CRIMINAL complicity has been dramatically changed by the combined decisions of the UK Supreme Court and the Privy Council in Jogee; Ruddock [2016] UKSC 8; [2016] UKPC 7; [2016] 2 W.L.R. 681. At least since the Accessories and Abettors Act 1861, it has been settled that a person (S) who has intentionally assisted or encouraged another (P) to commit a crime has been liable to be tried, convicted, and punished as if S was a principal. For decades, there has also been a much-debated, additional form of complicity where the accomplice was “parasitically” liable for further crimes committed by P beyond the scope of a common criminal purpose shared by S and P. For that kind of liability, the accomplice need not have assisted or encouraged the further crime but need only have foreseen that it was a possible incident of the common purpose. The effect of Jogee and Ruddock is that this further form of complicity, first recognised explicitly in the Privy Council decision of Chan Wing-Siu [1985] A.C. 168 and later endorsed by the House of Lords in Powell; English [1999] 1 A.C. 1, has been shorn off the criminal law. As a result, Chan Wing-Siu directions will no longer be given to juries.
Complicity is conceptually difficult. It can also be practically difficult to establish which parties did specific physical acts or whether the acts were done with fault, which makes assessments of liability hard. Jogee and Ruddock were just such cases, and show how the lower threshold for liability in parasitic complicity was so attractive to prosecutors. Jogee was “egging [P] on” to do “something” at around the time P fought and ultimately stabbed V fatally; the only witness who gave evidence was the deceased's girlfriend. P and Jogee were convicted of murder, Jogee on the basis that at the very least he was an accomplice to an attack on V and foresaw that P might stab V with intent to cause serious harm. Ruddock was prosecuted in Jamaica on the basis that he and P had executed a common intention to steal a car and to kill the victim. The judge directed the jury that they could find this common intention where each defendant “knew that there was a real possibility the other defendant might have a particular intention and … went on to take part in [the offence].” Ruddock denied knowledge of the murder and claimed he was merely getting a lift in the car. Both appeals were successful since the trial judges had relied on (the erroneous) Chan Wing-Siu. Jogee will now be retried. Ruddock might be too.
The Supreme Court and Privy Council held that the law of complicity is now the same for all fact patterns. S must physically assist or encourage every crime for which he is to be held liable (at [76]). This general statement pushes the work of defining “assist or encourage” onto the jury. By rejecting the Chan Wing-Siu principle, the court made clear that, where S participates in crimes within a common purpose with P, S does not automatically assist or encourage other crimes that P might commit; it will be a matter of evidence of precisely what S did. Further, “neither [S's] association nor presence is necessarily proof of assistance or encouragement; it depends on the facts” (at [11]). Finally, just as S can (depending on the circumstances) withdraw from P's crime, S will not be liable where an overwhelming supervening event, which no one in S's shoes might have contemplated, occurs and nullifies S's earlier contribution to P's offending. This focus on overwhelming supervening events replaces parasitic complicity's test of whether P's actions were “fundamentally different” from those foreseen by S (see Powell; English); it may also now encompass cases where P deliberately and significantly varied the common plan (e.g. Saunders and Archer (1575) 2 Plowden 473). The test is necessarily vague, which regrettably will lead to significant uncertainty, innovative pleadings, and appeals.
The court also clarified the fault requirements for complicity. For strict liability crimes, S is liable if he assists or encourages P, intending to assist or encourage P in doing what is in fact a crime (at [99]). For crimes requiring fault, S must also intend P to act with whatever fault element is required for the crime (at [10]). Where S did not intend P to act with the required fault element, S may still be liable for any strict, constructive, or objective liability offences which P committed. Such liability highlights how English (and Jamaican) law is structured with a focus on some forms of harm, such as death. For example, while S is not liable for murder if S did not intend to assist or encourage P to cause GBH to V with intent, S could still be liable for unlawful act manslaughter. It is unclear whether the court meant to say that S would be liable for manslaughter as an accessory, rather than as a principal (at [96]). The secondary liability route is preferable: S would have assisted or encouraged an unlawful act which caused death rather than have been a “cause” of death (P's decision to do whatever killed V would presumably break any causal link between S and V's death: Kennedy (No. 2) [2007] UKHL 38; [2008] 1 A.C. 269).
However, Jogee and Ruddock give us only a partial picture of the fault element in complicity. The court does not give an answer to the long-standing problem of what S's mental state must be in relation to whether P will commit a crime. For example, S might assist P, thinking that there are 99 things P might do which are lawful, and one which is unlawful. S might even intend that, if P does the one unlawful thing, P should do so with any fault element required for the offence. It is doubtful that S is as culpable as P if P does the one unlawful thing instead of any of the other 99 lawful things contemplated by S. The ideal position is that S would be liable if he believed that P would commit the relevant crime. This was largely the position in NCB v Gamble [1959] 1 Q.B. 11, a case approved as authority for S's ignorance of the criminal law not being relevant to his liability (at [9], [99]). Jogee also endorsed Maxwell [1978] 1 W.L.R. 1350, a case where S knew that a violent terrorist attack was contemplated without knowing specifically that a bomb would be involved. P's offence only needed to be “within the range of possible offences which [S] intentionally assisted or encouraged him to commit” (at [14]).
In most cases, the prosecution will continue to allege that S intended P to do the relevant acts, and in fact intended P to commit the crime. Prosecutors thus present a simpler narrative for the jury, where the defendants were “in it together”, even if the effect is that they prove more than is strictly required to make out S's liability. However, S need not intend P to commit the crime or be “interested” in whether P commits the crime (at [90]–[91]). Previously, Chan Wing-Siu had not required that S intend to assist or encourage each crime P might commit. It achieved this by focusing on S's mental state about whether the crime would happen: it erroneously treated foresight that an offence was possible as equivalent to authorising it. The significance of foresight thus changed from being mere evidence of a common purpose to commit crime to creating liability for crimes that went beyond the parties' common purpose. This elision was rejected in Jogee and Ruddock, making it clear that foresight of an offence will only ever be evidence of intention that the offence should be committed if necessary. However, many of the convictions the foresight standard had allowed might now be achieved by relying upon the concept of conditional intention (at [90]–[95]). Almost all mental states about the future are conditional. The paradigm instances are where a defendant (D) intends to do x or achieve y even or only if something otherwise unintended occurs. Examples are robbing a bank even if there is resistance, or taking a shotgun to shoot a guard only if he resists a robbery. Simply to foresee or “endorse” a possibility of this sort, and not be dissuaded from embarking on the common purpose, is not necessarily to intend it. If these states of mind are treated as the same as an intention, we risk dramatically expanding intention so that it subsumes recklessness, which is a lower level of fault defined by unjustifiable and conscious risk-taking. However, prosecutors will probably present foresight of a possibility as evidence of an intention that the consequence should happen in order to do x or achieve y. Juries will have to make difficult decisions when they have little direct evidence of S's state of mind. In addition, an intention by S that should P act with the required fault element if P commits the crime is only a conditional intention that P have a fault element. It is not a conditional intention that P should commit a crime.
The Supreme Court and Privy Council's leaner formulation of complicity places even greater importance on how juries read difficult fact situations, particularly those involving spontaneous violence. Nonetheless, the decision is to be welcomed. One form of complicity is easier to work with than two. The law is also now more principled: the test for an accomplice's liability is set at an appropriate threshold, which is no longer significantly lower than the principal's. No longer will it be so easy for S's involvement in a death to lead to a murder conviction and the mandatory life sentence; manslaughter will more easily be an appropriate alternative on the indictment.
