The Discretion of the Judge
The Right Hon. Lord Justice Bingham*

A judge of my acquaintance once told me that when, in the course of trying a
case, he encountered any problem of unusual difficulty, it was his practice to
glower at counsel in his most forbidding manner and demand "Is this not a matter
within my discretion?" On counsel agreeing that it was - which it seems they
readily did - he would sink back in his chair with relief, relaxed in the knowledge
that no matter what he decided his decision would be immune from successful
challenge on appeal.

The complaisance of counsel may seem surprising, given the consequence for
their clients of a decision being regarded as discretionary. But judges and
practitioners have, I think, habitually used the expression "judicial discretion" in
a variety of senses, and academic writers have used it in a different sense again.
This lack of consensus is also surprising. Since at latest 1581, Iauthoritative voices
on both sides of the Atlantic, Coke2, Mansfield3 and Marsha1l4 among them, have
urged that the discretionary powers of judges and justices be strictly limited and
controlled by the law to avoid the arbitrariness of an unpredictable personal
decision. Right down to this century, strong language has been used. "To remit
the maintenance of constitutional right to the region of judicial discretion", said
Lord Shaw of Dunfermline in 1913, "is to shift the foundations of freedom from
the rock to the sand."5 Or as Justice William Douglas put it, "Absolute discretion,
like corruption, marks the beginning of the end of liberty."6 Lord Simon of
Glaisdale expressed the traditional view when he said, speaking on the judicial
discretion to admit or exclude evidence, "And if it comes to the forensic
crunch ... it must be law, not discretion, which is in command."7

• The Royal Bank of Scotland Lecture, Oxford Universiry, printed by arrangement with Lord
Justice Bingham, and the Royal Bank of Scotland.

I. Lambarde, Eirenarcha, 58.
2. Prohibitions del Roy (1607) Co. Rep. 63, at 64-65.
3. R. v. Wilkes (1779) 4 Burr. 2527, at p.2539.
4. Osborn v. The Bank of the United States (1824) 22 U.S. 738, at p.866.
5. Scott V. Scott [1913] A.C. 417, at p.477.
6. State of New York v. United States (1951) 342 US 822, at p.884.
7.D. v.NSPCC [1978] A.C. 171, at p.239 G.

27



THE DENNING LAW JOURNAL

Any lack of certainty as to what judicial discretion is may also be thought
undesirable. For if, as these warnings suggest, judicial discretions are dangerous as
capable of leading to arbitrariness, it is as well judges should be quite clear when
they are exercising a discretion and when not, and if the exercise of a discretion is a
barrier (whether or not surmountable) to an appeal then appellate judges should
similarly recognise when the barrier exists and when it does not.

I am vividly aware that he who defines invites scholarly refutation, and to
attempt the task in this forum may reasonably be thought foolhardy. I shall
nonetheless proffer a definition, and briefly defend it. On the assumption that my
definition is broadly acceptable, I shall then suggest that the role of judicial
discretion is now narrowly confined. I shall further suggest, with appropriate
apologies to Mr Dunning, as he then was, that its role has decreased, is decreasing
and need not in general be much further diminished. The dragon of arbitrary
discretion has not been slain, but it has been domesticated and put on a short leash.

According to my definition, an issue falls within a judge's discretion if, being
governed by no rule of law, its resolution depends on the individual judge's
assessment (within such boundaries as have been laid down) of what it is fair and
just to do in the particular case. He has no discretion in making his findings of fact.
He has no discretion in his rulings on the law. But when, having made any
necessary finding of fact and any necessary ruling oflaw, he has to choose between
different courses of action, orders, penalties or remedies he then exercises a
discretion. It is only when he reaches the stage of asking himself what is the fair
and just thing to do or order in the instant case that he embarks on the exercise of a
discretion.

I believe this definition to be broadly consistent with the usage adopted in
statutes. There are of course numerous statutes which confer a discretion on the
court, describing it as such; many of these relate to the award of costs, the
imposition of criminal penalties and the exercise of procedural powers, all of them
pre-eminently discretionary fields. To some examples I shall return. But often a
discretion is conferred although not so described: " ... may, in accordance with
the rules of court, extend any such period to such extent and on such conditions as
it thinks fit ... ";8 "may order that such party be at liberty to inspect and take
copies of any entries in a banker's book ... ";9 "may ... as it thinks fit",
"may ... as the court thinks just"; "may impose such other condition as it thinks
fit";lO "On an application under this section the court may make or refuse to make
the declaration asked for ... ";11 and so on, almost ad infinitum. A discretion is
conferred whichever form of words is used.

While my exploration of the statute book is far from comprehensive, I have
encountered only two provisions which are inconsistent with my suggested

8. Maritime Conventions Act 1911, s.8.
9. Bankers' Books Evidence Act 1879, s.7.
10. Housing Act 1988, S.9(1)(2)(3).
II. Local Government Finance Act 1982, s.19.

28



THE DISCRETION OF THE JUDGE

definition. The first is in section 2 of Fox's Libel Act 1792 which provides "That,
on every such trial, the Court or Judge before whom such indictment or
information shall be tried, shall, according to their or his Discretion, give their or
his Opinion and Directions to the Jury on the Matter in Issue between the King
and the Defendant or Defendants, in like manner as in other Criminal Cases." I
would have to admit that the direction which a judge gives to a criminal jury on the
law is not now a matter of discretion within my definition. But the next section
uses the term in my sense: it provides that nothing shall "prevent the Jury from
finding a Special Verdict, in their Discretion, as in other Criminal Cases." The
second exception occurs in section 25 of the Children and Young Persons Act
1933, the product of an age happily innocent of teenage pop-stars and tennis
prodigies. It forbids anyone having the custody of a person under 18 to cause or
permit such person to go abroad for the purpose of singing, playing, performing or
being exhibited for profit without a licence from a police magistrate. The
magistrate may vary or revoke such a licence "for any cause which he, in his
discretion, considers sufficient."12 Now I would have no quarrel if the magistrate
were given power in his discretion to vary or revoke for sufficient cause, and that
may be what the sub-section means, but I could not accept that a cause could be
sufficient simply because the magistrate considered it so. The sub-section may,
however, only be intended to provide that the police magistrate's judgment on
sufficiency should be final, not an unusual provision when decisions are entrusted
to administrators. These exceptions may scratch the paintwork of my definition;
they do not, I think, hole it below the water-line.

It might be thought unnecessary to stress that the judge has no discretion in
making findings of fact. But judges do sometimes describe fact-finding as
discretionary. That learned and accurate judge, the late Sir Brian Mac Kenna (who
resembled Gibbon in nothing save his account of what he owed the University of
Oxford), referred to "the judge's other great discretionary power, that of finding
the facts when he tries a case alone." 13Lord Brightman referred to the existence or
non-existence of a fact as being left to the judgment and discretion of a public
body.l4 Justice Barak of the Israel Supreme Court has written "The first area of
judicial discretion deals with deciding the facts."15 Now it is one thing to say that
the responsibility of finding the facts is entrusted to a particular person or body, be
he judge, arbitrator, official or public authority, and that such finding is to be
treated as conclusive or virtually so. But it is quite another to describe that
function as discretionary. It is, I suggest, nothing ofthe kind. In finding the facts
the judge's job is to consider all the conflicting evidence this way and that and
decide as best he can where the truth lies. It is very much the task performed, for
instance, by the historian or the journalist as part of his stock in trade. The judge is

12. S. 25 (5)(a).
13. "Discretion", The Irish Jun'st , Vol. IX (new series) I, at p.9.
14. R. v. Hi/lingdon London Borough Council, ex p. Pulhofer [1986] A.C. 484, at p.518.
15.Judicial Discretion (1989), at p.l3.

29



THE DENNING LAW JOURNAL

of course constricted by formalities and rules of evidence which do not afflict
them. On the other hand, he has powers of compelling testimony which they
would envy. It is nonetheless essentially the same function. Yet to say of a historian
or a journalist that he exercised a discretion in reaching conclusions of fact would,
I suggest, be regarded as libellous. The judge must exercise judgment, not
discretion, in finding the facts, and it is usually the most difficult and often the
most exacting task which the civil trial judge has to undertake. It calls for a degree
of rigour which is disguised by references to choosing between competing
accounts of a disputed event or preferring the evidence of one witness to that of
another, and to speak of discretion in this context is to open the door to potentially
dangerous habits of thought. It can lead to such absurdities as assessing expert
evidence on the demeanour of the expert or such errors as finding a fact to be
established because it has been denied by a witness held to be unreliable. It can
encourage excessive reliance on the judge's hunch and intuition, neither of them
an invariably safe basis for decision. 16The judge must decide, in as objective a
manner as the materials permit, which version of a disputed event (if either) he
accepts as the more convincing; once he has done so he has no choice, whether that
conclusion makes the overall resolution of the case more difficult or less so. When
reference is made to the trial judge's discretion to decide the facts, what is really
meant is that appellate courts will usually be reluctant to interfere with his findings
because he, having seen and heard the witnesses, is in a better position to decide
whose evidence is reliable than anyone else. In cases turning largely on oral
evidence, this is doubtless very often true: the trial judge's immediate contact with
the witnesses and the unfolding drama of litigation gives him insights denied to
those who come later. It is the advantage which the journalist on the scene at the
time enjoys over the historian. And even if the judge may be wrong, no one else can
be sure of being right. But it is well, even in this class of case, to preserve a measure
of scepticism. As Lord Wilberforce has recently observed:

"English judges entertain the belief that they can tell if a man - or even a
woman - is speaking the truth. This is a Palladium: and it has comforting
consequences: 'The judge saw the witness in the box - observed his
demeanour'. 'He was disbelieved by the judge - or the jury'. 'We (the appeal
court) cannot interfere'. But there is not much scientific basis for this. Such
studies, as I know of, show that liars are believed as often as truth-tellers are
disbelieved. And one can test it with multiple tribunals - e.g. arbitrations,
whether all British or from different nations. I can give several instances
where exactly opposite views as to credibility were confidently given by
members of such tribunals - a fact which encourages people to avoid oral
evidence before them. Indeed, one often finds foreign arbitrators irritated
with the English style of examination and cross-examination - it is not a
good way of getting at the truth or persuading the tribunal."17

16. And seeCurrem Legal Problems (1985) 1, at p.7.
17. 5 Arbitration Imernational (1989), at p.349.

30



THE DISCRETION OF THE JUDGE

Further, the very implediacy of the trial judge's impressions can sometimes
cloud his judgment. There are, I thjnk, more cases than is generally acknowledged,
particularly those largely dependent on documentary or expert evidence, in which
the insights of the trial judge are less reliable than the more detached reflection of
an appellate court as those of the journalist sometimes are than those of the
historian. I fear I have digressed, but I would wish firmly to exclude the notion of
discretion from the very important area of factual decision.

In boldly asserting that a judge does not exercise discretion in giving his rulings
on the law I have, I appreciate, side-stepped a very high level philosophical debate
conducted by Professors Hart, Dworkin, MacCormick, Raz and others. 18 I fear
that my offering on the sergeant directed to take his five most experienced men on
patrol would be of small value anyway, but it is plain that decisions on the law fall
well outside my definition of judicial discretion. This is not in any way to criticise
the usage of others; the English language is a rich pasture open to all. But on this
point at least I think I can rely on the invariable usage of the practising profession
in this country. It is a distinction regularly drawn when leave is sought to appeal
against an interlocutory decision, the judge's usual practice being to grant leave if
he has decided a question of law and to refuse it if he conceives himself to have
exercised a discretion. There may well be uncertainty where discretion begins and
ends but there is unanimity that it falls short of legal rulings. I would, however,
offer one observation prompted by the philosophical debate. There are occasions
when judges think that they are required to elect between different legal solutions
and, in effect, create new law. Usually, even in cases that turn on the law, the
question is whether the given case falls within an established principle or which of
two established principles governs it. But there are occasions when existing lines of
authority fall short of the given case and the question is whether they should be
extended to cover it. Donoghue v. Stevenson19 is perhaps the most obvious example;
hence, no doubt, the divergence of opinion in that case. There are other, rarer,
occasions when a problem seems to occur in an authoritative desert, where the
usual aids - principle, precedent, dicta and the opinions of learned authors - are
virtually absent. The judge cannot then simply extend the line on an existing graph
of authority because there is none. It is surely true that when judges buttress their
conclusions with references to public policy, commercial good sense, certainty,
good industrial relations and so on it is because they are conscious of making a
choice and are, quite properly, concerned to justify the choice that they have made.

The role of discretion in the balancing exercises which the courts are, it seems
increasingly, invited to carry out is not altogether easy to determine. For instance,
in balancing the public interest in withholding official documents against the
public interest in the administration of justice,2° or in weighing the competing

18. See,e.g. , H. L.A. Hart, The Concept of Law (1961); R. Dworkin, Taking Rights Sen'ously (1977); N.
MacCormick,Legal Reasoning and Legal Theory (1978); J. Raz, The Authority of Law (1979); A. Barak,

Judicial Discretion (1989).
19. [1932] A.C. 562.
20. Conway v. Rimmer [1968] A.C. 910, at p.952 A.

31



THE DENNING LAW JOURNAL

public interests in disclosure and non-disclosure,21 or for and against publication22
or in the maintenance of confidence against disclosure,23 or between maintaining
professional confidences and protecting the public against possible violence,24 is
the judge exercising a discretion and if so to what extent? Some would, I think,
regard this exercise as largely if not wholly discretionary.25 But I do not think the
factors to be put into the scales are the subject matter of discretion. They are
matters established, or held to be established, by evidence, or more rarely matters
of which judicial notice may be taken. Discretion has nothing to do with that, as I
have already suggested. In the English Spycatcher26 litigation, for instance,
detailed evidence was adduced to show why publication would damage national
security; the trial judge discounted most of this, but in doing so he was not, at least
in my terms, exercising a discretion. In assessing the weight to be given to various
factors for and against any decision, much must of course turn on the judgment of
the individual assessor, and for this reason an appeal court will be slow to interfere
with a value judgment on which reasonable minds could differ. Strictly, however,
I think it is probably only at the last stage of striking a balance and granting or
refusing relief that one moves into the realm of pure discretion. For this view, I
hope I may enlist the aid of Lord Diplock. In Birkett v. James,27 speaking of
interlocutory decisions, he said:

"They are decisions which involve balancing against one another a variety of
relevant considerations upon which opinions of individual judges may
reasonably differ as to their relative weight in a particular case. That is why
they are said to involve the exercise by the judge of his 'discretion' ... when
leave is granted, an appellate court ought not to substitute its own
'discretion' for that of the judge merely because its members would
themselves have regarded the balance as tipped against the way in which he
had decided the matter . . ."

That may sound like somewhat doubtful aid. But on both occasions when Lord
Diplock, a stickler for accurate usage, spoke of discretion in that passage he put the
expression in inverted commas, to indicate, as I infer, that he was not referring to
an exercise of discretion properly so called.

The role of judicial discretion today is, I suggest, subject to one exception, fairly
narrowly confined. For that proposition I summon up the weighty support of Sir
Wilfred Greene MR:

21.D. v.NSPCC[1978] A.C.171,atp.219C.
22. British Steel Corporation Ltd. v. Granada Television Ltd. [1981] A.C. 1096, at p.1202 C.
23. A.G. v. Guardian Newspapers (No.2) [1990] A.C. 109, at p.282.
24. W. v. Egdell [1990] 2 W.L.R. 471, at p.491 A.
25. E.g., Barak,Judicial Discretion (1989), at p.68.
26. A.G. v. Guardian Newspapers (No 2)[ 1990] A.C. 109.
27. [1978] A.C. 297, p.317 D.

32



THE DISCRETION OF THE JUDGE

"In all discretionary remedies it is well known and settled that in certain
circumstances - I will not say in all of them, but in a great many of them - the
Court, although nominally it has a discretion, if it is to act according to the
ordinary principles upon which judicial discretion is exercised, must
exercise that discretion in a particular way, and if a judge at a trial refuses to
do so, then the Court of Appeal will set the matter right."28

That was said 50 years ago, but it is even truer today, as can (I think) be
demonstrated by reference to the areas in which discretion is most obviously
exercised.

I start with the award of costs. A series of statutes entrusts this to the discretion
of the court or arbitrator.29 Notably, section 51 of the Supreme Court Act 1981
provides that costs shall be in the discretion of the court which "shall have full
power to determine by whom and to what extent the costs are to be paid." So
special is this discretion that rights of appeal are specifically restricted,30 and its
width has been recently emphasised by the House of Lords.31 It is nevertheless
universally recognised as the primary principle governing courts and arbitrators in
the exercise of their discretion that costs should follow the event32 and to depart
from that rule without showing sufficient reason is to raise a rebuttable
presumption of error. 33It is of course necessary to identify the event, which in a
tangle of claims and cross-claims may not be straightforward, but a party's
entitlement to receive costs or responsibility to pay them is first to be judged by
reference to his success or failure in the litigation.34 This primary principle may
not be applied in its full rigour, for example where a party's conduct in the
litigation35 justifies a departure or where the successful party recovers less than
had earlier been offered or paid into court,36 but a well-established body of
authority37 shows that a judge is by no means free to indulge his personal whims or
prejudices when awarding costs. The reality is revealed by the traditional
exchange between the court and counsel for the unsuccessful party when the victor
asks for costs. "You can't resist that, can you, Mr X?" to which the answer, much
more often than not, is "No" although, at least in the Court of Appeal, he usually
adds "But I have an application to make." A discretion exists, but within a
compass which is well understood and has, I think, shrunk over the years.

The field Sir Wilfred Greene had in mind in the passage quoted was that of

28. R. v. Stafford Justices [1940] 2 K.B. 33, at p.43.
29. E.g., Arbitration Act 1950, s.18; Magistrates Courts Act 1980, s.64; Highways Act 1980, s.209;
County Courts Act 1984, s.45; Agricultural Holdings Act 1986, Sched. II, para 23.
30. Supreme Court Act 1981, s.18(1)(I).
31. Aiden Shipping Co. Ltd. v. Interbulk Ltd. [1986] A.C. 965, at pp.975, 979 D.
32. The Erich Scroeder [1984] I Lloyd's Rep. 192, at p.194.
33. Tramountana Armadora SA v. Atlantic Shipping Co. SA [1978] I Lloyd's Rep. 391, at p.394.
34. Archital Luxfer Ltd. v. Henry Boot Construction Ltd. [1981] I Lloyd's Rep. 642, at p.650.
35. Ibid., at p.650; Video Box Office Ltd. v. GST Holdings Ltd., [1990] The Independent, 27 April.
36. The los I [1987] I Lloyd's Rep. 321.
37. See Supreme Court Practice (1988) Voll, 6212/l0et seq.

33



THE DENNING LAW JOURNAL

remedies, and it is elementary that equitable remedies are discretionary and not a
matter of right. But over a century has now passed since Lord Blackburn said:38

"The jurisdiction of the Court of Equity to enforce the specific performance,
or to grant an injunction to prevent the breach of a covenant, is no doubt a
discretionary jurisdiction, but I perfectly agree with the view expressed by
your Lordships that the discretion is not one to be exercised according to the
fancy of whoever is to exercise the jurisdiction of Equity, but is a discretion
to be exercised according to the rules which have been established by a long
series of decisions, and which are now settled to be the proper guide to
Judges in Courts of Equity."

Even earlier Lord Kingsdown had said:39

"The rule I take to be clearly this: if a Plaintiff applies for an injunction to
restrain a violation of a common law right, if either the existence of the right
or the fact of its violation be disputed, he must establish the right at law; but
when he has established his right at law, I apprehend that unless there be
something special in the case, he is entitled as of course to an injunction to
prevent the recurrence of that violation."

The circumstances which may deprive a plaintiff of his remedy are well
established: there is no likelihood of repetition; damages will be an adequate
remedy; the plaintiff has acquiesced in the defendant's conduct; the plaintifPs
conduct has disentitled him from seeking relief; and so on. Whether these
impediments exist may involve an exercise of judgment, or findings of fact which
an appeal court may be slow to disturb, but will not involve an exercise of
discretion. If they do not exist, whether the remedy be injunction or any other
equitable remedy, there is virtually no ground for refusing relief. If they do, an
exercise of discretion will be called for but usually within a strictly confined area
and it will be readily reviewable. The recently developed field of Mareva
injunctions illustrates the point very clearly: the conditions for granting such an
injunction have been clearly laid down;40 where the affidavit evidence shows the
conditions to be met, the judge is almost bound to grant relief. Many a judge,
instinctively reluctant to grant this draconian relief, has in practice found it almost
impossible on a reasonably well-prepared application to find grounds for refusing
to do so.

At this point the well-informed schoolboy would doubtless interject that
remedies in public law at least must be truly discretionary, because the judges are

38. Doherry v.Allman (1878) 3 App. Cas. 709, at p.728.
39. Imperial Gas Light & Coke Co. v. Broadbent (1859) 7 H.L.e. 600, at p.612. See also Fullwood v.
Fullwood (1878) Ch. D. 176, at p.179.
40. See Gee, Mareva Injunctions & Anton Piller Relief, 2nd ed. (1990), at pp.10-11.

34



THE DISCRETION OF THE JUDGE

always saying so. He would be right. They are. And the judges are right too. It is
therefore possible, as Professor Wade says,41"that the court may find some act to
be unlawful but nevertheless decline to intervene." But there is a wide gap
between the rhetoric and the reality. The constraints of time and the limitations of
my own research do not enable me to discuss the subject in detail or pronounce
with authority. I have, however, a very strong hunch that in the decade since the
Crown Office List became a boom town there would turn out to be no more than a
handful of cases in which excess or abuse of power had been established but a
remedy refused. In these few cases, the reason for the refusal of relief would (I
suspect) usually turn out to be that the applicant had disentitled himself to relief
by his own conduct, that the illegality was of a technical nature,42 that delay in
application had made relieffutile or impracticable, or that the public authority was
willing to give effect to the decision without the formality of an order against it.
The much trumpeted principle that an applicant for judicial review must first
exhaust his other remedies43does not seem often in practice to have led to a denial
of relief.

In the procedural field at least one might expect the judge, as master of
proceedings in his court, to exercise an almost unfettered discretion. Many
statutes and rules of court confer apparently wide discretions, but each is quickly
confined between banks of practice and authority. The court may extend the
validity of a writ, says the rule;44but only for good reason say the cases,45which
also give guidance on what may and may not be a good reason.46 If certain
conditions are satisfied, the court may order the plaintiff to give such security for
the defendant's costs as it thinks just if having regard to all the circumstances of the
case it thinks it just to do SO;47but the principles on which the discretion is
exercised are in general so well understood that contests save as to amount are
relatively unusual. 48The court may stay an action on grounds of forum non
conveniens, but authoritative guidance on the exercise of the discretion is now
found in The Spiliada.49 If certain conditions are satisfied, the court may if it
thinks fit make an order for interim payment of such amount as it thinks just;50but
experience shows the mortality rate among such orders to be high.51The court
may, if of opinion that in the circumstances of the case undue hardship would
otherwise be caused, and on such terms if any as the justice of the case may require,

41. Wade, Administrative Law 6th ed. (1988), p.709.
42. E.g., R. v. Governors of Bacon's School, ex p. ILEA [1990] The Independent, 29th March.
43. Wade, supra n. 41, at p.714.
44. RSC O. 6 r. 8(2).
45. In panicular,Kleinwort Benson Ltd. v.Barorak Ltd. [1987] A.C. 597.
46. See generally, Supreme Coun Practice 1988, 6th Cum. Supp., p.9, para. 6/8/3.
47. RSC O. 23 r. 1(1).
48. The advent of the European Community has, however, raised new questions: Porzelack KG v.
Porzelack (UK) Ltd. [1987] I W.L.R. 420; De Bry v. Fitzgerald [1990] I All E.R. 560.
49. [1987] A.C. 460.
50. RSC O. 29 r. II, 12.
51. See, e.g. ,British & Commonwealth Holdings p.l.c. v. Quadrex Holdings Inc. [1989] Q.B. 842.

35



THE DENNING LAW JOURNAL

extend the time for commencing arbitration proceedings;52 but the judge called
upon to exercise this jurisdiction is well advised to have regard to the guidance
given by Mr Justice Brandon in The Jocelyne53 which earned the approval of Lord
Justice Brandon in The Aspen Trader.54 Perhaps no clearer example of how
practice develops can be found than in section 69 of the Supreme Court Act 1981
which requires certain specified causes of action to be tried by a jury and provides
that any other action shall be tried without a jury "unless the court in its discretion
orders it to be tried with a jury." Here, one might think, was a generously framed
discretion, capable of being exercised in favour of jury trial for a difficult personal
injury case. So, it would seem, the master and the judge thought in Ward v.
James,55 decided under the section's predecessor. Yet, said Lord Denning MR:56

"it is of the first importance that some guidance should be given - else you
would find one judge ordering a jury, the next refusing it, and no one would
know where he stood. It might make all the difference to the ultimate result
of the case. This would give rise to much dissatisfaction. It is an essential
attribute of justice in a community that similar decisions should be given in
similar cases, and this applies as much to mode of trial as anything else. The
only way of achieving this is for the courts to set out the considerations which
should guide the judges in the normal exercise of their discretion. And that is
what has been done in scores of cases where a discretion has been entrusted
to the judges."

So guidance was given, and jury trials save in the specified cases have vanished
from the civil scene, it would seem likely for ever.57

Recent experience prompts me to mention an example, drawn from quite a
different field, of an apparently wide discretion legislated almost out of existence
by judicial decision and finally overtaken by statute itself. Section 30 of the Law of
Property Act 1925 permits the trustee in bankruptcy of a bankrupt husband to
apply to the court for an order that property owned jointly by husband and wife be
sold, and the court may make such order as it thinks fit. This again would have
seemed to give the court a very wide discretion to do what seemed right to reflect
the respective interests of the creditors on the one hand and the wife and children
on the other. But a long line of cases held that the trustee was ordinarily entitled to
an order for sale in the absence of "very special circumstances" or "good reas'ons"

52. Arbitration Act 1950, s. 27.
53. [1977] 2 Lloyd's Rep. 121.
54. [1981] I Lloyd's Rep. 273.
55. [1966] I Q.B. 273.
56. Ibid , at p.293.
57. My own personal impression is confirmed by the experience of Master Warren QC, the Senior
Master of the Queen's Bench Division, who knows of no such case since Ward v.James except Hodges v.
Harland & Wolff Lrd. [1965] I All E.R. 1086, decided some 3 weeks later. And see, most recently Singh
v. London Underground Lrd. [1990] The Independenr, 25th April.

36



THE DISCRETION OF THE JUDGE

or "a substantial case of hardship" or, finally, "exceptional circumstances". 58And
now the opportunity to question whether a test of exceptional circumstances does
not unreasonably constrict the language of the Act has passed, for it has been
woven into the 1986 Insolvency Act.59

I suppose most of us would today regard the criminal trial as the real sanctuary
of judicial discretion. This has a large measure of truth, although even an
experienced criminal judge, on reading Rosemary Pattenden's book The Judge,
Discretion and the Criminal Trial, would (I think) be surprised at the wealth of his
(or her) endowment.6o The reason is not far to seek: during the trial the court's
management decisions are to a large extent immune from challenge; and once the
trial is over, all but the most obviously wrong exercises of discretion tend to be
superseded by the verdict of the jury. Pattenden, however, observes61 that over the
course of this century the attitude of the courts has changed and the court's
reluctance to interfere with an exercise of discretion has waned.

A glance at the more recent case law - quite apart from the recent decision
concerning the Irish conspirators to murder Mr Tom King - bears her out. 62Thus
over the last few months alone the court has interfered with discretionary decisions
to allow committal proceedings to continue,63 not to stop a case going to the jury,64
not to discharge a jury,65 to accede to a jury's request to be supplied with scales,66
to interrupt excessively,67 not to direct the jury that previous convictions for
dishonesty were relevant only to credibility68 and, on many occasions, to admit
admissible but prejudicial evidence. One recalls that twenty years ago the Court of
Appeal (Criminal Division) declined to interfere when a chairman of quarter
sessions had on repeated occasions during a defendant's case observed in a loud
voice "Oh, God", and then laid his head across his arm and made groaning and
sighing noises. Counsel had not, the Court held, been positively and actively
obstructed in the doing of his work and the chairman's conduct, if it might be
regarded as discourteous and as showing signs of impatience, disparaged only the
defendant's counsel, not his case.69 I do not think this decision was found
convincing even at the time, at any rate by advocates with personal experience of
the chairman in question, but I am quite confident that the judge's discretion to
conduct a criminal trial as he wishes would today be much more rigorously
scrutinised.

58. Re Holliday [1981] Ch 405, at pp.419 G, 420 B, E 415 F, 424 C, 425 H;ReLowrie [1981] 3 All E.R.
353, at p .355 j.
59. S. 356 (5).
60. Op. cit., Appendix, at p. I83.
61. Op. cit., at pp.21, 22.
62. R. v. Cullen and Others [1990] The Independent, 1 May.
63. R. v. Sunderland Magistrates Court, ex p. Z. [1989] Crim.L.R. 56.
64.R. v.Morley [1989] Crim.L.R. 566.
65. R. v.Jaquith, Emode [1989] Crim.L.R. 563.
66. R. v. Stewart SapplelOn [1989] Crim.L.R. 653.
67.R. v.Renshaw[1989]Crim.L.R. 811.
68. R. v. Prince [1990] Crim.L.R. 49.
69.R. v.Hircock, Farmer, Leggett [1970] 1 Q.B. 67.

37



THE DENNING LAW JOURNAL

The accelerating tendency towards a narrowing of discretion is nowhere better
illustrated than in the field of sentencing. As long ago as 1361 justices of the peace
were empowered to punish offenders "according to that which to them shall seem
best to do by their discretions and good advisement", 70and in later centuries the
sentencing discretion was often described in statutes as such.71 Interestingly, in
view of the new (and to my mind obnoxious) American practice of prescribing
maximum and minimum penalties within a very narrow band, one may note that
the Slave Trade Act 1824 imposed as the penalty for dealing in slaves,
transportation for a term not exceeding 14 years or hard labour for a term not
exceeding 5 years nor less than 3 years at the discretion of the court. The practice of
prescribing minimum sentences did not, happily, catch on. It does, however, seem
almost incredible, looking back, that there should for so many centuries have been
no effective judicial means of challenging the exercise of the sentencing discretion
on the ground that a penalty, though lawful, was excessive. The first statutory step
towards controlling the discretion of course came with the establishment of the
Court of Criminal Appeal in 1907, which could substitute such sentence as it
thought should have been passed at the trial (whether more or less severe).72 But
the Court got off to a slow start: such was the respect felt for the sentencer's
discretion that in 1908, when the Act was in operation for nine months, only 14
sentences were reduced73 - roughly the tally today on an average week-day
morning. It might be objected that giving a right of appeal against sentence was
merely to substitute the discretion of three judges for that of one, but any appellate
court is in particular constrained to build up a body of precedent and seek to
achieve a reasonable level of consistency, so that the clear result was progressively
to narrow the sentencer's discretion. But I think it is really only in relatively recent
times that the judge's discretion in passing sentence has been subjected to the
degree of discipline familiar in other fields. To this a number of causes have
contributed. One, without doubt, has been the entry into the field of distinguished
academic analysts, particularly Dr David Thomas, but also others. Linked with
that is the systematic reporting of decisions on sentence. Another cause is the work
of the Criminal Division itself. When, twenty years ago, Widgery LJ spoke of the
well-known duty of the Criminal Division "to lay down principles and guidelines to
assist sentencers of all grades in the application of the discretion which the
imposition of sentence requires", 74he was speaking no less than the truth. But it is
over the last decade that, for the first time, a serious attempt has been made to
provide detailed guidance on sentencing in a systematic, rational and explicit way.

70. Justices of the Peace Act 1361.
71. E.g., Offences against the Person Act 1861; Foreign Enlistment Act 1870.
72. The Criminal Appeal Act 1907, s.4(3).
73. Holdsworth, History of English Law, vol. I, at p.218.
74. R. v. Newsome and Browne (1970) 54 Cr. App. Rep. 485, at p.490.

38



THE DISCRETION OF THE JUDGE

One thinks of the landmark decisions in such fi~lds as rape,75 the importation of
drugs,76 theft in breach of trust, 77serious disorder,78 killing by dangerous driving,79
and others. For years it was customary to challenge sentences as being wrong in
principle but it was often far from easy to identify any principle. The introduction
of principle into this field will, I think, be seen as the enduring and in many ways
personal achievement of the present Lord Chief Justice. This is a development to be
whole-heartedly welcomed. Without explicit guidelines there can be no informed
public debate on sentencing practice, and the appropriate punishment of offenders
against society is a proper matter for consideration by society (if for purposes of
argument one assumes there to be such a thing). Such consideration is no more an
encroachment on the judges' independence than the prescription of maximum
penalties by statute, which has always of course been accepted. It is absurd to
suppose that the judges could properly have a sentencing policy of their own
independently of the society of which they form part. Another cause contributing to
the narrowing of discretion has been vocal public dissatisfaction with a sentencing
-regime which has been frequently portrayed, on occasion fairly, as arbitrary, even
whimsical, and inadequately controlled. This dissatisfaction has not in the main
arisen from sentences seen as unduly severe, which have after all been amenable to
review on appeal, and the cases show that the Criminal Division has been willing
(often to the fury of sentencers) to make quite minor adjustments of sentence when
it felt that justice required them. The real clamour has concerned sentences seen as
unduly lenient and the new power to refer such sentences to the Court80 must be
understood both as a response to that body of opinion and as a guarantee of the
judicial guidelines. The upshot of all these developments is certainly not that the
sentencer has lost his discretion. He has not. The guideline cases are only
guidelines. The facts of two cases and the personal circumstances of two defendants
are never the same. The primary responsibility of passing the appropriate sentence
remains with the sentencer, whose decision in the vast majority of cases is never the
subject of appeal. But I think it is undoubtedly true that this discretion has become _
much more judicial in the sense that it is exercised in a much more structured
framework and is subject to much narrower constraints and is much more readily
reviewable than it ever used to be.

In a quite different corner of the criminal field, I think we have a golden
opportunity to observe the organic development of a somewhat novel discretion in
something approaching laboratory conditions. The criminal judge has long
enjoyed a discretion to exclude evidence of which the prejudicial effect is thought
likely to outweigh the probative value. But the Police and Criminal Evidence Act
1984 moved the goalposts: it provided for the issuing of codes governing such

75. R. v. Bil/am (1986) Cr. App. R.(S) 48.
76. R. v.Aramah (1982) 4 Cr. App. R.(S)'407;R. v. Martinez (1984) 6 Cr. App. R.(S) 364.
77.R. v.Barrick (1985) 81 Cr. App. R. 78.
78. R. v. Keys (1986) 8 Cr. App. R. (S) 444.
79. R. v.Boswell (1984) 6 Cr. App. R. (S) 257.
80. Criminal Justice Act 1988, s. 36(1).

39



THE DENNING LAW JOURNAL

matters as the questioning of suspects,81 provided that the codes should be taken
into account in determining any question in proceedings to which they were
relevant,82 gave suspects a right under section 58 (albeit qualified) to take legal
advice and (relevant for present purposes) conferred a discretion on the court
under section 78(1) to "refuse to allow evidence on which the prosecution
proposes to rely to be given if it appears to the court that, having regard to all the
circumstances, including the circumstances in which the evidence was obtained,
the admission of the evidence would have such an adverse effect on the fairness of
the proceedings that the court ought not to admit it." How was this discretion to be
exercised? As was observed in a recent case, with some degree of under-statement,
"When the Act came into force the effect which the Courts would give to this
section was in doubt."83 One view, although not so crudely put, was that the
proceedings did not become unfair if evidence was admitted which led to the
conviction of a guilty defendant, whether or not a breach of the non-binding codes
had occurred. This was not much different from the approach previously taken to
breaches of the Judges' Rules. The other view was that since Parliament had
intended a defendant to enjoy certain safeguards the proceedings became unfair if
he was denied them and was convicted as a result. Plainly the risk existed that
different judges would adopt highly divergent practices, with the unacceptable
consequence that a decision of possibly crucial importance to the criminal
defendant would turn on the predilections of the individual decision-maker. One
could not, I think, claim that any wholly coherent approach to this very wide
discretion has yet been formulated, but if one looks at the cases where section 58 or
the codes have been breached and reliance is placed on section 78, in the relatively
very short period since the 1st January 1986 when the section came into force, the
elements of such an approach may perhaps be discerned. It has been repeatedly
stated that a breach of section 58 or the codes does not of itself require evidence of a
confession to be excluded,84 but there has nevertheless been a growing and
possibly even exaggerated tendency to exclude. In the cases where evidence has
been held to be rightly admitted, despite breaches, the reason has been given that
the defendant was well aware of his rights and access to a solicitor would not have
improved his position85 or more generally that in all the circumstances of the
particular case the breaches did not affect the fairness of the proceedings. 86But in
most of the cases where breaches have been established evidence has been
excluded or it has been held that it should have been. The right to legal advice has
been described as fundamental. 87 In a case where there had been wholesale

81. S. 66.
82. S. 67(1 I).
83. R. v. Keenan (1989) 90 Cr. App. R. I, at p.6.
84.R. v.Keenan (1990) 90 Cr. App. R. 1, at p.12;R. v. Walsh {1989] Crim.L.R. 822, at p.823;R. v.
Matthews {199O] Crim.L.R. 190.
85. R. v. Alladice (1988) 87 Cr. App. R. 380;R. v.Dunford {1990] Thelndependent, 30th March.
86. R. v. Waters {1989] Crim.L.R. 62;R. v. Mauhews [1990] Crim.L.R. 190.
87. R. v. Samuel [1988] Q.B. 615, at p.630;R. v. Beycan [1990] Crim.L.R. 185, at p.186.

40



THE DISCRETION OF THE JUDGE

breaches it was said that to admit the evidence would be to condone flouting of the
provisions designed to protect against confessions which were not genuine. 88 But it
is now clear that in the ordinary way evidence will only be excluded if the breaches
are significant and substantialB9 and the crucial consideration has usually been
whether, because of the breaches, the defendant gave answers he might not have
given90 or has otherwise been prejudiced in resisting the charge.91 It may be that
the pendulum has swung too far towards exclusion upon breaches being shown,
without adequate consideration of the effect on the fairness of the proceedings
which the Act requires.92 My point, however, is that within about five years of the
Act coming into force one sees a discretion drawn in very wide terms being defined
and regulated so that the lines upon which judges should exercise it are likely quite
soon to become clear and well-understood. While prediction is hazardous, the
general shape of the rule seems likely to be that the judge should ordinarily exclude
prosecution evidence if there have been significant and substantial breaches of
section 58 or the codes and the defendant has as a result given damaging answers
which he would not otherwise have given or has been substantially prejudiced in
resisting the charge. If this were the broad shape of the rule, difficult borderline
cases would no doubt arise, but in the great mass of cases it would be obvious to
judges and practitioners how the discretion should and would be exercised. One
would also expect that as the consequences of failure to comply with these highly
detailed provisions were increasingly borne in upon police forces, the incidence of
significant and substantial breaches would sharply decline.

Perhaps the last real stronghold of almost unreviewable discretion is where the
care and custody of children are concerned. This is not because demonstrable
errors in the judge's balancing exercise in this field will not be corrected: authority
shows that they can and should.93 It is because, first and most importantly, the
evidence is likely to be entirely oral and the issue is likely to turn on the judge's
assessment of the personal qualities and motives of the competing parents and
other members of the extended family whose capacity as carers is in question. In
the absence of some striking mis-judgment it will be almost impossible to show
that his conclusion is wrong. And the judge's discretionary decision enjoys a rare
inviolability, secondly, because there is in this field usually no satisfactory solution
and it is thus impossible for an appellate court to be confident that its view of the
less unsatisfactory solution, if different from that of the judge, is superior.94 The

88. R. v.Ismail [1990] Crim.L.R. 109, at p.1I0.
89.R. v.Absolam [1989] Cr. App. R. 232, at p.337;R. v.Keenan (1990) 90 Cr. App. R. I. at p.13;R. v .

. Mauhews [1990] Crim.L.R. 10, at p.191.
90. R. v. Samuel [1988] Q.B. 615, at p.630;R. v. Absolam (1989) 88 Cr. App. R. 332, at p.337;R. v.
Delaney (1989) 88 Cr. App. R. 338;R. v.Quayson [1989] Crim.L.R. 218;R. v. Walsh [1989] Crim.L.R.
822;R. v.Beycan [1990] Crim.L.R. 190.
91.R. v.Parris (1989) 89 Cr. App. R. 68;R. v.Keenan (1990) 90 Cr. App. R. I;R. v.Fennelley [1989]
Crim.L.R. 142;R. v.Britton [1989] Crim.L.R. 144;R. v.Ladlow [1989] Crim.L.R. 219.
92. R. v. Williams [1989] Crim.L.R. 66;R. v. Fogah [1989] Crim.L.R. 141.
93. Re F (a minor) (Wardship: appeal) [1976] Fam. 238; G. v. G. [1985] 2 All E.R. 225.
94. Clarke-Hunt v. Newcombe (1982) 4 F.L.R. 482, at p.488; G. v. G., supra n. 93, at p.228 b.

41



THE DENNING LAW JOURNAL

result is, that since the House of Lords endorsed this general approach in G v. G,95
challenges by the aggrieved parent to the trial judge's exercise of discretion have
almost always failed. It is not perhaps very happy that an unfettered right of appeal
should be effectively abrogated by judicial decision, nor that, in a field where
judicial decisions have a unique capacity to cause lasting misery, the trial judge's
decision should be effectively final. On the other hand, it would be very hard
indeed to suggest any guideline to govern the exercise of this discretion which was
not either so obvious or so heavily qualified as to be futile. It would seem that in
this limited field, for better or worse, reliance must be placed on the trial judge to
show the wisdom, sensitivity and insight of Solomon, although lacking the latter's
extra-judicial powers.

May I, in conclusion, touch - much too briefly - on the important issues raised
by Professor Atiyah and Professor Treitel in their inaugural lectures respectively
twelve and ten years ago?96 Atiyah, it will be recalled, drew attention to the
movement over the last century from clear, sharp-edged principles beloved of our
Victorian forbears to judgmental discretionary rules thought to be better fitted for
dispensing individualised ad hoc justice case by case. Treitel considered this
development further with particular reference to the law of contract, showing how
certain old rules had been diversified and qualified to cater for hard cases, but
suggesting that in the contractual field discretions had not, on the whole, up to
then operated so as to create an unacceptable degree of uncertainty. Now certain of
the instances discussed by the professors are certainly discretions within my
suggested meaning: for example, the rules relating to matrimonial property97 or
the provision in section 49(2) of the Law of Property Act 1925 enabling the court to
order the return of a deposit paid under a contract for the sale ofland. 98But other
instances, although posing judgmental tests, do not involve the exercise of any
discretion in my sense: for example, the pervading test of reasonableness in the law
oftort,99 the closest connection test in determining the proper law of a contract, I
the test of merchantability now contained in section 14(6) of the Sale of Goods Act
1979,2 any test based on the intention of the parties,3 the new test of common law
duress,4 or departure from the rule that damages for breach of contract are
invariably to be assessed by reference to the time of breach. 5 None of these permits
the judge to apply a simple rule of thumb test. Most require evidence followed by
an exercise of judgment, the correctness of which may give rise to differences of

95. Supra, n. 93.
96. P.S. Atiyah,From PrinciplestoPragmarism, 17th Feb. 1978;G.H. Treitel,Docrrine and Discrerion in
rhe Law of COn/racr, 7th March 1980.
97. Atiyah,op. cir., at p.ll.
98. Treitel,op. cir., at p.15.
99. Atiyah,op. cir., at p.ll.
I. Atiyah,op. cir., at p.13.
2. Atiyah,op. cir., at p.14.
3. Treitel,op. cir., at p.4.
4. Treitel, op. cir., at p.5.
5. Treitel,op. cir., at p.9.

42



THE DISCRETION OF THE JUDGE

opinion. I share to the full the professors' distrust of undirected and unreviewable
discretions, but most of the powers they discuss are not, I think, undirected nor,
even more importantly, are they unreviewable. We may perhaps take comfort in
the fact that even the New Zealand Illegal Contracts Act 1970,which may fairly be
regarded as the ultimate in conferment of wide, undefined discretions, is now
said,6 despite gloomy earlier prognostications,7 to have produced a coherent body
of decisions. So I, with Treitel, would view the present situation, if not with
complacency, at least with a reasonable degree of optimism and confidence.8

It is, I think, a deeply rooted instinct of any responsible body, whether a
company, a college, a club, a body of trustees, a trade union or anything else,
however wide its powers, to endeavour to act with a reasonable measure of
consistency. So the tendency to subject a wide discretion to more or less restrictive
rules is not a specifically legal phenomenon. Nor, certainly, is it an English
phenomenon: a discretion conferred by the New Zealand Matrimonial Property
Act 1963 was so mechanically applied by the judges as, in the views of some, to
subvert the object of the legislation altogether.9

But it is, as shown by that example and others I have discussed, a very marked
feature of judicial practice. I do not, therefore, think that whether discretion is
understood in my sense or as embracing also the judgmental open-textured rules
discussed by the professors there is in general any ground for concern that
arbitrary uncontrolled discretion is likely to run riot at the expense of clear
discernible principle. Nor do I think that any approach less responsive to the
circumstances of particular cases would in general be acceptable to the public as
consumers of the judicial product in an age when the public is more inclined to see
the law as an ass than as the embodiment of everything that's excellent. But this is,
I am conscious, a predictable view from one on the bureaucratic side of the
counter. After all, the Lord Chancellor of former days no doubt considered the
length of his foot a very convenient, reliable and serviceable measure, happily free
of what would now be called resource implications.

6. A Beck, "Illegality and the Court's Discretion: The New Zealand Illegal Contracts Act in Action",
(1989) 13N.Z.U.L.R. 389.
7. M.P. Furmston, "The Illegal Contracts Act 1970- An English View", 1972,5 N.Z. U.L.R. 151.
8. Op. cit., at p.20.
9. See Markesinis, "Comparative Law - A Subject in Search of an Audience", (I990) 53M.L.R. I, at
pp.12-13.

43