[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	The Law Commission
You are here: BAILII >> Databases >> The Law Commission >> Homicide: Murder, Manslaughter And Infanticide (Report) [2006] EWLC 304 (28 November 2006) URL: http://www.bailii.org/ew/other/EWLC/2006/304.html Cite as: [2006] EWLC 304

[New search] [Printable PDF version] [Help]

Laid before Parliament by the Secretary of State for Constitutional Affairs and Lord Chancellor pursuant to
section 3(2) of the Law Commissions Act 1965

Ordered by The House of Commons to be printed

28 November 2006

HC 30 London: TSO £xx.xx

The Law Commission was set up by the Law Commissions Act 1965 for the purpose of promoting the reform of the law.

The Law Commissioners are:

The Honourable Mr Justice Etherton, Chairman
Professor Hugh Beale QC, FBA
Mr Stuart Bridge
Professor Jeremy Horder
Mr Kenneth Parker QC

The Chief Executive of the Law Commission is Mr Steve Humphreys.

The Law Commission is located at:
Conquest House
37-38 John Street
Theobalds Road, London WC1N 2BQ.
The terms of this report were agreed on 1 November 2006.
The text of this report is available on the Internet at: http://www.lawcom.gov.uk

THE LAW COMMISSION

MURDER, MANSLAUGHTER AND INFANTICIDE

To the Right Honourable the Lord Falconer of Thoroton, Secretary of State for Constitutional Affairs and Lord Chancellor

PART 1

WHY IS A NEW HOMICIDE ACT NEEDED?

THE TERMS OF REFERENCE FOR THE REVIEW OF MURDER

1.1      In July 2005, the Government announced a review of the law of murder in England and Wales, with the following terms of reference:

(1) To review the various elements of murder, including the defences and partial defences to it, and the relationship between the law of murder and the law relating to homicide (in particular manslaughter). The review will make recommendations that:

(a) take account of the continuing existence of the mandatory life sentence for murder;

(b) provide coherent and clear offences which protect individuals and society;

(c) enable those convicted to be appropriately punished; and

(d) are fair and non-discriminatory in accordance with the European Convention of Human Rights and the Human Rights Act 1998.

(2) The process used will be open, inclusive and evidence-based and will involve:

(a) a review structure that will look to include key stakeholders;

(b) consultation with the public, criminal justice practitioners, academics, those who work with victims' families, parliamentarians, faith groups; and

(c) looking at evidence from research and from the experiences of other countries in reforming their law.

(3) The review structure will include consideration of areas such as culpability, intention, secondary participation etc inasmuch as they apply to murder. The review will only consider the areas of euthanasia and suicide inasmuch as they form part of the law of murder, not the more fundamental issues involved which would need separate debate. For the same reason abortion will not be part of the review.

How did the Law Commission take forward these terms of reference?

1.2      We did not review every issue that could, in theory, be regarded as falling within the scope of the review. The areas of law that seemed to us to give rise to real difficulty or anomalies have guided us in our focus.

1.3      Issues that we did not address, even though they fell within our terms of reference, included:

(1) The prohibited conduct element, including causation, the legal criteria governing when life begins and when life ends and child destruction (the offence of killing a child in the womb who was capable of being born alive).

(2) Justifications for killing, such as necessity and self-defence.^[1]

(3) The defences of insanity and intoxication.

(4) Aggravating features of an admitted murder, such as an especially evil motive or the fact that a child was intentionally targeted. We have not considered these because we believe that they were adequately addressed by Parliament in the guidelines contained in the Criminal Justice Act 2003.

1.4      Three issues warrant special mention. The first issue is the defence of duress.^[2] Duress, like insanity, is a complete defence to any crime to which it applies. At present, there are judicially created rules under which duress is no defence to murder or to attempted murder. We believed that consultees might favour reform of this rule, so we decided to consider the defence of duress in detail rather than leaving it to a separate review of duress as it applies across the board.

1.5      The second issue is 'mercy' killing. This fell within the scope of our terms of reference only in so far as it related to the grounds for reducing a more serious homicide offence to a less serious one. We considered it carefully in that context both in the Consultation Paper ("the CP") ^[3] and during the consultation process.

1.6      We have decided that a recommendation for a specific partial defence of 'mercy' killing should await a further and more detailed consultation exercise specifically concentrating on the issue. We quite simply did not have the time that we would have needed to conduct a full consultation with the necessary groups. Such a consultation would have needed to include, amongst others: the Department of Health; groups representing doctors and patients; care organisations; gerontologists; and groups both 'pro' and 'anti' euthanasia. We address the matter further in Part 7 and recommend that such a consultation exercise should be undertaken.

1.7      The third issue is the fact that there is no draft Homicide Bill reflecting our recommendations. The Law Commission commonly appends a draft Bill to accompany its reports but we have not done so in this report. This is because our report is only the first stage in the current review of the law of homicide. The Home Office will be undertaking the second stage of the review.^[4] Accordingly, it would have been inappropriate for the Commission to produce a draft Bill.

THE EXISTING LAW AND THE PROBLEMS WITH IT

1.8      The law governing homicide in England and Wales is a rickety structure set upon shaky foundations. Some of its rules have remained unaltered since the seventeenth century, even though it has long been acknowledged that they are in dire need of reform. Other rules are of uncertain content, often because they have been constantly changed to the point that they can no longer be stated with any certainty or clarity. At the end of the nineteenth century there was a valuable attempt at wholesale reform. This was thwarted largely by quite unconnected political problems. The consequence was that the Homicide Bill did not progress beyond its second reading in Parliament. Moreover, certain piecemeal reforms effected by Parliament, although valuable at the time, are now beginning to show their age or have been overtaken by other legal changes and, yet, have been left unreformed.^[5]

1.9      This state of affairs should not continue. The sentencing guidelines that Parliament has recently issued for murder cases^[6] presuppose that murder has a rational structure that properly reflects degrees of fault and provides appropriate defences. Unfortunately, the law does not have, and never has had, such a structure. Putting that right is an essential task for criminal law reform.

1.10      We will recommend that, for the first time, the general law of homicide be rationalised through legislation. Offences and defences specific to murder must take their place within a readily comprehensible and fair legal structure. This structure must be set out with clarity, in a way that will promote certainty and in a way that non-lawyers can understand and accept.

1.11      We will be going into these matters in much greater depth but the following is a brief explanation of the existing law and its flaws.

The current structure of offences

1.12      Two general homicide offences – murder and manslaughter – cover the ways in which someone might be at fault in killing. There are also a number of specific homicide offences, for example, infanticide and causing death by dangerous driving (the latter was not within our terms of reference for consideration).

1.13      Murder, which carries a mandatory life sentence, is committed when someone ("D") unlawfully kills another person ("V") with an intention either to kill V or to do V serious harm.

1.14      Manslaughter can be committed in one of four ways:

(1) killing by conduct that D knew involved a risk of killing or causing serious harm ('reckless manslaughter');

(2) killing by conduct that was grossly negligent given the risk of killing ('gross negligence manslaughter');^[7]

(3) killing by conduct taking the form of an unlawful act involving a danger of some harm to the person ('unlawful act manslaughter'); or

(4) killing with the intent for murder but where a partial defence applies, namely provocation, diminished responsibility or killing pursuant to a suicide pact.

The term 'involuntary manslaughter' is commonly used to describe a manslaughter falling within (1) to (3) while (4) is referred to as 'voluntary manslaughter'.

Problems with these offences

1.15      The current definitions of these offences (and, for the most part, of the provocation defence) are largely the product of judicial law making in individual cases over hundreds of years. They are not the products of legislation enacted after wide consultation and research into alternative possibilities. From time to time, the courts have tinkered with the definitions. New cases have then generated further case law to resolve ambiguities or new avenues for argument left behind by the last case.^[8]

1.16      We identified a number of problems with the homicide offences in the CP. The consultation process has confirmed that numerous problems exist.

The serious harm rule

1.17      Under the current law, D is liable for murder not only if he or she kills intentionally but also if he or she kills while intentionally inflicting harm which the jury considers to have been serious. In our view, the result is that the offence of murder is too wide. Even someone who reasonably believed that no one would be killed by their conduct and that the harm they were intentionally inflicting was not serious, can find themselves placed in the same offence category as the contract or serial killer. Here is an example:

D intentionally punches V in the face. The punch breaks V's nose and causes V to fall to the ground. In falling, V hits his or her head on the curb causing a massive and fatal brain haemorrhage.

1.18      This would be murder if the jury decided that the harm that D intended the punch to cause (the broken nose) can be described as 'serious'.^[9] Whilst it is clear that a person who kills in these circumstances should be guilty of a serious homicide offence, it is equally clear to the great majority of our consultees that the offence should not be the top tier or highest category offence.

1.19      As we explained in the CP,^[10] Parliament, when it passed the Homicide Act 1957, never intended a killing to amount to murder – at that time a capital offence – unless (amongst other things) the defendant ("D") realised that his or her conduct might cause death. The widening of the law of murder beyond such cases came about through an unexpected judicial development of the law immediately following the enactment of the 1957 legislation.^[11] More will be said about this in paragraphs 1.26 to 1.29.

1.20      The inclusion of all intent-to-do-serious-harm cases within murder distorts the sentencing process for murder. The fact that an offender only intended to do serious harm, rather than kill, is currently regarded as a mitigating factor that justifies the setting of a shorter initial custodial period as part of the mandatory life sentence.^[12] On the face of it, this seems perfectly reasonable. However, there is a strong case for saying that when an offence carries a mandatory sentence, there should be no scope for finding mitigation in the way in which the basic or essential fault elements come to be fulfilled.

1.21      We have been informed by research, carried out by Professor Barry Mitchell, into public opinion about murder.^[13] This shows that the public assumes that murder involves an intention to kill or its moral equivalent, namely a total disregard for human life.^[14] The latter may not be evident in a case where someone has intentionally inflicted harm the jury regards as serious, as when D intentionally breaks someone's nose. Indeed, some members of the public regarded deaths caused by intentionally inflicted harm that was not inherently life threatening as being in some sense "accidental".^[15]

1.22      Having said that, we do not recommend that killing through an intention to do serious injury^[16] should simply be regarded as manslaughter. Manslaughter is an inadequate label for a killing committed with that degree of culpability. In any event, to expand the law of manslaughter still further would be wrong because manslaughter is already an over-broad offence.^[17]

1.23      We will be recommending that the intent-to-do-serious-injury cases should be divided into two. Cases where D not only intended to do serious injury but also was aware that his or her conduct posed a serious risk of death should continue to fall within the highest category or top tier offence. This is warranted by the kind of total disregard for human life that such Ds show. They are morally equivalent to cases of intentional killing. Cases where D intended to do serious injury but was unaware of a serious risk of killing should fall (along with some instances of reckless killing) into a new middle tier homicide offence.

Reckless manslaughter

1.24      The scope of murder is both too broad and too narrow. Where the scope of murder is too narrow, the scope of manslaughter is correspondingly too broad. In particular, the law is too generous to some who kill by 'reckless' conduct, that is those who do not intend to cause serious harm but do realise that their conduct involves an unjustified risk of causing death. The law is too generous in treating all those who realise that their conduct poses a risk of causing death but press on regardless as guilty only of manslaughter. Again, the problems have arisen from the way that periodic judicial development of the law in individual cases, albeit well-intentioned, has changed the boundaries of homicide offences.

1.25      It is necessary to provide some brief background to the discussion. When the Homicide Act 1957 was passed, it was still accepted by both Parliament and by the courts that the archaic language of 'malice aforethought' governed the fault element in murder. But malice aforethought has never been a term with very clear boundaries and differences soon emerged between the courts and Parliament over how it should be understood (as indicated in paragraph 1.19 above).

1.26      At that time, the courts treated malice aforethought as covering cases in which the offender: either (a) intended to kill; or (b) intended to cause serious harm; or (c) had knowledge that the act which causes death will probably cause the death of or grievous bodily harm to some person.^[18] However, during the passage of the Homicide Act 1957, Parliament was led to believe that (b) was not a species of malice aforethought and that malice aforethought could not be established without D being proven to have at least been aware that the harm done was life-threatening. This was the basis upon which the Homicide Act 1957 was passed. Parliament's belief was founded on the Lord Chief Justice's evidence to the Royal Commission on Capital Punishment, whose report led to the passing of the Homicide Act 1957.^[19]

1.27      Defined either way, however, malice aforethought provided sufficient coverage to ensure that the worst kinds of reckless killer could be convicted of murder. Here is an example:

D, intending to cause fear and disruption, plants a bomb. D gives a warning which D believes might be sufficient to permit the timely evacuation of the area but probably would not be. In the ensuing explosion, someone is killed.

1.28      At the time of the enactment of the Homicide Act 1957, such a person would have been regarded by both Parliament and the courts as acting with 'malice aforethought' and would be guilty of murder. As Lord Chief Justice Cockburn, many years before, had put it:

If a man did an act, more especially if that were an illegal act, although its immediate purpose might not be to take life, yet if it were such that life was necessarily endangered by it, - if a man did such an act, not with the purpose of taking life, but with the knowledge or belief that life was likely to be sacrificed by it, that was not only murder by the law of England, but by the law of probably every other country.^[20]

1.29      Immediately after the passing of the Homicide Act 1957, however, the Court of Appeal indicated that only an intention to kill or to cause serious harm – ((a) and (b) in paragraph 1.26 above) – would suffice as proof of 'malice aforethought'.^[21] In 1975, there was what can be interpreted as an attempt by the House of Lords to reconcile this new view with the older and broader understanding of malice aforethought, that is, as including exposing others to a probable risk of serious harm or death.^[22] It was held that a jury could find that D intended to kill or to cause serious harm if he or she foresaw one or other of these results was a highly probable result of his or her conduct.^[23] However, developments did not stop there.

1.30      In 1985, the use of the label "malice aforethought" to describe the fault element for murder was overtly criticised by the House of Lords, even though it is at the heart of section 1 of the Homicide Act 1957.^[24] Further, the House of Lords made it clear beyond doubt that intention should not be construed as to automatically include the mere foresight of probable consequences. That development led to a series of further cases on the exact width of the law.^[25] What has emerged is that murder no longer includes killing by reckless risk-taking, as such, however heinous the killing. Such killings, although they can be encompassed by the woolly language of "malice aforethought", are not intentional. Consequently, from 1985 onwards, the hypothetical bomber described in paragraph 1.27 above could no longer be guilty of murder because he or she did not intend to kill or to cause serious injury. He or she could only be guilty of (reckless) manslaughter.^[26]

1.31      Backed by the vast majority of our consultees, we are recommending that the hypothetical bomber should be guilty of a homicide offence more serious than manslaughter. In Part 2, we give some other examples where the culpability of the offender is similarly so high that a manslaughter verdict is an inadequate label for the offence.^[27]

The 'two category' structure of general homicide offences

1.32      The distinction between murder and manslaughter is almost certainly over 500 years old. No further general category of homicide has been developed over the intervening period. So, over the centuries, the two categories of murder and manslaughter have had to bear the strain of accommodating changing and deepening understandings of the nature and degree of criminal fault and the emergence of new partial defences. They have also had to satisfy demands that labelling and sentencing should be based on rational and just principles.

1.33      Further, the existence of the death penalty and, then, its successor the mandatory life sentence for murderers, has meant that the argument over who should be labelled a 'murderer' has become identified with who should receive the mandatory sentence for murder. Whilst in some respects understandable, the link with sentencing can distort the argument about labels. For example, it is arguable, and we will recommend that, although a person who kills intentionally in response to gross provocation does not deserve a mandatory sentence, he or she should still be labelled a 'murderer'.

1.34      Our consultees almost all agreed that the two-category structure of the general law of homicide is no longer fit for purpose. Consequently, we are proposing to replace the two-tier structure with a three-tier structure. Such a structure will be much better equipped to deal with the stresses and strains on the law and with the issues of appropriate labelling and sentencing. The three tiers in descending order of seriousness would be first degree murder, second degree murder and manslaughter.

1.35      Under our recommendations, first degree murder would encompass:

(1) intentional killing; or

(2) killing through an intention to do serious injury with an awareness of a serious risk of causing death.

1.36      Second degree murder would encompass:

(1) killing through an intention to do serious injury (even without an awareness of a serious risk of causing death); or

(2) killing where there was an awareness of a serious risk of causing death, coupled with an intention to cause either:

(a) some injury;

(b) a fear of injury; or

(c) a risk of injury.

1.37      Second degree murder would also be the result when a partial defence of provocation, diminished responsibility or killing pursuant to a suicide pact is successfully pleaded to first degree murder.

1.38      Manslaughter would encompass:

(1) where death was caused by a criminal act intended to cause injury, or where the offender was aware that the criminal act involved a serious risk of causing injury; or

(2) where there was gross negligence as to causing death.

Our recommendations for manslaughter build upon previous Law Commission recommendations and Home Office proposals.^[28]

Complicity in murder committed by another person

1.39      A serious problem has arisen in murder cases with regard to what lawyers call the doctrine of 'complicity', that is, involving oneself in a criminal enterprise with another who commits murder. A typical example might be where A, B and C see V (a supporter of a rival football team), walking home alone. They set upon V punching and kicking him. When V falls over, B produces a knife and stabs V through the heart. A and C say that (i) although they knew that B sometimes carried a knife, they did not know he would use it on this occasion, and (ii) they did not have murder in mind when attacking V.

1.40      In this example, B is likely to be found guilty of murder as a principal offender. The question is: are A and C involved in the killing in some way that is sufficiently culpable to warrant being guilty of a homicide offence? The law currently answers this question by telling the jury to ask itself whether there was a 'fundamental difference' between what A and C thought might happen and what B did.^[29] If the jury thinks that there was such a fundamental difference, then A and C are guilty of no homicide offence at all. A and C will be guilty of assaulting V, but assault is obviously a much lesser offence than murder or manslaughter.

1.41      We believe that this represents a gap in the law. It should be possible to convict A and C of a homicide offence. This gap appears especially significant when people in A and C's position knew from the start of the joint criminal venture that the person who eventually commits the murder was armed. In such circumstances, the mere fact that they did not foresee that the armed participant would actually commit murder should not absolve A and C from all responsibility for the homicide.

1.42      In the CP,^[30] we put forward a proposal for filling this gap. It was that A and C should be guilty of manslaughter if:

(1) they were engaged in a joint criminal venture with B; and

(2) it should have been obvious to them that B might commit first or second degree murder in the course of that joint criminal venture.

This would mean that A and C could not escape responsibility for the homicide simply by denying that they knew B might commit murder if this should have been obvious to them. This is especially likely to be the case when they knew that the eventual killer was already armed. The vast majority of consultees supported the proposal. In Part 4, we set out and explain the recommendation that we are making.

1.43      Moreover, we will also recommend that if A, B and C are engaged in a joint criminal enterprise and A and C do realise that B might commit murder in the fulfilment of the joint venture, or intend that B does so, then A and C should be guilty of murder along with B. The crucial difference is A and C's awareness of what B may or will do. This awareness makes it justifiable to convict A and C of murder, not just manslaughter. Again, the issue is discussed in more depth in Part 4.

Partial defences

1.44      In this review, we are mainly concerned with partial defences, for example provocation, rather than with complete defences, such as self-defence. Currently, there are three partial defences to murder: provocation, diminished responsibility and killing in pursuance of a suicide pact. If successfully pleaded, they do not result in a complete acquittal but in a conviction of manslaughter rather than murder.

1.45      There is also what might be called a 'concealed' partial defence, created by legislation as a specific offence. This is the offence of infanticide which is committed when a mother whose balance of mind is disturbed kills her baby when the baby is less than 12 months old.^[31] Infanticide is both an offence and a partial defence. A mother may be charged with this offence. Alternatively, she may be charged with murder and plead infanticide as a partial defence to murder.

Problems with these partial defences

PROVOCATION

1.46      The partial defence of provocation is a confusing mixture of judge-made law and legislative provision. The basic rule has been clear enough for a long time: if D kills, having been provoked to lose his or her self-control, in circumstances in which an ordinary person might also have done so, it is manslaughter not murder. However, the higher courts have disagreed with one another on a number of occasions about the scope of the defence. Consequently, the scope of the defence has become unclear. There appears to be little prospect of the courts resolving this disagreement.

1.47      A particular anomaly is that D is entitled to have evidence that he or she was provoked to lose self-control put before the jury no matter how unlikely it is that the defence will succeed. Thus, if D claims that he was provoked to lose his self-control by V's failure to cook his steak medium rare as ordered, the defence has to be put to the jury even though it has no merit and ought to be rejected. By way of contrast, if instead of being provoked, D's killing was a fear-driven over-reaction to a threat of future serious violence, he or she has no defence to murder at all, however well founded the fear. The courts have declined to create or extend a partial defence to cover such cases. Accordingly, reform of this area now depends upon legislative action by Parliament.

1.48      In 2004 we recommended reform of the partial defence of provocation. We set out how we thought the defence should be reformed to create greater certainty and to correct the lop-sided character of the law.^[32] During the current consultation, consultees have again broadly agreed that the defence should be reformed along the lines we are recommending. We return to this topic in Part 5.

DIMINISHED RESPONSIBILITY

1.49      The introduction of the partial defence of diminished responsibility in 1957 was a welcome reform. However, medical science has moved on considerably since then and the definition of diminished responsibility is now badly out of date. We are recommending an improved definition which we have drawn up with the help of the Royal College of Psychiatrists and other expert consultees. The new definition has had wide support amongst consultees. We believe that the new definition has the flexibility to accommodate future changes in diagnostic practice, whilst ensuring that the public remains well protected from those mentally disordered offenders who pose a continuing threat. Detailed discussion can be found in Part 5.

KILLING IN PURSUANCE OF A SUICIDE PACT

1.50      Section 4 of the Homicide Act 1957 makes the survivor of a suicide pact who took part in the killing of another person in the pact guilty of manslaughter and not murder. This provision was meant to allow the jury to take pity on those desperate enough to seek to take their own lives along with that of another person or persons. In the CP,^[33] we discussed whether the provision should be repealed. We provisionally concluded that it should be. However, in the light of our decision not to take the issue of 'mercy' killing further at this stage, we are not now recommending the repeal of section 4. Its merits should be considered as a part of the broader question of whether there should be a partial defence of 'mercy' killing. The issue is considered in Part 7.

INFANTICIDE

1.51      Where the offence of infanticide is concerned, the problem is not so much the definition but, rather, the procedure for ensuring that evidence of a mother's mental disturbance at the time of the killing is heard at trial. A mother may be 'in denial' about having killed her infant. She may, therefore, be unwilling to submit to a psychiatric examination if the point of this examination seems to her to be to find out why she did it. This is because she cannot accept that she did do it. In such circumstances, she is unlikely to have another defence and is, therefore, likely to be convicted of murder. This is not in the public interest.^[34] However, this is not an easy problem to solve. We recommend the adoption of a post-trial procedure designed to do justice in these cases. The procedure is explained in Part 8.

Missing defences

1.52      Whereas there has recently been controversy over whether provocation should continue to be a partial defence to murder, other strong claims for mitigation of the offence of murder have failed to gain legal recognition. Judges have decided that they would prefer Parliament to decide whether there should be new partial defences to murder but Parliament has not had the time to consider the matter.

Excessive Force in Defence

1.53      We have already mentioned the need for a partial defence when D, fearing serious violence from an aggressor, overreacts by killing the aggressor in order to thwart the feared attack. We are recommending that D's fear of serious violence should be the basis for a partial defence to murder through reform of the provocation defence.^[35] This has been almost unanimously approved by consultees.

Duress

1.54      Circumstances involving duress arise when D becomes involved in the killing of an innocent person but only because D is personally threatened with death or with a life-threatening injury and the only way to avoid the threat is to perpetrate or participate in the killing.^[36] At present, however, duress is no defence to murder at all. Indeed, the current guidelines do not even mention it as a mitigating factor in sentencing for murder.^[37] This is not right. Little, if any, blame may attach to someone's decision to take part in a killing under duress. Take the following example:

A taxi driver has his vehicle commandeered by a gunman who holds a gun to the driver's head and tells him to drive to a place where the gunman says he may shoot someone. The taxi driver does as the gunman demands and the gunman goes on to shoot and kill someone.

1.55      Under the existing law the taxi driver is complicit in the killing and stands to be convicted of murder. Just like the gunman, he will receive the mandatory life sentence. In certain circumstances, a claim of duress should provide a defence to first degree murder in this and in other cases. We explain the circumstances in which duress should be a defence in Part 6.

1.56      Almost all our consultees were agreed that duress should be a defence to murder in some manner or form. We had provisionally proposed that duress should have the effect of reducing first degree murder to second degree murder. That proposal had majority support amongst consultees. However, we have concluded that, whatever the merits of that solution considered in isolation, it would lead to undue complexity or even anomaly in other areas of the law of homicide. There could be no principled and straightforward way of applying or dis-applying the defence of duress to second degree murder, attempted murder and manslaughter. For this reason it is not a solution that we now recommend. Instead, we are recommending that duress should be a full defence to first degree murder, second degree murder and attempted murder.

1.57      However, stringent conditions will have to be satisfied before the defence can succeed. The burden of proof will be on D to show that he or she was threatened with death or life-threatening harm,^[38] had no realistic opportunity to seek the police's protection and had not already unjustifiably exposed him or herself to the risk of being threatened. Further, the jury must judge that a person of ordinary courage might have responded as D did by committing or taking part in the commission of the crime.

Sentencing and sentencing reform

1.58      All persons convicted of murder must be sentenced to imprisonment for life. A life sentence commonly consists of three periods or phases.

(1) The first phase is the 'minimum term': this is the period that the offender must spend in prison before he or she is eligible for release. Its length is meant to reflect the seriousness of the offence and hence the demands of retribution and deterrence. The length of the minimum term is set by the trial judge. In deciding upon the length of that term, the judge must refer to guidelines that Parliament has provided in the Criminal Justice Act 2003.

(2) When that minimum term has expired, the second phase begins (assuming that the offender is not released immediately). The second phase is the period in custody during which the offender may be considered for parole: the decision whether or not to release an offender is made by reference to considerations of public protection. The Parole Board will not release the offender if he or she still poses a danger to the public. Offenders may, therefore, spend considerably longer in prison than the minimum term recommended by the judge at trial.

(3) Finally, there is the third phase: being 'out on licence'. When the offender is deemed safe to release, he or she is released on licence until the end of his or her life. That means that he or she must comply with the conditions of the licence – conditions that may involve, for example, staying away from certain places – or risk recall to prison.

1.59      The guidelines that judges follow in deciding what the first phase – the minimum term – should be are relatively new and very important. The guidelines are found in schedule 21 to the Criminal Justice Act 2003.

1.60      Under these guidelines, the length of the minimum term will depend upon the gravity of the murder. There are three starting points for adult offenders:

(1) a whole life term for exceptionally serious cases, such as premeditated killings of two or more people, sexual or sadistic child murders or politically motivated murders;

(2) 30 years' minimum for serious cases such as murders of police or prison officers, murders involving firearms, sexual or sadistic killings or killings aggravated by racial or sexual orientation; and

(3) 15 years' minimum for murders not falling within the two higher categories.^[39]

1.61      The issue of sentencing is not within our remit. However, we do not intend our proposals for reform to undermine or weaken the existing guidelines. For example, the use of a firearm to kill a police officer on duty may be regarded as an aggravating factor so serious that the same sentence (life imprisonment) and the same minimum amount of time in custody (thirty years) may be appropriate whether the offender is convicted of first or second degree murder. We expect that guidelines for sentencing in second degree murder cases will be set down by Parliament as a part of any reforms to the law. Some of the issues are addressed in more detail in Appendix A.

1.62      What is clear to us is that the setting down of different recommended minimum terms for custody, depending on the nature or circumstances of a murder, makes it vital to rationalise and clarify the law of murder itself.

AN OVERVIEW OF THE STRUCTURE THAT WE ARE RECOMMENDING

1.63      We recommend that there should be a new Homicide Act for England and Wales. The new Act should replace the Homicide Act 1957. The new Act should, for the first time, provide clear and comprehensive definitions of the homicide offences and the partial defences. In addition, the new Act should extend the full defence of duress to the offences of first degree and second degree murder and attempted murder, and improve the procedure for dealing with infanticide cases.

1.64      In structuring the general homicide offences we have been guided by a key principle: the 'ladder' principle. Individual offences of homicide should exist within a graduated system or hierarchy of offences. This system or hierarchy should reflect the offence's degree of seriousness, without too much overlap between individual offences. The main reason for adopting the 'ladder' principle is as Lord Bingham has recently put it (in a slightly different context):

The interests of justice are not served if a defendant who has committed a lesser offence is either convicted of a greater offence, exposing him to greater punishment than his crime deserves, or acquitted altogether, enabling him to escape the measure of punishment which his crime deserves. The objective must be that defendants are neither over-convicted nor under-convicted… .^[40]

1.65      The 'ladder' principle also applies to sentencing. The mandatory life sentence should be confined to the most serious kinds of killing. A discretionary life sentence should be available for less serious (but still highly blameworthy) killings.

1.66      Partial defences currently only affect the verdict of murder. This is because a verdict of murder carries a mandatory sentence. That sentence is not appropriate where there are exceptional mitigating circumstances of the kind involved in the partial defences. These mitigating circumstances necessitate a greater degree of judicial discretion in sentencing. The law creates this discretion by means of the partial defences which reduce what would otherwise be a verdict of murder, which carries a mandatory sentence, to manslaughter, which does not. Therefore, our recommended scheme does not extend the application of the partial defences to second degree murder or manslaughter. These offences would permit the trial judge discretion in sentencing and they therefore lack the primary justification for having partial defences.^[41]

The structure of offences

1.67      We believe that the following structure would make the law of homicide more coherent and comprehensible, whilst respecting the principles just set out above:

(1) First degree murder (mandatory life penalty)

(a) Killing intentionally.

(b) Killing where there was an intention to do serious injury, coupled with an awareness of a serious risk of causing death.

(2) Second degree murder (discretionary life maximum penalty)

(a) Killing where the offender intended to do serious injury.

(b) Killing where the offender intended to cause some injury or a fear or risk of injury, and was aware of a serious risk of causing death.

(c) Killing in which there is a partial defence to what would otherwise be first degree murder.

(3) Manslaughter (discretionary life maximum penalty)

(a) Killing through gross negligence as to a risk of causing death.

(b) Killing through a criminal act:

(i) intended to cause injury; or

(ii) where there was an awareness that the act involved a serious risk of causing injury.

(c) Participating in a joint criminal venture in the course of which another participant commits first or second degree murder, in circumstances where it should have been obvious that first or second degree murder might be committed by another participant.

Partial Defences reducing first degree murder to second degree murder

1.68      The following partial defences would reduce first degree murder to second degree murder:

(1) provocation (gross provocation or fear of serious violence);

(2) diminished responsibility;

(3) participation in a suicide pact.

Other specific homicide offences

1.69      There will remain a number of specific homicide offences, such as infanticide, assisting suicide and causing death by dangerous driving.

CONCLUSION

1.70      The Criminal Justice Act 2003 ("the 2003 Act") one of the most important pieces of legislation in the history of criminal justice reform, brought in a new sentencing regime for murder.^[42] However, the radical reforms effected by the 2003 Act stand upon shaky foundations because the offence of murder, and the partial defences to it, do not have defensible definitions or a rational structure. Unfortunately, although twentieth century legislation on murder brought many valuable reforms, the definitions of murder and the partial defences remain misleading, out-of-date, unfit for purpose, or all of these. Quite simply, they are not up to the task of providing the kind of robust legal support upon which the viability of the 2003 Act depends.

1.71      As we indicated in the CP,^[43] it is worth noting that many of the problems we discuss were identified by a Parliamentary Select Committee as long ago as 1874. The Committee said:

If there is any case in which the law should speak plainly, without sophism or evasion, it is where life is at stake; and it is on this very occasion that the law is most evasive and most sophistical.^[44]

1.72      A few years later, former Prime Minister W. E. Gladstone indicated his willingness to promote the enactment of a Homicide Act to rationalise the law, based on what the Criminal Law Commissioners had at that time proposed. However, nothing was done.^[45] That led one criminal lawyer to remark, at the beginning of the twentieth century, that a belief that a criminal code would be passed in the House of Commons was as naïve as "expecting to find milk in a male tiger".^[46]

1.73      We hope that, at the beginning of the twenty-first century, an expectation that the law of homicide should be rationalised by statute is not quite that naïve. In the Parts that follow, we will explain in detail why we have decided to recommend a structure for the law of homicide in this form.

Note 1    These defences are defences to many crimes other than homicide. They, therefore, need to be looked at as part of a review of the general law rather than specifically in a homicide context.    [Back]

Note 2    Paras 1.54 to 1.57 below and Part 6.    [Back]

Note 3    A New Homicide Act for England and Wales? (2005) Consultation Paper No 177.    [Back]

Note 4    This second stage will involve a public consultation on the broader areas of public policy which any review of the law of homicide needs to address.     [Back]

Note 5    Some of the historical background on attempts to reform the law of murder is given in Part 1 of the CP.    [Back]

Note 6    Criminal Justice Act 2003, s 269 and sch 21.    [Back]

Note 7    It is sometimes argued that manslaughter by recklessness and by gross negligence form one single category of manslaughter with two alternative fault requirements. We will be recommending that any reform of the law should adopt this approach: see Part 3.    [Back]

Note 8    Eg, on murder see Woollin [1999] 1 AC 82 and on manslaughter see Adomako [1995] 1 AC 171 and Smith (Morgan) [2001] 1 AC 146.    [Back]

Note 9    In a different context, it has been held that a jury is entitled to find that an intentional punch breaking someone’s nose involves the intentional infliction of ‘serious’ harm: Saunders [1985] Criminal Law Review 230. In a case of non-fatal injury, that conclusion may be acceptable.    [Back]

Note 10    Paras 1.119 to 1.123.    [Back]

Note 11    Vickers [1957] 2 QB 664.    [Back]

Note 12    Criminal Justice Act 2003, s 269 and sch 21.    [Back]

Note 13    We included Professor Mitchell’s research in Appendix A to the CP.    [Back]

Note 14    Above, paras A.7 to A.8.    [Back]

Note 15    Above, para A.7.    [Back]

Note 16    We are recommending use of the term ‘injury’ in place of ‘harm’: see Part 2.    [Back]

Note 17    The need to narrow the crime of involuntary manslaughter has already been accepted by Government: Home Office, Reforming the Law on Involuntary Manslaughter: The Government’s Proposals (2000).    [Back]

Note 18    Sir James Stephen, Digest of the Criminal Law (1877), Art 223(b).    [Back]

Note 19    See the CP, paras 1.119 to 1.121.    [Back]

Note 20    Desmond, The Times, April 28, 1868. So far as other countries are concerned, Lord Cockburn CJ was guilty of a little exaggeration. In France an intention to kill was required at that time (and it still is) if the offender was to be guilty of the most serious homicide offence: Sir James Stephen, History of the Criminal Law of England (1883), vol III, 90 to 91.    [Back]

Note 21    Vickers [1957] 2 QB 664, 670, by Lord Goddard CJ.    [Back]

Note 22    Hyam [1975] AC 55.     [Back]

Note 23    It must be said that the exact basis for the decision of the majority in Hyam has always been unclear. There is some indication that, whether or not the state of mind in (c) in para 1.26 above was a sufficient basis for finding an intention to kill or cause serious harm, it constituted malice aforethought, as Stephen had indicated a century before: n 18 above.    [Back]

Note 24    Moloney [1985] AC 905, 920, by Lord Bridge. The Homicide Act 1957, s 1 abolishes so-called ‘constructive malice’ (felony-murder), but assumes that ‘malice’ is still the term for the fault element in murder.    [Back]

Note 25    Eg Hancock and Shankland [1986] AC 455; Nedrick [1986] 1 WLR 1025; Woollin [1999] 1 AC 82.    [Back]

Note 26    Unless he or she thought that someone was virtually certain to be killed in the explosion, a (foreseen) probability of death being no longer sufficient in itself. Even if a death in the explosion was foreseen as virtually certain to occur, however, the jury would still only be asked whether it was willing to find an intention to kill or cause serious injury. The jury would not be required to find that D intended to kill or cause serious injury.    [Back]

Note 27    Para 2.98 below.    [Back]

Note 28    Involuntary Manslaughter (1996) Law Com No 237; Home Office, Reforming the Law on Involuntary Manslaughter: The Government’s Proposals (2000).    [Back]

Note 29    Powell and Daniels, English [1999] 1 AC 1.    [Back]

Note 30    Para 5.83.    [Back]

Note 31    Infanticide Act 1938.    [Back]

Note 32    Partial Defences to Murder (2004) Law Com No 290, para 1.13.    [Back]

Note 33    Part 8, paras 8.19 to 8.35.    [Back]

Note 34    Kai-Whitewind [2005] EWCA Crim 1092, [2005] 2 Cr App R 31.    [Back]

Note 35    Part 5 below.    [Back]

Note 36    Alternatively, D may be threatened that members of his or her family will be killed or seriously harmed.     [Back]

Note 37    Criminal Justice Act 2003, s 269 and sch 21.    [Back]

Note 38    Or that he or she was threatened that a person for whom he or she reasonably felt responsible would be killed or seriously harmed.    [Back]

Note 39    For further analysis see, Andrew Ashworth, Sentencing and Criminal Justice (4th ed 2005) pp 116 to 118.    [Back]

Note 40    Coutts [2006] UKHL 39, [2006] 1 WLR 2154 at [12].    [Back]

Note 41    This argument is pursued further in Part 2 where we will explain why a secondary justification for partial defences – the labelling argument – is insufficiently weighty to justify permitting partial defences to reduce homicide offences below first degree murder to offences further down the ladder.    [Back]

Note 42    Section 269 and sch 21 - see paras 1.58 to 1.62 above.    [Back]

Note 43    Paras 1.129 to 1.131.    [Back]

Note 44    Special Report from the Committee on the Homicide Law Amendment Bill (1874) 314.    [Back]

Note 45    K J M Smith, Lawyers, Legislators and Theorists: Developments in English Criminal Jurisprudence 1800-1957 (1998) p 149.    [Back]

Note 46    Above.    [Back]