BAILII [Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]

United Kingdom House of Lords Decisions


You are here: BAILII >> Databases >> United Kingdom House of Lords Decisions >> Hasan, R v [2005] UKHL 22 (17 March 2005)
URL: http://www.bailii.org/uk/cases/UKHL/2005/22.html
Cite as: [2005] 2 AC 467, [2005] 2 WLR 709, [2005] UKHL 22, [2005] 4 All ER 685

[New search] [Context] [Printable version] [Buy ICLR report: [2005] 2 WLR 709] [Buy ICLR report: [2005] 2 AC 467] [Help]


JISCBAILII_CASES_CRIME
Judgments - Hasan (Respondent) (On Appeal from the Court of Appeal (Criminal Division)) (formerly Regina v. Z (2003) (On Appeal from the Court of Appeal (Criminal Division))

HOUSE OF LORDS

SESSION 2004-05
[2005] UKHL 22
on appeal from: [2003] EWCA Crim 191





OPINIONS

OF THE LORDS OF APPEAL

FOR JUDGMENT IN THE CAUSE

Regina

v.

Hasan (Respondent) (On Appeal from the Court of Appeal (Criminal Division))

(formerly Regina v. Z (2003) (On Appeal from the Court of Appeal (Criminal Division))

ON

THURSDAY 17 MARCH 2005

The Appellate Committee comprised:

Lord Bingham of Cornhill

Lord Steyn

Lord Rodger of Earlsferry

Baroness Hale of Richmond

Lord Brown of Eaton-under-Heywood


HOUSE OF LORDS

OPINIONS OF THE LORDS OF APPEAL FOR JUDGMENT

IN THE CAUSE

Regina v. Hasan (Respondent) (On Appeal from the Court of Appeal (Criminal Division))

(formerly Regina v. Z (2003) (On Appeal from the Court of Appeal (Criminal Division))

[2005] UKHL 22

LORD BINGHAM OF CORNHILL

My Lords,

  1.   This appeal by the Crown against the decision of the Criminal Division of the Court of Appeal (Rix LJ, Crane J and Judge Maddison: [2003] EWCA Crim 191, [2003] 1 WLR 1489, sub nom R v Z) raises two questions. The first concerns the meaning of "confession" for the purposes of section 76(1) of the Police and Criminal Evidence Act 1984. The second concerns the defence of duress.

    Anonymity

  2.   At trial in the Central Criminal Court, the name of the defendant Aytach Hasan ("the defendant") and the names of the main participants in the proceedings were given in open court. But two of those participants (Frank Sullivan and Claire Taeger) were then awaiting trial and the trial judge, His Honour Judge Paget QC, properly made an order under section 4(2) of the Contempt of Court Act 1981 prohibiting the publication of their names or any information concerning them or their forthcoming trial. This trial has now taken place and the order has been discharged. In the Court of Appeal the names of the defendant, then the appellant, and the main participants were anonymised and the case was reported as R v Z. An order was made under section 11 of the 1981 Act prohibiting any publication, save in a complete report of the judgment or in a legal journal, of the fact that the defendant had spoken to a police officer about Sullivan and Taeger and of that officer's report of the conversation. Having invited submissions from the parties, I am of the clear opinion that the Court of Appeal had no power under section 11 to restrain publication of evidence given in open court and referred to openly in the judge's summing up. In this opinion the names of the main participants will accordingly be used.

    The facts

  3.   In brief summary, the relevant facts are these. The defendant had worked as a driver and minder for Claire Taeger, who ran an escort agency and was involved in prostitution. In about July or August 1999, according to the defendant, Sullivan became Taeger's boyfriend and also her minder in connection with her prostitution business. He had, the defendant said, the reputation of being a violent man and a drug dealer.

  4.   The prosecution alleged that on 29 August 1999 a man living in Croydon telephoned Taeger's agency asking for the services of a prostitute. The defendant went to the address with a prostitute. But the client had changed his mind and claimed that he had not made a telephone call. The defendant insisted that a 50 cancellation fee be paid, and forced his way into the house, producing a knife and demanding payment. The client went upstairs and opened a safe, whereupon the defendant took some 4000 from it and ran from the house. This incident founded the first count of aggravated burglary in the indictment later preferred against the defendant. But his account of the incident was quite different. He said that he had been given the 50 fee without any threat and had taken nothing from the safe. But he said that after this incident he had reported the existence of the safe and its contents to Taeger in the presence of Sullivan.

  5.   According to the defendant, his work for Taeger fell off with the arrival of Sullivan, who urged Taeger to get rid of him. There was a row in October or November 1999 and he stopped working for Taeger. But she lived in a flat which the defendant let to her, and she owed him outstanding rent. As security for this, he said, Sullivan made a red Rover car available to him, which he parked outside this flat. The next day it was gone, and he assumed that Taeger had a key and had taken it.

  6.   According to the defendant's evidence at trial, he saw Sullivan shortly before Christmas 1999. Sullivan said he was short of cash as he was doing a big cocaine deal. He wanted the key to the Rover, which the defendant said he would look for. Just after Christmas 1999, the defendant said, Sullivan visited him again. He again spoke of a cocaine deal, giving the defendant to believe he had killed two dealers. He also spoke of killing another man by injecting him with a heroin overdose. He offered to show the defendant the body of a man, Bryan Davies, in the boot of the Rover.

  7.   The second count of aggravated burglary in the indictment against the defendant related to an incident on 23 January 2000, involving the same house and the same victim as the earlier incident. The defendant admitted at trial that he had forced his way into the house on this occasion, armed with a knife, and had attempted to steal the contents of the safe, but claimed that he had acted under duress exerted by Sullivan, who had fortified his reputation for violence by talking of three murders he had recently committed. On the day in question, the defendant claimed, he had been ambushed outside his home by Sullivan and an unknown black man whom he described as a "lunatic yardie". Sullivan demanded that the defendant get the money from the safe mentioned on the earlier occasion, and told the defendant that the black man would go with him to see that this was done. Sullivan said that, if the defendant did not do it, he and his family would be harmed. The defendant claimed that he had no chance to escape and go to the police. The black man drove the defendant to the house and gave him a knife, saying that he himself had a gun. The defendant then broke into the house and tried unsuccessfully to open and then to remove the safe. The black man was in the vicinity throughout, and drove him away when the attempt failed.

  8.   Bryan Davies had died of a heroin overdose on 16 December 1999. On 14 April 2000 his body was discovered in the boot of the Rover, and the police believed that he had been injected with a fatal dose. Sullivan and Taeger were arrested and when interviewed said that the defendant had had the Rover in December 1999. They were awaiting trial at the time of the defendant's trial.

  9.   On 5 June 2000 the defendant was arrested and interviewed in relation to the two burglaries. He denied any involvement in either. The victims of the second burglary then identified him on an identification parade. He was charged and produced a note which began "I rely on duress". He gave no detailed particulars.

  10.   On 26 June 2000 the defendant was interviewed, in the presence of his solicitor, by police investigating the death of Bryan Davies. He made a witness statement, describing his relationship with Sullivan and Taeger and explaining how the Rover had come to be outside his flat, where Taeger lived, before Christmas 1999. He then had an off-the-record conversation with the police, which my noble and learned friend Lord Steyn has described in paras 45 and 46 of his opinion.

  11.   By a defence statement dated 4 August 2000 the defendant gave further details of his defence of duress, claiming that he had been coerced into committing the second burglary by Sullivan.

  12.   The defendant's trial on two counts of aggravated burglary began on 30 January 2001 and ended on 9 February. The jury acquitted him on the first count but convicted him on the second. He was sentenced to 9 years' imprisonment.

  13.   Lord Steyn has recounted the course of the trial and summarised the trial judge's ruling on the confession issue under section 76 of PACE, and has quoted the judgment of the Court of Appeal on this question: see paras 47-49 and 60 of his opinion. I am in complete agreement with his reasoning, and I share his conclusion. I shall therefore confine this opinion to the issue of duress.

  14.   On that issue the judge put four questions to the jury:

      "Question 1: Was the defendant driven or forced to act as he did by threats which, rightly or wrongly, he genuinely believed that if he did not burgle [the] house, his family would be seriously harmed or killed? If you are sure that he was not forced by threats to act as he did, the defence fails and he is guilty. But if you are not sure go on to question 2. Would a reasonable person of the defendant's age and background have been driven or forced to act as the defendant did? If you are sure that a reasonable person would not have been forced to act as the defendant did, then the defence fails and he is guilty. If you are not sure, then go on to question 3. Could the defendant have avoided acting as he did without harm coming to his family? If you are sure he could, the defence fails and he is guilty. If you are not sure go on to question 4. Did the defendant voluntarily put himself in the position in which he knew he was likely to be subjected to threats? If you are sure he did, the defence fails and he is guilty. If you are not sure, he is not guilty. Those four questions are really tests."

    The first of these questions repeated in substance a question the judge had already framed for the consideration of the jury. In his earlier direction he had explained the second question somewhat more fully:

      "The second question is: Would a reasonable person, of the defendant's age and background, have been forced and driven to act as the defendant did? That question is necessary because everybody has to be judged by the same standards. The reactions of a reasonable person may or may not be the same as the reactions of any particular defendant. You represent society and you set the standards of what is reasonable. In judging what a reasonable person would do, you are not expected to imagine a saint and that is why I say a reasonable person of the defendant's age and background. What, in your judgment, as judges of the facts, would such a person have done in the circumstances? Would he have felt compelled to act as he did?

      If you are sure that a reasonable person would not have been forced to act as the defendant did, again, the defence fails and the defendant would be guilty. But if you are not sure if a reasonable person might have been forced to act as the defendant did, then you go on to the third question."

    He had earlier directed the jury on the third question as follows:

      "The third question is: Could the defendant have avoided acting as he did without harm coming to his family? In fact, as we know, having broken in, he left empty handed. No harm apparently has resulted. I will remind you of the evidence in due course but Mr Sullivan, according to the defendant, accepted that position.

      If he had left as soon as the alarm went off and as soon as [the victim] started telephoning the police, would it have been any different? Could he have pretended that he could not find the house? You will remember some of the questions that he was asked on this topic by [prosecuting counsel]. Could he have pretended to the minder - if there was a minder - that there was no answer when he rang? All those are matters for you to consider. If you are sure that he could have avoided acting as he did without harm coming to his family, again the defence fails and he is guilty. But if you are not sure that he could have avoided acting as he did without harm coming to his family, then there is one final question."

    Then the judge had turned to the fourth question:

      "Question 4: Did the defendant voluntarily put himself in the position, in which he knew he was likely to be subjected to threats? You look to judge that in all the circumstances. If he had stopped associating with Frank Sullivan after the August 1999 incident, would he have ever found himself in this predicament?

      It is for you to decide. It is right to say he says he did stop associating but Sullivan kept finding him. It may not be wholly straightforward. It is for you to consider and it is a relevant consideration because if someone voluntarily associates with the sort of people who he knows are likely to put pressure on him, then he cannot really complain, if he finds himself under pressure. If you are sure that he did voluntarily put himself in such a position, the defence fails and he was guilty. If you are not sure and you have not been sure about all of the other questions, then you would find him not guilty."

  15.   On his appeal to the Court of Appeal the defendant criticised the judge's directions on the third and fourth questions. With regard to the third question Rix LJ, giving the judgment of the court, said (in para 49 of the judgment):

      "We think that the direction on this third question was a misdirection. There never was any suggestion that the appellant could have avoided the effect of the threat against him, assuming one had ever been made, by going to the police or simply refusing to carry out the robbery. On analysis the issues raised under this third question collapse into the issues raised under questions one and two. We therefore think that there is a danger that the jury may have been confused by being asked an additional question on matters already covered by the first two questions."

    Having considered a number of authorities, the Court of Appeal also concluded (paras 72-77) that there was a misdirection in the judge's formulation of question 4

      "and that he should have directed the jury to consider whether the [defendant] knew that he was likely to be subjected to threats to commit a crime of the type [with] which he was charged."

  16.   Having upheld the defendant's ground of appeal on the confession issue, and found two misdirections on the duress issue, the court considered the defendant's conviction on the second count to be unsafe and quashed it. In this appeal to the House, the Crown seek to establish that the judge's directions on the third and fourth questions involved no misdirection, and they suggest that his direction on the first question was favourable to the defendant. It is necessary to consider the law on duress in a little detail.

    Duress

  17.   The common sense starting point of the common law is that adults of sound mind are ordinarily to be held responsible for the crimes which they commit. To this general principle there has, since the 14th century, been a recognised but limited exception in favour of those who commit crimes because they are forced or compelled to do so against their will by the threats of another. Such persons are said, in the language of the criminal law, to act as they do because they are subject to duress.

  18.   Where duress is established, it does not ordinarily operate to negative any legal ingredient of the crime which the defendant has committed. Nor is it now regarded as justifying the conduct of the defendant, as has in the past been suggested: Attorney-General v Whelan [1934] IR 518, 526; Glanville Williams, Criminal Law, The General Part (2nd ed, 1961), p 755. Duress is now properly to be regarded as a defence which, if established, excuses what would otherwise be criminal conduct: Director of Public Prosecutions for Northern Ireland v Lynch [1975] AC 653, 671, 680, 710-711; Hibbert v The Queen (1995) 99 CCC (3d) 193, paras 21, 38, 47, per Lamer CJC.

  19.   Duress affords a defence which, if raised and not disproved, exonerates the defendant altogether. It does not, like the defence of provocation to a charge of murder, serve merely to reduce the seriousness of the crime which the defendant has committed. And the victim of a crime committed under duress is not, like a person against whom a defendant uses force to defend himself, a person who has threatened the defendant or been perceived by the defendant as doing so. The victim of a crime committed under duress may be assumed to be morally innocent, having shown no hostility or aggression towards the defendant. The only criminal defences which have any close affinity with duress are necessity, where the force or compulsion is exerted not by human threats but by extraneous circumstances, and, perhaps, marital coercion under section 47 of the Criminal Justice Act 1925.

  20.   Where the evidence in the proceedings is sufficient to raise an issue of duress, the burden is on the prosecution to establish to the criminal standard that the defendant did not commit the crime with which he is charged under duress: R v Lynch, above, p 668. In its Report "Legislating the Criminal Code. Offences against the Person and General Principles" (1993, Law Com. No 218, Cm 2370, paras 33-34), the Law Commission recommended that a legal burden of proof, on the balance of probabilities, be placed on a defendant to establish a defence of duress. It was not suggested in argument that this was a change which should be made, and there must be real doubt whether it is a change which the House in its judicial capacity could properly make even if persuaded of the merits of doing so. Imposition of a reverse legal burden on the defendant would in any event require very careful consideration. But it must be accepted, as the Law Commission pointed out in para 33 of this Report, that the defence of duress is peculiarly difficult for the prosecution to investigate and disprove beyond reasonable doubt. As Professor Sir John Smith QC observed in his commentary on R v Cole [1994] Crim LR 582, 584, with reference to the Law Commission proposal,

      "duress is a unique defence in that it is so much more likely than any other to depend on assertions which are peculiarly difficult for the prosecution to investigate or subsequently to disprove."

    The prosecution's difficulty is of course the greater when, as is all too often the case, little detail of the alleged compulsion is vouchsafed by the defence until the trial is under way.

  21.   Having regard to these features of duress, I find it unsurprising that the law in this and other jurisdictions should have been developed so as to confine the defence of duress within narrowly defined limits. Most of these are not in issue in this appeal, but it seems to me important that the issues the House is asked to resolve should be approached with understanding of how the defence has developed, and to that end I shall briefly identify the most important limitations:

        (1)  Duress does not afford a defence to charges of murder (R v Howe [1987] AC 417), attempted murder (R v Gotts [1992] 2 AC 412) and, perhaps, some forms of treason (Smith & Hogan, Criminal Law, 10th ed., 2002, p 254). The Law Commission has in the past (eg. in "Criminal Law. Report on Defences of General Application" (Law Com No 83, Cm 556, 1977, paras 2.44-2.46, [1977] EWLC 83 )) recommended that the defence should be available as a defence to all offences, including murder, and the logic of this argument is irresistible. But their recommendation has not been adopted, no doubt because it is felt that in the case of the gravest crimes no threat to the defendant, however extreme, should excuse commission of the crime. It is noteworthy that under some other criminal codes the defence is not available to a much wider range of offences: see, for example, section 20(1) of the Tasmanian Criminal Code, section 40(2) of the Criminal Code Act of the Northern Territory of Australia, section 31(4) of the Criminal Code Act Compilation Act 1913 of Western Australia, section 17 of the Canadian Criminal Code and section 24 of the Crimes Act 1961 of New Zealand.

        (2)  To found a plea of duress the threat relied on must be to cause death or serious injury. In Alexander MacGrowther's Case (1746) Fost. 13, 14, 168 ER 8, Lee CJ held:

      "The only force that doth excuse, is a force upon the person, and present fear of death."

    But the Criminal Law Commissioners in their Seventh Report of 1843 (p 31, article 6) understood the defence to apply where there was a just and well-grounded fear of death or grievous bodily harm, and it is now accepted that threats of death or serious injury will suffice: R v Lynch, above, p 679; R v Abdul-Hussain (Court of Appeal (Criminal Division), 17 December 1998, unreported).

    (3)  The threat must be directed against the defendant or his immediate family or someone close to him: Smith & Hogan, above, p 258. In the light of recent Court of Appeal decisions such as R v Conway [1989] QB 290 and R v Wright [2000] Crim LR 510, the current (April 2003) specimen direction of the Judicial Studies Board suggests that the threat must be directed, if not to the defendant or a member of his immediate family, to a person for whose safety the defendant would reasonably regard himself as responsible. The correctness of such a direction was not, and on the facts could not be, in issue on this appeal, but it appears to me, if strictly applied, to be consistent with the rationale of the duress exception.

    (4)  The relevant tests pertaining to duress have been largely stated objectively, with reference to the reasonableness of the defendant's perceptions and conduct and not, as is usual in many other areas of the criminal law, with primary reference to his subjective perceptions. It is necessary to return to this aspect, but in passing one may note the general observation of Lord Morris of Borth-y-Gest in R v Lynch, above at p 670:

      ".. it is proper that any rational system of law should take fully into account the standards of honest and reasonable men. By those standards it is fair that actions and reactions may be tested."

    (5)  The defence of duress is available only where the criminal conduct which it is sought to excuse has been directly caused by the threats which are relied upon.

    (6)  The defendant may excuse his criminal conduct on grounds of duress only if, placed as he was, there was no evasive action he could reasonably have been expected to take. It is necessary to return to this aspect also, but this is an important limitation of the duress defence and in recent years it has, as I shall suggest, been unduly weakened.

    (7)  The defendant may not rely on duress to which he has voluntarily laid himself open. The scope of this limitation raises the most significant issue on this part of this appeal, and I must return to it.

  22.   For many years it was possible to regard the defence of duress as something of an antiquarian curiosity, with little practical application. Sir James Stephen, with his immense experience, never knew or heard of the defence being advanced, save in the case of married women, and could find only two reported cases: A History of the Criminal Law of England (1883), vol II, p 106. Edwards, drawing attention to the absence of satisfactory modern authority, inferred that the defence must be very rare: "Compulsion, Coercion and Criminal Responsibility" (1951) 14 MLR 297. Professor Hart described duress as a defence of which little is heard: Punishment and Responsibility (1960), p 16. This has changed. As Dennis correctly observed in "Duress, Murder and Criminal Responsibility" (1980) 96 LQR 208,

      "In recent years duress has become a popular plea in answer to a criminal charge."

    This is borne out by the steady flow of cases reaching the appellate courts over the past 30 years or so, and by the daily experience of prosecutors. As already acknowledged, the House is not invited in this appeal to recast the law on duress. It can only address, piecemeal, the issues which fall for decision. That duress is now regularly relied on as a complete defence to serious criminal charges does not alter the essential task which the House must undertake, but does give it additional practical importance. I must acknowledge that the features of duress to which I have referred in paras 18 to 20 above incline me, where policy choices are to be made, towards tightening rather than relaxing the conditions to be met before duress may be successfully relied on. In doing so, I bear in mind in particular two observations of Lord Simon of Glaisdale in R v Lynch above (dissenting on the main ruling, which was reversed in R v Howe, above):

      ".. your Lordships should hesitate long lest you may be inscribing a charter for terrorists, gang-leaders and kidnappers." (p 688).

      "A sane system of criminal justice does not permit a subject to set up a countervailing system of sanctions or by terrorism to confer criminal immunity on his gang."(p 696).

    In Perka v The Queen [1984] 2 SCR 232, 250, Dickson J held that

      "If the defence of necessity is to form a valid and consistent part of our criminal law it must, as has been universally recognised, be strictly controlled and scrupulously limited to situations that correspond to its underlying rationale."

    I agree. I also agree with the observation of the Supreme Court of Canada in R v Ruzic (2001) 153 CCC (3d) 1, para 59, although in that case the presence and immediacy requirements in section 17 of the Canadian Criminal Code were struck down as unconstitutional:

      "Verification of a spurious claim of duress may prove difficult. Hence, courts should be alive to the need to apply reasonable, but strict standards for the application of the defence."

    If it appears at trial that a defendant acted in response to a degree of coercion but in circumstances where the strict requirements of duress were not satisfied, it is always open to the judge to adjust his sentence to reflect his assessment of the defendant's true culpability. This is what the trial judge did in R v Hudson and Taylor, below, where he ordered the conditional discharge of the defendants.

    The judge's direction to the jury on questions 1 and 2
  23.   The appellant did not challenge the judge's direction to the jury on questions 1 and 2. Save in one respect those directions substantially followed the formulation propounded by the Court of Appeal (Criminal Division) (Lord Lane CJ, Taylor and McCullough JJ) in R v Graham [1982] 1 WLR 294, 300, approved by the House of Lords in R v Howe above, at pp 436, 438, 446, 458-459. It is evident that the judge, very properly, based himself on the JSB's specimen direction as promulgated in August 2000. That specimen direction included the words, adopted by the judge, "he genuinely believed". But the words used in R v Graham and approved in R v Howe were "he reasonably believed". It is of course essential that the defendant should genuinely, ie. actually, believe in the efficacy of the threat by which he claims to have been compelled. But there is no warrant for relaxing the requirement that the belief must be reasonable as well as genuine. There can of course be no complaint of this departure from authority, which was favourable to the defendant.

    The judge's direction to the jury on question 3

  24.   As recorded in para 15 above, the Court of Appeal held that the judge had misdirected the jury on question 3 because, it was held, there was no suggestion that the defendant could have taken evasive action. This may, or may not, on the facts, be so, and this suggested misdirection does not feature in the question on duress certified for the opinion of the House. It is true, as the Court of Appeal recognised in its judgment, that there may be an area of overlap between questions 2 and 3: a reasonable person of a defendant's age and background would not have been forced and driven to act as the defendant did if there was any evasive action reasonably open to him to take in order to avoid committing the crime. But the third question put by the judge, and regularly put in such cases, whether or not correctly put on the facts of this case, in my opinion focuses attention on a cardinal feature of the defence of duress, and I would wish to warn against any general notion that question 3 "collapses" into or is subsumed under questions 1 and 2.

  25.   In the draft Criminal Code prepared by the Criminal Law Commissioners in 1879, section 23, a defence was provided in the case of "Compulsion by threats of immediate death or grievous bodily harm from a person actually present at the commission of the offence". The requirement of immediacy is reflected in the criminal codes of several other jurisdictions. Section 67(1) of the Queensland Criminal Code refers to "immediate death or grievous bodily harm threatened by someone else able to carry out the threat". Section 20(1) of the Tasmanian Code refers to "compulsion by threats of immediate death or grievous bodily harm, from a person actually present at the commission of the offence". Section 31(4) of the Western Australian Code, section 17 of the Canadian Code and section 24(1) of the New Zealand Code use very much the same language. In Scotland where, as in England and Wales, the defence of coercion has recently enjoyed something of a vogue after a long period of dormancy, the law is clear that a threat, to found the defence, must be of immediate and not future death or serious injury: Hume's Commentaries, vol i, p 53; Thomson v HM Advocate 1983 JC 69, 72-73, 75, 80; Cochrane v H M Advocate 2001 SCCR 655, 656, 659-661. In Perka v The Queen [1984] 2 SCR 232, 251, 259, a decision directed to the analogous defence of necessity, Dickson J identified the necessary conditions as including "urgent situations of clear and imminent peril" in which "compliance with the law [would be] demonstrably impossible". In Hibbert v The Queen (1995) 99 CCC (3d) 193, para 49, Lamer CJC quoted with approval the reference by Horder ("Autonomy, Provocation and Duress" [1992] Crim LR 706, 709) to taking "the necessary evasive action".

  26.   The recent English authorities have tended to lay stress on the requirement that a defendant should not have been able, without reasonably fearing execution of the threat, to avoid compliance. Thus Lord Morris of Borth-y-Gest in R v Lynch, above, at p 670, emphasised that duress

      "must never be allowed to be the easy answer of those who can devise no other explanation of their conduct nor of those who readily could have avoided the dominance of threats nor of those who allow themselves to be at the disposal and under the sway of some gangster-tyrant."

    Lord Simon of Glaisdale gave as his first example of a situation in which a defence of duress should be available (p 687):

      "A person, honestly and reasonably believing that a loaded pistol is at his back which will in all probability be used if he disobeys .."

    In the view of Lord Edmund-Davies (p 708) there had been

      "for some years an unquestionable tendency towards progressive latitude in relation to the plea of duress."

  27.   In making that observation Lord Edmund-Davies did not directly criticise the reasoning of the Court of Appeal in its then recent judgment in R v Hudson and Taylor [1971] 2 QB 202, but that was described by Professor Glanville Williams as "an indulgent decision" (Textbook of Criminal Law, 2nd ed, 1983, p 636), and it has in my opinion had the unfortunate effect of weakening the requirement that execution of a threat must be reasonably believed to be imminent and immediate if it is to support a plea of duress. The appellants were two teenage girls who had committed perjury at an earlier trial by failing to identify the defendant. When prosecuted for perjury they set up a plea of duress, on the basis that they had been warned by a group, including a man with a reputation for violence, that if they identified the defendant in court the group would get the girls and cut them up. They resolved to tell lies, and were strengthened in their resolve when they arrived at court and saw the author of the threat in the public gallery. The trial judge ruled that the threats were not sufficiently present and immediate to support the defence of duress but was held by the Court of Appeal to have erred, since although the threats could not be executed in the courtroom they could be carried out in the streets of Salford that same night. It was argued for the Crown that the appellants should have neutralised the threat by seeking police protection, but this argument was criticised as failing to distinguish between cases in which the police would be able to provide effective protection and those when they would not. The Court of Appeal placed reliance on the decision of the Privy Council in Subramaniam v Public Prosecutor [1956] 1 WLR 965. That case, however, involved a defendant who sought at trial to advance a defence of duress under a section of the Penal Code of the Federated Malay States which provided that, with certain exceptions,

      "nothing is an offence which is done by a person who is compelled to do it by threats, which, at the time of doing it, reasonably cause the apprehension that instant death to that person will otherwise be the consequence .."

    The appeal was allowed because evidence relied on by the appellant to show that he had had a reasonable apprehension of instant death was wrongly excluded. It is hard to read that decision as authority for the Court of Appeal's conclusion. I can understand that the Court of Appeal in R v Hudson and Taylor had sympathy with the predicament of the young appellants but I cannot, consistently with principle, accept that a witness testifying in the Crown Court at Manchester has no opportunity to avoid complying with a threat incapable of execution then or there. When considering necessity in R v Cole [1994] Crim LR 582, 583, Simon Brown LJ, giving the judgment of the court, held that the peril relied on to support the plea of necessity lacked imminence and the degree of directness and immediacy required of the link between the suggested peril and the offence charged, but in R v Abdul-Hussain, above, the Court of Appeal declined to follow these observations to the extent that they were inconsistent with R v Hudson and Taylor, by which the court regarded itself as bound.

  28.   The judge's direction on question 3 was modelled on the JSB specimen direction current at the time, and is not in my opinion open to criticism. It should however be made clear to juries that if the retribution threatened against the defendant or his family or a person for whom he reasonably feels responsible is not such as he reasonably expects to follow immediately or almost immediately on his failure to comply with the threat, there may be little if any room for doubt that he could have taken evasive action, whether by going to the police or in some other way, to avoid committing the crime with which he is charged.

    The judge's direction to the jury on question 4

  29.   The judge's direction to the jury on question 4 is quoted in para 14 above and, as recorded in para 15, the Court of Appeal ruled that this was a misdirection because the judge had not directed the jury to consider whether the defendant knew that he was likely to be subjected to threats to commit a crime of the type of which he was charged. It is this ruling which gives rise to the certified question on this part of the case, which is:

      "Whether the defence of duress is excluded when as a result of the accused's voluntary association with others:

      (i)  he foresaw (or possibly should have foreseen) the risk of being subjected to any compulsion by threats of violence, or

      (ii)  only when he foresaw (or should have foreseen) the risk of being subjected to compulsion to commit criminal offences, and, if the latter,

      (iii)  only if the offences foreseen (or which should have been foreseen) were of the same type (or possibly of the same type and gravity) as that ultimately committed."

        The Crown contend for answer (i) in its objective form. The defendant commends the third answer, omitting the first parenthesis.

  30.   In their definition of duress the Criminal Law Commissioners of 1879 included a proviso:

      "Provided also, that he [the defendant] was not a party to any association or conspiracy the being party to which rendered him subject to such compulsion."

    A qualification to very similar effect is to be found in the criminal codes of Queensland (section 67(3)(b) and (c)), Tasmania (section 20(1)), the Northern Territory of Australia (section 41(2)), Western Australia (section 31(4)), the Commonwealth of Australia (section 10.2(3)), the Australian Capital Territory (section 40(3)), Canada (section 17), New Zealand (section 24(1)) and no doubt others. But its implications were not for many years examined in the British courts.

  31.   The issue might have been raised in R v Lynch, above, where the appellant claimed to have been press-ganged by the IRA, but the argument in that case was largely directed to the question whether the defence of duress was open to a defendant charged as a secondary party to murder. It was in R v Fitzpatrick [1977] NI 20, another IRA case, that the Court of Criminal Appeal in Northern Ireland had occasion to consider the matter in depth. The ratio of the decision is found in the judgment of the court delivered by Lowry LCJ at p 33:

      "A person may become associated with a sinister group of men with criminal objectives and coercive methods of ensuring that their lawless enterprises are carried out and thereby voluntarily expose himself to illegal compulsion, whether or not the group is or becomes a proscribed organisation ..

      .. if a person voluntarily exposes and submits himself, as the appellant did, to illegal compulsion, he cannot rely on the duress to which he has voluntarily exposed himself as an excuse either in respect of the crimes he commits against his will or in respect of his continued but unwilling association with those capable of exercising upon him the duress which he calls in aid."

  32.   That statement was no doubt drafted with the peculiar character of the IRA in mind. R v Sharp [1987] QB 853 arose from criminal activity of a more routine kind committed by a gang of robbers. The trial judge's direction which was challenged on appeal is fully quoted in R v Shepherd (1987) 86 Cr App R 47, 51, and was to this effect:

      ".. but in my judgment the defence of duress is not available to an accused who voluntarily exposes and submits himself to illegal compulsion.

      It is not merely a matter of joining in a criminal enterprise; it is a matter of joining in a criminal enterprise of such a nature that the defendant appreciated the nature of the enterprise itself and the attitudes of those in charge of it, so that when he was in fact subjected to compulsion he could fairly be said by a jury to have voluntarily exposed himself and submitted himself to such compulsion."

    The Court of Appeal (Lord Lane CJ, Farquharson and Gatehouse JJ) upheld that direction in R v Sharp, expressing the principle at p 861:

      ".. where a person has voluntarily, and with knowledge of its nature, joined a criminal organisation or gang which he knew might bring pressure on him to commit an offence and was an active member when he was put under such pressure, he cannot avail himself of the defence of duress."

    In R v Shepherd, above, the criminal activity was of a less serious kind: the question which the jury should have been (but were not) directed to consider (p 51) was "whether the appellant could be said to have taken the risk of P's violence simply by joining a shoplifting gang of which he [P] was a member".

  33.   R v Ali is summarised at [1995] Crim LR 303, but the ratio of the decision more clearly appears from the transcript of the judgment given by the Court of Appeal (Lord Taylor of Gosforth CJ, Alliott and Rix JJ) on 14 November 1994. The appellant claimed to have become involved in drug dealing and to have become indebted to his supplier, X, who (he said) had given him a gun and told him to obtain the money from a bank or building society the following day, failing which he would be killed. The appellant accordingly committed the robbery of which he was convicted. In directing the jury on the defence of duress advanced by the defendant the trial judge had said:

      "The final question is this: did he, in obtaining heroin from Mr X and supplying it to others for gain, after he knew of Mr X's reputation for violence, voluntarily put himself in a position where he knew that he was likely to be forced by Mr X to commit a crime?"

    It was argued by the appellant that the judge should have said "forced by Mr X to commit armed robbery", but this was rejected, and the court held that by "a crime" the jury could only have understood the judge to be referring to a crime other than drug dealing. The principle stated by the court on p 7 of the transcript was this:

      "The crux of the matter, as it seems to us, is knowledge in the defendant of either a violent nature to the gang or the enterprise which he has joined, or a violent disposition in the person or persons involved with him in the criminal activity he voluntarily joined. In our judgment, if a defendant voluntarily participates in criminal offences with a man 'X', whom he knows to be of a violent disposition and likely to require him to perform other criminal acts, he cannot rely upon duress if 'X' does so."

    (In this case, as in R v Cole, above, it would seem that the defence of duress should in any event have failed, for lack of immediacy, since the threat was not to be executed until the following day, and therefore the defendant had the opportunity to take evasive action).

  34.   In its Working Paper No 55 of 1974, the Law Commission in para 26 favoured

      "a limitation upon the defence [of duress] which would exclude its availability where the defendant had joined an association or conspiracy which was of such a character that he was aware that he might be compelled to participate in offences of the type with which he is charged."

    This reference to "offences of the type with which he is charged" was, in substance, repeated in the Law Commission's "Report on Defences of General Application" (Law Com No 83, [1977] EWLC 83) of 1977, [1977] EWLC 83 , paras 2.38 and 2.46(8), in clause 1(5) of the draft bill appended to that report, in clause 45(4) of the draft bill appended to the Law Commission's Report on "Codification of the Criminal Law" (Law Com No 143) of 1985, as explained in para 13.19 of the Report, and in clause 42(5) of the Law Commission's draft "Criminal Code Bill" (Law Com No 177, [1989] EWLC 177_1_1, [1989] EWLC 177_1_1 ) published in 1989. But there was no warrant for this gloss in any reported British authority until the Court of Appeal (Roch LJ, Richards J and Judge Colston QC) gave judgment in R v Baker and Ward [1999] 2 Cr App R 335. The facts were very similar to these in R v Ali, above, save that the appellants claimed that they had been specifically instructed to rob the particular store which they were convicted of robbing. The trial judge had directed the jury (p 341):

      "A person cannot rely on the defence of duress if he has voluntarily and with full knowledge of its nature joined a criminal group which he was aware might bring pressure on him of a violent kind or require him if necessary to commit offences to obtain money where he himself had defaulted to the criminal group in payment to the criminal group."

    This was held to be a misdirection (p 344):

      "What a defendant has to be aware of is the risk that the group might try to coerce him into committing criminal offences of the type for which he is being tried by the use of violence or threats of violence."

    At p 346 this ruling was repeated:

      "The purpose of the pressure has to be to coerce the accused into committing a criminal offence of the type for which he is being tried."

    The appeals were accordingly allowed and the convictions quashed.

  35.   Counsel for the defendant in the present case contends (as the Court of Appeal accepted) that this ruling was correct and that the trial judge in the present case misdirected the jury because he did not insist on the need for the defendant to foresee pressure to commit the offence of robbery of which he was convicted.

  36.   In R v Heath (Court of Appeal: Kennedy LJ, Turner and Smedley JJ, 7 October 1999, [2000] Crim LR 109) the appellant again claimed that he had become indebted to a drug supplier, and claimed that he had been compelled by threats of physical violence to collect the consignment of drugs which gave rise to his conviction. His defence of duress failed at trial, rightly as the Court of Appeal held. In its judgment, Kennedy LJ said:

      "The appellant in evidence conceded that he had put himself in the position where he was likely to be subjected to threats. He was therefore, in our judgment, not entitled to rely on those same threats as duress to excuse him from liability for subsequent criminal conduct."

    The court found it possible to distinguish R v Baker and Ward, observing:

      "It is the awareness of the risk of compulsion which matters. Prior awareness of what criminal activity those exercising compulsion may offer as a possible alternative to violence is irrelevant."

    The facts in R v Harmer (Court of Appeal: May LJ, Goldring and Gross JJ, 12 December 2001, [2002] Crim LR 401) were very similar to those in R v Heath, which the court followed. It does not appear from the court's judgment given by Goldring J whether R v Baker and Ward was directly cited, but it would seem that counsel for the appellant did not rely on it. He argued that the appellant did not foresee that he might be required to commit crimes for the supplier. But the court did not accept this argument:

      "We cannot accept that where a man voluntarily exposes himself to unlawful violence, duress may run if he does not foresee that under the threat of such violence he may be required to commit crimes. There is no reason in principle why that should be so."

  37.   The principal issue between the Crown on one side and the appellant and the Court of Appeal on the other is whether R v Baker and Ward correctly stated the law. To resolve that issue one must remind oneself of the considerations outlined in paras 18-22 above. The defendant is seeking to be wholly exonerated from the consequences of a crime deliberately committed. The prosecution must negative his defence of duress, if raised by the evidence, beyond reasonable doubt. The defendant is, ex hypothesi, a person who has voluntarily surrendered his will to the domination of another. Nothing should turn on foresight of the manner in which, in the event, the dominant party chooses to exploit the defendant's subservience. There need not be foresight of coercion to commit crimes, although it is not easy to envisage circumstances in which a party might be coerced to act lawfully. In holding that there must be foresight of coercion to commit crimes of the kind with which the defendant is charged, R v Baker and Ward mis-stated the law.

  38.   There remains the question, which the Court of Appeal left open in para 75 of their judgment, whether the defendant's foresight must be judged by a subjective or an objective test: i.e. does the defendant lose the benefit of a defence based on duress only if he actually foresaw the risk of coercion or does he lose it if he ought reasonably to have foreseen the risk of coercion, whether he actually foresaw the risk or not? I do not think any decided case has addressed this question, and I am conscious that application of an objective reasonableness test to other ingredients of duress has attracted criticism: see, for example, Elliott, "Necessity, Duress and Self-Defence" [1989] Crim LR 611, 614-615, and the commentary by Professor Ashworth on R v Safi [2003] Crim LR 721, 723. The practical importance of the distinction in this context may not be very great, since if a jury concluded that a person voluntarily associating with known criminals ought reasonably to have foreseen the risk of future coercion they would not, I think, be very likely to accept that he did not in fact do so. But since there is a choice to be made, policy in my view points towards an objective test of what the defendant, placed as he was and knowing what he did, ought reasonably to have foreseen. I am not persuaded otherwise by analogies based on self-defence or provocation for reasons I have already given. The policy of the law must be to discourage association with known criminals, and it should be slow to excuse the criminal conduct of those who do so. If a person voluntarily becomes or remains associated with others engaged in criminal activity in a situation where he knows or ought reasonably to know that he may be the subject of compulsion by them or their associates, he cannot rely on the defence of duress to excuse any act which he is thereafter compelled to do by them. It is not necessary in this case to decide whether or to what extent that principle applies if an undercover agent penetrates a criminal gang for bona fide law enforcement purposes and is compelled by the gang to commit criminal acts.

  39.   I would answer this certified question by saying that the defence of duress is excluded when as a result of the accused's voluntary association with others engaged in criminal activity he foresaw or ought reasonably to have foreseen the risk of being subjected to any compulsion by threats of violence. I would answer the other certified question as proposed by Lord Steyn.

    Conclusion

  40.   The judge's direction to the jury on question 4 involved no misdirection. It was based on the JSB specimen direction current at the time, save that it omitted the qualification made to reflect the erroneous ruling in R v Baker and Ward. The ruling was, on the law as I have stated it, too favourable to the defendant, but he cannot complain of that. It is desirable that the content, and perhaps even the order, of the current JSB directions should be reconsidered in the light of this opinion, but that is not a task which the House should undertake. I would accordingly answer the certified question as indicated, allow the Crown's appeal, set aside the Court of Appeal's order, restore the defendant's conviction and remit this matter to the Court of Appeal so that the defendant may surrender to his bail.

    LORD STEYN

    My Lords,

        I. The Proceedings in the Central Criminal Court

        and in the Court of Appeal

  41.   In early 2001 the defendant (the respondent on this appeal) stood trial before a judge and jury on two counts of aggravated burglary contrary to section 16(1) of the Theft Act 1968. On 9 February 2001 the jury acquitted the defendant on the first of these counts and convicted him on the second. On an appeal by the defendant to the Court of Appeal (Criminal Division) against his conviction on the second count two important questions of law arose, namely (1) whether a statement by the defendant used in evidence at the trial was a confession within the meaning of section 76(1) of the Police and Criminal Evidence Act 1984 and should have been excluded and (2) whether in the circumstances the defence of duress was available to the defendant. On both these issues the Court of Appeal ruled in favour of the defendant and held that his conviction was unsafe. This decision was reported under the name: R v Z [2003] 1 WLR 1489.

    II. Duress

  42.   The Crown now appeals to the House of Lords on the two grounds on which the Court of Appeal found in favour of the defendant.

  43.   My noble and learned friend Lord Bingham of Cornhill has explained the background and analysed the law on duress and its application to this case. Lord Bingham has demonstrated that the excuse of duress was not available in the circumstances of this case. I am in full agreement with the opinion of Lord Bingham. I have nothing to add on the point of duress. But I will examine the second point arising on the appeal, namely whether the statement of the defendant used at trial was a confession under section 76(1) which required the Crown to discharge the burden placed on it by section 76(2).

    III. The Facts Relevant to the Confession Issue

  44.   The two burglaries took place at the same premises: the first on Sunday 29 August 1999 and the second on Sunday 23 January 2000. The first count on which the defendant was acquitted can now be put to one side. The defendant was undoubtedly involved in the second burglary. He denied guilt. On 5 June 2000 the police arrested the defendant in respect of inter alia the second burglary. On 7 June 2000 the victims of the second burglary identified the defendant as the perpetrator of that burglary. On the same day the defendant for the first time raised in terms of extreme generality the issue of duress as an excuse for his participation in the second burglary.

  45.   On 26 June 2000 the defendant had an "off the record" interview with police officers who were involved in a separate murder enquiry. The reason for the confidential interview was that the defendant said that he was in fear of a notorious criminal called Sullivan. The police agreed not to question the defendant about the burglaries. He was not cautioned. There was no tape recording. The police prepared a report of the interview. In the context of the murder enquiry, the defendant said that Sullivan only told him about the murder in late February or early March 2000. When made the report of the confidential interview contained nothing adverse to the defendant's interest in respect of the second burglary. It was either entirely exculpatory or entirely neutral in effect.

  46.   Eventually there were important differences between what the defendant had said during this confidential interview and what he was to say at his trial. In the confidential interview the defendant did not say that he had taken part in the second burglary because of threats made by Sullivan against himself and his family. In accordance with the police report of the confidential interview the threats had not been made until late February or early March 2000, that is after the second burglary.

    IV. The Ruling on the Confidential Statement at Trial

  47.   At the criminal trial the prosecution relied on the confidential statement for two purposes. First to assert that the defendant was a dishonest witness and secondly as evidence of the statement's truth, namely an admission that the defendant had not become aware of Sullivan's claims that he had killed somebody until after the second burglary. The questions arose whether the confidential statement was a confession under section 76 of PACE or ought to be excluded under section 78. Faced with section 76, read with section 82(1), and the decision of the Court of Appeal in R v Sat-Bhambra (1989) 88 Cr App R 55, counsel for the defendant conceded at trial that an exculpatory or neutral statement was not a confession within the meaning of section 76 and that accordingly it could not be excluded under that particular statutory provision. The judge acted on this basis. Counsel for the defendant did, however, invite the judge to exclude the statement under section 78 on the basis that the admission of it would have such an adverse effect on the fairness of the proceedings that the court ought not to admit it. Taking into account that at the time of the trial Sullivan was in prison and could no longer be a danger to the defendant, the judge ruled that section 78 did not prevent the Crown from cross-examining the defendant on differences between the confidential statement and his account at trial. The prosecutor apparently did so with some effect.

  48.   The jury had before it the evidence of the victims of the second burglary as well as the account of the defendant, viewed against the impact of the confidential statement on which the prosecutor cross-examined. The jury convicted the defendant of the second burglary.

    V. The Decision of the Court of Appeal on the Confidential Statement

  49.   On appeal the Court of Appeal upheld the judge's decision under section 78. There is no appeal on that aspect of the case. The Court of Appeal concluded however that the confidential statement was a confession: R v Z [2003] 1 WLR 1489. The Court of Appeal based this conclusion on the impact of section 3 of the Human Rights Act 1998, and the effect of Saunders v UK (1997) 23 EHRR 313, para 71.

    VI. The Certified Question

  50.   The relevant point of law of general public importance certified by the Court of Appeal reads as follows:

      "Whether a 'confession' in section 76 of the Police and Criminal Evidence Act 1984 includes a statement intended by the maker to be exculpatory or neutral and which appears to be so on its face, but which becomes damaging to him at the trial because, for example, its contents can then be shown to be evasive or false or inconsistent with the maker's evidence on oath."

    VII. The Provisions of PACE

  51.   It is necessary to set out the relevant provisions of PACE. So far as it is material section 76 provides as follows:

      "(1)  In any proceedings a confession made by an accused person may be given in evidence against him in so far as it is relevant to any matter in issue in the proceedings and is not excluded by the court in pursuance of this section.

      (2)  If, in any proceedings where the prosecution proposes to give in evidence a confession made by an accused person, it is represented to the court that the confession was or may have been obtained -

        (a) by oppression of the person who made it; or

        (b) in consequence of anything said or done which was likely, in the circumstances existing at the time, to render unreliable any confession which might be made by him in consequence thereof,

        the court shall not allow the confession to be given in evidence against him except in so far as the prosecution proves to the court beyond reasonable doubt that the confession (notwithstanding that it may be true) was not obtained as aforesaid.

        (3) In any proceedings where the prosecution proposes to give in evidence a confession made by an accused person, the court may of its own motion require the prosecution, as a condition of allowing it to do so, to prove that the confession was not obtained as mentioned in subsection (2) above.

        (Emphasis supplied)

    Section 82(1), an interpretative provision of PACE, provides in respect of section 76 that:

      ". . . 'confession' includes any statement wholly or partly adverse to the person who made it, whether made to a person in authority or not and whether made in words or otherwise;"

    A partly adverse statement is commonly described as a "mixed statement". In R v Sharp [1988] 1 WLR 7 the House of Lords held that such a statement is evidence of the self-serving as well as the incriminating parts.

  52.   It is necessary to read section 76(1), as interpreted in accordance with section 82(1), together with section 78 which provides for the exclusion of unfair evidence. Section 78, so far as it is material, provides:

      "(1)  In any proceedings the court may 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 evidence would have such an adverse effect on the fairness of the proceedings that the court ought not to admit it.

      (2)  Nothing in this section shall prejudice any rule of law requiring a court to exclude evidence."

    Subject to the discretion of a trial judge under the common law to exclude evidence where its likely prejudicial effect outweighs its probative value (see section 82(3); R v Sang [1980] AC 402) the provisions of section 76, read with section 82(1), and section 78, constitute a part codification of the law governing criminal evidence.

    VIII. Four Preliminary Observations

  53.   Four preliminary observations about the framework of these provisions of PACE must now be noted. First, section 76 owes its origin to the Eleventh Report of the Criminal Law Revision Committee (Cmnd 4991 (1972)). The Committee had concluded that reliability was the most appropriate basis for determining the admissibility of confessions but considered that the use of oppression should also result in exclusion. That is the rule which is contained in the two parts of section 76(2) of PACE. The rationale of the two-pronged rule in section 76(2) was explained in Lam Chi-ming v The Queen [1991] 2 AC 212 by Lord Griffiths as follows [220E-F]:

      "Their Lordships are of the view that the more recent English cases established that the rejection of an improperly obtained confession is not dependent only upon possible unreliability but also upon the principle that a man cannot be compelled to incriminate himself and upon the importance that attaches in a civilised society to proper behaviour by the police towards those in their custody."

    Secondly, it is necessary to consider the meaning of "oppression" in section 76(2)(a). Section 76(8) provides non-exhaustively that "'oppression' includes torture, inhuman or degrading treatment, and the use or threat of violence (whether or not amounting to torture)". In R v Fulling [1987] QB 426 Lord Lane CJ, giving the judgment of the Court of Appeal, concluded that:

      "'oppression' in section 76(2)(a) should be given its ordinary dictionary meaning. The Oxford English Dictionary as its third definition of the word runs as follows: 'Exercise of authority or power in a burdensome, harsh, or wrongful manner; unjust or cruel treatment of subjects, inferiors, etc.; the imposition of unreasonable or unjust burdens.' One of the quotations given under that paragraph runs as follows: 'There is not a word in our language which expresses more detestable wickedness than oppression.'

      We find it hard to envisage any circumstances in which such oppression would not entail some impropriety on the part of the interrogator."

    Thirdly, the width of section 78(1) is of critical importance. Although it is formally cast in the form of a discretion ("the court may") the objective criterion whether "the evidence would have such an adverse effect on the fairness of the proceedings" in truth imports a judgment whether in the light of the statutory criterion of fairness the court ought to admit the evidence. Fourthly, any evidence obtained by the police by oppression is liable to be excluded under section 78. It would cover the case where the police by oppression obtained a wholly exculpatory but plainly false statement from an accused such as to damage his credibility at trial. That would be unfair under section 78. It is therefore clear that section 76, as read with section 82(1), and section 78, are designed to provide in a coherent and comprehensive way for the just disposal of all decisions about statements made by accused persons to the police. There is no gap in the procedural safeguards of the relevant provisions of PACE.

    IX. The Decisions in Sat-Bhambra (1988) and Park (1993)

  54.   In two decisions the question whether a wholly exculpatory or neutral statement can be a confession was considered by the Court of Appeal (Criminal Division). In Sat-Bhambra (1989) 88 Cr. App R 55 Lord Lane CJ observed (at 61):

      "First, were the answers given by the appellant upon the interviews properly to be described as a confession or confessions? Section 82(1) of the Act defines confession as follows: ''confession' includes any statement wholly or partly adverse to the person who made it, whether made to a person in authority or not and whether made in words or otherwise.'

      His answers upon the interviews, the tapes of which the jury heard, were, as his counsel described, exculpatory. Their principal damaging effect was to demonstrate that the appellant was evasive and prevaricating and that many of the statements which he made proved eventually to be false.

      The question therefore arises: can a statement be described as wholly or partly adverse to the person making it, when it is intended by the maker to be wholly exculpatory and appears to be so on its face, but becomes damaging at the trial because, for example, its contents can by then be shown to be evasive or false or inconsistent with the maker's evidence on oath?

      . . . The words of the section do seem prima facie to be speaking of statements adverse on the face of them. The section is aimed at excluding confessions obtained by words or deeds likely to render them unreliable, i.e. admissions or partial admissions contrary to the interests of the defendant and welcome to the interrogator. They can hardly have been aimed at statements containing nothing which the interrogator wished the defendant to say and nothing apparently adverse to the defendant's interests. If the contentions of the appellant in the present case are correct, it would mean that the statement 'I had nothing to do with it' might in due course become a 'confession', which would be surprising, with or without section 82(1).

      We are inclined to the view that purely exculpatory statements are not within the meaning of section 82(1). We are supported in this view by the learned author of Cross on Evidence, 6th ed., p 544. The same view is taken by Andrews and Hirst on Criminal Evidence, paragraph 19.04. They cite the words of Lord Widgery CJ in Pearce (1979) 69 Cr App R 365, where he says 'A denial does not become an admission because it is inconsistent with another denial.'

      In so far as they express a contrary view we respectfully dissent from the views of the Supreme Court of Canada in Piche v R (1970) 11 DLR 700, and of Chief Justice Warren in Miranda v Arizona 384 U.S. 436, 477 (1975), where he said that such statements 'are incriminating in any meaningful sense of the word.'

      However in the light of what we have to say hereafter, we do not need to come to any firm conclusion on this aspect of the case . . ."

    (The observation in Pearce which is attributed to Lord Widgery CJ is contained in a judgment prepared by Lloyd J.) The observations by Lord Lane CJ in Sat-Bhambra, although technically obiter dicta, were characteristically analytical.

  55.   In Park (1994) 99 Cr App R 270 a defendant had been stopped by police officers whilst driving a car which contained property stolen in burglaries. The question arose whether a statement was a confession. The court applied the interpretation of section 82(1) which had been suggested in Sat-Bhambra: at 274. Kennedy LJ added the following observation (at 274):

      "In the current edition of Archbold (1993) at paragraph 15-293, dealing with this particular section and that authority, it is said that section 82(1) was not aimed at statements which the maker intended to be exculpatory and which were exculpatory on their face, but which could later be shown to be false or inconsistent with the maker's evidence on oath. It seems to us that that is precisely the situation here in relation not only to the answers in which the appellant denied ownership of certain items but also in relation to those answers where he accepted ownership of certain items, and accordingly, in our judgment, neither the conversation at the roadside nor, when we come to it, the conversation in the police station yard amounted to a confession."

    It may well be that the statement made by the defendant in Park was in fact a mixed statement, i.e. partly adverse to the defendant. But the Court of Appeal concluded that a wholly exculpatory statement falls outside the scope of section 82(1). It is, however, on the reasoning in Sat-Bhambra that one is principally dependent.

    X. Section 82(1)

  56.   In the present case the Court of Appeal did not disagree with the interpretation adopted in Sat-Bhambra but concluded that under section 3(1) of the Human Rights Act 1998, which came into force on 2 October 2000, that decision must be reconsidered. Before this view can be examined it is necessary to consider the interpretation of section 76, read with section 82(1), and viewed in the context of section 78, as a matter of ordinary statutory interpretation. That is necessary because counsel for the accused submitted that the words in section 82(1) "'confession' includes any statement wholly or partly adverse to the person who made it" provide a gateway to bringing wholly exculpatory or neutral statements within the scope of section 76. Counsel emphasised that the legislature could in section 82(1) have used the straightforward definition that "confession" means a wholly or partly adverse statement. That, he conceded, would have left no room for doubt. But, he said, "includes" is an expansive concept. In my view this argument attaches too much importance to this choice of language. The explanation for the drafting technique is probably that the word "includes" was selected because the core meaning of "confession", i.e. a wholly adverse statement, is at the forefront. Section 82(1) then extends that core meaning to partly adverse statements. This restates the effect of R v Harz and Power [1967] 1 AC 760. In other words, in terms of admissibility no distinction is to be made between a full confession of guilt and admissions falling short of guilt. But, in any event, it is wholly implausible that the draftsman would have made express reference only to wholly or partly adverse statements if he also had in mind covering under the definition of "confession" wholly exculpatory statements. There is no support in the preceding case law for such a view: R v Harz and Power, supra. Neither the Eleventh Report of the Criminal Law Revision Committee nor any other external aid to PACE give any assistance to such an argument. The plain meaning of the statute is against such a strange interpretation. And it is inconceivable, on policy grounds, that the legislature would have introduced such a fundamental change in the law by leaving the question whether an exculpatory statement is a confession to depend on developments at trial.

  57.   There is nothing in the statutory context which compels a strained interpretation of section 82(1). After all, as has been pointed out, section 78 is wide enough to permit the court to exclude wholly exculpatory statements which were obtained by oppression, e.g. in order to fabricate a false exculpatory account to the detriment of the defendant. In these circumstances the House ought now to affirm the interpretation suggested in Sat-Bhambra.

  58.   Properly construed section 76(1), read with section 82(1), requires the court to interpret a statement in the light of the circumstances when it was made. A purely exculpatory statement (e.g. "I was not there") is not within the scope of section 76(1). It is not a confession within the meaning of section 76. The safeguards of section 76 are not applicable. But the safeguards of section 78 are available.

    XI. Section 3 of the Human Rights Act 1998

  59.   In the Court of Appeal counsel for the defendant relied on the decision of the European Court of Human Rights in Saunders v UK (1997) 23 EHRR 313. He emphasised what the European Court of Human Rights said (at para 71):

      "In any event, bearing in mind the concept of fairness in article 6, the right not to incriminate oneself cannot reasonably be confined to statements of admission of wrongdoing or to remarks which are directly incriminating. Testimony obtained under compulsion which appears on its face to be of a non-incriminating nature - such as exculpatory remarks or mere information on questions of fact - may later be deployed in criminal proceedings in support of the prosecution case, for example to contradict or cast doubt upon other statements of the accused or evidence given by him during the trial or to otherwise undermine his credibility. Where the credibility of the accused must be assessed by a jury the use of such testimony may be especially harmful. It follows that what is of the essence in this context is the use to which evidence obtained under compulsion is put in the course of the criminal trial."

    Relying on Saunders, and section 3(1) of the Human Rights Act 1998, counsel for the defendant invited the court to reconsider Sat-Bhambra. On the other hand, counsel for the prosecution submitted that Sat-Bhambra was compatible with section 3(1) and was good law.

  60.   Rix LJ observed (at para 37):

      "In our judgment, the Human Rights Act 1998 and in particular its section 3(1), which provides that 'So far as it is possible to do so, primary legislation and subordinate legislation must be read and given effect to in a way which is compatible with the Convention rights', require us to reconsider this issue. The discussion in Sat-Bhambra already indicates that two views are possible as to what amounts to an 'adverse' (or, more generally, an incriminating) statement and Saunders's case shows that the ECtHR has adopted for itself the view expressed by the Supreme Courts of Canada and the USA rather than that of our courts. The definition of 'confession' is an inclusive one and clearly intended to be a broad one. The question in any event arises: at what time is the judgment, whether a statement is or is not a confession, whether it is or is not adverse, to be made? Sat-Bhambra indicates that the decision is to be made at the time of the statement; but prima facie one would have thought that the test is to be made at the time when it is sought to give the statement in evidence. That is, to our mind, confirmed by the underlying rationale of section 76. We do not agree that it is primarily to prevent verballing. That is now the function of Code C, and in any event verballing is a danger whether an accused speaks voluntarily or not. Section 76 goes back to an earlier time when the concern was that an accused, who has a right of silence, may be prevailed upon both to surrender his right and to make unreliable statements by reason of either 'oppression' or 'anything said or done . . . likely . . . to render unreliable' what he says (section 76(2)). In such circumstances the prosecution bear the criminal burden of proving that the confession was not obtained in such circumstances. If therefore an accused is driven to make adverse statements by reason of oppression, why should he lose the protection of section 76(2) just because, although he may have sought to exculpate himself, in fact he damned himself?

      We therefore think that the confidential statement was, at the time it had to be considered, a confession. . . ."

    On appeal to the House counsel for the defendant supported this interpretation.

  61.   It is now necessary to examine this reasoning. The reliance by the Court of Appeal on the decision in Saunders was misplaced. As the cited passage from the judgment in Saunders shows, the ECtHR was solely concerned with evidence obtained under compulsion or under threat of a legal penalty. The ECtHR did not make any pronouncement on all statements made to investigators during a criminal investigation, in whatever context. The ECtHR did not attempt to define what might amount to a confession for the purposes of section 76 of PACE. The Saunders decision is of no assistance in the present context. (Saunders was discussed in Brown v Stott (Procurator Fiscal, Dunfermline) [2003] 1 AC 681: it is not necessary to cover the same ground in this case.)

  62.   That brings me to the reliance by the Court of Appeal on section 3(1) of the 1998 Act. Undoubtedly there is a strong obligation under section 3(1) to interpret legislation compatibly with Convention rights. There is a strong rebuttable presumption in favour of an interpretation consistent with Convention rights: Ghaidan v Godin-Mendoza [2004] 2 AC 557. Rix LJ held that the interpretation of section 76(1), read with section 82(1), which was suggested in Sat-Bhambra, would be incompatible with a Convention right. The House must, however, consider whether in truth any Convention right is engaged. While it is not spelt out in the judgment of the Court of Appeal, Rix LJ presumably had in mind that article 6 is the particular Convention right in question. There is, however, nothing in the text of article 6 or in the corpus of European jurisprudence which supports the view that sections 76(1) and 82(1) create any incompatibility with article 6. Given the unrestricted capability of section 78 to avoid injustice by excluding any evidence obtained by unfairness (including wholly exculpatory or neutral statements obtained by oppression), sections 76(1) and 82(1) are in my view compatible with article 6. The decision of the Court of Appeal to the contrary was wrong.

    XII. Postscript

  63.   In the present proceedings the defendant was cross-examined on his earlier statement under section 4 of the Criminal Procedure Act 1865 (commonly referred to as Lord Denman's Act). The provision of section 119 of the Criminal Justice Act 2003, governing previous inconsistent statements, contain changes but are not yet in force. The effects of the changes are a matter for future debate. The House was told that this provision will come into force on 5 April 2005.

    XIII. Disposal

  64.   I would answer the certified question, which is set out in para 50 above, in the negative.

  65.   For the reasons given by Lord Bingham and the reasons I have given, I would also allow the appeal and make the order which Lord Bingham proposes.

    LORD RODGER OF EARLSFERRY

    My Lords,

  66.   I have had the privilege of considering in draft the speeches of my noble and learned friends, Lord Bingham of Cornhill and Lord Steyn. For the reasons they give I too would allow the appeal, restore the respondent's conviction and remit the matter to the Court of Appeal.

    BARONESS HALE OF RICHMOND

    My Lords,

  67.   In 1993 I set my name to the Law Commission's Report, Legislating the Criminal Code: Offences against the Person and General Principles (Law Com No 218, 1993, [1993] EWLC 218 ). This followed wide and expert consultations on two earlier drafts of a Criminal Code, the first drafted by an eminent academic team, Report to the Law Commission on the Codification of the Criminal Law (Law Com No 143, 1985), and the second by the Commission with extensive help from the code team, A Criminal Code for England and Wales (Law Com No 177, [1989] EWLC 177_1_1, [1989] EWLC 177_1_1, 1989), and a further consultation paper, Legislating the Criminal Code: Offences against the Person and General Principles (LCCP 122, 1992). In relation to duress, the draft codes expressed in code style the recommendations of the Law Commission's earlier work on defences of general application, Defences of General Application, Working Paper No 55, 1974, and Report on Defences of General Application, Law Com No 83, [1977] EWLC 83, 1977. These took a largely subjective view of the requirements of the defence, but the 1993 Report carried this even further.

  68.   The overall result was a proposed defence which had "as its guiding principle the reasonable reaction of the defendant in the circumstances as he or she believed them to be" (Law Com No 218, [1993] EWLC 218 , para 29.7). Thus (i) there had been almost no support on consultation for the approach in Graham that the defendant's belief that a threat had been made had to be reasonable; consistently with the defence of self-defence, therefore, the defendant should be judged on the facts as she honestly believed them to be (paras 29.8 to 29.10); (ii) the substantial balance of opinion on consultation had been that the defence should be available to a defendant who honestly believed that official protection would be ineffective (paras 29.3 to 29.7); and (iii) there remained strong support for the view that the defence should apply where the particular defendant in question could not reasonably have been expected to resist the threat (paras 29.11 to 29.14). The draft Bill reflected this consistently subjective approach to the defence.

  69.   Alongside this, the Commission adhered to the view that duress should be a complete defence to all crimes, not simply a matter of mitigation:

      "31.4  We believe that if it is wrong even in respect of murder to condemn the defendant for not acting heroically rather than reasonably, it would be even more unjust to condemn defendants for lesser acts done under the same conditions. To censure and punish defendants who found themselves in such circumstances would bring the law into disrepute. To take a recent example, it was confirmed in Lewis [The Times, 19 November 1992] that a threat of a reprisal that it is unreasonable to expect the witness to resist is a defence to a charge of contempt in respect of a refusal to give evidence. It would, in our view, be intolerable if, for instance, a wife whose husband threatened her with serious injury or death, and who as a result reasonably refused to give evidence against him, had nonetheless to be convicted of the offence of contempt."

    The Commission also saw practical difficulties in the way of treating duress as mitigation:

      "31.7   . . . If duress is rejected as a defence, that must be either because the defendant who acts under duress is at some way at fault, albeit it only by not behaving heroically; or because there is some public policy reason for convicting him even though he is not at fault. If he is at fault, the law should mark his fault by a penalty, or at least should not assume that in no case will an effective penalty be imposed. If the reasons for rejecting duress as a defence are ones of public policy, it is hard to see that that policy is forwarded by a regime that assumes that convictions are to be purely nominal in nature; or, even more, that assumes that in some cases the law will not be enforced at all."

  70.   As Professor Andrew Ashworth (in Principles of Criminal Law, 4th edition 2003, p 228) points out, there are other policy problems with relying on duress as a mitigating factor:

      "Mitigation may be right if 'desert' is the basis for sentence, but supporters of deterrent sentencing have a particular problem. Their general approach is to maintain that the stronger the temptation or pressure to commit a crime, the stronger the law's threat should be in order to counter-balance it. The law and its penalties should be used to strengthen the resolve of those under pressure."

    That is, indeed, a common approach to sentencing: in drug smuggling cases, for example, the 'mule' may well have been subjected to intense pressure to carry the goods into the United Kingdom, but heavy sentences are imposed, not only to deter others from succumbing to such pressures, but also to deter the barons from using them. Mr Perry, for the Crown, argued that it was doing the vulnerable no favours to expand the scope of duress for their benefit, as this would merely encourage their duressors to exploit them. As Professor Ashworth continues:

      "The difficulty with this analysis is that it suggests heavy deterrent sentences for all cases except the most egregious, where it prescribes no penalty at all - a distinction with momentous effects but no clear reference point."

  71.   The Commission was, of course, thoroughly aware of the practical difficulties caused by the fact that duress is most likely to arise in terrorist, gang or other organised crime offences and that, particularly in such circumstances, "the defence of duress is so easy to raise and may be so difficult for the prosecution to disprove beyond reasonable doubt, the facts of necessity being as a rule known only to the defendant himself" (Law Com No 218, [1993] EWLC 218 , para 30.15, quoting Lord Lane CJ in R v Howe [1986] QB 626, 641, to which Lords Bridge and Griffiths attached particular weight in the House of Lords [1987] AC 417, 438 and 444).

  72.   The Commission's solution, strongly supported by the judiciary and most practitioners, was to place the persuasive burden of proving duress, on the balance of probabilities, on the defence (paras 30.16, 33.1 to 33.16). Duress, in their view, was different from other defences, in that the facts on which it is founded are not part and parcel of the incident during which the offence was committed. They will characteristically have happened well before, and quite separately from, the actual commission of the offence that the prosecution must know about and must prove. The difficulty of the prosecution disproving the unilateral claims of the defendant made it "hardly surprising that . . . judges and others should express lack of enthusiasm about the defence of duress as a whole."

  73.   This solution, coupled with the Commission's "sell-out to subjectivism", has been strongly criticised: see Jeremy Horder, "Occupying the moral high ground? The Law Commission on duress" [1994] Crim LR 334. The moral basis of the defence remains a hot topic of debate: see, for example, Professor William Wilson, "The Structure of Criminal Defences" [2005] Crim LR 108. I accept that even the person with a knife at her back has a choice whether or not to do as the knifeman says. The question is whether she should have resisted the threat. But, perhaps because I am a reasonable but comparatively weak and fearful grandmother, I do not understand why the defendant's beliefs and personal characteristics are not morally relevant to whether she could reasonably have been expected to resist. No doubt unduly influenced by Professors Sir John Smith, Edward Griew and Ian Dennis, therefore, I remain attracted by the Law Commission's proposals. The real reasons for the unpopularity of the defence are those given by Lord Lane CJ in Howe: that it is readily raised by the least deserving of people but difficult for the prosecution to disprove. We are told by Mr Perry that, perhaps because of advances in forensic science which have made crimes easier to detect and more difficult to defend, duress is now very frequently raised, often late in the day, by defendants up and down the country.

  74.   If we are not to have legislation to alter the burden of proof, and I agree that it is not open to us to do it ourselves, then I understand your lordships' desire to maintain the objective standards set by Lord Lane in Graham. But it seems to me that the best counter to Lord Lane's concerns is the Fitzpatrick doctrine which is the issue in this case. Logically, if it applies, it comes before all the other questions raised by the defence: irrespective of whether there was a threat which he could not reasonably be expected to resist, had the defendant so exposed himself to the risk of such threats that he cannot now rely on them as an excuse? If even on his own story he had done so, then the defence can be withdrawn from the jury without more ado; if that issue has to be left to the jury, but they resolve it against him, there is no need for them to consider the other questions.

  75.   But how far does this principle go? The 1985 draft code (Law Com No 143, clause 45(4)) and the 1989 draft (Law Com 177, volume 1, clause 42(5) as both provided that the defence of duress "does not apply to a person who has knowingly and without reasonable excuse exposed himself to the risk of such a threat." The code team believed that this was to the same effect as the 1977 draft (Law Com No 83, [1977] EWLC 83, draft Criminal Liability (Duress) Bill, clause 1(5)) which had referred to someone who "knew he would or might be called upon to commit the offence with which he is charged or any offence of the same or a similar character". They must therefore have thought that the words "such a threat" encompassed not only the harm threatened but also the reasons why the threat was made. Similarly, the draft Criminal Law Bill (annexed to Law Com No 218) provided in clause 25(4):

      "This section does not apply to a person who knowingly and without reasonable excuse exposed himself to the risk of the threat made or believed to have been made."

  76.   I agree, of course, that there was nothing in the case law before R v Baker and Ward to limit the kinds of crime which the defendant should have foreseen that he might be compelled to commit. I also agree that the limitation is unworkable in practice and difficult to justify in principle. The principle is that someone who voluntarily accepts the risk of being placed in the "do it or else" dilemma should not be allowed to use that dilemma as an excuse (even if in some circumstances it might amount to mitigation). There are, however, two other questions.

  77.   The first is that the cases tend to talk about exposing oneself to the risk of "unlawful violence". That, it seems to me, is not enough. The foreseeable risk should be one of duress: that is, of threats of such severity, plausibility and immediacy that one might be compelled to do that which one would otherwise have chosen not to do. The battered wife knows that she is exposing herself to a risk of unlawful violence if she stays, but she may have no reason to believe that her husband will eventually use her broken will to force her to commit crimes. For the same reason, I would say that it must be foreseeable that duress will be used to compel the person to commit crimes of some sort. I have no difficulty envisaging circumstances in which a person may be coerced to act lawfully. The battered wife knows very well that she may be compelled to cook the dinner, wash the dishes, iron the shirts and submit to sexual intercourse. That should not deprive her of the defence of duress if she is obliged by the same threats to herself or her children to commit perjury or shoplift for food.

  78.   But this brings me to a concern which I have had throughout this case. It is one thing to deny the defence to people who choose to become members of illegal organisations, join criminal gangs, or engage with others in drug-related criminality. It is another thing to deny it to someone who has a quite different reason for becoming associated with the duressor and then finds it difficult to escape. I do not believe that this limitation on the defence is aimed at battered wives at all, or at others in close personal or family relationships with their duressors and their associates, such as their mothers, brothers or children. The Law Commission's Bills all refer to a person who exposes himself to the risk "without reasonable excuse". The words were there to cater for the police infiltrator (see Law Com No 83, [1977] EWLC 83, para 2.37) but they are also applicable to the sort of association I have in mind. The other elements of the defence, narrowly construed in accordance with existing authority, are more than adequate to keep it within bounds in such cases.

  79.   The certified question on this part of the appeal was:

      "Whether the defence of duress is excluded when as a result of the accused's voluntary association with others: (i) he foresaw (or possibly should have foreseen) the risk of being subjected to any compulsion by threats of violence; or (ii) only when he foresaw (or should have foreseen) the risk of being subjected to compulsion to commit criminal offences, and, if the latter, (iii) only if the offences foreseen (or which should have been foreseen) were of the same type (or possibly of the same type and gravity) as that ultimately committed."

    As will be apparent, I would have chosen option (ii), together with the further explanation of the concept of "voluntary association with others" given in paragraph 78 above. It follows that I too would allow the Crown's appeal on this part of the case.

  80.   On the other part of the case, I would also allow the Crown's appeal, for the reasons given in the opinion of my noble and learned friend, Lord Steyn, with which I agree.

    LORD BROWN OF EATON-UNDER-HEYWOOD

    My Lords,

  81.   I have had the advantage of reading in draft the speeches of my noble and learned friends Lord Bingham of Cornhill and Lord Steyn. For the reasons they give, with which I am in full agreement, I too would answer the certified questions in the manner suggested, allow the Crown's appeal and make the orders proposed by Lord Bingham.


BAILII: Copyright Policy | Disclaimers | Privacy Policy | Feedback | Donate to BAILII
URL: http://www.bailii.org/uk/cases/UKHL/2005/22.html