The limits of the forensic process can be exposed in circumstances where it is clear that one member of a limited group of persons was responsible for killing or hurting V, but it is impossible to prove which member of the group it was and, furthermore, impossible to prove that each member of the class was either the principal offender or an accomplice in the offence. An example of this difficulty, one that occurs with regrettable frequency, is where a small child is killed in a domestic setting and it is clear that the child was killed by either her mother, her father, or another member of the household but, because of difficulties of proof, it is impossible to surmount the forensic challenge.

We welcome any attempt to increase the protection of vulnerable victims in domestic settings. This is the admirable intent behind section 5 of the Domestic Violence, Crime and Victims Act 2004. Moreover, any criticism of this new offence must be offered in the awareness that a perfect solution is not to hand. There is an irresolvable tension between protecting the vulnerable in a domestic setting and protecting other members of the household from wrongful conviction.
Under section 5, an offence of causing or allowing the death of a child (a person under 16) or vulnerable adult (a person made vulnerable through age or disability) occurs if the following is proved:

(1) V, a child or vulnerable adult, has died as a result of an unlawful act of a person who was a member of the same household as the victim and who had frequent contact with him;
(2) there was a significant risk of serious physical harm being caused to V by the unlawful act of such a person;
(3) D was a member of the household in frequent contact with V at the time of the act;
(4) either D was the person whose act caused the death of V or D was, or ought to have been aware, of the risk specified in (2) and failed to take such steps as he could reasonably have been expected to have taken to protect V from the risk;
(5) the killing of V occurred in circumstances which D foresaw or ought to have foreseen.
The offence carries a maximum sentence of 14 years imprisonment. Several features require comment.

Negligence as mens rea
Although less serious than manslaughter, this new form of homicide is a grave, stigmatic offence. Accordingly, it is worrying that the default form of culpability for this offence is simple negligence. D may be implicated on the basis of facts that he should have foreseen and for failure to take steps he should have taken. Not only is this a lesser culpability than the gross negligence required for the cognate form of manslaughter, the new offence imposes duties to act on a much wider basis than manslaughter. To be liable for manslaughter on the basis of omission, one must be a member of that very limited class of persons on whom duties of care are imposed to protect others or, otherwise, a person who has voluntarily assumed a duty of care for V. For this new offence, it is enough to be 16 years of age or over, a member of the same household, and have frequent contact with V. A potential defendant could well be V’s teenage sibling, enduring the same chaotic and violent household as the even more unfortunate V.

Because of the range of likely defendants, it is vital that courts, when assessing what steps D should have taken to prevent the killing of V, should be realistic about the lived-in environment and capacities of D and not, with the benefit of hindsight, impose a standard of vigilance to be expected of persons of competence from stable backgrounds. Many of the auguries are not good [S&S2 p. 152-4]. In Stone and Dobinson [1977] QB 354, no account was taken of the low intelligence and exceedingly limited social competencies of the defendants when finding that they had been grossly negligent in failing to do more to help V. In Elliot and C [1983] 2 All ER 1005, the young age and psychological vulnerability of D did not deflect a finding of Caldwell recklessness. With a standard of simple negligence for a serious offence it is vital that courts should determine what D herself could and should have done in the light of her age and competencies in the social environment in which she was placed. If past practice is a guide, this contextualised approach to the negligence standard may well not be taken.

Omissions and causation
Where D is not the perpetrator, it must be found that he allowed V’s death. To say of someone that they allowed something to occur implies at least a social responsibility to prevent the outcome. If this requirement is taken seriously it should reduce the range of potential defendants. It is important that courts do not infer this responsibility merely from the fact that D is over 16 and shares the same household as V. Additionally, it is important that courts should rule that D will only “allow” a death if the actions he should have taken would, beyond any reasonable doubt, have prevented it [S&S2 p. 104]. Unfortunately, in the case of omissions, the need for a causal nexus can be attenuated or overlooked even in situations where a causal link between the omission and the relevant harm is clearly required. In Stone and Dobinson, the defendants were convicted of manslaughter because of their failure to arrange help for the anorexic V. Given the nature of that condition, it is far from obvious that V’s life would have been saved had the defendants done all that they should have done. The matter was not examined; there was merely an assumption that their respective omissions played a causally sufficient part in V’s death. A likely scenario for this new offence would be a failure on the part of D to report, to the police or social services, violence by her partner toward one or more of their children. This failure is, of itself, insufficient for liability. It must be shown that reporting her partner would have led to life-saving intervention by others.

Right to silence
It is contemplated that this new crime will, on occasion, be used in tandem with charges of murder and manslaughter. Section 6 of the Domestic Violence, Crime and Victims Act 2004 provides that, where an adverse inference could be taken from D’s silence under section 35 of the Criminal Justice and Public Order act 1994 in respect of causing or allowing the death of a child or vulnerable adult, an adverse inference may also be taken in respect of any charge of murder or manslaughter arising from the same facts. This could put a defendant at risk of, say, a conviction for murder in circumstances where there is no case to answer in relation to that charge. That position is certain to be challenged as infringing the right to a fair trial guaranteed by Article 6 of the ECHR. There are also likely to be challenges under Article 7 if the offence is used indiscriminately. Mere membership of a violent household does not put one on fair notice that one will incur serious criminal liability in respect of the violent acts of other members of the household.

An emotive issue
There is no question that domestic violence against children and other vulnerable persons is a real social problem. The Government was entirely correct in seeking by legislation to address this violence. Indeed, it could have gone further: it is notable that the new offence concerns causing and allowing death, and does not extend to cases of serious injury. There are many more cases of serious injury arising from violence in domestic settings than there are fatalities. The understandable emotion occasioned by the death of a child makes for a readier public acceptance of the kind of measures we have criticised here. But the end does not justify the means. A more balanced and focused offence, meeting the concerns identified here and extended to cases of serious injury in households as well as deaths, would offer more effective protection to victims and greater fairness to defendants.

Download pdf file