When determining the defence of duress, the objective nature of the enquiry restricts the availability of the defence. The test in Graham is effectively a three-phase test. Firstly, the defendant’s perception that he is being threatened, the defendant’s perception of the magnitude of the threat, and lastly, the defendant’s fortitude in the face of the threat. There is a good case for saying that the first question should be purely subjective, that is, D should be judged on the basis of what he actually believed, and what he actually feared. However, this is made partly objective by the requirement of a ‘reasonable’ belief, unduly restricting the first and second points. . In Graham (1982) it was stated that D is not to be judged on the facts as he believed them to be, but whether as a result of what D ‘reasonably’ believed that the duressor had said or done, D had good cause to fear death or serious injury. Therefore, unlike other defences such as self-defence, an honest but mistaken belief will not excuse the defendant. In the last stage of the test a threat of death or serious injury excuses the accused whose resistance it overcomes only if the threat might have overcome the resistance of a ‘sober person of reasonable firmness, sharing the characteristics of the accused.’
In Bowen (1996) the defendant had a low IQ falling short of mental impairment. It was decided that this could not be said to be a characteristic that made those who had it less courageous and less able to withstand threats and pressure. The Court of Appeal explained which characteristics of the defendant are admissible. The jury is entitled to take account of age, sex and physical health, and some mental characteristics so long as they do not bolster the claim of the defence. Psychiatric evidence may be admissible to show that the accused is suffering from a recognised psychiatric condition, provided persons generally suffering from such a condition may be more susceptible to pressure and threats. However, individual characteristics that have a direct bearing on the ability of D to resist the threats are to be excluded. Consequently, if D is more pliable, vulnerable, timid or susceptible to threats than a normal person, these characteristics will not be legitimate to invest the ordinary person for the purpose of considering the objective test. In Horne (1994) the judge stated that if the standard for comparison is a person of reasonable firmness, it must be irrelevant for the jury to consider ant characteristics of D, which show that he is not a person of reasonable firmness. ‘It would be a contradiction in terms to ask the jury this question and then ask them to take into account, as one of D’s characteristics, (mental instability).’ Following Flatt, self induced characteristics are not relevant in determining whether a person of ordinary firmness would similarly have succumbed to a threat like the defendant did, therefore this would include voluntary intoxication.
A major restriction of duress is that it does not provide a defence where the defendant is on a charge of murder or attempted murder, Howe (1987) and Gotts (1992). These cases confirmed the decision made over a hundred years earlier, in Dudley and Stephens (1884). In this case D and S had been shipwrecked with another man and a cabin boy. After several days without food or water they decided to kill and eat the boy who was the weakest of the four. Four days later they were rescued. They pleaded duress of circumstances but the courts rejected this. In Howe it was stated that even though a defendants life may be threatened unless he kills another, the defence of duress is not available and he must be prepared to die himself. In this case the House of Lords gave a number of reasons for not allowing the defence to apply to murder. The first of which is that D should act as a hero. However, it can be argued that this does not follow the two stage test in Graham, as the ‘reasonable man’ may have killed in the same circumstances, therefore if this is satisfied then the law should not expect a person to be a hero. Secondly, The law requires that a person take the lesser of two evils; this is not satisfied if an innocent life is taken. However, duress, by its very nature is based on a concession to human frailty, therefore issues of proportionality are merely of evidential importance, not a prerequisite. If the Graham test is satisfied then it is not necessary that the lesser of two evils should apply. There is also the situation where D is required to kill one person in order to save the life of more than one person, then he can be said to have chosen the lesser of two evils.
The Lords also argued that by not implementing the Law Commission’s recommendations in 1977 it must be assumed that Parliament did not want duress to apply to murder. However, a counter argument here would be that the legislation in Lynch two years previous to this, which allowed duress for accessories to murder, was not overruled, so it could be argued that Parliament agreed with this. This argument effectively proves very little, as the Government may not have given Parliament an opportunity to consider the issue for a number of reasons, for example, lack of time, or having more important legislation to pass. Their fourth argument was that it should be left up to the Parole Board to decide on possible early release for defendants in this situation. However, this does leave the problem that morally innocent people are still convicted and labelled as murderers and are left to the discretion of the Parole Board.
The Lords also argued that the defence was ‘so easy to raise and may be difficult for the prosecution to disprove. This is contradicted by the decision in Shah not to allow the defence or voluntary involvement in terrorist organisations or gangs. Also, the objective test in Graham limits the defence and its availability. This would also apply to most defences and ignores the fact that in Howe, like Lynch the jury rejected the defence and convicted.
As just argued, the defence is also restricted, mainly due to public policy considerations, by the fact that it is not available where D has voluntarily exposed himself to the risk. This is clearly explained in the case of Sharp (1986). D attempted an armed robbery, which resulted in a death, he argued that he only took part because he had been threatened with death if he did not co-operate. The judge withdrew the defence and the Court of Appeal upheld the conviction. Lord Lane identified the features that deny the defence in this situation. Firstly, D must voluntarily join a criminal organisation or gang, he must have ‘knowledge of its nature,’ he must know that other organisation or gang members ‘might bring pressure on him to commit an offence, and lastly, he must have been ‘an active member when he was put under such pressure.’ In the case of Shepherd in the following year, D was a member of a gang of non-violent shoplifters. D was charged with theft and pleaded duress on the basis that he had tried to leave the gang but was threatened with violence against him and his family. The trial judge withdrew the defence from the jury but on appeal his conviction was quashed, as the question should have been put to the jury of whether D knew of a propensity to violence when he joined the gang. Restrictions also apply where D associates with people known to be violent criminals such as drug dealers, Baker and Ward (1999), Heath (1999).
Due to the restrictions analysed in the defence of Duress, the Law Commission recommended the abolition of the objective test in the first part of the Graham test, which would make it easier for the defendant to gain the defence. However, to compensate for this it has been suggested that the burden of proof be reversed so that D has to prove his honest belief of the threat. However, this has been controversial as it effectively goes against a whole line of authority in the criminal law (Woolmington), and might been seen as too high a price to pay. It has also been suggested that the objective test be abolished in the second stage of the test as well, with the jury being able to take into account all circumstances that affect the gravity of the threat to the defendant, including factors such as suggestibility. This would bring the defence in line with the special defence of provocation following the decision in Smith (2000). This may possible leave the defence of duress open to abuse as a result, however the jury would still have the ultimate decision.