Sydney Law Review
The “Australian Position” Concerning Criminal Complicity: Principle, Policy or Politics?
This article examines the differences that have recently emerged between the United Kingdom Supreme Court and the High Court of Australia concerning the law of criminal complicity. It contends that, if we are accurately to analyse the decisions of those Courts in, respectively, R v Jogee  AC 387 and Miller v The Queen  HCA 30; (2016) 259 CLR 380, we must acknowledge the extra-legal considerations that influenced these highly-respected tribunals. To criticise what Justice Keane has called ‘the Australian position’ is to reveal a partial truth. Certainly, that position is questionable. Indeed, here it is argued that the ‘change of normative position’ justification for the extended joint criminal enterprise doctrine does not withstand critical scrutiny. Nevertheless, the divergent results in Jogee and Miller probably owe more to public opinion, politics and widely-held judicial views about when an ultimate court of appeal is entitled to reverse an established common law rule, than they do to any fundamental differences between London and Canberra concerning principle and/or policy.
On 18 February 2016, a joint sitting of the United Kingdom (‘UK’) Supreme Court and the Privy Council in R v Jogee unanimously held that the ‘parasitic accessory liability’ doctrine (‘PAL’) no longer forms part of English law. Lords Hughes and Toulson, writing for the Court, concluded that
there is no doubt that the Privy Council laid down a new principle in Chan Wing-Siu when it held that if two people set out to commit an offence (crime A), and in the course of it one of them commits another offence (crime B), the second person is guilty as an accessory to crime B if he foresaw it as a possibility, but did not necessarily intend it.
This principle, their Lordships continued, was unjustified: it made the common law more severe — on the basis of ‘an incomplete, and in some respects erroneous, reading of the previous case law, coupled with generalised and questionable policy arguments’.
In light of these conclusions, it was hardly surprising that the High Court of Australia was soon asked to reconsider what is known in New South Wales (‘NSW’) and South Australia (‘SA’) as the doctrine of extended joint criminal enterprise (‘EJCE’). But, contrary to some commentators’ hopes, on 24 August 2016, their Honours, with Gageler J dissenting, held in Miller v The Queen that ‘the principle of extended joint criminal enterprise liability stated in McAuliffe should remain part of the common law of Australia’.
Two questions immediately come to mind. What is the explanation for the differences that have emerged between these highly-respected tribunals? And, more starkly, which Court was right?
Commentators so far have focused more on the second of these questions than the first. The suggestion appears to be that, once it is determined which Court was correct in principle, the reasons for the different decisions are either obvious (the other court simply was wrong) or unimportant. But, in this article, I argue that it is important to understand why divergent outcomes were reached — and that, if we are to do so, it is not enough merely to contend that the Court with which we disagree failed properly to understand the relevant precedents, or deployed suspect reasoning concerning moral culpability. That is, while I agree with those commentators who have criticised the ‘Australian position’, I also believe that, if we are to establish the truth about Jogee and Miller, we must look beyond the respective Courts’ actual decisions and acknowledge the extra-legal factors that influenced both judgments. In short, the answer to the puzzle of how two such distinguished judicial panels could reach such diametrically opposed conclusions seems to lie more in public opinion and politics than it does in law.
The article is structured as follows. In Part II, I discuss the Australian common law position concerning accessorial liability, the ‘plain vanilla’ doctrine of joint criminal enterprise (‘JCE’), and EJCE. One of my aims is to show that, while, properly viewed, accessorial liability and JCE have the same doctrinal basis as one another, the same cannot accurately be said of accessorial liability and EJCE.
If EJCE is to be justified, this might be on the basis of the ‘change of normative position’ rationale advanced by Simester and adopted by the High Court plurality both in Clayton v The Queen and in Miller. Certainly, it cannot be justified in the manner suggested by Keane J in Miller; namely, that with EJCE, the other participants have authorised the actual perpetrator to commit the further crime. The very example that his Honour uses to support this analysis shows it to be misconceived: the defendant in Gillard v The Queen did not give authority to Preston to kill; rather, he was Preston’s ‘thick and simple’ ‘errand boy.’
In Part III of the article, however, I contend that Simester’s change of normative position justification is fallacious. Contrary to his claim, the person who continues in a criminal enterprise despite his/her foresight that during that enterprise a cooffender might commit murder, is not sufficiently culpable to warrant being convicted of that offence if it results. In so arguing, I do not contend that the EJCE murderer’s culpability is glaringly lower than that of all other offenders convicted of murder in SA and NSW. Such reasoning ignores the constructive murder rule’s continued existence in both of those states and, partly for that reason, is no more persuasive than the similarly commonly deployed argument that ‘[t]he common law has developed ordinarily to insist that justice requires that a primary party become criminally liable only by acting with intention.’ Rather, my position is that a person’s agreement with another/others to commit a foundational crime is not as relevant as Simester thinks it is to his/her culpability regarding a death that a co-offender causes once the enterprise has been embarked upon. Similar reasoning to Simester’s has been used to justify the constructive murder rule, too. However, as Krebs notes, it is ‘descriptive in nature, rather than explanatory’: it is really no more than an after-the-fact rationalisation of the practice of enhancing the punishment of the dangerous with the aim, real or pretended, of achieving general deterrence.
In Part IV of the article, I argue that, even though the Jogee approach is more normatively desirable than the Miller one, it is, to an extent, understandable why the plurality in the latter case acted as it did. That EJCE was not as ‘highly controversial’ in SA and NSW as PAL had become in the UK was one factor obstructing the type of change that some commentators were urging upon the Court.
Accordingly, we ought to be both more and less critical of courts’ decisions in controversial cases such as this. We should be more critical in the sense that, when assessing judicial reasoning, we should consider not merely the reasoning itself, but rather all of the factors that influenced the court to decide the case as it did (including the social context in which the relevant case was decided). We should acknowledge, that is, that the courts are political actors. We ought to be less critical in the limited sense that we should acknowledge the pressure that courts face to avoid creating any perception that they are political. The courts are necessarily mindful of their reputation. It is easy enough for judges to use creative reasoning where this produces an outcome that is consistent with public opinion; but when judicial decisions try to lead society, claims of judicial activism tend to follow.
This is not to say that their Honours in Miller should have decided that case as they did. Ultimately, it is hard to justify the Court’s retention of this unjust common law doctrine. Rather, it is to recognise that the High Court was in a difficult position. And it is to suggest that we might have a better chance of influencing future decisions in cases such as Miller if we address all of the factors that judges take into account when deciding them.
In Part V of the article, I conclude that, for as long as EJCE (and the constructive murder rule) remains part of the law in NSW and SA, the law will remain unfair and unprincipled. But it is likely to remain for some time yet. Parliament has shown no great appetite for reforming the law in this area, and the High Court has now twice upheld the McAuliffe principle. Unless public opinion becomes as strongly hostile to EJCE as UK opinion apparently was to PAL in the years immediately before Jogee, it is difficult to imagine either Parliament or the courts changing their approach.
The Australian common law position is that there are three ways in which a person can be complicit in another person’s criminal conduct.
First, a person can be convicted as an accessory to another person’s crime. It is no longer necessary to maintain the old common law distinction between accessories before the fact (those who aided, abetted, counselled or procured the principal to commit an offence but were absent from the scene of the crime) and principals in the second degree (those who provided such aid etc and were present when the crime was committed). Whether present or absent from the scene, the person who intentionally assists or encourages (or procures) an offence — which he/she can only do if, at the time of giving the assistance or encouragement, he/she knows of the principal’s intention to perform the relevant conduct with the requisite mens rea — can be convicted of it. Whether the person is in fact convicted, however, depends on whether the Crown is able first to prove the principal’s guilt. In other words, accessorial liability is derivative, not primary.
Second, a person can be convicted on the basis of the JCE doctrine of an offence that another person has actually perpetrated. Such liability attaches when the Crown proves that: (i) the accused and another or others expressly or tacitly agreed to commit a crime; (ii) one or more of the parties to the agreement, in accordance with that agreement, performed all of the acts necessary for the commission of the crime; (iii) the accused participated in the joint enterprise; and (iv) at the time of entering the agreement, the accused had the requisite mental state for the relevant offence (where this is so, the state of mind ‘continues unless the accused withdraws from the agreement’).
In Osland, McHugh J seemed additionally to require the Crown to prove that the accused was present when the actual perpetrator performed the relevant act(s). But it does not appear that there is such a requirement. In Likiardopoulos, the High Court held, at the very least, that it is unnecessary for a JCE passive participant to have been present throughout the whole of a fatal assault. There is, moreover, a significant amount of intermediate appellate court authority for the proposition that such a participant need not have been present at all while the conduct was being performed.
Further, since Osland it has been clear that JCE liability is not derivative, but primary. That is, to secure the conviction of an offender on this basis, the Crown need not first prove the actual perpetrator’s guilt of the crime that the parties have agreed to commit. Rather, the actual perpetrator’s acts are attributed to the other parties to the agreement. Provided that such parties individually had the requisite mens rea, and can successfully raise no defence, they will be guilty of the relevant offence.
Third, as has already been noted, a person can be convicted of an offence that: (i) he/she did not actually perpetrate; and (ii) was different from that which was the object of a JCE to which he/she was a party. In Johns v The Queen, the plurality upheld Street CJ’s statement in the Court below that
an accessory before the fact bears, as does a principal in the second degree, a criminal liability for an act which was within the contemplation of both himself and the principal in the first degree as an act which might be done in the course of carrying out the primary criminal intention — an act contemplated as a possible incident of the originally planned particular venture.
That remains the law — although Street CJ’s references to accessorial liability should now be ignored. In both McAuliffe and Miller, the High Court substituted for such language the language of JCE: a person will be liable for a crime other than that which he/she and his/her co-offenders agreed to if the participants jointly contemplated that crime as a possible incident of the execution of their agreement. But, whatever is this principle’s precise content — and the better view is that, despite referring to joint contemplation, Johns in fact required the Crown to prove that the passive participant foresaw the further offence and assented to its commission if the occasion arose — it is seemingly never used. This is because of the Court’s acceptance in McAuliffe that the party who does not assent to the further crime’s commission will still be liable for it if he/she individually foresaw that another participant in the enterprise might commit it — and, despite that foresight, continued to participate in the venture (EJCE liability). In cases involving the commission of a crime other than that which was the object of a JCE, it makes no sense for the Crown to rely on the Johns joint foresight/assent principle instead of the less stringent McAuliffe individual foresight standard.
1 Accessorial Liability and JCE
In England, it has long been debated whether accessorial liability and EJCE have a different doctrinal basis. I believe that they must; but before I explain this view, it is necessary to compare accessorial liability with ‘plain vanilla’ JCE. Properly viewed, are accessorial liability and JCE separate forms of liability, and do they have a different rationale from one another?
Certainly, there are differences between accessorial liability and JCE. First, although the Crown will usually be able to proceed against an offender, alternatively, as an accessory and on the basis of JCE, there are some cases where, despite there being no agreement between the offenders — and therefore no JCE liability — accessorial liability will be established. An example is Attorney-General’s Reference, where the English Court of Appeal held that a defendant who surreptitiously laced another person’s drinks with spirits could be convicted as an accessory to that person’s resulting offence of driving with a blood alcohol concentration above the prescribed limit. ‘Aiding and abetting almost inevitably involves a situation in which the secondary party and the main offender are together at some stage discussing the plans which they may be making in respect of the alleged offence,’ Lord Widgery CJ observed, but it is unnecessary that there be such a ‘meeting of minds’.
Are there cases where JCE liability arises, but accessorial liability does not? Odgers appears to answer this question in the negative:
Jogee ... demonstrated that joint enterprise liability may be seen as simply a subset of [accessorial liability] ... After all, where A has agreed with B that crime X should be committed, A has plainly intended to encourage B to commit that crime.
Simester, however, has recently offered a different view. While conceding that ‘the practical gap between “plain vanilla” joint enterprises and abetment is small,’ he points to the case where, rather than there being only offenders A and B, a person has joined a ‘large group whose plans are already formed and where [the actual perpetrator] ... is unaware of ... [that person’s] actions’. Here, the person might have given the actual perpetrator no assistance. Certainly, there has been no encouragement. But the problem with this example seems to be that the person has not agreed with the actual perpetrator to commit the crime. As in Attorney-General’s Reference, the actual perpetrator was oblivious to the person’s conduct. Moreover, to anticipate the discussion in the next paragraph, that the person has provided no assistance or encouragement to the actual perpetrator seems to mean that he/she has not participated in any JCE with him/her.
Second, it has been said that, whereas with accessorial liability the Crown must prove assistance, encouragement or procuring, in the case of JCE it must prove that the accused ‘participated in the criminal act of another’. In reality, however, this is not a way in which accessorial and JCE liability differ from one another. It is true that, as noted above, the Crown must prove that the accused participated in the JCE. But J C Smith was surely right to observe that the only way in which a person can participate in another person’s crime is by assisting or encouraging him/her. Indeed, consistently with this view, the NSW Court of Criminal Appeal (‘NSWCCA’) in Tangye v The Queen insisted that:
A person participates in [a] joint criminal enterprise either by committing the agreed crime itself or simply by being present ... when the crime is committed, and (with knowledge that the crime is to be or is being committed), by intentionally assisting or encouraging another participant in the joint criminal enterprise to commit that crime.
Of course, this passage suggests that another way in which JCE and accessorial liability can be differentiated from one another is that presence is required for the former. However, as we have seen, this is seemingly not so. Nevertheless, as also noted above, there is one more thing distinguishing these two types of liability: whereas the accessory’s liability is derivative, JCE offenders are principals in the first degree.
Do the differences between JCE and accessorial liability demonstrate that they have a different doctrinal basis from one another and are supported by different rationales? I do not think so.
To deal with the second difference first, conspicuously absent from McHugh J’s analysis in Osland is any principled justification for the primary nature of JCE. In other words, it is not as though anything in the nature of the doctrine compelled the High Court to hold that the actual perpetrator’s acts are attributed to those who agreed with him/her to commit the relevant offence. Instead, the law’s fictitious insistence that the actual perpetrator’s acts count as those of all members of the relevant enterprise appears to be based purely on pragmatic considerations. If the liability of passive JCE participants depended on that of the actual perpetrator, the door would be opened to ‘scandalously unmeritorious acquittals’. To use one example, the offender who agreed with a mentally ill person to commit a crime could not be held liable for that crime if the co-offender actually perpetrated it at a time when he/she was within the M’Naghten rules. It is better for the law to state that:
Once the parties have agreed to do the acts which constitute the actus reus of the offence ... the criminal liability of each ... depend[s] upon the existence or non-existence of mens rea or upon their having a lawful justification for the acts, not upon the criminal liability of the actual perpetrator.
The same concern underlies Lord Hobhouse’s claim that JCE is a ‘different principle’ from accessorial liability. Writing extra-judicially, his Lordship said:
[I]f the state of mind of D is criminally more serious than that of A, the question arises whether D should be convicted of the crime corresponding to his own state of mind or should only be convicted of the lesser crime corresponding to A’s state of mind. This forces the law to choose whether to recognize the agency principle or to ... convict only upon the complicity principle as an accessory.
Lord Hobhouse denies that there is any ‘artificiality’ in the law’s attribution of another’s acts to the person who, by agreeing that such acts be done, explicitly or implicitly authorised their performance. But, as I have just argued, to treat a person as though he/she performed conduct that he/she has in fact not performed is obviously both artificial and fictitious. What has really happened is that this offender has intentionally encouraged or assisted the actual perpetrator to commit a crime. In short, it might be convenient to hold that JCE liability is primary; but this should not obscure the fact that we hold the JCE offender liable for the same reasons as we do the accessory.
Unfortunately, in IL only Bell and Nettle JJ were willing squarely to acknowledge that, as with accessorial liability, ‘[t]he object of the [JCE] doctrine is to fix with complicity for the crime committed by the perpetrator those persons who encouraged, aided or assisted him.’
In IL, the appellant participated in a JCE to manufacture a large commercial quantity of methylamphetamine. During this enterprise, she and her co-offender boiled mixtures of raw methylamphetamine and an inflammable solvent, acetone, for the purpose of extracting impurities from the drug. The co-offender was killed as a result of a fire in the suburban bathroom in which this evaporation process was being conducted. The Crown identified the act causing death as the lighting of a ring burner in the bathroom. But it could not prove that the appellant performed that act. Nevertheless, it charged IL not only with the drugs offence, but also with murder or, alternatively, manslaughter.
The Crown’s murder case was based on JCE and the constructive murder rule. It followed from Osland, it contended, that the actual perpetrator’s act of lighting the ring burner was attributed to IL. Accordingly, so it was said, her act caused the death of another person (her co-offender) and she could be convicted of murder because such an act was performed during her or an accomplice’s commission of an offence punishable by 25 years’ imprisonment (the drugs offence).
The NSWCCA unanimously accepted that a jury would be entitled to convict IL of murder, but a High Court majority disagreed. This was not because it was persuaded by the appellant’s arguments, which essentially amounted to an attack on the constructive murder rule (it was argued that, for IL to be guilty of murder, the Crown had to prove that she foresaw the possibility that death would result from the drug manufacturing enterprise). Rather, during an extraordinary hearing, four of their Honours surprised both parties by advancing their own arguments concerning why the prosecution could not succeed.
It is submitted that the argument that ultimately appealed to Bell and Nettle JJ is the correct one. As noted above, their Honours emphasised JCE’s object. That object is not to facilitate prosecutions such as that attempted in IL where, because the actual perpetrator killed only himself, there was nothing even resembling a murder in which IL was complicit. It is to allow for the conviction of those who intentionally assist or encourage another person to commit a crime. Certainly, their Honours implied, there are unusual cases such as Osland where the actual perpetrator has committed no crime, but where JCE liability nevertheless arises. But in those cases the perpetrator has performed the actus reus of an offence. As I have argued, because the passive participant (a) has agreed with him/her to do so; (b) possesses the requisite mens rea; and (c) has no defence available to him/her, it would be unjust if he/she were acquitted.
The other majority Justices also appear to have been concerned about the prospect of IL being convicted of murder even though she had merely participated in an act of self-killing. But the reasoning that their Honours used to avoid such an outcome is dubious and contrived. Chief Justice Kiefel, Keane and Edelman JJ considered that Osland required them to hold that the act of lighting the ring burner was, for legal purposes, IL’s act. But their Honours held that ‘[t]he offences of murder and manslaughter in s 18 of the Crimes Act 1900 (NSW) require that one person kill another person.’ Because IL was a self-killer’s accomplice, their Honours continued, s 18 was not engaged. As Gageler and Gordon JJ pointed out in their respective dissenting judgments, one problem with this reasoning is that, if it is accepted that the act causing death is to be treated as IL’s, there was a killing of another. IL’s act had caused the deceased’s death. Accordingly, it would surely have been better for Kiefel CJ, Keane and Edelman JJ to accept, as Bell and Nettle JJ essentially did, that no JCE liability for either murder or manslaughter could arise here, because — unlike Mrs Osland — IL had not participated in the actus reus of those offences. In so doing, their Honours would have explicitly acknowledged what they tend implicitly to recognise through their deployment of suspect reasoning to ensure that IL’s appeal succeeded. As with accessorial liability, JCE’s proper function is merely to allow for the conviction of those who intentionally assist or encourage another to (a) commit a crime, or (in unusual cases such as Osland)
(b) perform an actus reus.
We must now deal with the other difference between JCE and accessorial liability; namely, that agreement is not co-extensive with intentional encouragement, assistance or procuring. Certainly, a small number of offenders can be prosecuted as accessories, but not on the basis of JCE. But this does not mean that these two forms of liability have a different rationale or doctrinal basis from one another. To repeat: viewed properly, JCE and accessorial liability’s shared aim is to convict those who are complicit in another person’s crime (or, failing that, actus reus). Accessorial liability merely appears to achieve this objective more perfectly than does JCE.
2 Accessorial Liability and EJCE
By contrast, for so long as it remains part of the law, EJCE liability has a different doctrinal basis from accessorial liability. An indication that this is so is that both the actus reus and mens rea requirements differ significantly as between these two types of liability. Concerning the actus reus, the EJCE offender need not provide any assistance or encouragement to the actual perpetrator to commit the further offence. It is enough that he/she agreed with that person to commit the foundational crime. Concerning the mens rea, there is no need for the passive participant to intend that the further crime be committed if the occasion arises. As stated above, it is enough that he/she foresaw that offence’s commission as a possible incident of the joint criminal venture. So, while Smith and Hogan denies that there are doctrinal (or normative) differences between PAL (and thus EJCE) and accessorial principles, it also states: ‘Any doctrinal differences that now seem to separate joint enterprise from basic secondary liability are the result of judicial development of the former.’ I agree; except I would omit the word ‘Any’ from the above quotation and substitute ‘The’, and I would also delete the words ‘seem to’. I do not think that it can be denied that there are doctrinal differences between EJCE and accessorial liability. But that is a different thing entirely from supporting the jurisprudential developments that mandate this conclusion.
If a person can be convicted of a crime even though he/she provided no intentional assistance or encouragement to the actual perpetrator, it must be explained why this is. In Miller, Keane J provided an explanation that seems clearly wrong. I have referred above to Lord Hobhouse’s view that, once two or more persons agree to commit a crime, the actual perpetrator acts with the other participants’ authorisation and as their agent. It is for this reason that, according to this theory, his/her acts are the acts of all parties to the agreement. In Miller, Keane J applied a similar theory to EJCE:
Where parties commit to a joint criminal enterprise, each participant becomes, by reason of that commitment, both the principal and the agent of the other participants: for the purposes of that enterprise they are partners in crime. Each participant also necessarily authorises those acts which he or she foresees as possible incidents of carrying out the enterprise in which he or she has agreed, and continues, to participate.
Later in his judgment, his Honour returned to this theme:
[W]here a joint criminal enterprise is in the nature of a business activity on the part of the participants, as was the case in Gillard ... it is not sound policy to minimise the criminal responsibility of those who organise crime, and in so doing create the foreseen risk of an incidental crime, merely because they are able to engage others as their agents for that purpose.
A number of things can be said about this reasoning. First, as suggested above, the agency justification is problematic enough concerning JCE; it is singularly inapt to explain EJCE liability. Smith and Hogan’s criticisms of Lord Hobhouse’s reasoning are cogent. Certainly, the contract killer can, without any artificiality, be described as the agent of those who hire him/her. But it is inaccurate to describe in such terms the dominant individual who is driven to the scene of a murder by a weak-willed and unintelligent person, who then waits in the car while the dominant offender perpetrates the killing. Where the weak-willed offender did not even agree with the actual perpetrator that murder be committed, but merely foresaw that crime as a possible incident of the armed robbery that they had agreed to commit, it is surely an even greater abuse of language to say that, when the actual perpetrator did proceed to commit the further offence, he/she was acting as the passive participant’s agent.
Second, it follows from what I have just said that the appellant in Gillard certainly did not authorise the actual perpetrator to commit murder, or ‘engage [him] ... as [his] agent ... for that purpose’. It is strange that Keane J should use that case to exemplify his point.
Third, and relatedly, Keane J’s reasoning does not come to grips with the seemingly unanimous disfavour with which commentators have regarded Sir Robin Cooke’s well-known contention in Chan Wing-Siu that the
principle whereby a secondary party is criminally liable for acts by the primary offender of a type which the former foresees but does not necessarily intend ... turns on contemplation or, putting the same idea in other words, authorisation, which may be express but is more usually implied.
As J C Smith said as long ago as 1990, ‘“contemplation” is not the same thing as “authorisation”. One may contemplate that something will be done by another without authorising him to do it.’ Justice Keane’s denial that Sir Robin ‘equate[d] contemplation with authorisation’ is baffling, as is his Honour’s insistence that his Lordship was really arguing that ‘participation in the commission of a crime, with foresight of the risk of the incidental crime, establishes authorisation of the incidental crime’. Is Keane J saying that the law, by a fiction, deems the EJCE offender to have authorised what in fact he/she has not authorised? Perhaps so, because his Honour proceeds to hold that such a person can be held to have intended the killing that he/she has foreseen; and such an approach is surely entirely fictitious.
If the actual perpetrator cannot properly be regarded as the EJCE offender’s agent, or as acting with his/her authority, how else might EJCE be justified? Simester has advanced the most influential such justification. For him, the person who, for example, sells a jemmy foreseeing that the buyer might use it to commit a burglary should not be convicted of burglary if it results. Rather, to be liable as an accessory, one should be proved to have known when providing the relevant assistance or encouragement, of the principal’s intention to commit the particular crime. While the jemmy-seller deserves ‘moral reproof’, to criminalise his/her conduct would be to interfere impermissibly with citizens’ liberty to pursue their own activities. But, Simester thinks, the same considerations are inapplicable when a JCE participant is held liable for a crime that he/she foresaw might result from the enterprise. Such criminalisation does not cause individuals to ‘sacrifice valuable ways of life’; it simply requires them not to become involved in criminal enterprises. Moreover, such offenders are sufficiently culpable to be convicted of murder where they have foreseen that crime’s possible commission and it has resulted. We have seen that Simester regards as culpable the person who provides assistance to a principal — through otherwise innocent conduct such as selling a jemmy — while suspecting that he/she intends to commit an offence. The same is of course true of the individual who continues with a criminal enterprise despite his/her suspicion that a co-offender will commit murder. But, for Simester, it is not this latter person’s foresight that makes him/her blameworthy enough justifiably to be labelled a murderer. Indeed, he implies elsewhere that such a person exhibits enough culpability to be convicted of murder despite ‘only’ foreseeing the possibility of its commission. Rather, it is the offender’s entry into a criminal agreement that makes the difference:
Through entering into a joint enterprise, S changes her normative position. She becomes, by her deliberate choice, a participant in a group action to commit a crime. Moreover, her new status has moral significance: she associates herself with the conduct of the other members of the group in a way that the mere aider and abettor, who remains an independent character throughout the episode, does not.
That is, for Simester, what makes it fair to hold the EJCE offender liable for murder even though he/she has merely foreseen the possibility of that offence’s commission, is that he/she has joined a group ‘that has set itself against the law and order of society at large’. By agreeing to commit a burglary, for instance, the person has ‘passed over [a] ... moral threshold’ and can be held liable for any consequences — including any intentional killings — that he/she foresaw as possible incidents of the enterprise. The question, however, is whether Simester is right. If not, the justification for EJCE that the High Court has twice endorsed disappears; and there is seemingly nothing to replace it.
Before discussing the merits of Simester’s theory, I must deal with one commonly deployed, but, by itself, inadequate, argument for the view that the EJCE offender is insufficiently culpable to be convicted of murder. In Miller, Gageler J referred to two related criticisms of EJCE. ‘[M]aking the criminal liability of the secondary party turn on foresight when the criminal liability of a principal party turns on intention,’ his Honour observed, ‘creates an anomaly’. In turn, this anomaly ‘highlights the disconnection between criminal liability and moral culpability’. McNamara makes essentially the same point, but notes that EJCE’s mental element is also inconsistent with the mens rea for both JCE and accessorial liability. According to this argument, then, to demonstrate that the EJCE offender is not culpable enough justifiably to be convicted of murder is as easy as showing that: (i) a principal can be convicted of murder only if the Crown can prove that he/she either intended to kill or cause grievous bodily harm or foresaw the probability that his/her conduct would result in death (in NSW) or death or grievous bodily harm (in SA); and (ii) accessorial or JCE liability only attaches if the Crown can prove that the accused, respectively, intentionally assisted or encouraged the principal, or agreed to commit the crime.
The first problem here is that a principal offender’s murder liability does not always ‘turn on [proof of] intention’ or the high degree of recklessness just mentioned. The constructive murder rule’s continued existence in NSW and SA is fatal to such a claim. Accordingly, once we acknowledge that a person can be convicted of murder if it is proved that he/she killed someone, however unintentionally, during (etc) his/her commission of a serious offence, we can no longer plausibly contend that the EJCE offender is subject to an anomalously low mens rea standard when compared with those convicted of murder as principals.
Second, and more importantly, this reasoning overlooks Simester’s change of normative position justification for EJCE. As stated above, Simester’s point is that the EJCE offender is culpable enough to be convicted of murder not because of, but despite, his/her mere foresight of the possibility of the further crime. Simply to compare the foresight of probability threshold for reckless murder — and the intention/agreement requirements for accessorial liability and JCE — with the lower foresight of possibility EJCE standard, is to miss that point. What instead must be dealt with is Simester’s claim — accepted twice by a High Court majority — that the EJCE offender’s agreement with another/others to commit the foundational offence is of crucial relevance to his/her culpability regarding a foreseen death that results from that enterprise. Only if that claim fails can we use any anomalies that exist in the law of murder to support a contention that the McAuliffe principle catches those who lack the requisite culpability.
In my view, Simester’s claim does fail. To explain this view, I will consider the constructive murder rule.
Interestingly, Binder has sought to justify this rule with reasoning that resembles Simester’s concerning EJCE. The State may, Binder thinks, hold a person liable for murder even though he/she did not foresee that the relevant conduct might result in death, if two conditions are satisfied. First, the person’s conduct must have created a foreseeable risk of death. Second, the offender must have engaged in this objectively dangerous conduct for a felonious purpose that is independent of injury to the deceased. In other words, consistently with Simester’s argument concerning EJCE, a lower mens rea threshold is said to be justified for constructive murder, because the offender’s embarkation on a foundational offence such as robbery has ‘changed the normative meaning’ of the resulting homicide, ‘aggravat[ing] the wrong of causing harm negligently’.
One problem with the change of normative position reasoning concerns the EJCE accused who has agreed to commit a relatively trivial foundational offence, such as larceny, with foresight that a hot-tempered co-offender might commit murder during the enterprise. Has there really been a change of normative position here, justifying the imposition of murder liability on the person if his/her co-offender proceeds to commit that offence? It is immaterial that it is rare for a person to be convicted of murder because of EJCE where the foundational offence is not particularly serious. What is relevant is that, for Simester, a person’s normative position changes as soon as he/she agrees to commit a crime, however trivial it is. What is also relevant is that the position is different concerning the constructive murder rule. We have seen that Binder thinks that it is only an independent felonious purpose that changes the normative meaning of a negligent killing. Similarly, in SA and NSW, it is only when an unintentional killing has occurred during (etc) the commission of a crime punishable by, respectively, 10 and 25 years’ imprisonment, that that killing is, or can be, murder. It is unclear why the constructive murderer must engage in more serious criminality than the EJCE offender before his/her normative position changes.
Another problem concerns the type of offender contemplated by Lanham. Lanham’s hypothetical offender knows and approves of B’s plan to commit an armed robbery. He/she also realises that B might commit murder during the robbery. He/she nevertheless assists B by, without B’s knowledge, ‘causing a bomb scare which occupies the attention of a large number of local police officers’. B does commit murder during the armed robbery. Is the offender who has assisted him/her really less culpable regarding the relevant killing than is the party to an agreement to commit theft who continues with such an enterprise despite foreseeing murder? Because there has been no agreement between the two offenders to commit armed robbery, Simester would say ‘yes’. But this does not seem right.
A final point is this. We have seen that the constructive murderer crosses the moral threshold at a different point from where the EJCE murderer crosses it. He/she must commit a serious offence before he/she can be held liable for an unintended killing; it is enough that the EJCE offender agrees to do something criminal. But this is not the only relevant difference between the constructive murder rule and EJCE. It is also noteworthy that, concerning the former, Binder contends that the offender can justifiably be held liable for killings that were merely reasonably foreseeable. This can be contrasted with Simester’s insistence that a JCE participant can only be convicted of murder when he/she has actually foreseen its possible commission. This is puzzling. Why does a person’s embarkation on a crime sometimes mean that he/she is culpable enough only to be held liable for foreseen consequences, but on other occasions mean that he/she is sufficiently blameworthy to be held liable for foreseeable consequences?
Consistently with Ashworth’s position, ‘[n]o good reason’ has been given to support the view that individuals should be held liable for murder where a killing is the unintended consequence of their intentional or knowing criminal conduct. Indeed, this becomes apparent when we consider Binder’s and Simester’s reasons for supporting the rules that they do. Binder simply argues that the constructive murderer’s bad motives are what counts. Such a person imposes a foreseeable risk of death for a far from worthwhile end. But if, for example, the racially-motivated offender who kills unintentionally, is to be convicted of manslaughter, why do other malign motives ‘change the normative meaning’ of an offender’s conduct sufficiently so as to justify his/her being labelled a murderer? Simester, by contrast, focuses more on the dangerousness of criminal groups. ‘Criminal associations’, he says,
tend to encourage and escalate criminality. They present a threat to public safety that ordinary criminal prohibitions, addressed to individual actors, do not entirely address. ... A group is a form of society, and a group constituted by a joint unlawful enterprise is a form of society that has set itself against the law and order of society at large.
This passage is important for two reasons. First, those who form criminal agreements might be more dangerous than those who engage in criminal conduct by themselves; but that is irrelevant to their culpability regarding unintended harms that flow from their wrongdoing. Second, however, concerns about the dangerousness of criminal groups certainly do underlie EJCE (just as concerns about selfish risk-creation might well explain the constructive murder rule where it exists). In this connection, there is, for example, Lord Hutton’s recognition in
R v Powell of ‘the need to give effective protection to the public against criminals operating in gangs’. Might the change of normative position reasoning really be no more than a rationalisation of the law’s response to such policy concerns? The problems identified above indicate that this is so. In other words, it seems that the EJCE murderer’s normative position is said to change at a different point from where the constructive murderer’s does, simply because that is consistent with what the law actually says. Likewise, the insistence that objective fault regarding the killing is enough for constructive murder, but not for EJCE murder, seems to owe more to an attempt to justify the existing rules than it does to principle. It is true that, both at common law and under many statutory formulations of the constructive murder rule, even objective fault is unnecessary before a person can be convicted of ‘felony murder’. But usually where a person kills in the course or furtherance of committing a serious offence, the conduct causing the death will be objectively dangerous. There is also Stephen J’s famous statement in R v Serné that felonious conduct should have to be ‘likely in itself to cause death’ before it can found a murder conviction. Certainly, of course, the law says that the EJCE offender must have foreseen the possibility of murder before he/she can be convicted of that offence.
If a person’s commitment to a foundational offence is not as relevant as Simester and Binder think it is to his/her culpability regarding a resulting unintentional killing, what follows? The answer is that we are left with anomalously low mens rea standards for those alleged to be guilty of murder on the basis of EJCE and/or the constructive murder rule. That is, once we dispense with the change of normative position justification for both EJCE and the constructive murder rule, the argument from anomaly can be used to support the view that persons convicted of murder on either basis are insufficiently culpable to be convicted of that offence.
It is true that there have been some recent statements that all such offenders are sufficiently blameworthy to be convicted of murder. Bindon, for example, argues that an offender such as the appellant in Ryan v The Queen, though he/she might neither have intended to kill or inflict grievous bodily harm nor have foreseen any risk of death, possesses the requisite culpability because he/she has ‘brought about a situation of critical danger or violence’. But this analysis rather seems to overlook the uproar that followed the House of Lords’ ill-advised decision in
DPP v Smith. Was Dixon CJ mistaken in Parker v The Queen when he declared DPP v Smith’s objective standard to be ‘misconceived and wrong’? Bindon would apparently say ‘yes’. Indeed, she would seemingly have it that a person is culpable enough to be convicted of murder even if his/her conduct involved no objective danger; a mere act of violence would be enough.
If we take Parker seriously, then so we must the High Court’s reasoning in
R v Crabbe. There it was observed that the person who performs conduct knowing that death or grievous bodily harm will probably result has a state of mind that is ‘comparable with’ an intention to cause the relevant consequence. But the Court also held that the same is not true of the person who merely foresees that death or grievous bodily harm might flow from his/her actions. Provided that we first accept that an accused’s agreement to commit a foundational offence increases his/her culpability concerning a foreseen murder no more than does any other reprehensible motive for creating risk, it follows inexorably from this reasoning that persons caught by the McAuliffe principle, too, lack the requisite culpability.
If McAuliffe does not withstand critical scrutiny, why did the Miller plurality uphold it? Of course, it might simply be that their Honours, however erroneously, agree strongly with McAuliffe and would in no circumstances have overruled it. But a close reading of the plurality’s decision, and a consideration both of the social context in which Miller was decided and widely-held judicial views about when an ultimate court of appeal may depart from an established common law rule, makes me believe that their Honours refused to abandon McAuliffe mainly because there was insufficient public support for such abandonment. Judges will rarely reverse or depart from settled law unless it seems clear that they are acting consistently with community values. This was so in the case of Jogee. The same would not have been true in Miller.
At this point, I must make it clear what I am, and am not, arguing. I am not seeking to prove that Miller would have been decided differently had social conditions been the same in SA and NSW as they were in the UK when Jogee was decided. Like any counterfactual, any such claim is insusceptible of formal proof. Neither am I even arguing that, in my opinion, Miller would certainly have turned out differently if EJCE had been as publicly controversial as PAL became. While many senior judges have indicated that a court may not override an established common law doctrine unless such action would be accepted by the public, some have also stated that, even if judges are satisfied that the community supports (or would be indifferent to) change, they must proceed cautiously. As Doyle CJ has suggested, it might be better for judges to leave the relevant matter to Parliament if the decision would have ‘unpredictable’ or ‘far reaching’ consequences, or where ‘the subject matter clearly calls for investigations which the Court cannot make’.
What I am arguing is that, in my view, the Court probably would have abandoned McAuliffe if public opinion had been as hostile to it as UK opinion was to Chan Wing-Siu. I base myself not only on the judicial and extra-judicial statements just noted, but also on similar remarks made by one member of the Miller plurality, Bell J, in a lecture that her Honour delivered about ten months before that case was decided. Crucially, Bell J also considered the McAuliffe principle. Her Honour’s remarks indicate that she conceived of the main issue in Miller as being not McAuliffe’s correctness, but whether it was appropriate for the Court to intervene.
‘Judges,’ her Honour stressed, ‘are concerned with how the law is applied in actual cases.’ This, she was inclined to see as a strength of the common law method. Why? Her Honour’s answer was essentially that this method allows the courts to ensure that just results are reached, at least in some instances — even if some judicial ‘creativity’ is necessary before this can occur. Justice Bell used R v Mullen to illustrate her point. That case, which was decided three years after Woolmington v DPP, concerned the burden of proof in a criminal trial. To escape liability for murder, did Mr Mullen have to prove that the killing was ‘authorised or justified or excused by law’ within the meaning of s 291 of the Queensland Criminal Code? Or was the Crown required to prove that he did not kill the deceased accidentally? As Bell J noted, in Mullen, Dixon J observed that the Code’s text supported the former view. Moreover, s 291 had widely been considered to operate in this way. But Dixon J nonetheless
concluded that [the] Criminal Code ... did not necessarily imply a principle that the burden was on the prisoner to prove accident ... [V]alues that had come to be fundamental over the course of the last century [were thus] read into the Griffith Code.
So, according to Bell J, courts will sometimes even use questionable reasoning to achieve the ‘right’ result. But it is important to understand when her Honour believes they are entitled to do so. In the above quotation, Bell J hints at this too, and she makes the point explicitly shortly afterwards. Referring to Brennan J’s remarks in Dietrich about when judges may change a common law rule, Bell J said that his Honour ‘had in mind’ circumstances where such action accords with the ‘relatively permanent values of the Australian community’.
That is, for Bell J, the courts may modify the law only where this is consistent with community values. And before they will do so, they will check — insofar as this is possible — that what they perceive to be the ‘relatively permanent’ ‘values of ... society’ really do have this status. As her Honour later suggests, to take a more expansive approach would be to invite claims of ‘judicial activism’; it would risk undermining the courts’ reputation for impartiality.
It is interesting to view Jogee and Miller in light of these remarks. As the Supreme Court noted in the former case, PAL was ‘highly controversial’ in the UK; indeed, it was so controversial that the Court had ‘set up an appeal’, using ‘Jogee and Ruddock as vehicles for that task’. In the years following the formation in 2010 of the campaigning group ‘Joint Enterprise Not Guilty by Association’ (‘JENGbA’), public concern about ‘joint enterprise’ appears to have increased significantly. In early 2012, the House of Commons Justice Select Committee requested data concerning how regularly the Crown was using PAL. By the end of 2014, it was calling for an ‘urgent review of the law of joint enterprise in murder cases’. Concerns were raised both before the Committee and in the press about the disproportionate number of mixed-race, black and working-class defendants who were being convicted because of PAL. There were press reports focused on cases, such as Jordan Cunliffe’s, in which PAL was alleged to have caused substantial injustice. Claims were made that it was a ‘lazy law’ that cast the net too widely. The media gave attention to research that exposed the extremely widespread use of PAL.
On 6 July 2014, ‘Common’, a television drama written by Jimmy McGovern, aired on BBC One. It did not portray PAL favourably. The following night, ‘Guilty by Association’, was broadcast — also on BBC One. Dr Sally Halsall and her daughter, Charlotte Henry, featured in this programme. They argued that Dr Halsall’s son and Charlotte’s brother, Alex Henry, had unjustly been convicted of murder as a parasitic accessory. Alex himself claimed in the press that he had not foreseen that the principal offender might commit murder. Other defendants claimed that they had suffered miscarriages of justice. Alex Henry’s sister and mother continued to campaign articulately against PAL.
Of course, when the Supreme Court was abandoning PAL, it helped that it could point to a legal justification. The careful doctrinal argument in Jogee has rightly been praised. But would this argument, however persuasive it is, have caused their Lordships to change the law had they not been satisfied that PAL was widely perceived to be causing injustice? Judges’ reluctance to reverse a settled common law rule, and their related concern to secure legal certainty, causes me to doubt this. Indeed, certainty is not the only relevant consideration. Questions of legitimacy also arise. As McHugh J has said:
An important limitation on judicial law-making arises from the need for the judge’s view to be accepted by the community ... If a change in the common law would be rejected by the community, it should not be made, however much the judge thinks that the change is in the community’s interest.
Had PAL not been as obviously controversial as it was, might their Lordships have feared such community rejection? Might this have caused them to exercise restraint? Justice Bell’s extra-judicial remarks about McAuliffe, which I deal with below, fortify my belief that this is a real possibility.
In SA and NSW before Miller, there was nothing approaching the sort of opposition to EJCE that there was in the UK to PAL before Jogee. There were no press articles condemning EJCE. There were no documentaries or television dramas. Parliamentarians had not expressed concerns. And there were no widely-ventilated claims of injustice in particular cases. The Miller plurality made explicit reference to this last fact:
Of course, were the law stated in McAuliffe to have led to injustice, any disruption occasioned by departing from it would not provide a good reason not to do so. However, here, as in Clayton, the submissions are in abstract form and do not identify decided cases in which it can be seen that extended joint criminal enterprise liability has occasioned injustice.
Indeed, a careful reading of the plurality’s reasons strengthens the conclusion that it was probably the absence in SA and NSW of community opposition to EJCE — and not their Honours’ unreserved support for the doctrine — that caused the case to be decided as it was.
It is noteworthy that, in the passage just set out, their Honours’ focus is categorically not on whether EJCE is justified in an abstract sense. It is instead on whether that doctrine, principled or not, has caused practical injustice. This is reminiscent of Lord Steyn’s observation in R v G that ‘[t]he surest test of a new legal rule is not whether it satisfies a team of logicians but how it performs in the real world.’ Because EJCE had performed adequately — that is, because, unlike in the UK, there had been no vociferous campaign against it — there was no need to risk causing the ‘inconvenience’ that might well flow from its abandonment. But would their Honours have been more receptive to the idea of dispensing with McAuliffe if the appellants — as the UK campaigners had purported to do — had identified ‘decided cases’ in which EJCE ‘occasioned injustice’? The suggestion in the above passage is that they would have.
Certainly, as I have noted, their Honours did, elsewhere in their judgment, endorse Simester’s principled defence of EJCE. But they did so in a manner that inspires less than complete confidence that they would have taken the same approach had they been faced with the kinds of pressures that confronted the Supreme Court in Jogee. So, after acknowledging the disagreement between J C Smith and Simester concerning the doctrinal basis of JCE and EJCE, and after noting that the Clayton majority accepted the latter’s views, their Honours said:
Acknowledgement of the sui generis nature of the secondary liability that arises from participation in a joint criminal enterprise may be thought to resolve at least some of the anomalies that are suggested to arise from allowing foresight of the possible commission of the incidental offence by a co-venturer as a sufficient mental element of liability.
High Court Justices sometimes refer disapprovingly to those who use the passive voice, so why did five of them use it here? Was it to create some distance between their Honours and what they were saying? Certainly, the Miller endorsement of Simester’s views seems more muted than the Clayton one.
Furthermore, their Honours did approve J C Smith’s early view, which he later doubted, that the person who continues to participate in a criminal enterprise despite his/her foresight that another participant might commit murder, is sufficiently culpable to be convicted of that offence. But, again, it is perhaps instructive to consider how their Honours chose to express such approval:
It is not self-evident ... that the policy of the law should be against the imposition of liability for murder in such a case. Certainly, [the person’s] moral culpability is not less than that of the secondary party in a case such as Johns.
Plainly, it is not ‘self-evident’ that the accused should avoid murder liability in such a case. If it were, the McAuliffe principle would not have been created in the first place. But that is a different thing entirely from giving unqualified support to that principle. Concerning Johns, their Honours were not saying that the McAuliffe offender is as culpable as the individual who has been proved to have ‘conditional[ly] inten[ded] that the incidental offence be committed’. Rather, they were contending that the person with individual foresight of the possibility of a further crime has acted in just as blameworthy a fashion as he/she would have had he/she and his/her co-offender jointly foreseen the commission of that offence and yet continued with the foundational enterprise. To say this is different from arguing that foresight — whether joint or individual — is undoubtedly enough for murder liability to be justified.
In short, their Honours’ point was seemingly that there is room for argument about whether all of those convicted of murder due to McAuliffe are culpable enough justifiably to be convicted of that offence. As much is demonstrated, their Honours suggested, by the existence of some distinguished academic support for EJCE. Accordingly, what reason would there be to cause the ‘disruption’ that would flow from McAuliffe’s abandonment — unless EJCE had been shown, or at least had widely been perceived, to have ‘occasioned injustice’? Indeed, disruption would not be the only adverse consequence that such a decision would cause. Courts that act on their own perceptions of what justice requires — as opposed to altering the law only in accordance with the community’s ‘reasonably permanent values’ — tend quickly to suffer reputational damage.
Justice Bell’s remarks about the McAuliffe principle, in the lecture to which I have referred, support the analysis just presented. Her Honour said:
Contrary to the more extreme views as to the rapaciousness of the judiciary, there are areas of the general part of the criminal law in which the High Court has stayed its hand. Many commentators consider that the principles of extended joint criminal enterprise stated by the High Court in McAuliffe ... impose liability too widely. The Court has declined to reconsider McAuliffe ... Kirby J would have ... confined the liability of the secondary participant. His Honour set out cogent reasons in favour of that view.
What is noteworthy here is Bell J’s evident anxiety that the High Court not be perceived to be ‘rapacious’. Justice Bell is aware, her Honour tells us, that there are those who consider the Court to have arrogated to itself powers that are properly exercised only by the other branches of government. But her Honour dismisses these as ‘extreme’ views — and she can cite a case, Clayton, that she thinks illustrates her point. In other words, however ‘cogent’ the reasons for abandoning EJCE might have been, it was also important for their Honours to avoid acquiring a reputation for taking it upon themselves to act legislatively or for whimsically departing from the Court’s precedents. By rejecting the appellant’s argument, the Court could say, as Bell J did of the Clayton majority, that it had ‘stayed its hand’ in the interests of democracy.
Indeed, a democratic sensitivity is evident in her Honour’s final remarks about EJCE. After referring to the NSW Law Reform Commission’s 2010 recommendation that the doctrine’s mental element be made more stringent,
Bell J said:
It is almost five years since the Commission’s Report was published and no action has been taken on this or the other recommendations in it. Some would say this argues for the Court to be less timorous and that it should have agreed to revisit McAuliffe given legislative lethargy when it comes to the orderly reform of the criminal law. Presumably democrats would say that the legislature’s silence connotes acceptance of the law as the Court has stated it.
As her Honour implies here, the Court in Miller had a choice. On one hand, it could have overturned McAuliffe. Certainly, an argument against its doing so was that, if Parliament had wished to alter or dispense with EJCE, it had had many opportunities so to act, but had availed itself of none of them. But such arguments are not always successful. In R v G, for example, the House of Lords had not been persuaded by such considerations to uphold the R v Caldwell rule that an accused would be ‘reckless’ within the meaning of s 1(1) of the Criminal Damage Act 1971 (UK) if he/she could be proved merely to have created an obvious risk that property would be destroyed or damaged without adverting at all, at the time of the relevant conduct, to the possibility of there being such a risk. On the other hand, the Court could have upheld McAuliffe. In taking this approach, the plurality used very similar reasoning to that suggested by Bell J in the above quotation:
In the decade since Clayton ... [t]he New South Wales Law Reform Commission undertook a review of the law of complicity. ... The Parliament of New South Wales has to date not chosen to act on the Commission’s recommendations. The Parliament of South Australia has also not chosen to reform the law as stated in McAuliffe.
In light of this history, it is not appropriate for this Court to now decide to abandon extended joint criminal enterprise liability ...
Why did the plurality take this latter course, when, as R v G shows, it did not have to do so? It is submitted that, as argued above, public opinion in NSW and SA was not sufficiently hostile to EJCE for the Justices to be satisfied that that doctrine should be abandoned. If their Honours had considered that there existed a widespread community perception that EJCE operated unjustly — if they had thought, that is, that the problems raised by the appellants were less ‘abstract’ — it is quite possible that they would not have observed so scrupulously the requirements of democratic principle. Certainly, Bell J’s extra-judicial remarks indicate that her Honour was far from persuaded of McAuliffe’s justifiability. To ‘stay ... [one’s] hand’ in the face of ‘cogent’ criticisms is not to support. And as I have argued, the plurality’s reasoning, upon close inspection, is redolent not of unconditional curial approval but, rather, of judicial restraint.
Further, the reasoning in R v G is illuminating. Their Lordships took into account a number of considerations when deciding to overturn Caldwell. Most significantly, their Lordships found that, contrary to the majority’s decision in that earlier case, Parliament, when it enacted the Criminal Damage Act, had plainly intended to provide that a person would only be ‘reckless’ for the purposes of the s 1 offences if he/she had actually realised that his/her conduct might cause the relevant consequence(s). But Lord Bingham held that even this obvious misinterpretation was not determinative of the appeal. If this mistake had ‘offended no principle and given rise to no injustice’, his Lordship thought, ‘strong grounds’ would have existed for not intervening. It was only because the error was ‘offensive to principle and [was] apt to cause injustice’, that the need to correct it was ‘compelling’. In truth, however, it was not decisive that the rule was apt to operate unjustly. What instead appears to have persuaded the Court to overrule Caldwell was that, in R v G, their Lordships had before them a case where the relevant principle had been demonstrated to cause such unfairness. How had this been demonstrated? Lords Bingham and Steyn placed great emphasis here on its being ‘evident’ that the trial judge’s Caldwell direction had ‘offended the jury’s sense of fairness’. For Lord Bingham, ‘[t]he sense of fairness of 12 representative citizens sitting as a jury ... is the bedrock on which the administration of criminal justice in this country is built. A law which runs counter to that sense must cause concern.’
Again, if there had been such evidence of community unease about EJCE, the High Court might well have considered parliamentary inactivity to be no obstacle to the overruling of McAuliffe.
Upon her Honour’s retirement from the NSW Supreme Court, Bell J said of her new role as a High Court Justice:
The prospect of my new role has had an unsettling effect on me which is hard to understand since, as early as my days at the Redfern Legal Centre, I had no difficulty in perceiving the errors of principle made by the High Court and in seeing how readily they could be corrected.
While her Honour was delivering a joke, there is nevertheless an insight here. That insight is relevant to the argument that I have made in this article.
As I have sought to demonstrate, the High Court in McAuliffe did make an ‘error of principle’. The JCE participant who foresees, without necessarily agreeing to, the commission of murder should not be convicted of that offence if it results. Certainly, the change of normative position reasoning that has been used to support EJCE (and the constructive murder rule) does accurately describe what John Gardner has called ‘the law’s own moral outlook’. But it does not justify that outlook. Moreover, if, as I think, accessorial liability and the basic JCE doctrine have the same doctrinal foundation as one another — the purpose of both being to hold liable only those who intentionally assist or encourage another person to perform criminal conduct — it is difficult to see why a JCE participant should be liable for a further crime, committed by a co-participant, unless he/she has intentionally assisted or encouraged him/her to perform the relevant acts. The Jogee insistence that such a person can only be held liable for the further crime if he/she intended that it be committed should the occasion arise, is therefore more normatively desirable than the McAuliffe position — whatever pragmatic considerations and policy concerns might have tempted courts away from accepting such logic.
Once an error of principle is made, however, it is not ‘readily corrected’. There is perhaps a tendency among commentators merely to point to courts’ mistakes, as though all that it will take for those courts to correct those mistakes is a strong dose of reason. Justice Bell suggests that, before becoming a judge, she was inclined to think that cases were decided purely on the basis of such rational legal and moral argumentation. But the reality is that, because the courts are concerned to maintain their legitimacy, they must consider in such cases not only matters of principle, but also how the public and press will respond to any decision to reverse a line of precedents. I have argued here that such concerns probably influenced in no small way the plurality’s decision in Miller. The same seems true of Jogee.
This is certainly not to defend Miller. That decision leaves undisturbed a rule that sacrifices fairness to accused persons for the illusion of community protection. It does so even though their Honours must have known that neither the NSW nor the SA Parliament is at all likely to intervene in this area in the foreseeable future. Of course, the High Court’s reputation must be maintained. Relatedly, it is understandable that their Honours wish to do their work as unobtrusively as possible: while it is important that justice is seen to be done, it is also important that the workings of the justice system are not seen, or remarked upon, too frequently. If the UK press’s response to Jogee is any indication, the Australian press would certainly have noticed a decision to abandon EJCE. But might not the Court’s apparent fear of being perceived as ‘rapacious’ have been at least slightly exaggerated? And, given the injustice that EJCE can produce (as demonstrated by the UK justice system’s experience with PAL), might it not have been worth using some of the Court’s ‘reputational capital’ to correct the McAuliffe error before a litigant is able to satisfy a court that he/she has suffered less ‘abstract’ injustice than was brought to their Honours’ attention in Miller?
Nevertheless, if we are accurately to analyse that case, we must recognise that it probably amounts more to a statement about the circumstances in which the plurality Justices are willing to overturn established common law rules, than it does about whether they support McAuliffe. If we do recognise that courts are reluctant to change even unjust common law rules unless there is clear public support for their doing so, we will also maximise our chances in the future of developing arguments that persuade them to take a less cautious approach in important cases such as Miller.
[∗] Colin Phegan Lecturer, University of Sydney Law School, Sydney, Australia; Deputy Director Sydney Institute of Criminology.
  AC 387 (‘Jogee’).
 Ibid 396 .
 Ibid 412  citing Chan Wing-Siu v The Queen  UKPC 27;  AC 168 (‘Chan Wing-Siu’).
 Ibid 414 .
 Ibid 415 . For commentary on Jogee, see David Ormerod and Karl Laird, ‘Jogee: Not the End of a Legal Saga but the Start of One?’  (8) Criminal Law Review 539.
 Following the Victorian Parliament’s recent decision to place the law of complicity largely on a statutory footing, SA and NSW are the only two Australian jurisdictions in which the common law of complicity applies. For the Victorian position, which was introduced after the delivery of the Weinberg Report, but which, to an extent, departs from its recommendations, see Crimes Act 1958 (Vic) ss 323–324C. See also Justice Mark Weinberg, Victorian Department of Justice and Judicial College of Victoria, ‘Simplification of Jury Directions Project: A Report to the Jury Directions Advisory Group’ (Report, Government of Victoria, August 2012) 17–102 [2.1]–[2.306].
 See, eg, Stephen Odgers, ‘McAuliffe Revisited Again’ (2016) 40(2) Criminal Law Journal 55; Sarah Pitney, ‘Undoing a “Wrong Turn”: The Implications of R v Jogee; Ruddock v The Queen for the Doctrine of Extended Joint Criminal Enterprise in Australia’ (2016) 40(2) Criminal Law Journal 110, 115.
  HCA 30; (2016) 259 CLR 380 (‘Miller’).
 Ibid 388 . In HKSAR v Chan Kam Shing  HKCFA 87; (2016) 19 HKCFAR 640, the Hong Kong Court of Final Appeal (‘HKCFA’) came to the same conclusion as the High Court. While my focus is on English and Australian law, the HKCFA’s unanimous decision not to follow Jogee might well have been influenced by the types of factors that appear to have influenced the Miller plurality.
 See, eg, Beatrice Krebs, ‘Accessory Liability: Persisting in Error’ (2017) 76(1) Cambridge Law Journal 7; Stephen Odgers, ‘The High Court, the Common Law and Conceptions of Justice’ (2016) 40(5) Criminal Law Journal 243; A P Simester, ‘Accessory Liability and Common Unlawful Purposes’ (2017) 133 (January) Law Quarterly Review 73.
 Miller  HCA 30; (2016) 259 CLR 380, 426 , 429–30 – (Keane J).
 Brown v The State (Trinidad and Tobago)  UKPC 10 (29 January 2003) .
 A P Simester, ‘The Mental Element in Complicity’ (2006) 122 (October) Law Quarterly Review 578, 598–601; see also A P Simester et al, Simester and Sullivan’s Criminal Law: Theory and Doctrine (Hart Publishing, 6th ed, 2016) 248–50; Jeremy Horder and David Hughes, ‘Joint Criminal Ventures and Murder: The Prospects for Law Reform’ (2009) 20(3) Kings Law Journal 379, 398–9; Laura Stockdale, ‘The Tyranny of Small Differences: Culpability Gulf between Subjective and Objective Tests for Extended Joint Criminal Enterprise in Australia’ (2016) 90(1) Australian Law Journal 44, 49.
  HCA 58; (2006) 81 ALJR 439, 444  (‘Clayton’).
  HCA 30; (2016) 259 CLR 380, 398 .
 Ibid 426 , 427–8 , –.
 Ibid 429 .
  HCA 64; (2003) 219 CLR 1 (‘Gillard’).
 Ibid 7 .
 EJCE applies to all crimes, not just to murder; however, in most of the leading cases the doctrine was used to fix liability for that crime, and it is regarding murder that it is most controversial. Accordingly, this article is mainly concerned with that offence.
 Cf Miller  HCA 30; (2016) 259 CLR 380, 419 – (Gageler J).
 Crimes Act 1900 (NSW) s 18(1)(a); Criminal Law Consolidation Act 1935 (SA) s 12A.
 Miller  HCA 30; (2016) 259 CLR 380, 419  (Gageler J). His Honour conceded that, ‘in the case of manslaughter special considerations apply’. But the common law has accepted the sufficiency of objective fault for many other crimes, too. Indeed, as a NSW textbook observes, the presumption in favour of a subjective mental element where a statutory offence is silent concerning mens rea, arising from He Kaw Teh v The Queen  HCA 43; (1985) 157 CLR 523, has only rarely remained unrebutted: David Brown et al, Criminal Laws: Materials and Commentary on Criminal Law and Process of New South Wales (Federation Press, 6th ed, 2015) 203–4.
 Guyora Binder, ‘The Culpability of Felony Murder’ (2008) 83(3) Notre Dame Law Review 965, 1036, 1045.
 Beatrice Krebs, ‘Joint Criminal Enterprise’ (2010) 73(4) Modern Law Review 578, 599.
 Jogee  AC 387, 416 .
 See, eg, Fardon v A-G (Qld)  HCA 46; (2004) 223 CLR 575, 593  (Gleeson CJ).
 See, eg, Andrew Dyer, ‘(Grossly) Disproportionate Sentences: Can Charters of Rights Make a Difference?’  MonashULawRw 6; (2017) 43(1) Monash University Law Review 195, 212–13.
 Certainly, judges must usually not intervene when an established common law rule has been challenged, lest they create a perception that they are legislating: Chief Justice John Doyle,
‘Do Judges Make Policy? Should They?’ (1998) 57(1) Australian Journal of Public Administration 89, 94. But would such a perception have been created had Miller been decided differently? Public opinion had not set itself against EJCE, but there were no strident calls for it to be retained either. Moreover, there is force in Kirby J’s contention in Clayton  HCA 58; (2006) 81 ALJR 439, 461  that EJCE had, recently, been expressed by judges and could ‘therefore be re-expressed by them’, and in Gageler J’s related claim in Miller  HCA 30; (2016) 259 CLR 380, 423  that EJCE’s excision would not have had far-reaching consequences such as to make it proper for the Court to leave the matter to Parliament: cf text accompanying below nn 168–9.
 As noted by the Miller plurality, neither the SA nor the NSW Parliament has chosen to override the McAuliffe principle:  HCA 30; (2016) 259 CLR 380, 401 . In the UK before Jogee, there was similar parliamentary passivity: see, eg, William Wilson and David Ormerod, ‘Simply Harsh to Fairly Simple: Joint Enterprise Reform’  (1) Criminal Law Review 3, 26–7; Matthew Dyson, ‘The Future of Joint-Up Thinking: Living in a Post-Accessory Liability World’ (2015) 79(3) Journal of Criminal Law 181, 193–4. But in the traditionally liberal state of Victoria, Parliament has recently acted: see above n 6.
 Clayton  HCA 58; (2006) 81 ALJR 439; Miller  HCA 30; (2016) 259 CLR 380. See also Gillard  HCA 64; (2003) 219 CLR 1, 36–7 – (Hayne J) and 15  (Gummow J).
 McAuliffe v The Queen  HCA 37; (1995) 183 CLR 108, 117–18 (‘McAuliffe’).
 This distinction is referred to in Osland v The Queen  HCA 75; (1998) 197 CLR 316, 341–2  (McHugh J) (‘Osland’).
 A-G’s Reference (No 1 of 1975)  EWCA Crim 1;  QB 773, 779–80 (‘Attorney-General’s Reference’).
 Giorgianni v The Queen  HCA 29; (1985) 156 CLR 473, 503–8; Stokes v The Queen (1990) 51 A Crim R 25, 37–9.
 Likiardopoulos v The Queen  HCA 37; (2012) 247 CLR 265, 276–7  (‘Likiardopoulos’); Osland  HCA 75; (1998) 197 CLR 316, 341–2 .
 Osland  HCA 75; (1998) 197 CLR 316, 342–3  (McHugh J); McAuliffe  HCA 37; (1995) 183 CLR 108, 113–4; Miller  HCA 30; (2016) 259 CLR 380, 388 .
 McAuliffe  HCA 37; (1995) 183 CLR 108, 114.
 Osland  HCA 75; (1998) 197 CLR 316, 342–3  (McHugh J), quoting with approval R v Lowery [No 2]  VicRp 63;  VR 560, 560; McEwan v The Queen  VSCA 329; (2013) 41 VR 330, 337  (‘McEwan’).
 Huynh v The Queen  HCA 6; (2013) 87 ALJR 434, 439 , 442 –; Tangye v The Queen (1997)
92 A Crim R 545, 556–7.
 McEwan  VSCA 329; (2013) 41 VR 330, 337 .
 Ibid 337 .
 Osland  HCA 75; (1998) 197 CLR 316, 342 , 343 , 344 , 345 ; cf 350 .
  HCA 37; (2012) 247 CLR 265, 274 .
 See, eg, Dickson v The Queen  NSWCCA 78; (2017) 94 NSWLR 476, 489–90 ; Youkhana v The Queen  NSWCCA 41; (2015) 249 A Crim R 424, 427 –; Sever v The Queen  NSWCCA 135 (25 June 2010) 36 .
  HCA 75; (1998) 197 CLR 316.
 Ibid 350  (McHugh J).
  HCA 3; (1980) 143 CLR 108, 130–1 (‘Johns’).
 R v Johns  1 NSWLR 282, 290.
 McAuliffe  HCA 37; (1995) 183 CLR 108, 115.
 Miller  HCA 30; (2016) 259 CLR 380, 388 .
 Compare Jogee  AC 387, 407–8 – with Miller  HCA 30; (2016) 259 CLR 380, 393 .
 Johns  HCA 3; (1980) 143 CLR 108, 131–2. See also Editorial, ‘McAuliffe Revisited’ (2004) 28(1) Criminal Law Journal 5, 6; Stephen J Odgers, ‘Criminal Cases in the High Court of Australia: McAuliffe and McAuliffe’ (1996) 20(1) Criminal Law Journal 43, 45.
 As foreseen by Stephen Gray, ‘I Didn’t Know, I Wasn’t There: Common Purpose and the Liability of Accessories to Crime’ (1999) 23(3) Criminal Law Journal 201, 205, 213.
 McAuliffe  HCA 37; (1995) 183 CLR 108, 117–8.
 Compare, for example, John C Smith, ‘Criminal Liability of Accessories: Law and Law Reform’ (1997) 113 (July) Law Quarterly Review 453, 461–2 with Simester, ‘The Mental Element in Complicity’, above n 13, 592–5.
 Accordingly, in a case where the alleged accessory is claimed to have intentionally assisted the principal, the Crown need not prove that the principal knew of the assistance: R v Lam  VSCA 109; (2008) 185 A Crim R 453, 477  citing with approval R v Lam  VSC 294; (2005) 159 A Crim R 448, 472 . The position is different, however, concerning encouragement: R v Stringer  EWCA Crim 1396;  QB 160, 171–2 .
  EWCA Crim 1;  QB 773.
 Ibid 779.
 Odgers, above n 7, 58.
 Simester, above n 10, 77.
 See above n 59.
 R v Stewart  EWCA Crim 3;  3 All ER 159, 165 (Hobhouse LJ) (‘Stewart’).
 Smith, above n 58, 462.
 (1997) 92 A Crim R 545, 557.
 Sir John Smith, ‘Joint Enterprise and Secondary Liability’ (1999) 50(2) Northern Ireland Legal Quarterly 153, 157–8.
 David Lanham, ‘Primary and Derivative Criminal Liability: An Australian Perspective’  (September) Criminal Law Review 707, 707.
 Matusevich v The Queen  HCA 30; (1977) 137 CLR 633.
 Osland  HCA 75; (1998) 197 CLR 316, 350 .
 Stewart  EWCA Crim 3;  3 All ER 159, 165.
 Lord Hobhouse, ‘Agency and the Criminal Law’ in Francis D Rose (ed) Lex Mercatoria: Essays on International Commercial Law in Honour of Francis Reynolds (LLP, 2000) 39, 47.
 Ibid 46.
 Three High Court Justices have recently denied that ‘the notion of attribution of an act ... involve[s] a fiction that the act was undertaken by the [passive participant(s)]’: IL v The Queen  HCA 27; (2017) 91 ALJR 764, 776  (Kiefel CJ, Keane and Edelman JJ) (‘IL’). Rather, the act is merely treated for legal purposes as having been performed by the passive participant(s): at 773 , 776 . This distinction is excessively subtle. Because the law proceeds as though the passive participant performed an act (in IL, lighting a ring burner) that he/she did not perform, it does uphold a fiction.
 David Ormerod and Karl Laird, Smith and Hogan’s Criminal Law (Oxford University Press, 14th ed, 2015) 260; Jogee  AC 387, 415 ; R v Mendez  EWCA Crim 516;  QB 876, 882 .
  HCA 27; (2017) 91 ALJR 764.
 Ibid 783 , quoting Johns  HCA 3; (1980) 143 CLR 108, 125.
 IL  HCA 27; (2017) 91 ALJR 764, 778 .
 Ibid 779 .
 Ibid 768 .
 IL v The Queen  NSWCCA 51 (8 April 2016).
 For an analysis of IL’s arguments, see Andrew Dyer, ‘IL v The Queen: Joint Criminal Enterprise and the Constructive Murder Rule: Is This Where Their “Logic Leads You”?’  SydLawRw 11; (2017) 39(2) Sydney Law Review 245.
 Kiefel CJ, Bell, Nettle and Edelman JJ: Transcript of Proceedings, IL v The Queen  HCA Trans 65 (4 April 2017).
 IL  HCA 27; (2017) 91 ALJR 764, 782 .
 Ibid 783–4 .
 Ibid 781–5 –.
 Ibid 768 , 773 , 774 .
 Ibid 768 .
 Ibid 773 , .
 Ibid 794  (Gageler J), 798  (Gordon J).
 The Australian courts have not yet determined whether liability arises where a person is merely an accessory to an actus reus: Likiardopoulos  HCA 37; (2012) 247 CLR 265, 276–7 . But it is hard to believe that, if ever the matter arises for decision, they will prevent the prosecution of an accused who intentionally assisted/procured an actus reus, but did not agree with the actual perpetrator that it be performed. Certainly, the English courts have accepted that a procurer of an actus reus can be convicted: R v Millward  Crim LR 527; R v Wheelhouse  Crim LR 756; R v Pickford  QB 203, 213; DPP v K  EWCA Civ 808;  1 Cr App R 36, 44–5. This development has surely been prompted by the same concern as underlay Osland: the desire to ensure that culpable passive participants are convicted even though the actual perpetrator, for some reason special to him/herself, is acquitted: see, eg, DPP v K  EWCA Civ 808;  1 Cr App R 36, 45.
 Simester, ‘The Mental Element in Complicity’, above n 13, 593–5.
 Cf Jogee  AC 387, 418 –.
 The Miller plurality held that the accused must foresee more than a ‘fanciful’/‘negligible’ possibility of the further crime’s commission:  HCA 30; (2016) 259 CLR 380, 401–2 –.
 Ormerod and Laird, above n 76, 260.
 Cf Justice Margaret Beazley, ‘Extended Joint Criminal Enterprise in the Wake of Jogee and Miller’ (Speech delivered at the NSW Office of the DPP, 7 March 2017) 24 ; but see also 12–13 –.
  HCA 30; (2016) 259 CLR 380, 427 .
 Ibid 429 .
 Ormerod and Laird, above n 76, 260.
 Those were the facts of Gillard  HCA 64; (2003) 219 CLR 1.
 In Gillard, it was held to be open to the jury so to find: ibid 11–12 , 14  (Gleeson CJ and Callinan J), 30–1 – (Kirby J).
  HCA 30; (2016) 259 CLR 380, 429 .
  UKPC 27;  AC 168, 175.
 J C Smith, ‘R v Wakely’  (February) Criminal Law Review 119, 121. See also, eg, K J M Smith, A Modern Treatise on the Law of Criminal Complicity (Clarendon Press, 1991) 220; Simon Bronitt and Bernadette McSherry, Principles of Criminal Law (Lawbook Co, 4th ed, 2017) 441 [7.135].
  HCA 30; (2016) 259 CLR 380, 428 .
 Odgers, above n 10, 244.
  HCA 30; (2016) 259 CLR 380, 428 .
 Cf Gageler J’s reasoning at ibid 419 .
 Clayton  HCA 58; (2006) 81 ALJR 439, 444 ; Miller  HCA 30; (2016) 259 CLR 380, 398 .
 Simester, ‘The Mental Element in Complicity’, above n 13, 589–91.
 Ibid 591.
 Ibid 600.
 Ibid 598–601.
 Simester et al, Simester and Sullivan’s Criminal Law, above n 13, 249.
 Simester, ‘The Mental Element in Complicity’, above n 13, 598–9.
 Ibid 600.
 Krebs, above n 26, 598.
 Miller  HCA 30; (2016) 259 CLR 380, 419 .
 Ibid 419 .
 Luke McNamara, ‘A Judicial Contribution to Over-Criminalisation?: Extended Joint Criminal Enterprise Liability for Murder’ (2014) 38(2) Criminal Law Journal 104, 111.
 Ibid 113.
 Crimes Act 1900 (NSW) s 18(1)(a); R v Crabbe  HCA 22; (1985) 156 CLR 464, 467–8 (‘Crabbe’).
 Royall v The Queen  HCA 27; (1991) 172 CLR 378, 395 (Mason CJ), 405 (Brennan J), 415–7 (Deane and Dawson JJ), 430–1 (Toohey and Gaudron JJ), 455 (McHugh J).
 Crabbe  HCA 22; (1985) 156 CLR 464, 468–9.
 Crimes Act 1900 (NSW) s 18(1)(a); Criminal Law Consolidation Act 1935 (SA) s 12A.
 For a list of qualifying offences in NSW, see New South Wales Law Reform Commission, Complicity, Report No 129 (2010) app C. There are also many qualifying offences in SA — although, there, the act causing death must additionally have been an intentional act of violence: Criminal Law Consolidation Act 1935 (SA) s 12A.
 McNamara does note that the Clayton majority (endorsing Simester) said that, whereas ‘liability as an aider and abettor is grounded in the secondary party’s contribution to another’s crime ... in joint enterprise cases, the wrong lies in the mutual embarkation on a crime’:  HCA 58; (2006) 81 ALJR 439, 444 . But his response is simply that, concerning EJCE, ‘[t]he crime mutually embarked upon was a different lesser crime’: McNamara, above n 124, 114. Simester does not argue, however, that the EJCE offender has embarked upon the further crime. As noted above, he instead argues that such an offender is blameworthy enough to be convicted of murder because of his/her embarkation on the foundational offence with foresight of that further crime.
 Commentators have noted the resemblance between EJCE and the constructive murder rule: see, eg, Sanford H Kadish, ‘Reckless Complicity’ (1997) 87(2) Journal of Criminal Law and Criminology 369, 376.
 Guyora Binder, Felony Murder (Stanford University Press, 2012) chs 1–4.
 Ibid 9.
 Binder, above n 25, 1036.
 Robert Hayes and F L Feld, ‘Is the Test for Extended Common Purpose Over-Extended?’  UNELawJl 2; (2009) 4(2) University of New England Law Journal 17, 17–18, 20–1.
 Criminal Law Consolidation Act 1935 (SA) s 12A (other than abortion); Crimes Act 1900 (NSW) s 18(1)(a).
 See above n 130.
 David Lanham, ‘Johns v The Queen’  MelbULawRw 7; (1980) 12(3) Melbourne University Law Review 425, 426.
 For similar views, see Krebs, above n 26, 599; Simon Gardner, ‘Case Comment: Joint Enterprise’ (1998) 114 (April) Law Quarterly Review 202, 205.
 Simester, ‘The Mental Element in Complicity’, above n 13, 599.
 Andrew Ashworth, ‘A Change of Normative Position: Determining the Contours of Culpability in Criminal Law’ (2008) 11(2) New Criminal Law Review 232, 255.
 Binder, above n 25, 1036.
 Binder, above n 133, 9–11.
 Binder, above n 25, 1036.
 Though similar reasoning does also feature in Binder’s account: Binder, above n 133, 75–7.
 Simester, ‘The Mental Element in Complicity’, above n 13, 599–600 (citations omitted).
 See, eg, R v Jarmain  KB 74, 80.
  1 AC 1, 25 (‘Powell’).
 R v van Beelen (1973) 4 SASR 353, 403; R v Ryan  VicRp 76;  VR 553, 563–4.
 See, eg, Crimes Act 1900 (NSW) s 18(1)(a).
 See, eg, Ryan v The Queen  HCA 2; (1967) 121 CLR 205.
 (1887) 16 Cox CC 311, 313.
 See, eg, R v Sharah (1992) 30 NSWLR 292.
  HCA 2; (1967) 121 CLR 205.
 Prue Bindon, ‘The Case for Felony Murder’  FlinJlLawRfm 9; (2006) 9(2) Flinders Journal of Law Reform 149, 180. See also David Crump, ‘Reconsidering the Felony Murder Rule in Light of Modern Criticisms: Doesn’t the Conclusion Depend upon the Particular Rule at Issue?’ (2009) 32(3) Harvard Journal of Law and Public Policy 1155, 1165–1170, 1174–5.
  AC 290. Concerning this furore, see, eg, Lord Irvine, ‘Intention, Recklessness and Moral Blameworthiness: Reflections on the English and Australian Law of Criminal Culpability’  SydLawRw 1; (2001) 23(1) Sydney Law Review 5, 7–10.
  HCA 14; (1963) 111 CLR 610, 633.
 A further problem with requiring subjective fault, Bindon thinks, is that some who are intuitively murderers — for example, the appellant in R v Moloney  UKHL 4;  AC 905 — cannot be proved actually to have realised that their conduct carried a risk of causing death: Bindon, above n 158, 169. But Bindon’s intuitions are not as universal as she thinks. Contrary perhaps to what she implies,
Mr Moloney was ultimately not convicted of murder; rather, the House of Lords substituted a manslaughter conviction. On remission, the Court of Appeal ordered his immediate release from custody. He had been imprisoned for 3½ years: Glanville Williams, ‘The Mens Rea for Murder: Leave it Alone’ (1989) 105 (July) Law Quarterly Review 387, 396.
  HCA 22; (1985) 156 CLR 464.
 Ibid 469.
 As Chief Justice J J Doyle, observes, two exceptions are probably Mabo v Queensland [No 2]  HCA 23; (1992) 175 CLR 1 and Dietrich v The Queen  HCA 57; (1992) 177 CLR 292 (‘Dietrich’): ‘Judicial Law Making —
Is Honesty the Best Policy?’  AdelLawRw 5; (1995) 17 Adelaide Law Review 161, 197–9. But both decisions were highly controversial (see, eg, Justice Dyson Heydon, ‘Judicial Activism and the Death of the Rule of Law’  OtaLawRw 2; (2004) 10(4) Otago Law Review 493, 510), and since Sir Anthony Mason’s retirement as Chief Justice in 1995, the High Court has been far less willing to risk causing such controversy: see, eg, Chief Justice Murray Gleeson, ‘Judicial Legitimacy’ (2000) 20(1) Australian Bar Review 4; Fiona Wheeler and John Williams, ‘“Restrained Activism” in the High Court of Australia’ in Brice Dickson (ed), Judicial Activism in Common Law Supreme Courts (Oxford University Press, 2007) 19, 57–8.
 Lord Devlin, ‘Judges and Lawmakers’ (1976) 39(1) Modern Law Review 1, 8–10; Lord Bingham, ‘The Judge as Lawmaker: An English Perspective’ in Lord Bingham, The Business of Judging: Selected Essays and Speeches (Oxford University Press, 2000) 25, 31, quoting Lord Reid, ‘The Judge as Law Maker’ (1972) 12 Journal of Society of Public Teachers of Law 22, 23; Lord Dyson, ‘Are the Judges Too Powerful?’ (Speech delivered at the Bentham Association, London, 12 March 2014); Dietrich  HCA 57; (1992) 177 CLR 292, 319–20 (Brennan J); Justice Michael McHugh, ‘The Law-Making Function of the Judicial Process — Part II’ (1988) 62(2) Australian Law Journal 116, 122;
Sir Anthony Mason, ‘Future Directions in Australian Law’  MonashULawRw 6; (1987) 13(3) Monash University Law Review 149, 158–9. Of course, despite these statements, Sir Anthony Mason and some of his colleagues were occasionally willing to intervene where they lacked clear community support, but this seems not to be true of the current High Court: see above n 165.
 Lord Devlin, above n 166, 9–10.
 Jaensch v Coffey  HCA 52; (1984) 155 CLR 549, 599–600 (Deane J); Justice Michael McHugh, ‘The Judicial Method’ (1999) 73(1) Australian Law Journal 37, 48.
 Chief Justice Doyle, above n 165, 207. Cf n 30, however.
 Above nn 166–9.
 Justice Virginia Bell, ‘Keeping the Criminal Law in “Serviceable Condition”: A Task for the Courts or the Parliament?’ (2016) 27(3) Current Issues in Criminal Justice 335, 342.
 Ibid 339.
 Ibid 340.
  HCA 12; (1938) 59 CLR 124 (‘Mullen’).
  AC 462.
  HCA 12; (1938) 59 CLR 124, 136.
 Bell, above n 171, 340.
  HCA 57; (1992) 177 CLR 292, 319.
 Bell, above n 171, 340.
 See text accompanying below nn 221–2. See also, eg, Chief Justice Doyle, above n 30, 94.
 The courts are arguably being no less ‘activist’ when they change the law in a manner that is consistent with community values; however, the press and public are less likely to perceive activism when they agree with a decision: see, eg: McHugh, above n 168, 42–3.
 Jogee  AC 387, 416 .
 Sir Richard Buxton, ‘Jogee: Upheaval in Secondary Liability for Murder’  (5) Criminal Law Review 324, 327. For a persuasive response to some of Sir Richard’s other claims, see Matthew Dyson, ‘Letter to the Editor’  (9) Criminal Law Review 638.
 For JENGbA’s website, see JENGbA, JENGBA – Joint Enterprise: Not Guilty by Association <http://www.jointenterprise.co/> .
 House of Commons Justice Committee, Joint Enterprise, Paper No 11, Session 2010–12 (2012) 10–11 –, .
 House of Commons Justice Committee, Joint Enterprise: Follow-Up, Paper No 4, Session 2014–15 (2014) 20 .
 Ibid 12 .
 See, eg, Emeka Egbuonu, ‘Joint Enterprise Law Criminalises Young, Black Men. It Urgently Needs Reform’, The Guardian (online), 22 September 2015 <https://www.theguardian.com/society/2015/
 See, eg, Tom Whitehead, ‘Garry Newlove Killer to Have Murder Conviction Reviewed amid Concern Over Joint Enterprise Role’, The Telegraph (online), 29 December 2014 <https://www.telegraph.co.uk/news/uknews/law-and-order/11316174/Garry-Newlove-killer-to-have-murder-conviction-reviewed-amid-concern-over-joint-enterprise-role.html>.
 ‘Is Joint Enterprise a “Lazy Law”?’, BBC (online), 24 April 2013 <http://www.bbc.com/news/uk-15299605> .
 Clive Coleman, ‘“Joint Enterprise” Prosecution Figures Released’, BBC (online), 1 April 2014 <http://www.bbc.com/news/uk-26721796> citing research in Maeve McClenaghan, Melanie McFadyean and Rachel Stevenson, ‘Joint Enterprise: An Investigation into the Legal Doctrine of Joint Enterprise in Criminal Convictions’ (Report, The Bureau of Investigative Journalism, April 2014).
 ‘Common: Jimmy McGovern Takes on “Group” Crime Rule’, BBC (online), 6 July 2014 <http://www.bbc.com/news/uk-england-28148073> .
 Camilla Horrox, ‘Alex Henry Family Condemn Joint Enterprise Law in BBC Documentary’, Getwest London (online), 8 July 2014 <http://www.getwestlondon.co.uk/news/local-news/alex-henry-family-condemn-joint-7392330> .
 Alex Henry, ‘I Might Have Been Naive, but that Doesn’t Make Me a Murderer’, Mirror (online),
27 October 2014 <http://www.mirror.co.uk/news/uk-news/joint-enterprise-law-my-brother-4520061> .
 See, eg, Lena Corner, ‘Joint Enterprise: The Legal Doctrine Which Critics Say has Caused Hundreds of Miscarriages of Justice’, Independent (online), 15 December 2014 <http://www.independent.co.uk/
 See, eg, Ben Rossington, ‘Joint Enterprise Law: ‘My Brother Was Wrong to Throw Punches but He Didn’t Murder Anyone’, Mirror (online), 27 October 2014 <http://www.mirror.co.uk/news/uk-news/joint-enterprise-law-my-brother-4520061> .
 Sir Anthony Mason, ‘Legislative and Judicial Law-Making: Can We Locate an Identifiable Boundary?’  AdelLawRw 4; (2003) 24(1) Adelaide Law Review 15, 26. See also John v Federal Commissioner of Taxation  HCA 5; (1989) 166 CLR 417, 438.
 Practice Statement (Judicial Precedent)  1 WLR 1234; Dietrich  HCA 57; (1992) 177 CLR 292, 320 (Brennan J); Lord Reid, above n 166, 23; B V Harris, ‘Final Appellate Courts Overruling Their Own “Wrong” Precedents: The Ongoing Search for Principle’ (2002) 118 (July) Law Quarterly Review 408, 413–4; J W Harris, ‘Towards Principles of Overruling — When Should a Final Court of Appeal Second Guess?’ (1990) 10(2) Oxford Journal of Legal Studies 135, 157–161.
 Justice J D Heydon, ‘Limits to the Powers of Ultimate Appellate Courts’ (2006) 122 (July) Law Quarterly Review 399, 405; Chief Justice Gleeson, above n 165, 7; Chief Justice A M Gleeson, ‘Individualised Justice — The Holy Grail’ (1995) 69(6) Australian Law Journal 421, 432; Justice Ronald Sackville, ‘Courts and Social Change’  FedLawRw 13; (2005) 33(3) Federal Law Review 373, 382.
 McHugh, above n 166, 122.
 See text accompanying below nn 221–30.
 Miller  HCA 30; (2016) 259 CLR 380, 400  (citations omitted).
  UKHL 50;  1 AC 1034, 1063 .
 Miller  HCA 30; (2016) 259 CLR 380, 400 .
 Ibid 398 .
 Gypsy Jokers Motorcycle Club Inc v The Commissioner of Police (2008) 234 CLR 532, 559 .
 The Clayton majority expressed the confident view that Simester had ‘demonstrate[d]’ that accessorial and JCE liability have different bases:  HCA 58; (2006) 81 ALJR 439, 444 .
 Smith, ‘R v Wakely’, above n 106, 121.
 Smith, above n 58, 464–5.
 Miller  HCA 30; (2016) 259 CLR 380, 399 .
 Ibid 394 .
 Such participants, it was said, ‘must be taken to have authorised or assented to ... [the further crime’s] commission even if it is their preference that it be avoided’: ibid 399  (emphasis added).
 Ibid 400  (citations omitted).
 In Miller, there was not only a risk that the Court would be criticised for being self-aggrandising if it overruled McAuliffe. As Gageler J suggested, their Honours might well also have found ‘troubling’ the ‘prospect of criticism of a court system which could proceed on an erroneous view of the common law for more than 20 years’: ibid 424 –.
 Bell, above n 171, 342.
 Whether the plurality in that case was exercising any restraint is questionable: see, eg, Clayton  HCA 58; (2006) 81 ALJR 439, 443 . See also above n 213.
 New South Wales Law Reform Commission, above n 130, 129 [4.235], 135 [4.265]–[4.267].
 Bell, above n 171, 342.
  UKHL 50;  1 AC 1034.
 There, the Crown had submitted that this was one reason why the House should not overrule Caldwell: ibid 1055 .
  UKHL 1;  AC 341, 353–4.
 The Caldwell majority additionally held that inadvertent recklessness was a sufficient mens rea for the more serious s 1(2) offence: ibid 354–5.
 Miller  HCA 30; (2016) 259 CLR 380, 401 –.
  UKHL 50;  1 AC 1034, 1055–6 – (Lord Bingham), 1058–9  (Lord Steyn).
 Ibid 1054  (Lord Bingham), 1060  (Lord Steyn).
 Ibid 1056  (Lord Bingham).
 Ibid 1055 ; see also 1063  (Lord Steyn).
 Ibid 1056  (Lord Bingham).
 Justice Virginia Bell, ‘Farewell Ceremony for the Honourable Justice Virginia Bell Upon the Occasion of Her Retirement as a Judge of the Supreme Court of New South Wales’ (Speech Delivered at the Supreme Court of New South Wales, 19 December 2008) <http://www.supremecourt.
 John Gardner, Offences and Defences: Selected Essays in the Philosophy of Criminal Law (Oxford University Press, 2007) 246.
 The difficulty of proving such ‘conditional intent’ is one commonly cited reason for adopting a less stringent mens rea for passive participants to JCEs that result in further crimes: Miller  HCA 30; (2016) 259 CLR 380, 398–9  citing Powell  1 AC 1, 14.
 See text accompanying above nn 150–51.
 To return to McHugh J’s statement (see text accompanying above n 206), even though there was no campaign against EJCE, was there any evidence that the community would reject a decision to abandon it? See also above n 30.
 See text accompanying above nn 186–202; Ben Crewe et al, ‘Joint Enterprise: The Implications of an Unfair and Unclear Law’  (4) Criminal Law Review 252.
 Shai Dothan, Reputation and Judicial Tactics: A Theory of National and International Courts (Cambridge University Press, 2015) 2.