This article contends that there are compelling reasons for reconceptualising the contours of self-defence, and for the introduction of a bespoke partial defence complemented by jury directions and the admissibility of social framework evidence to assist vulnerable offenders who kill their abusers in a desperate attempt to protect themselves. The New Zealand Law Commission in 2016 recently recommended, inter alia, that self-defence be re-categorised and broadened to allow victims of family violence who kill to potentially claim a defence in the absence of an imminent threat of harm, standardised on an “all or nothing” perspective. In truth, a far wider contextualisation needs to apply, beyond the limited and constrained terms of reference before the Commission. The contours of self-defence applicability ought to extend to extra-familial vulnerable offenders, encompassing individuals subjected to human trafficking and/ or modern slavery, those trapped by ostensible gang membership, and those experiencing third-party abuse who respond with lethal force. It is our assertion, after a comparative review of the theoretical and doctrinal precepts of a number of alternative legal systems, that the full and partial defence schema should be more nuanced. Extant laws fail to appropriately recognise the need for a de novo partial defence template and reflective individuated culpability thresholds.
This article advances an alternative approach to the New Zealand Law Commission’s (the Commission) recent recommendations for victims of family violence who commit homicide. New Zealand criminal law is dangerously “out of step internationally in how it responds to victims of family violence who kill”,2 and the reform recommendations advanced by the Commission are designed to combat the issue.3 The report recommends, inter alia, that self-defence be modified to ensure that victims of family violence who kill are eligible to claim the defence in the absence of an “imminent” threat.4 This change is designed to be complemented by reforms to the Evidence Act 2006 and Sentencing Act 2002 to ensure that a broad range of family violence evidence is admissible during trial in support of the defence, in addition to constituting relevant mitigation at the sentencing stage. Other measures include potential changes to Prosecutorial Guidelines and the “three strikes” law5 in order to provide a more holistic approach to reform. Unfortunately the promulgated reforms are flawed in failing to consider extra-familial vulnerable offenders who kill in self-defence, and in advocating that revised self-defence provisions should operate on an “all-or-nothing” basis, whereby the defence either succeeds or it fails.
The proposals advanced herein draw upon experience of self-defence, duress and partial defence provisions across New Zealand, Victoria (Australia), Canada, the United States, and England and Wales. Importantly the recommendations canvassed reject the New Zealand Law Commission’s argument that a lower threshold self-defence test should apply to victims of abuse who kill their abuser only if a familial link is established. The narrow focus on this discrete category of vulnerable offender under the New Zealand Law Commission’s terms of reference meant the Commission was unable to “consider the law in respect of other defendants who may be less blameworthy in a comparative sense”.7 Real and hypothetical scenarios are advanced to demonstrate the extent to which individuals subjected to human trafficking and/or modern-day slavery, those trapped by ostensible gang membership and those experiencing third-party abuse may similarly respond to that abuse with lethal force. In contrast to the Commission’s recommendations, compelling arguments are advanced for general as opposed to specific reforms to self-defence in the context of vulnerable offenders who kill an abuser.
The justificatory basis for self-defence is revisited, comparing affirmative self- defence in Victoria which is available in the context of family violence where the threat is not imminent and the force used is excessive, and the position in several US states where a partial imperfect self-defence provision operates in this context. This comparative analysis reveals that the full and partial defence schema should be more nuanced than the New Zealand Law Commission’s report suggests. Even if the Commission’s broader self-defence provisions are accepted by the Ministry of Justice, they will not assist extra-familial victims of abuse and for intra-familial victims, self-defence will not always be available on the facts. That is not to say a defence should be available axiomatically. A partial defence ought to be an option, and an appropriate, bespoke self-preservation defence is advanced herein.
The partial defence is designed to sit directly beneath self-defence. It would operate to reduce a murder conviction to manslaughter where the defendant kills in response to a fear of serious abuse from the victim against the defendant or another identified individual, but unlike affirmative self-defence the lack of an imminent threat and the use of excessive force would not necessarily negate the defence. The absence of imminence and proportionality requirements are justified on the basis that self-preservation is a partial rather than a complete defence. In cases where the defendant claims to have held a particular belief as regards the circumstances, the reasonableness or otherwise of that belief is relevant to the question whether D genuinely held it; there must be an intelligible basis for the belief; if it is determined that D did genuinely hold it, and there was an intelligible basis for doing so, D is entitled to rely on it for the purposes of the partial defence, whether or not it was mistaken, or (if it was mistaken) the mistake was a reasonable one to have made. Importantly, the defence does not automatically apply where self-defence fails on grounds that the threat was not imminent or the force was excessive, otherwise the defence would be overly broad in ambit and subject to similar criticisms that were levelled at defensive homicide in Victoria.8 Appropriate threshold filter mechanisms operate to prevent the defence from being available in unmeritorious cases. The defence does not apply where the defendant intentionally incited serious violence or acted in a considered desire for revenge, and is qualified by a normal person test which requires that a person of the defendant’s age with a normal degree of tolerance and self-restraint might have reacted in the same or a similar way in the circumstances. Psychiatric conditions may be relevant to the normal person test in limited circumstances where the condition is especially probative, but evidence of voluntary intoxication remains irrelevant. In all cases, the trial judge may decline to leave the defence to the jury on the basis that no jury properly directed could reasonably conclude that the defence might apply.