What are the essential elements of Duress, Necessity and Provocation and how are these differentiated in practice? Suggest ways in which the law might be made coherent
Duress, Necessity and Provocation each operate as defences in which the crime in question has been committed as a result of the negation of the free will of the defendant by external factors which are sufficient to remove (or, in the case of provocation, merely diminish) the culpability of the accused. While each of the defences is distinct, they share many common factors an examination of which supports the argument for the development of a unified, consistent and coherent approach.
Duress is a defence to all crimes except murder, attempted murder and (archaically) certain forms of treason. To succeed in the defence, the defendant must show that he committed the crime in response to threats of death or serious harm.
Necessity is a defence in similar circumstances which is less susceptible to succinct definition but which is commonly agreed to apply in a situation in which the defendant has been confronted with an inescapable choice between two harmful outcomes and has chosen that calculated to do the least harm.
Provocation is a defence only to murder. It is only a partial defence: if successful, the defendant will still be found guilty of manslaughter. The defendant must show that he was provoked by something said or done to lose his self control and that a reasonable person would have done the same in the circumstances.
The constituent elements of each should be considered in greater detail.
The essential elements of the defence of duress were set out by Lane LCJ in Graham and approved by the House of Lords in Howe . Was the Defendant impelled to act as he did because, as a result of what he reasonably believed another party had said or done, he had good cause to fear that if he did not so act that party would kill him or cause him serious physical injury? If so, have the prosecution established so that the jury is sure that a "sober person of reasonable firmness", sharing the characteristics of the Defendant, would not have responded to whatever he reasonably believed had been said or done by committing the offence. Thus there are three objective components:
- The reasonable belief of the defendant;
- The fact that this gave him good cause to be fearful; and
- Whether the response was to be expected of a sober person of reasonable firmness.
The objective nature of this test for duress placed it on a par with self-defence which at the time of Lord Lane's decision was also understood to be an objective test. However, shortly thereafter Gladstone Williams held that an unreasonable belief, if honestly held, may provide the basis for self-defence. Smith submits that the decision in Graham lays down too strict a rule:
"D should surely be judged on the basis of what he actually believed and what he actually feared. If his actual fear was such that no person of reasonable person could be expected to resist it, he should be excused. He may have been unduly credulous or stupid, but he is no more blameworthy than a person whose fear is based on reasonable grounds."
It is essential that there is a direct causal connection between the threat and the action. Thus in DPP v Bell , the defence failed in a situation in which the Defendant committed the offence of driving having consumed excess alcohol because he drove away in a car while in genuine fear of his life. It was held that although the defence would have been available in the immediate circumstances of the escape, the prosecution would nonetheless succeed because he had continued to drive beyond the point where the perceived preservation of his life required it and had therefore committed the offence. The threat must be one of death or serious harm, any other kind of threat, however impelling, will not suffice: in DPP v Hicks a defendant who drove while unfit through drink in order to obtain medication for his daughter who was ill but not mortally so did not escape prosecution. Note, however, that there is a specific statutory defence in s.5(2)(b) of the Criminal Damage Act 1971 which relates to lesser threats in relation to the particular offence of criminal damage. However, as is implied by Hicks (supra), the threat does not have to be directed at the Defendant.
In Pommell , a charge of unlawful possession of a firearm was defeated when the defendant established that he had removed the gun from a person who he feared "was going to do some damage to some people". It is submitted that this is a potentially unsatisfactory decision given the vague nature of the parties in danger but is perhaps justified by the demonstrably grave consequences of use of a firearm. More helpfully, in Abdul-Hussain , the Court of Appeal held that the threat had to be toward the defendant himself or some person for whom he had responsibility. The threat cannot be self-generated: the risk of suicide is not a defence to escaping from prison . A genuine belief in the threat is sufficient even if the threat subsequently proves not to have been real. Graham (supra) makes it clear that a "reasonable belief" in the threat will allow duress as a defence.
Determining the requisite characteristics of the "reasonable person" is problematic. In Bowen , Stuart-Smith LJ reviewed earlier authorities and derived a number of principles:
- The fact that the particular accused is more than usually susceptible to threats does not satisfy the "reasonable person" test;
- However, particular conditions such as tender years and pregnancy with a consequent fear of safety for the unborn child should be taken into account;
- Characteristics which may be appropriately considered in relation to the defence of provocation (see below) such as homosexuality will not necessarily allow duress to succeed since there is no reason to suppose that a homosexual person is less robust in response to threats of death or serious harm than the norm;
- Self-induced abuse of drugs or alcohol does not give rise to a condition compatible with that of the reasonable person;
- Recognised psychiatric conditions giving rise to vulnerability may be taken into account since the jury is entitled to consider the position of the "reasonable person" suffering from such a condition.
The defendant must take any reasonably available steps to escape from the threat such as seeking police protection . The threat must be of imminent harm although there is no necessity for the harm to be virtually immediate. Thus in Abdul-Hussain (supra), the hijacking of an aeroplane by asylum-seekers in order to avoid return to a jurisdiction in which they legitimately expected torture and/or death was held to be a potential defence of duress which ought to have been left to a jury. The defendant must not have put himself in the position in which he might be threatened in such a way: in cases such as Sharp , it has been held that where defendants join criminal gangs and are subsequently threatened if they refuse to carry out the criminal activity contemplated, the defence of duress will not be available provided that they were aware that they had joined a violent gang and aware of the consequent likelihood of threat if they subsequently refuse to participate in the anticipated criminal acts.
The courts have begun to recognise a species of duress which might be described as "duress of circumstances" this might be likened to the defence of necessity (see below). This emerged in Willer in which the defendant was charged with reckless driving after he had driven on a pavement in order to escape a gang of youths which threatened violence against him and his passengers. The judge at first instance refused to leave the defence of necessity to the jury. The ensuing conviction was quashed by the Court of Appeal on the ground that the defence of duress was available. This represented an extension of previously understood principles of duress since the offence committed was not at the direct instigation of the threatening parties: contrast the type of duress in which gang members procure the commission of a specific act by threats as discussed above with this situation in which the decision to drive recklessly was forced upon the defendant but not specifically dictated. Similarly, in Conway a driver drove recklessly because he feared that his passenger was being pursued by individuals who had shot at her previously and she instructed him to drive off. In fact, the pursuing parties were police officers who wished to interview the passenger. The defendant's conviction was quashed on the ground that the defence of "duress of circumstances" had not been left to the jury. Willer and Conway were followed in Martin (Colin) where the circumstances were that the defendant drove whilst disqualified because his wife threatened to commit suicide if he did not get her son to work on time. It is arguable, however, that it was not necessary to apply the extended defence of duress of circumstances to this situation because there was a clear threat by the wife which would have resulted in death or serious harm and which might therefore have been brought under already established principles. While (as will be seen below) there are clearly elements common to the defences of duress of circumstances and necessity, the two cannot be regarded as indistinguishable because duress is not a defence to murder but necessity may be and while only threats of death or serious bodily harm may give rise to duress, as will be seen, threats of a lesser nature may establish the defence of necessity.
Herring points out the confusion that has arisen over the terminology used in relation to this defence arguing that the term has been used in at least three different senses:
- Where the defendant was in a situation in which whatever course of action he chose, some harm would be done and he chose that calculated to cause the least harm ("the lesser of two evils");
- Duress of circumstances (see above);
- As a generic term to describe the overarching doctrine to be applied where the defendant pleads an emergency of some kind and argues that he did what was necessary to avoid a harm encompassing, self-defence, duress and "the lesser of two evils".
In order to differentiate necessity from duress and provocation herein, it is proposed to deal with the doctrine in the first and arguably most restrictive sense described above. In Reniger v Forgossa , the judges of the Exchequer Chamber agreed that breaking laws might be justified "where the words of them are broken to avoid greater inconveniences , or through necessity, or by compulsion…" There is a deep-rooted judicial distrust of this doctrine as expressed by Edmund Davies, LJ in Southwark London Borough v Williams :
"…the law regards with the deepest suspicion any remedies of self-help, and permits those remedies to be resorted to only in very special circumstances. The reason for such circumspection is clear - necessity can very easily become simply a mask for anarchy."
In any event, precise formulation is problematic. In R v Bournewood Community and Mental Health Trust, ex p L , Lord Goff acknowledged the existence of the defence and used it to justify the detention of a mentally ill patient in a situation in which the statutory provisions of mental health law did not apply and the defence was historically applied in relation to the force-feeding of prisoners in Leigh v Gladstone where to do so was necessary to preserve serious damage to their health or even the loss of their lives. However, in S (D) it was stated that Abdul-Hussain (supra) "reflects other decisions which have treated the defence of duress and necessity as being part of the same defence…[and] as being different labels for essentially the same thing". Again, it is argued that this should not be regarded as accurate. As observed above, while duress cannot be a defence to murder, it has now established that necessity may be. In the highly publicised case of the conjoined twins . A hospital wished to perform an operation which might save the life of one twin but which would inevitably lead to the death of the other. Ward LJ acknowledged that the death of one twin in such circumstances would amount to murder by the doctors but proceeded to consider whether there would in the circumstances be any available defence. He arrived at the conclusion that "the carrying out of the operation will be justified as the lesser evil and no unlawful act would be committed. It has been argued that if in Martin (Colin) (supra) the threat by the wife had not been to commit suicide but merely to leave the husband with disastrous effects upon the family, while the defence of duress would clearly not be available since there was no threat of death or serious injury, the defence of necessity might be. A further distinction to be drawn between duress and necessity is that the latter may arise from a positive duty to act. Thus in F v West Berkshire Health Authority it was held to be lawful to carry out a sterilisation operation upon a woman who lacked the necessary mental capacity to consent because failure to due so might result in her becoming pregnant which would have disastrous psychological consequences. Similarly, in the landmark case of Gillick , the provision of contraceptive advice or treatment to a girl under 16 (which would contravene the criminal law then prevailing) was held to be lawful on the basis that failure to do so might lead to serious consequences for her physical or mental health.
It is doubtful whether the defence of necessity can be raised where the statute which creates the criminal offence would admit of a consideration of such circumstances. In DPP v Harris the court took the view that necessity could never be a defence to the offence of driving without due care and attention since the use of the word "due" to qualify the requisite care allowed the court to take into account the prevailing circumstances in determining whether the care exercised in the particular instance was appropriate. There is, however, a potential inconsistency here: as has been seen above, the defence of duress was allowed in the case of Conway above in respect of the more serious offence of reckless driving and its exclusion from use in the case of the lesser offence would appear illogical.
The difficulty of allowing a defence of necessity in respect of murder is illustrated by the old authority of Dudley & Stevens . The accused and a boy were shipwrecked and adrift in an open boat. It was suggested that the boy should be killed and eaten. After some prevarication the boy was killed but by this time he had in any event become seriously weakened and would probably have died shortly in any event. By eating him, the defendants survived. Although it was found as fact that without these acts of murder and cannibalism the defendants were virtually certain to have perished they were nonetheless convicted (albeit then subjected to a commuted sentence). In this case the "necessity" to kill and eat the boy was undeniable but the conviction can probably be justified on the basis that where there were a number of people in the boat, although necessity dictated a murder, it did not dictate the murder of that particular individual.
Other jurisdictions do not appear to have encountered the same degree of difficulty in acknowledging a general defence of necessity. For example, Art.3, s.3.02 of the American Model Penal Code provides:
"Conduct which the actor believes to be necessary to avoid a harm or evil to him or to another is justifiable, provided that:
(a) the harm or evil sought to be avoided by such conduct is greater than that sought to be prevented by the law defining the offence charged…"
This effectively enacts the "lesser of two evils" principle describes above and, it should be noted, encompasses not only situations of necessity such as those illustrated above but would also be applicable to the vast majority of the examples of duress described in the preceding section. While there is therefore considerable overlap between justifiable findings of duress and necessity and, in the case of necessity of circumstances, arguably no true distinction to be drawn, the character of the defence of provocation is somewhat different.
Provocation must be very carefully distinguished from the above defences. It is available only as a defence to murder. There is a further distinction in that even when found provocation cannot operate as a complete defence but, because it negates the mens rea required for murder, has the effect of reducing the homicide to manslaughter. The defence is regulated by s.3 of the Homicide Act 1957:
"Where on a charge of murder there is evidence on which the jury can find that the person charged was provoked (whether by things done or by things said or by both together) to lose his self-control, the question whether the provocation was enough to make a reasonable man do as he did shall be left to be determined by the jury; and in determining that question the jury shall take into account everything both done and said according to the effect which, in their opinion, it would have on a reasonable man."
Thus there is a compelling similarity in this test with that applied in situations of duress with a common reliance upon consideration of the effect of circumstances upon a reasonable man. There are a number of essential limbs to such a defence. The defendant must be shown to have been provoked. Thus in Acott there was clearly a loss of self control when the defendant killed his mother in a frenzied attack but since there was no evidence of a provocative occurrence, it was held that the defence should not have been left to the jury. In that case, Lord Steyn cited the example of a loss of temper caused by slow-moving traffic as insufficient in order to emphasise the principle that the provocation must be as a result of "something said or done". It becomes difficult to establish the necessary connection between the provocation and the loss of control where there has been a lapse of time. In Ahluwalia , a woman who had suffered violence and abuse during many years of an arranged marriage was again threatened with attack by her husband. The same evening, after the husband had gone to bed, she killed him by pouring petrol over him and setting light to it. On appeal, the issue arose as to whether the requisite "sudden and temporary" loss of control could be established in circumstances in which there was such a delay between provocation and action. The Court of Appeal refused to regard the trial judge as incorrect in nonetheless insisting that the loss of control had to be "sudden and temporary" observing that he had not also insisted that the last provocative word or act of the deceased must be followed immediately be the act. The court expressed concern at the suggestion that the law should be relaxed in its insistence upon "sudden and temporary":
"There are important considerations of public policy which would be involved should provocation be redefined so as possibly to blur the distinction between sudden loss of self control and deliberate retribution."
It is therefore possible immediately to identify a distinction between the requirements for duress where the necessary causal connection becomes significantly more difficult to establish with the passage of time and those in respect of duress and necessity where an ongoing situation is quite capable of supplying a defence provided that the court is satisfied that the defendant in acting as he did had no reasonable alternative available to him.
However, just as duress does not have to emanate from the victim, the provocation need not be directed at the defendant nor come from the victim. In Pearson , an older brother who helped his younger brother who had been the victim of abuse kill their father, the abuser, was able to rely upon the defence. There is greater controversy, however, where the provocation might be regarded as self-induced. In Edwards , the Privy Council took the view that provocation could not be relied upon where this had been induced by the actions of the defendant. However, in Johnson , the Court of Appeal held that self-induced provocation might be available in the rare circumstances in which a jury was able to conclude that a reasonable person, albeit having induced the provocation, might nonetheless react as the defendant did. This decision stands in stark contrast to the refusal of the courts to countenance self-induced duress as discussed above.
However, similarities with the law relating to duress are to be found in the decisions which illustrate the "reasonable man" test in relation to provocation. In Camplin , the 15 year-old defendant was raped and then taunted by his attacker. The House of Lords agreed that a minor of this age could not be expected to demonstrate the same degree of self-control as the mature "reasonable man". Lord Diplock not only propounded this principle but proceeded to admit that other characteristics might be taken into account in determining the reasonableness or otherwise of the defendant:
"…the reasonable man referred to in the question is a person having the power of self control to be expected of an ordinary person of the sex and age of the accused, but in other respects sharing such of the accuser's characteristics as [the jury ] think would affect the gravity of the provocation to him; and that the question is not merely whether such a person would in like circumstances be provoked to lose his self-control but also whether he would react to the provocation as the accused did."
This reasoning has been developed in Smith (Morgan) in which the defendant killed in an apparently trivial argument over some carpentry tools and then argued that provocation should nonetheless apply because he suffered from a depressive illness which resulted in him having a lower threshold of self control than might ordinarily be expected. The Court of Appeal certified the question for consideration by the House of Lords of whether characteristics other than the age and sex of the accused attributable to the reasonable man by virtue of s.3 of the Homicide Act 1957 were relevant not only to the gravity of the provocation to him but also to the standard of control to be expected. The House of Lords held that the jury was entitled to take any characteristic of the accused into account in determining whether his reaction was reasonable but that this did not supply a complete exemption. Lord Slynn suggested that it was not enough for a defendant to argue that because he was depressive he must not be expected to have any self control but rather that the jury, while being allowed to take account of the relevant characteristics, still had a responsibility to determine the degree of self control which was still reasonably to be expected of a person possessing those characteristics.
This development brings the law into line with s.169(2) of the New Zealand Crimes Act which states that anything said or done is provocation if "in the circumstances of the case it was sufficient to deprive a person having the power of self control of an ordinary person, but otherwise having the characteristics of the offender [emphasis supplied] of the power of self control." It would thus appear that the "reasonable man" of section 3 has been made redundant. Confirmation of this might be found in the proposal of the Criminal Law Review Commission that the question for the jury should be "whether on the facts as they appeared to the defendant, the provocation can reasonably be regarded as a sufficient ground for the loss of self control leading the defendant to react against the victim with murderous intent and that in answering this question the defendant should be judged with due regard to all the circumstances including any disability, physical or mental, from which he suffered.
The Law Commission has attempted to tidy up the uncertainty caused by the evolution of the test by recommending that the yardstick should be "a person of the defendant's age and of ordinary temperament" but that in deciding whether a person of ordinary temperament in the circumstances of the defendant might have acted in a same or similar way, the court should take into account not only the defendant's age but all the circumstances of the defendant other than matters whose only relevance to the defendant's conduct is that they bear simply upon his or her general capacity for self control. While there then follows a helpful recommendation to the effect that a person should not be treated as having acted in considered desire for revenge if there was a fear a serious violence merely because he or she was also angry toward the deceased as a result of their conduct. This assists in determining the relationship between a legitimate response to provocation and simple retribution discussed above. Regrettably, there then follows (at para. 1.13(5) a missed opportunity:
"The partial defence should not apply to a defendant who kills or takes part in the killing of another person under duress of threats by a third person".
It is regrettable that the ambit of the Law Commission report was restricted to partial defences to murder and thus considered provocation along with defences such as diminished responsibility. While it must be conceded that there are compelling similarities between the above two defences in that they both involve an involuntary loss of control, it might also be argued that provocation has more in common with duress and necessity. In the case of these three, the unifying factor is that the compulsion to act in an illegal way is generated by external circumstances. A further anomaly in the existing law is the fact that while provocation is recognised as a valid defence because the defendant has been caused to lose control and should not in principle therefore held to be accountable for his actions, he will nonetheless remain subject to criminal liability and sanctions as a result of the mandatory conviction for the reduced offence of manslaughter. In this regard, it might be argued that the insistence of this principle is coloured by a perception of the supreme gravity of the crime of homicide. This is a view which does not inevitably find favour in other jurisdictions with the "crime of passion" being recognised as in French law as a complete defence. It is submitted that this is an illogical approach: if there is found to have been a loss of control so extreme as to cause a defendant to kill, it is hard to understand why this should not be allowed to operate so as completely to exonerate him and that he should nonetheless be punished for an act over which by definition the court has found that he was incapable of exercising control.
As has been demonstrated above, the justification for maintaining a distinction between duress and necessity is similarly difficult to ascertain. On occasion, as in cases of so-called "necessity of circumstance" the dividing line is so fine as to be virtually invisible and indeed appears to have been inadvertently crossed by the courts more than once. Rather than focussing upon the perceived differences between the three doctrines and artificially maintaining separate codes and differentiating between situations in which one or other should be applied, it may be more productive to identify the elements common to each. In each of the three cases, there is the commission of an offence by an individual brought about by external factors which are such to cause that defendant to act in a way which he objectively knows or ought to have known to be illegal. In each case, if the circumstances are sufficiently extreme as to allow the court to conclude that the defendant was not exercising free will in such a way as he would if he were not exposed to such pressures, the conduct will be excused (except in the case of provocation in which it is not excused entirely). In each case, the actions of the defendant are tested against the standard of response which would be expected of an ordinary person whose reactions would otherwise be consistent with the norms of acceptable behaviour. While controversy has raged as to the extent to which the concept of the "reasonable man" should be adjusted to take account of the particular vulnerabilities or susceptibilities of the individual in question, there is now a high degree of agreement (not least in respect of provocation) that the test cannot be absolute and unyielding and should be modified to allow for the different reactions which might emanate from persons of different ages, sex and even mental health and temperament. The difficulties of applying a test such as that of the reasonable man are not so insuperable as to make it inappropriate to extend it to all of the situations presently covered by each of the three defences. Accordingly, it might be argued that there should be a generally available defence of "involuntary action caused by external circumstances". The phrase "external circumstances" may then be construed sufficiently widely so as to encompass the whole spectrum of situations illustrated above from killing in anger (which will always remain morally repugnant) to the outcomes of the sort of agonising choices which are inflicted upon entirely blameless individuals such as the medical experts in the case of the conjoined twins (which will always attract a great deal of public sympathy at the dilemma faced by those who would according to the ordinary progress of the law be accused of murder). Such a reform would streamline and clarify the existing law and provide a single code by which human reaction in a wide variety of situations may be judged to be criminal or legally as well as morally defensible.
Clarkson, C., Necessary Action: a new defence, (2004) Crim LR 13
Cooper, J., Courts already respect the right to self defence, The Times, November 30, 2004
Herring J., Criminal Law, Text Cases and Materials, (1st Ed., 2004)
Law Commission, Partial Defences to Murder, (Final Report, August 2004)
Reed, A., Necessity, Lawful Excuse and the Offence of Criminal Damage, (2004) 146 Crim Law 1
Smith, J., Smith & Hogan Criminal Law (10th Ed., 2003)
Wilson, W., The Structure of Criminal Defences, (2005) Crim LR 108