join them
if invited to advise them on a question of law. If the clerk is not so invited,
any bias on his part will ordinarily have no influence on the outcome of the
proceedings; though if he has any interest in the outcome, it is obviously
undesirable that he should be acting at all in the capacity of clerk in relation
to those proceedings, in case his advice is called for. If however he is
invited to give the magistrates advice, it is open to the court to infer that,
having regard to the insidious nature of bias, there is a real likelihood of the
clerk’s bias infecting the views of the magistrates adversely to the applicant.

I have had the opportunity of reading in draft the speech of my noble
and learned Lord Woolf, and it follows from what I have said that I am in
agreement with his conclusions both about the effect of the Sussex Justices and
Camborne Justices cases, and that the only special category of case, in which
it is unnecesary to enquire whether there was any real likelihood of bias,
relates to circumstances where a person acting in a judicial capacity has a
direct pecuniary interest in the outcome of the proceedings.

In Reg. v. Barnsley Licensing Justices [1960] 2 Q.B. 167 at p. 187,
Devlin L.J. also preferred the real likelihood test, considering that the term
“real likelihood of bias” is not used to import the principle in Rex. v. Sussex
Justices, which had been invoked by Salmon J. at first instance [1959] 2 Q.B.
276, 286. It is, I think, desirable that I should quote the relevant passage
from the judgment of Devlin L.J. in full (see pp. 186-187):

“Here is an application by the co-operative society and there is sitting
to decide it a bench which is wholly composed of members of the
society and one woman whose husband was a member of the society,
presided over by a chairman who had interested himself actively in the
conduct of the affairs of the society or was desirous of doing so. Is
there, in those circumstances, a real likelihood of bias? I am not quite
sure what test Salmon J. applied. If he applied the test based on the
principle that justice must not only be done but manifestly be seen to
be done, I think he came to the right conclusion on that test. I cannot
imagine anything more unsatisfactory from the public point of view
than applications of this sort being dealt with by a bench which was so
composed, and, indeed, it is conceded that steps will have to be taken
to rectify the position. But, in my judgment, it is not the test. We
have not to inquire what impression might be left on the minds of the
present applicants or on the minds of the public generally. We have
to satisfy ourselves that there was a real likelihood of bias – not merely
satisfy ourselves that that was the sort of impression that might

-9-

reasonably get abroad. The term ‘real likelihood of bias’ is not used,
in my opinion, to import the principle in Rex v. Sussex Justices to
which Salmon J. referred. It is used to show that it is not necessary
that actual bias should be proved. It is unnecessary, and, indeed,
might be most undesirable, to investigate the state of mind of each
individual justice. ‘Real likelihood’ depends on the impression which
the court gets from the circumstances in which the justices were
sitting. Do they give rise to a real likelihood that the justices might
be biased? The court might come to the conclusion that there was
such a likelihood, without impugning the affidavit of a justice that he
was not in fact biased. Bias is or may be an unconscious thing and a
man may honestly say that he was not actually biased and did not
allow his interest to affect his mind, although, nevertheless, he may
have allowed it unconsciously to do so. The matter must be
determined upon the probabilities to be inferred from the
circumstances in which the justices sit.”

It is plain from this passage that Devlin L.J. was concerned to get away from
any test founded simply upon suspicion – “the sort of impression that might
reasonably get abroad” – and to focus upon the actual circumstances of the
case in order to decide whether there was in those circumstances a real
likelihood of bias. His question – do the circumstances give rise to a real
likelihood that the justices might be biased? – suggests that he was thinking of
likelihood as meaning not probability, but possibility; the noun probability is
not aptly qualified by the adjective “real”, and the verb “might” connotes
possibility rather than probability. Such a reading makes the real likelihood
test very similar to a test requiring a real danger of bias. It is true that, at the
conclusion of the passage which I have quoted, Devlin L.J. stated that the
matter must be determined “upon the probabilities”. I do not however think
that he meant “on the balance of probabilities”, but rather that he was
emphasising that the question was to be answered by reference to the relevant
circumstances.

However nine years later, in Metropolitan Properties Co. (F.G.C.)
Ltd. v. Lannon [1969] 1 QB 577, the law took a different turn. The case
was concerned with a decision by a rent assessment committee, when
determining fair rents for a block of flats in London. The rent so determined
was substantially below the rent suggested even by the expert called by the
tenants. The landlord sought to quash the decision on the ground that the
chairman of the committee was a solicitor who had been concerned with
advising tenants of flats in another comparable block of flats. The Court of
Appeal, allowing the appeal from a Divisional Court, held that the facts were
such as to give rise to an appearance of bias on the part of the chairman, and
on that ground they quashed the decision of the committee, even though there
was no actual bias on his part. In so holding, the court rejected the argument
of counsel for the committee, who invited the court to proceed on the basis
of the real likelihood test. Lord Denning M.R. and Edmund Davies L.J. both
invoked the much quoted dictum of Lord Hewart C.J. in Rex. v. Sussex

– 10 –

Justices, and declined to follow Devlin L.J.’s approach in Reg. v. Barnsley
Licensing Justices. Lord Denning M.R. stated the law as follows (at p. 599):

“In Reg. v. Barnsley Licensing Justices, Ex parte Barnsley and District
Licensed Victuallers’ Association, Devlin L.J. appears to have limited
that principle considerably, but I would stand by it. It brings home
this point: in considering whether there was a real likelihood of bias,
the court does not look at the mind of the justice himself or at the
mind of the chairman of the tribunal, or whoever it may be, who sits
in a judicial capacity. It does not look to see if there was a real
likelihood that he would, or did, in fact favour one side at the expense
of the other. The court looks at the impression which would be given
to other people. Even if he was as impartial as could be, nevertheless
if right-minded persons would think that, in the circumstances, there
was a real likelihood of bias on his part, then he should not sit. And
if he does sit, his decision cannot stand: see Reg v. Huggins [1895]
1 Q.B. 563 and Rex v. Sunderland Justices, per Vaughan Williams
L.J. [1901] 2 K.B. 357, 373. Nevertheless there must appear to be a
real likelihood of bias. Surmise or conjecture is not enough: see Reg.
v. Camborne Justices, Ex. parte Pearce [1955] 1 Q.B. 41, 48-51 and
Reg. v. Nailsworth Licensing Justices, Ex parte Bird [1953] 1 W.L.R.
1046. There must be circumstances from which a reasonable man
would think it likely or probable that the justice, or chairman, as the
case may be, would, or did, favour one side unfairly at the expense of
the other. The court will not inquire whether he did, in fact, favour
one side unfairly. Suffice it that reasonable people might think he did.
The reason is plain enough. Justice must be rooted in confidence: and
confidence is destroyed when right-minded people go away thinking:
‘The judge was biased.'”

Edmund Davies L.J. said (at p. 606) that it was enough if “there is reasonable
suspicion of bias on the part of one or more members of the adjudicating
body”; and the third member of the court, Danckwerts L.J., appears to have
proceeded, despite some doubt, upon a similar basis (at pp. 601-602).

I shall return to this case in a moment, but I have to say that it left a
legacy of some confusion behind it. In two cases, Reg. v. Uxbridge Justices,
Ex parte Burbridge, The Times, 21 June 1972, and Reg. v. McLean, Ex parte
Aikens (1974) 139 L.G.Rev. 261, Lord Widgery C.J. was prepared to proceed
on the basis of the reasonable suspicion test, though in neither case was the
choice of test decisive. However, in Reg. v. Altrincham Justices, Ex parte N.
Pennington [1975] Q.B. 549, Lord Widgery did not feel able to decide
whether the real likelihood test or the reasonable suspicion test was
appropriate. In that case the appellants were convicted of offences of having
sold vegetables by weight and having delivered a lesser weight to two county
schools. The presiding justice at the trial was a member of the education
committee, and was a governor of two schools, though not of those in

– 11 –

question. A Divisional Court quashed the convictions on the ground that the
presiding justice should have disqualified herself from hearing a case where
she had an active interest in the schools which were the victims of the offence.
In so holding, Lord Widgery referred to both the real likelihood test and the
reasonable suspicion test. However it was not clear to him from Lannon
which of those tests fell to be applied. Furthermore, in Reg. v. Liverpool City
Justices, Ex parte Topping [1983] 1 W.L.R. 119, in which justices became
aware of other unrelated charges against the defendant whose case they were
about to consider, the Divisional Court applied a form of the reasonable
suspicion test derived from the judgment of Lord Widgery in Ex parte
Burbridge; but they prefaced their choice of this test with the observation
that, in agreement with a view expressed by Cross L.J. in Hannam v.
Bradford City Council [1970] 1 W.L.R. 937, 949, there was little if any
difference between the real likelihood test and the reasonable suspicion test,
because if a reasonable person with the relevant knowledge thinks that there
might well be bias, then there is in his opinion a real likelihood of bias – a
view which appears to assume that real likelihood of bias means no more than
a real possibility of bias.

I have already quoted passages from the judgments of Lord Denning
M.R. and Edmund Davies L.J. in Lannon [1969] 1 QB 577, 599, 606,
which show that they did not in fact state the same test, Lord Denning’s test
being really no more than an adaptation of the real likelihood test, and only
Edmund Davies L.J. enunciating a test founded upon real suspicion of bias.
Furthermore Lord Denning, while purporting to differ from Devlin L.J. in the
Barnsley Licensing Justices case [1960] 2 Q.B. 167, in fact differed very little
from him. Thus, both considered that it was not necessary that actual bias
should be proved, the court having therefore to proceed upon an impression
derived from the circumstances; and that the question is whether such an
impression reveals a real likelihood of bias. The only difference between
them seems to have been that, whereas Devlin L.J. spoke of the impression
which the court gets from the circumstances, Lord Denning looked at the
circumstances from the point of view of a reasonable man, stating that there
must be circumstances from which a reasonable man would think it likely or
probable that the justice, or chairman, was biased. Since however the court
investigates the actual circumstances, knowledge of such circumstances as are
found by the court must be imputed to the reasonable man; and in the result
it is difficult to see what difference there is between the impression derived
by a reasonable man to whom such knowledge has been imputed, and the
impression derived by the court, here personifying the reasonable man. It is
true that Lord Denning expressed the test as being whether a reasonable man
would think it “likely or probable” that the justice or chairman was biased.
If it is a correct reading of his judgment (and it is by no means clear on the
point) that it is necessary to establish bias on a balance of probabilities, I for
my part would regard him as having laid down too rigorous a test. In my
opinion, if, in the circumstances of the case (as ascertained by the court), it
appears that there was a real likelihood, in the sense of a real possibility, of

– 12 –

bias on the part of a justice or other member of an inferior tribunal, justice
requires that the decision should not be allowed to stand. I am by no means
persuaded that, in its original form, the real likelihood test required that any
more rigorous criterion should be applied. Furthermore the test as so stated
gives sufficient effect, in cases of apparent bias, to the principle that justice
must manifestly be seen to be done, and it is unnecessary, in my opinion, to
have recourse to a test based on mere suspicion, or even reasonable suspicion,
for that purpose. Finally there is, so far as I can see, no practical distinction
between the test as I have stated it, and a test which requires a real danger of
bias, as stated in Reg. v. Spencer [1987] AC 128. In this way, therefore,
it may be possible to achieve a reconciliation between the test to be applied
in cases concerned with justices and other members of inferior tribunals, and
cases concerned with jurors.

I turn therefore to the cases concerned with jurors; and here the
relevant authorities support the view which I have just expressed. It is true
that, after Lannon, there were cases in which the reasonable suspicion test was
adopted (see e.g., Reg. v. Pennington [1985] 81 Cr. App. R. 217). However,
it is appropriate to turn straight to the leading authority, which is the decision
of your Lordships’ House in Reg. v. Spencer [1987] AC 128. In that case
the defendants, who were members of the nursing staff at a secure hospital,
were convicted in two separate trials of ill treating patients at the hospital,
contrary to section 126 of the Mental Health Act 1959. On appeal, the
principle issue was one of corroboration. But in addition a question arose
with regard to one of the jurors at the first trial. He had clearly demonstrated
in the course of the trial that he was biased against the defendants. At first
the judge, having consulted counsel, decided to take no action. However, it
then transpired that the juror’s wife worked at another mental hospital which
figured in the evidence at the trial. The judge, fearing that the juror might
have heard things from his wife which it would be better if he had not heard,
decided to discharge him; but, discovering that the juror was in the habit of
giving three other members of the jury a lift home, warned the members of
the jury that they should not discuss the case further with him. On the
following morning, however, defence counsel submitted that the remainder of
the jury should be discharged; but the judge decided, in the exercise of his
discretion, not to do so. Counsel for the prosecution had submitted that the
test which the judge should apply was that the jury should not be discharged
unless it could be shown that there was a very high risk that the apparently
biased jury had influenced any of his fellow jurors. Lord Ackner (with whom
Lord Brandon of Oakbrook and Lord Mackay of Clashfern agreed) however
held that the correct test was that stated by the Court of Appeal in Reg. v.
Sawyer [1980] 71 Cr. App. R. 283, 285, viz. whether there was a real danger
that the appellant’s position had been prejudiced in the circumstances. This
was the test which had in fact been applied by the Court of Appeal, but they
had concluded that there was no realistic chance that the three jurors who had
travelled in the car had been prejudiced or biased by what they had heard. On
this point, however, Lord Ackner found himself unable totally to dismiss that
possibility, and he concluded, with the remainder of the Appellate Committee,

– 13 –

that the verdict was unsafe and the appeal must be allowed [1987] AC 128,
146. Subsequently, the test so established in Reg. v. Spencer was applied by
the Court of Appeal in Reg. v. Putnam (1990) 93 Cr. App. R. 281. I should
add that in Reg. v. Morris (Otherwise Williams) (1990) 93 Cr. App. R. 102,
in which the reasonable suspicion test was applied, it appears that Reg. v.
Spencer was not cited to the court. In the light of the conclusion which I have
reached, I do not think that it is necessary for me to consider any more of the
earlier cases concerned with allegation of bias on the part of jurors. I only
wish to say that Reg. v. Box [1964] 1 Q.B. 430, to which some criticism was
directed in the course of argument, appears to have been concerned primarily
with an allegation of actual bias, and to have reasserted the principle that
knowledge by a juror of a defendant’s character or previous convictions is not
an automatic disqualification.

There are however two features of jury cases to which I will briefly
draw attention. The first is that the possibility of bias on the part of a juror
may, as in the case of Spencer itself, come to the attention of the judge in the
course of the trial. In such circumstances the judge, in deciding whether to
exercise his discretion to discharge one or more members of the jury, should
apply the same test as falls to be applied on appeal by the Court of Appeal,
viz. whether there is a real danger of bias affecting the mind of the relevant
juror or jurors. Even if the judge decides that it is unnecessary to do more
than issue a warning to the jury or to a particular juror, and thereby isolate
and neutralise any bias that might otherwise occur, the effect of his warning
is not merely to ensure that the jurors do not allow any possible bias to affect
their minds, but also to prevent any lack of public confidence in the integrity
of the jury. It is unnecessary for me to say any more on this subject, to
which no argument was addressed in the present case. Second, if any
question of bias on the part of a juror arises on appeal, the Court of Appeal,
having applied the real danger test, will then proceed in the light of its
conclusion on that test to exercise its powers under section 2 of the Criminal
Appeal Act 1968, in the normal way, as was done by your Lordships’ House
in Spencer.

I wish to add that in cases concerned with allegations of bias on the
part of an arbitrator, the test adopted, derived from Ex parte Topping [1983]
1 W.L.R. 119, has been whether the circumstances were such that a
reasonable man would think that there was a real likelihood that the arbitrator
would not fairly determine the issue on the basis of the evidence and
arguments adduced before him (see Ardahalian v. Unifert International S.A.
(The Elissar) [1984] 2 Lloyd’s Rep. 84, and Bremer Handelsgesellschaft
m.b.H. v. Ets. Soules et Cie. [1985] 1 Lloyd’s Rep. 160, [1985] 2 Lloyd’s
Rep. 199). Such a test is, subject to the introduction of the reasonable man,
consistent with the conclusion which I have reached, provided that the
expression “real likelihood” is understood in the sense I have described, i.e.
as meaning that there is a real possibility or, as I would prefer to put it, a real
danger of bias. It would appear to have been so understood by Mustill J. (as

– 14 –

he then was) in Bremer [1985] 1 Lloyd’s Rep. 160, 164, where he referred
to “an evident risk” of bias.

In conclusion, I wish to express my understanding of the law as
follows. I think it possible, and desirable, that the same test should be
applicable in all cases of apparent bias, whether concerned with justices or
members of other inferior tribunals, or with jurors, or with arbitrators.
Likewise I consider that, in cases concerned with jurors, the same test should
be applied by a judge to whose attention the possibility of bias on the part of
a juror has been drawn in the course of a trial, and by the Court of Appeal
when it considers such a question on appeal. Furthermore, I think it
unnecessary, in formulating the appropriate test, to require that the court
should look at the matter through the eyes of a reasonable man, because the
court in cases such as these personifies the reasonable man; and in any event
the court has first to ascertain the relevant circumstances from the available
evidence, knowledge of which would not necessarily be available to an
observer in court at the relevant time. Finally, for the avoidance of doubt, I
prefer to state the test in terms of real danger rather than real likelihood, to
ensure that the court is thinking in terms of possibility rather than probability
of bias. Accordingly, having ascertained the relevant circumstances, the court
should ask itself whether, having regard to those circumstances, there was a
real danger of bias on the part of the relevant member of the tribunal in
question, in the sense that he might unfairly regard (or have unfairly regarded)
with favour, or disfavour, the case of a party to the issue under consideration
by him; though, in a case concerned with bias on the pan of a magistrates’
clerk, the court should go on to consider whether the clerk has been invited
to give the magistrates advice and, if so, whether it should infer that there was
a real danger of the clerk’s bias having infected the views of the magistrates
adversely to the applicant.

It follows from what I have said that the Court of Appeal applied the
correct test in the present case. On that test, it was accepted by Mr Hytner
that there was no ground for disturbing the jury’s verdict. I would therefore
dismiss the appeal.

LORD ACKNER

My Lords,

I have had the advantage of reading in draft the speech prepared by my
noble and learned friend Lord Goff of Chieveley, and for the reasons he
gives, I, too, would dismiss the appeal.

– 15 –

LORD MUSTILL

My Lords,

I have had the advantage of reading in draft the speech prepared by my
noble and learned friend Lord Goff of Chieveley, and for the reasons he
gives, I, too, would dismiss the appeal.

LORD SLYNN OF HADLEY
My Lords,

I have had the advantage of reading in draft the speech prepared by my
noble and learned friend Lord Goff of Chieveley, and for the reasons he
gives, I, too, would dismiss the appeal.

LORD WOOLF

My Lords,

I have had the advantage of reading in draft the speech of Lord Goff
of Chieveley and I agree that this appeal should be dismissed for the reasons
which he gives. In particular, I agree that the correct test to adopt in deciding
whether a decision should be set aside on the grounds of alleged bias is that
given by Lord Goff, namely, whether there is a real danger of injustice having
occurred as a result of the alleged bias.

The test to be applied in each case has as its source the maxim that
nobody may be judge in his own cause. No distinction arises in the
application of the test because it is the clerk to the justices rather than the
justices themselves who are alleged to be biased. A clerk to the justices is
part of the judicial process in the magistrates court. This is accepted by Lord
Hewart C.J., when he said in his judgment in the Rex. v. Sussex Justices, Ex
parte McCarthy [1924] 1 KB 256, 259, that the clerk’s position “was such
that he could not, if he had been required to do so, discharge the duties which
his other position involved. His twofold position was a manifest
contradiction.” (The other position, being a member of the firm of solicitors
acting for the other driver who was involved in the accident which gave rise
to the prosecution.)

This is also made clear in the judgment in Reg. v. Camborne Justices,
Ex parte Pearce [1955] 1 Q.B. 41, where the facts were very similar to those

– 16 –

in the Sussex Justices case. The Camborne Justices case also involved a
justice’s clerk. The proceedings before the justices were the result of an
information under the Food and Drugs Act 1938 laid on behalf of the County
Council. The clerk to the justices was at the time a member of the council,
but not a member of the council’s health committee responsible for laying the
information. At the hearing he was sent for to advise the justices on a point
of law, but according to the evidence put before the Divisional Court he did
not discuss the facts of the case and having given his advice returned to the
Court. Unlike the Sussex Justices case, where the argument appears to have
been limited (the applicant was not called upon to address the court) and the
judgment was not reserved, in the Camborne Justices case the matter was fully
argued, Sir Reginald Manningham-Buller Q.C., S.-G. and J.P. Ashworth
appearing as amici curiae and a reserve judgment of the court was given by
Slade J. on behalf of a Divisional Court which was presided over by Lord
Goddard C.J. That judgment described the question which the court had to
decide, at p. 47, as being:

“What interest in a judicial or quasi judicial proceeding does the law
regard as sufficient to incapacitate a person from adjudicating or
assisting in adjudicating on it upon the ground of bias or appearance
of bias?”

To that question the court gave the answer (at p. 51):

“that to disqualify a person from acting in a judicial or quasi judicial
capacity upon the ground of interest (other than pecuniary or
proprietary) in the subject matter of the proceeding, a real likelihood
of bias must be shown.”

As the court concluded on the facts, that there was no real likelihood of bias
application was dismissed. However, for present purposes the importance of
the case is that the court did not consider they were dealing with a special
category of case and applied a test which I regard as being the equivalent of
the real danger test.

The problem created by the Sussex Justices case [1924] 1 KB 256
arises because Lord Hewart preceded his celebrated remark, at p. 259: “it is
not merely of some importance but is of fundamental importance that justice
should not only be done, but should manifestly and undoubtedly be seen to be
done”, with the comment, at pp. 258-259:

“It is said, and, no doubt, truly, that when that gentleman retired in
the usual way with the justices, taking with him the notes of the
evidence in case the justices might desire to consult him, the justices
came to a conclusion without consulting him, and that he scrupulously
abstained from referring to the case in any way.”

– 17 –

and later added: “speaking for myself, I accept the statements contained in the
justices’ affidavit”. If these passages in his judgment are taken at face value,
then they are consistent with the court in the Sussex Justices case coming to
the conclusion that there was no risk of actual bias and the court was therefore
applying some different test from the real danger test when deciding that the
decision had to be quashed. A similar situation arises in relation to the
comment of Lord Campbell in the Dimes v. Grand Junction Canal (1853) 3
H.L. Cas. 759, 793, case when he, alone among the members of the House
of Lords:

It could well be that too much attention should not be attached to the
remarks made as to the bona fides of the Lord Chancellor in the Dimes case
and the justices’ clerk in the Sussex Justices case, although, no doubt the Lord
Chancellor and the clerk respectively found them comforting. It must be
remembered that except in the rare case where actual bias is alleged, the court
is not concerned to investigate whether or not bias has been established.
Whether it is a judge, a member of the jury, justices or their clerk, who is
alleged to be biased, the courts do not regard it as being desirable or useful
to inquire into the individual’s state of mind. It is not desirable because of the
confidential nature of the judicial decision making process. It is not useful
because the courts have long recognised that bias operates in such an insidious
manner that the person alleged to be biased may be quite unconscious of its
effect.

It is because the court in the majority of cases does not inquire whether
actual bias exists that the maxim that justice must not only be done but seen
to be done applies. When considering whether there is a real danger of
injustice, the court gives effect to the maxim, but does so by examining all the
material available and giving its conclusion on that material. If the court
having done so is satisfied there is no danger of the alleged bias having
created injustice, then the application to quash the decision should be
dismissed. This, therefore, should have been the result in the Sussex Justices
case if Lord Hewart’s remarks are to be taken at face value and are to be
treated as a finding, and not merely an assumption, that there was no danger
of the justices’ decision being contaminated by the possible bias of the clerk.

The Dimes case, 3 H.L. Cas. 759, is different because it involved
direct pecuniary or proprietary interest on the part of the Lord Chancellor in
the subject matter of the proceedings and this creates a special situation, as
was pointed out at the beginning of the judgment in the Camborne Justices
case [1955] 1 Q.B. 41,47:

– 18 –

“… any direct pecuniary or proprietary interest in the subject matter
of proceeding, however small, operates as an automatic disqualification. In
such a case the law assumes bias.”

It was because Lord Hewart C.J’s judgment in the Sussex Justices case
[1924] 1 KB 256, 258-259, has created difficulties that in the Camborne
Justices case, where exactly the same issue was involved, the court warned
against the misuse of Lord Hewart’s judgment since it was being “urged as a
warrant for quashing convictions or invalidating orders upon quite
unsubstantial grounds and, indeed, in some cases upon the flimsiest pretext of
bias”, (pp. 51-52). As the court pointed out the continued citation of Lord
Hewart’s maxim may lead to the erroneous impression that “it is more
important that justice should appear to be done than that it should, in fact, be
done.”

I therefore suggest that the Sussex Justices case [1924] 1 KB 256
neither creates nor should it be placed in a separate category. The proper test
which Lord Goff has identified should have been applied in that case as it was
in the Camborne Justices case [1955] 1 Q.B. 41. There is only one
established special category and that exists where the tribunal has a pecuniary
or proprietary interest in the subject matter of the proceedings as in Dimes,
3 H.L. Cas. 759. The courts should hesitate long before creating any other
special category since this will immediately create uncertainty as to what are
the parameters of that category and what is the test to be applied in the case
of that category. The real danger test is quite capable of producing the right
answer and ensure that the purity of justice is maintained across the range of
situations where bias may exist.