Those who assist or encourage another to commit a crime may be liable as secondary parties and guilty of the same offence as the principal. The doctrine of secondary liability is long established and, as a matter of principle, relatively uncontroversial. Its application, however, has in practice repeatedly troubled the courts. Notwithstanding this, the Law Commission’s Review on Homicide will not seek to make proposals to reform the law relating to complicity.
The Court of Appeal’s judgment in the case of R v McLeary and Francis [2026] EWCA Crim 903 was handed down in June 2026. Francis had been present in a flat in Nottingham when McLeary fatally stabbed the deceased to the heart. There was evidence that Francis had stood in a doorway and had put his arm out when an eyewitness to the incident had tried to leave. The Court of Appeal held that the case against Francis should never have been left to the jury. Fraser LJ returned to Jogee and its rejection of guilt by mere association. On the facts of the case, Francis’ conduct in standing in the doorway was incapable, without more, of establishing that he had assisted or encouraged the murder perpetrated by McLeary. His conviction was, accordingly, quashed.
A decade has now passed since the Supreme Court’s seminal decision in R v Jogee [2016] UKSC 8. Though the doctrine is clear in the abstract, the evidential boundary between presence and participation remains remarkably difficult to draw in practice. McLeary and Francis provides a recent reminder of what is at stake. Against that background, one might ask whether the Law Commission can sensibly reform the law of homicide whilst leaving the underlying law of complicity untouched.
What Jogee changed
In 2016, the Supreme Court took the opportunity to re-appraise what had become known as joint enterprise, an expression which the court itself acknowledged “occasions some public misunderstanding.” Central to its reasoning was a rejection of guilt by mere association: such a concept, the court made clear, had no proper place in the common law.
There followed a discernible shift in terminology, giving way to more precise definitions. The CPS amended its own guidance, preferring ‘secondary liability’, whilst the Law Commission speaks of ‘complicity’. Yet the term ‘joint enterprise’ has proved resilient in practice and remains common parlance in the robing rooms across the country.
In Jogee the court acknowledged that the law had taken a wrong turn. There was a danger that joint enterprise could be misunderstood as implying liability through mere association or presence. Before Jogee, following the cases of Chan Wing-Siu [1985] AC 168 and Powell & English [1999] 1 AC 1, a second defendant could be liable for the offence committed by the first defendant where the second defendant merely foresaw the possibility that the first defendant might commit it.
Foresight that the principal might commit the offence is evidence from which such an intention might reasonably be inferred; it is not itself the necessary intention. The real and practical difficulty would be to identify when the surrounding evidence permits the jury to infer that presence has become intentional participation.
The first question was whether a defendant was, in fact, a participant, either by assisting or encouraging the commission of the crime. Participation may take many forms and may include providing support by contributing to the force of numbers in a hostile confrontation. This example is particularly relevant, it seems, to considerations of group-related street violence.
The second issue is whether the accessory intended to encourage or assist the principal to commit the crime, acting with the requisite mens rea. The thorny issue of conditional agreement is worthy of further consideration.
R v Bagge & Anor [2025] EWCA Crim 402
The 2025 decision in R v Bagge & Anor [2025] EWCA Crim 402 is another important post-Jogee authority on secondary liability for homicide in the context of group violence involving weapons.
In the early hours of 13 April 2023, a group of young men including the two applicants congregated at Pirate Studios in Tottenham where a party was taking place. There was tension at the party and weapons were shown. When the party ended, people emerged into the street at 4 am; a gun was produced in an ambush and the victim was shot in the head. Before people ran off, another person attacked the prone body with a sword-type knife, causing a gaping wound. The pathology evidence indicated that while the knife wound could have been fatal, the gunshot was treated as the substantial cause of death, likely to have caused immediate death.
The prosecution’s case rested on the existence of a conditional agreement amongst several people that, should the need arise, one or more of them would use unlawful violence against others intending at least to cause really serious harm. The shooting and the stabbing were acts carried out in furtherance of the agreement. By entering the agreement, two secondary parties assisted or encouraged the principal offender. All three men were convicted of murder.
The central issue on appeal was whether the trial judge should have upheld the submissions of no case to answer, given the alleged insufficiency of evidence from which a reasonable jury could conclude that a conditional agreement existed or that either applicant was party to it. The judge had ruled that the question in circumstantial cases was whether there was evidence on which a jury could infer guilt and it was not her function to decide whether she could reject all reasonable possibilities consistent with innocence. Matters of assessment and weight of the evidence were for the jury and not the judge. This did not shift the Galbraith test, as the judge only asked herself if a reasonable jury could, on one view of the evidence, draw the inference of guilt.
As to the principles of secondary participation, the trial judge had distilled the conduct elements in assisting or encouraging the principal’s offending and the mental element of intending to assist or encourage an attack with the intent to kill or cause really serious harm.
Returning to Jogee, secondary liability does not require an agreement and people may become secondaries via spontaneous supportive presence, words or deeds. The defendants may have joined the agreement at different points over the course of the evening. This recognised the fact that group violence agreements are often contingent, with the group arming itself and standing ready to inflict serious violence if a perceived trigger occurred, rather than having a fixed plan to attack a specific person at a specific moment.
It may be, as in the present case, that there was evidence supporting the existence of a conditional agreement, such as a mutual awareness of weaponry, and inferences could be drawn from members of the group being together during the evening; waiting together outside the venue at the end and, after the attack, running off together.
There was also an issue in this appeal as to whether the use of a gun broke the chain of liability as it represented an overwhelming supervening act. That could be recognised, as it was in Jogee, to be an act by the perpetrator which nobody could have contemplated. If other members of the group were aware that one of their number carried a gun, or some other dangerous weapon, then there could be evidence that they did contemplate its use.
Secondary liability does not require that a defendant physically or verbally assists or encourages at the moment of the principal offence, i.e. an overt act. Where a conditional agreement to use serious violence with the requisite intent is proved, the making of this agreement could satisfy the conduct element within an armed group moving in unison towards a confrontation.
R v Odenewu [2026] EWCA Crim 444
The Court of Appeal gave judgment in Odenewu in March 2026. This was a case arising out of serious multi-handed violence with knives. One defendant was seen as the principal: he did the stabbing – killing one and seriously injuring another.
The secondary parties had argued that they had neither assisted nor encouraged the stabbing. They contended that these were spontaneous individual acts rather than the product of any common purpose. There had been a previous skirmish between the rival groups only two days before and the court, rejecting a narrow view of the incident, said that the evidence as a whole should be considered.
The defendants’ conduct as part of the group and their possession of knives during the preceding days led to their participation in what was a developing confrontation. It would be open to the jury to infer their conditional intention to participate in serious violence, especially as their movements took them from Dagenham, where they had their first skirmish, to Bournemouth, where the murder and wounding occurred.
The contrast between McLeary and Francis and Odenewu seems to illustrate the evidential difficulty in the wake of Jogee. Presence even at the scene of a murder is not participation, yet participation does not need to be demonstrated by an identifiable act at the moment of the killing. In a multi-handed fight, the jury may be invited to consider the defendant’s conduct as part of a group in the context of the whole of the evidence and infer from it participation and conditional intent. The difficulty lies in determining where, evidentially, one ends and the other begins.
R v Durrell-Smith – UKSC/2026/0060
Durrell-Smith and Nubie exchanged gunfire outside a nightclub in Birmingham, with Durrell-Smith firing first. When Nubie returned fire, one of his bullets passed through Durrell-Smith and struck an innocent bystander. Durrell-Smith was convicted of section 18 wounding of the bystander despite the fact that the bullet had been fired by Nubie at Durrell-Smith. This was in the alternative to attempted murder.
The prosecution had described the exchange as a “shoot-out”, where both parties were armed and shooting at one another. At the very least, there was an implied agreement to shoot and be shot at and an intention to kill or cause serious injury, such that Durrell-Smith was assisting or encouraging Nubie in the shooting of an innocent bystander.
There was evidence adduced from inside the nightclub where the appellant, armed with a gun, was wearing gloves and wanting to avoid gunshot residue, and CCTV outside showed the parties pointing their guns towards one another as they manoeuvred.
The court's starting point was R v Gnango [2011] UKSC 59. The court considered that the 2011 decision had been handed down at a time when the law on joint enterprise had taken a wrong turn, before its correction in Jogee.
Post-2016 appeals seen in R v Morgan [2021] EWCA Crim 895 and R v Seed & Others [2024] EWCA Crim 650 fell into line with Jogee. In the latter appeal it was suggested that referring to the situation as a ‘shoot-out’ may not be helpful, as this phrase, coined out of convenience, may imply reciprocity – which is a feature of the agreement which must be proved to a jury.
In R v ARU [2024] EWCA Crim 1101 the Court of Appeal took the opportunity to confirm the analysis in Seed and Morgan, where there had been group fighting with knives. The requirement of knowledge of the other party’s possession of weapons and their intention to use them had to be a ‘virtual certainty’ for the necessary agreement to be inferred from the circumstances.
The court also envisaged that reciprocity may be more readily inferred in cases involving the exchange of gunfire as opposed to an incident involving the use of knives. The court also highlighted Lord Dyson’s earlier judgment in Gnango, where he drew a distinction between an agreement in the nature of a duel and a mere fight.
In July 2026, the Court of Appeal, in dismissing the appeal, certified that this decision involved a point of law of general public importance which ought to be considered by the Supreme Court.
Although this is not a homicide case, given the circumstances pertaining to the use of firearms it could easily have resulted in a fatality. The underlying principles are the same, and the case demonstrates that, some ten years after Jogee, the Supreme Court is being asked to consider the boundaries of secondary liability arising from reciprocal armed violence.
Gang Narratives
All of these examples of serious gang-related violence highlight the sensitivities and potential unfairness of gang narratives woven through the evidence.
R v Seed & Others [2024] EWCA Crim 650 is another recent appeal on ‘joint enterprise’. The case involved two rival groups in Brent: the Reds and the Blues, who engaged in a shoot-out. The prosecution could not identify the individual on the Blues side who fired the fatal shot killing a 27-year-old Red.
The three surviving Reds were convicted of the murder of their own associate on the basis of Gnango reasoning, i.e. “shoot at or be shot at”. Those exchanging fire were encouraging the opponent to return fire. The court held that the Gnango reasoning survived Jogee where the necessary agreement can be inferred from engaging in, assisting or encouraging shooting at one another. This agreement need not be express or made beforehand, as it could arise spontaneously and be inferred from their conduct.
Goodall, Walters and Wilson were 20, 19 and 20 years old when they were convicted in August 2017 for the joint enterprise murder of Abdul Hafideh, aged 18, in Manchester. The case concerned a longstanding rivalry between two groups in Moss Side.
Hafideh’s mistake was to have entered the rival group’s territory and, when he was spotted, a large group pursued him, set upon him and killed him. Goodall, Walters and Wilson were not convicted on the basis that either of them inflicted the fatal stab wounds, but on the basis that their pursuit and collective attack was something which they intentionally participated in by assisting or encouraging the violence.
A total of 13 defendants were tried in two separate trials. Some defendants were convicted of murder and others of manslaughter. Evidence of association and affiliation with gangs was a prominent feature of the case, going to the issue of whether they intended to assist or encourage an attack involving at least really serious harm.
This issue mattered post-Jogee, as foresight was no longer the test for secondary liability. The Crown had to establish intentional assistance or encouragement, although foresight may have been evidence upon which intention was inferred.
This case has been referred to the Criminal Cases Review Commission. The case highlights the need for safeguards to protect defendants against the risk of unfairness arising from a too-readily adopted gang narrative based on stereotypes and inappropriate labelling.
Those seeking social justice point to the unfairness encountered by black defendants facing evidence of their affiliation with gangs. The challenge in this case appears not to be the doctrine laid down in Jogee, but rather the evidential route taken by the Crown to prove intent and the obvious danger of reviving guilt by association under another name.
The Law Commission will have sound reasons for keeping general complicity outside a review confined to homicide. Its proposed tiers of homicide may allow verdicts and sentences to reflect differences in culpability more accurately. But, before the law can distinguish fairly between degrees of homicide, it must establish with equal care why each defendant is responsible for the death.
The case of McLeary and Francis shows the limits of presence, whilst Bagge and Odenewu each show how participation and conditional intent may emerge from a developing confrontation. Durrell-Smith tests the reach of liability in a reciprocal gunfire scenario. The Goodall referral sounds a warning about the evidential weight given to gang labels.
There will always need to be careful scrutiny of the route from evidence to inference: what, precisely, did this defendant do to assist or encourage this offence, and with what intent? Without that discipline, the language of Jogee may be observed while guilt by association could gradually return in practice.
Copyright & permissions
