It is equally unsurprising that most of the comment has been so positive, given that the common law ‘joint enterprise’ doctrine the court was removing had been the subject of sustained and cogent criticism from academics, practitioners, members of the judiciary, campaign groups and others.

The decision clarifies both the bases of liability and the individual elements that must be required for proof of D’s liability as a secondary party.

Bases of liability

It is refreshing to have clarity on so many fundamental issues in a relatively concise and unanimous judgment:

  1. the same principles govern every form of secondary liability [76];
  2. there are only three ways in which a defendant can be liable in this context: (a) as a principal or joint principal where D has played a part in the commission of the actus reus of the offence; (b) as an accessory under the Accessories and Abettors Act 1861, s 8 where D has aided, abetted, counselled or procured P in the commission of the acts from which the crime was constructed; (c) under the Serious Crime Act 2007;
  3. there is no longer a separate category of parasitic accessory/joint enterprise liability. The Privy Council in Chan Wing-Siu v R [1985] AC 168 took a ‘wrong turn’ [87] in allowing foresight of the offence committed by the principal (P) to be sufficient mens rea in itself for the liability of the accessory (D) for that offence, under what became known as parasitic accessory liability (R v Powell; R v Daniels [1999] 1 AC 1 was also therefore wrongly decided);
  4. in secondary liability, the same principles apply whether D, by agreement with P, aids and abets P in committing one crime, or more; without prior agreement D aids and abets P in committing one or more crimes; or where, with or without prior agreement, D aids and abets P to commit one crime and P also commits a further crime in the course of doing so.

The Supreme Court was influenced to make this change for many reasons. As a matter of policy, the court was not satisfied that over the last 30 years the harsher ‘parasitic accessory liability’ regime had served as a deterrent. (See Wilson and Ormerod ‘Simply Harsh to Fairly Simple’ [2015] Crim LR 3). As a matter of principle, it was wrong to treat D’s foresight of what P might do as anything more than evidence from which a jury could infer the presence of a requisite intention. It was this which steered the Privy Council into its wrong-turning in Chan Wing-Siu. The Privy Council had been wrong to adopt D’s foresight of P’s likely offences as being, of itself, sufficient mens rea for D. As a matter of law, the correct approach is to treat D’s foresight of P’s likely conduct as evidence of D’s intent. Moreover, as a matter of practice, the law was continuing to create difficulty for trial judges and to generate appeals. Nor was the Supreme Court convinced by the soundness of the policy arguments which had proved influential in R v Powell; R v Daniels [1999] 1 AC 1 [74 and 75].

Elements of D’s liability

In all secondary liability cases, D’s actus reus is satisfied by proof that he did acts to encourage and/or assist P to commit the offence [8]. That conduct may take many forms [89]; it is not necessary to prove D’s conduct caused P to commit the offence [12]. It needs to be made clear to the jury what conduct it is that D is alleged to have participated in and how. D’s conduct in assisting, encouraging, or causing P to commit the crime may take different forms. It will usually be in the form of words and/or conduct. Merely associating with P or being present at the scene of P’s crime will not be enough; but if D intended by associating with P or being present at the scene to assist/encourage/cause P to commit the crime (eg, by contributing to the force of numbers in a hostile confrontation, or letting P know that D was there to provide back-up if needed) then D would be guilty [11], [78], and [89].

D’s mens rea is satisfied by proof that: (i) D intended to assist or encourage P; and (ii) D intended that P would have the mens rea required for the offence; and (iii) D must have knowledge of any other ‘existing facts necessary’ for P’s conduct/intended conduct to be criminal [9], [16].

Points deserving emphasis or clarification

  • Interpreting intention: In relation to proof of D’s mens rea, the significant shift is from it being sufficient for murder that D ‘…foresaw that P might intentionally cause grievous bodily harm (GBH) or kill if the circumstances arose’ to requiring that D ‘knew/intended that P will intentionally cause GBH or kill if the circumstances arose’. Proof of intention is what is now required. As elsewhere in the criminal law that is not limited to cases where D ‘desires’ or has as his ‘purpose’ the result that P commits the offence [90]. Most importantly, intention is not to be equated with foresight: ‘Foresight may be good evidence of intention but it is not synonymous with it’ [73]. It is submitted that D’s foresight of what P will do is relevant to all forms of intent be it direct, conditional, etc. Care must be taken to avoid undoing the Supreme Court’s good work in simplifying the law by embroiling the courts in an extended debate as to the relevance of various levels of foresight (foresight of a possibility, a probability, a virtual certainty, etc – see Woollin [1999] 1 AC 82. A certified question in Jogee’s case related to the level of foresight that was required in relation to the Chan Wing-Sui test – ie, foresight of harm that was more likely than not, rather than just some non-trivial foresight. Unsurprisingly in light of its conclusions, the Supreme Court did not address the question in its judgment). In light of this, it is likely that juries will be directed simply that the greater the level of risk of ‘X’ that they are sure that D foresaw, the more weight they may place on that fact in deciding whether D intended ‘X’. The evidential strength depends on the context [26]. Clarification of this from the Court of Appeal would be welcome. In interpreting intention, the court did not overrule or disapprove the decision in National Coal Board v Gamble [1959] 1 QB 11, in which Devlin J said: ‘If one man deliberately sells to another a gun to be used for murdering a third, he may be indifferent whether the third man lives or dies and interested only in the cash profit to be made out of the sale, but he can still be an aider and abettor.’ In this example, the seller was indifferent as to the murder being committed but it would, we assume, be open to a jury to convict him as having intended the offence based on his foresight.
  • Knowledge of weapons: D’s ‘knowledge or ignorance that weapons generally, or a particular weapon, is carried by P will be evidence going to what the intention of D was, and may be irresistible evidence one way or the other, but it is evidence and no more’ [26], [98]. The focus is on intent not on the knowledge of weapons [98].
  • Jury burden: One potential difficulty introduced by Jogee is there is a greater burden on the jury to take responsibility for delineating murder from manslaughter. Under the old law, the jury was explicitly directed that D’s foresight of a real possibility that P might intentionally kill or do GBH made him a murderer. Now it must be sure that D intended that P would intentionally kill or do GBH. Foresight is simply a matter for them to consider.
  • How precise must D’s knowledge/intent of P’s crime be? The principle in Maxwell v DPP for NI [1978] 1 WLR 1350 remains but is probably narrowed. D will be liable where D encourages or assists P knowing or intending that P will kill or cause GBH, not knowing which method P will employ – bomb, knife, gun, etc. The express statements from the court are that D will be liable only if he intentionally assists/encourages/causes P to commit one of a range of offences which D intends P will commit, so long as P commits an offence within that range [10], [14] and [90]. That represents a narrowing from the position under Maxwell which had been interpreted to mean that it was sufficient that P commits one of a range of offences that D had in mind as possibilities.
  • Proof that a murder occurred: The jury must be sure that someone committed the principal offence. The fact that the jury cannot be sure which of the members of a group delivered the fatal blow does not prevent murder convictions for all or any members. The judgment of the House of Lords in Rahman [2008] UKHL 45 remains good law in this respect.
  • Where P’s intent is graver than D intended: In the case of murder it is sufficient that D intended to assist or encourage P intentionally to commit GBH [95], [98]. Thus, if D intends that P intentionally cause GBH, but P intentionally kills, D is still liable for murder. This principle, from Rahman [2008] UKHL 45, is ,we submit, unaffected by the decision in Jogee, particularly given the retreat from the ‘fundamentally different’ acts test first introduced by R v English [1999] 2 AC 1.
  • Fundamentally different acts by P: Whether P committed the crime in a fundamentally different manner from what D foresaw is usually no longer of any special relevance. What matters is whether D intended that P commit the offence with the relevant mens rea. Where P commits the offence in a manner different to that which D intended, then only where P’s act amounts to ‘some overwhelming supervening act…which nobody in the defendant’s shoes could have contemplated might happen and is of such a character as to relegate his acts to history’ will D not be liable for it [97]–[98]. An easy example might be the facts of English, where E and his co-accused W attacked a police officer with fencing posts intending to cause GBH; E was chased off by another officer, at which point W produced a knife (which it was accepted E did not know he had) and used it to fatally stab the officer. A more difficult example is perhaps where P deliberately changes the intended victim of the offence. It is easy to see the problems that may arise in this scenario. Does this only apply if D has a specific victim in mind, for example? If D intends P to murder whoever is in the house and D murders the policeman who comes to investigate the burglary, is D guilty? How indeterminate can the class of victim be? It is hoped the Supreme Court’s restatement of the law is soon clarified yet further by the Court of Appeal.
  • The nature of D’s encouragement: The prosecution does not have to prove that D’s encouragement had a positive effect on P’s conduct or on the outcome: R v Calhaem [1985] QB 808 remains good law. This is sensible, as in many cases it would be impossible to prove. There might, for example, have been many supporters encouraging P so that the encouragement of a single one of them could not be shown to have made a difference.
  • Withdrawal remains uncertain: The rules applicable to withdrawal from secondary liability remain complex. For example, there is uncertainty over whether the same principles apply to spontaneous as distinct from pre-planned violence (compare Rajakumar [2013] EWCA Crim 1512 with Robinson [2000] EWCA Crim 8). Once again, the jury bears the burden of drawing boundaries. Clarification from the Court of Appeal of the terms of the ‘defence’ would be welcome.
  • Manslaughter retains a wide reach: D can be liable for manslaughter when P kills, provided D has intentionally participated in an offence which a sober and reasonable person would see as carrying some risk of some physical harm even if D has not or was incapable of personally seeing any risk of injury therefrom.


The overarching issue that arises from the judgment is what to do about those individuals who were convicted under the old law. The Supreme Court states in emphatic terms that it does not follow from the fact an individual was convicted under the old law that his conviction will now be quashed. As the Supreme Court points out, when a conviction is based upon the law as it applied at the time, the only option available to the defendant is to apply for exceptional leave to appeal. As the name suggests, this is not a routine matter. In the words of the Supreme Court, the Court of Appeal may grant leave ‘if substantial injustice be demonstrated, but it will not do so simply because the law applied has now been declared to have been mistaken’ [100]. In many cases an individual convicted under the old law may still be guilty of manslaughter. While the Supreme Court’s analysis of the law is undoubtedly correct in theory, it is clear the Court of Appeal will have to move swiftly to resolve these matters in practice. If the Court of Appeal is confident that a particular applicant would have been found guilty of manslaughter had the jury been directed under Jogee rather than the old law, the question remains whether there is a ‘substantial injustice’ based on being labelled and sentenced as a murderer. Under Sch 21 of the Criminal Justice Act 2003, the sentences for murder are much higher than they would be for manslaughter and impact for life on the offender. Will that significant difference in label and sentence constitute the ‘substantial injustice’ required for the grant of exceptional leave? To avoid a vast number of appeals the Court of Appeal may conclude not. ●

The views expressed are the authors’ personal views and not those of the Law Commission.

Contributors Prof David Ormerod QC is Law Commissioner for England and Wales and Karl Laird is lecturer in law at St John’s College Oxford.