The re-affirmation of criminal joint enterprise by the UKSC in Jogee

Joint enterprise increasingly permeates policing. The Code of Practice A which governs the police use of the power to stop and search individuals in public places, requires the police to have reasonable suspicion that a person is engaging in criminal activity before that person can be searched. In the 2015 edition, section 9A extends the meaning of ‘reasonable suspicion’ that those stopped and searched might be engaged in criminal activity to suspicion that their targets might be associated with a group another member of which might commit a crime. That provision opens the possibility of a major extension of stop and search usage (which can in itself lead to major public order disturbances), and the creation of apparent criminal association at the lowest and least visible level of police activity.

‘Joint Enterprise’ after Jogee: Reconsidering Law and Policy Conference 010916

Dr. Andrew Green (University of Sheffield), who is responsible for any expression of opinion in this article

The re-affirmation of criminal joint enterprise by the UKSC in Jogee

The joint enterprise doctrine (which holds that all who engage together in a criminal enterprise are held equally responsible for crimes committed within the scope of that enterprise) has been widely criticised in general for over criminalising some individuals convicted through its application, and in particular by the present authors for permitting and encouraging the use of tenuous evidence and unwarranted inferences to bring about the convictions of innocent people. The UKSC judgement in Jogee reverses decisions by the Privy Council in 1985 and the House of Lords in 1999 to permit the conviction of secondary participants in joint criminal enterprises on the basis of foresight of what crime a principal might commit rather than the previous intentional giving of assistance or encouragement to the principal. The change has been made to make the law consistent, and follows extensive campaigning by organisations of which Joint Enterprise: Not Guilty By Association (JENGBA) has been the most vocal. But it will reduce over criminalisation in no more than a few cases. We argue that in response to the continuing influence of police, media and government, the UKSC has found it expedient to reaffirm all the elements of the joint enterprise doctrine which make it such a powerful tool in securing the convictions of groups of people perceived to pose an existential threat to civilised society. Consequently the joint enterprise doctrine is more powerful than ever, and the injustices it causes and protests against it are likely to continue.

[This ‘abstract’ was prepared well in advance of the conference and needs changing to bring it into line with the paper that was actually delivered. References may be incomplete, for which I apologise.]

25 mins



  • Policy and law


This conference concerns both law and policy. Why policy? Because the law on joint criminal enterprise is a doctrine, in other words a supplement to the law on accessorial liability defined in the Accessories and Abettors Act 1861 and s.44 of the Serious Crime Act 2007. From 1985 until the decision of the UKSC earlier this year, appellate court judges acknowledged that the joint enterprise doctrine is necessary to implement policy.

So the obvious questions are

  1. What is the policy which joint enterprise prosecutions pursue?
  2. How is the policy implemented?
  3. Has the UKSC decision in Jogee affected the policy?


  • What is the policy which joint enterprise prosecutions pursue?


There are two possible sources of policy: justice and policing.


  • Justice


The joint enterprise law or doctrine can be applied in a wide variety of circumstances, and holding a secondary participant responsible for a crime actually committed by another may on occasion appear just – the gang boss who lets it be known he would like an enemy to be killed appears as culpable as the gang member who carries out the leader’s wish – but contested joint enterprise cases are rarely concerned with such situations.

So in what circumstances could equal responsibility for a crime committed by a primary offender be justly attributed to a secondary participant?  Before the UKSC handed down its judgement in Jogee, the minimum requirement was encouragement or assistance provided to the principal along with foresight of what crime the principal might commit; now it’s intentional encouragement or assistance. But there is no requirement that the secondary participant caused the principal to kill or cause serious injury, nor that without the assistance or encouragement the crime would not have taken place. Indeed the principal need know nothing of the assistance or encouragement provided by the secondary participant for the latter to be found guilty of committing the principal’s crime. The focus of a prosecution of a defendant alleged to be secondary participant is entirely on that person.

The culpability of the secondary participant in such circumstances is an entirely subjective one, a moral failing for which he may be justly punished with Draconian severity. This argument first appeared in Chan Wing-Siu (1985) – the case in which the joint enterprise law took a ‘wrong turn’: ‘The test of mens rea here is subjective. It is what the individual accused in fact contemplated that matters’ (p.7), along with the actus reus of participation in the criminal enterprise. The secondary participant may be justly punished for the crime committed by another because he or she intended that that crime be committed (an intention signalled by the lending of assistance or encouragement). This position was further elucidated in the Law Commission’s paper Participating in Crime where a secondary participant’s decision to join the group committed to a joint criminal enterprise was described as ‘a change of normative position’ from that of normal civilised society to that of the criminal gang. Claire McGourlay and I critiqued this concept in the article we published last year, ‘The wolfpacks in our midst.’ We also argued that the ‘change of normative position theory’ has been adopted within the criminal justice system (Mitchell [2008] EWCA Crim 2552; Green & McGourlay 2015, 290; accessed 27 August 2016). Punishment of secondary participants of a severity equal to that meted out to principals is held to be just because of their similar subjective moral position.



  • Policing


But the argument that joint enterprise law has been developed because its outcomes are just is not the justification generally given. When Lord Steyn said in Powell & English,the criminal justice system exists to control crime …’ Lord Hutton concurred: ‘The rules of the common law  … relate to practical concerns and, in relation to crimes committed in the course of joint enterprises, to the need to give effective protection to the public against criminals operating in gangs.’

Judges often express the danger in strong terms: ‘We have reached the stage where gangs of young people think that they are untouchable by the law,’ said Richard Wakerley J trying Pinnock & others in 2003. It is not an exaggeration to point out that … these gun carrying criminals are challenging the rule of law itself,’ said the President of the Queen’s Bench in Davis & Ellis, 2006.

In our article on joint enterprise, we traced the genealogy of the idea that judges bear the responsibility for countering such perceived existential threats, which includes the theories of Thomas Hobbes and Carl Schmitt and the decisionist tradition of English law. We argued that the modern developments of the joint enterprise doctrine have been designed to counter a perceived threat to the security of UK cities and towns from violent gangs or mobs.

The policy implemented by appellate court decisions is overwhelmingly for policing rather than justice.

We described some of the enormous practical benefits that the joint enterprise doctrine brings to policing.Since responsibility for a serious crime may be ascribed to a suspect whose alleged encouragement or assistance to the principal may have no effect on the latter, it is not necessary for the police to prove who the principal was. Rather, all that needs to be proved against a suspect is membership of a group, one of whose members may have committed a crime. That is easily proved with evidence of association. Contrast the ease with which such evidence may be found with the expensive and resource-intensive investigations and scientific crime scene analyses required for solving more conventional serious crimes. Further, there is a disincentive to making a discovery of who the perpetrator actually was, in that knowledge of who is alleged to be the perpetrator may lead to a requirement that conventional proof against them be found, and provides potential defence opportunities for dissociating other defendants from the alleged perpetrator.

According to the Bureau of Investigative Journalism:

The benefits to police and prosecutors are demonstrated by additional numbers of convictions for serious crimes which boost clear up and conviction rates. In consequence, the number of joint enterprise prosecutions and the numbers of individuals convicted appear to be rising rapidly. In the period 2005 to March 2014 up to 4,590 people have been prosecuted for homicide under joint enterprise, at least 17.7 per cent of all homicide prosecutions in this period (M. McClenaghan, M. McFadyean and R. Stevenson, ‘Revealed: Thousands Prosecuted Under Controversial Law of Joint Enterprise’, available at  (accessed 27 August 2016); see also Green and McGourlay 2015, 288-289).

Joint enterprise prosecutions are a police tactic which is part of the current strategy of public order policing. The policy behind this strategy is for policing, to which appellate court decisions contribute moderating evidential requirements designed to show that individual defendants convicted as secondary participants deserve severe punishment because of the normative position they have chosen to adopt. The ‘justice’ element of the policy appears to be little more than the legitimation of a policing tactic often perceived as oppressive and the cause of injustice, along with the stop and search practice to which it is now linked.


  • How is the policy implemented?
  • evidence


Those of us who work in the field of post appeal case review and investigation are frequently told by those who seek help in challenging joint enterprise convictions that they are unable to understand how the prisoner they support came to be convicted. What they cannot understand is not so much the law and doctrine relating to secondary participation, but how the evidence on which they were convicted could be considered sufficient to prove their guilt. In our wolfpacks article we gave examples of the kinds of tenuous evidence typically found in joint enterprise prosecutions: phone calls (made between alleged group members, not their content, which is unknown), the attribution and geographical location of phone handsets, CCTV recordings showing vehicle journeys made, use of cars associated with the crime at some other time, acquaintance with co-defendants, clothing or gestures interpreted as signifiers of covert gang membership (see Elliott [2010] EWCA Crim 2378; (2011) 175 JP 39), failure to provide an exculpatory explanation of presence at the scene of a crime and silence in interviews (Significantly supported by the data analysis in Patrick Williams and Becky Clarke, Dangerous associations: Joint enterprise, gangs and racism, Centre for Crime and Justice Studies 2016, 18 available at accessed 06/09/2016 and in the ’Key Findings’ extract (made available at the Conference) at

These types of evidence may prove either presence at a crime scene or association with other defendants, or both. According to the UKSC judgement in Jogee (77), Joint enterprise ‘is understood (erroneously) by some to be a form of guilt by association or of guilt by simple presence without more.’ In our Wolfpacks article, we observed:

Jurors are routinely instructed that evidence of a defendant’s presence at the scene of a crime is not in itself evidence of participation in the crime, and equally routinely assured that they may draw inferences of participation and foresight from presence at the scene. Typically, Judge Richard Wakerley told the jury trying Rahman, ‘… if you find that a particular defendant was on the scene and intended and did by his presence alone encourage the others to attack Tyrone Clarke [the murder victim], that would amount to participation in [the murder]’ (Green & McGourlay 2015, 286).

As for association, the UKSC assures us that ‘[i]t is important to emphasise that guilt of crime by mere association has no proper part in the common law’ (77) (although ‘[b]oth association and presence are likely to be very relevant evidence on the question whether assistance or encouragement was provided (11)). There appears to be a serious disjuncture here between theory and practice. Many examples of successful prosecutions which rely on proof of association can be found (Green & McGourlay 2015, 284).

The reason why evidence of association can be sufficient for a finding of guilt is first, because multiple meanings can be attributed to single items of evidence, and second, because of the inferential structures which are made possible by the evidential requirements in joint enterprise prosecutions.

[For example, Pinnock and others concerns a drive-by shooting in Sheffield. Three weeks afterwards the police stopped a car (for an unrelated reason) and recovered from it a phone handset. The driver of the car, Craig Brooks, denied ownership or use of the handset, and said it must have been left there by a passenger in the car. Records of calls made on and received by the handset associated it with co-defendants, and cell site analysis positioned it at the scene of the crime around the time when the murder was committed. Brooks was convicted as a secondary participant.]

I’ll give just one example. In Smith and others, Leonard Wilkins was convicted of the murder of a club doorman in Birmingham. ‘Can we look first at the evidence of connections between defendants?’ Judge Mitting asked the jury (16C). Wilkins had been in telephone contact with another defendant, Michael Christie, including two calls on the night of the murder, ‘but we do not know the location of Wilkins’ mobile telephone because his telephone was not seized and his records are not available.’ (21G) So, association established, but not presence. Here is the evidence of presence at the scene: Wilkins owned a BMW with the registration number J261MCT. DC Wood said that he could make out the shape and colour of a BMW in relevant CCTV footage and could ‘make out the number plate as blank, blank, blank, WCT or MCT or KCT.’ The Police National Computer revealed only two possible matches with cars, one of which was with Wilkins’ car. That is all the evidence against Leonard Wilkins.


  • Inferential structures


From that evidence, the Crown inferred (and the jury accepted) ‘that the BMW, J261 MCT, belonged to Wilkins, that it was in central Birmingham between 3.19 and 3.39 a.m. on 20 November, and that Wilkins was in it’ (17 C-E), and hence Wilkins was in the group which included Christie, and so he must have known that the group included someone who was armed with a gun and prepared to use it to inflict serious injury.

There are in these cases first level inferences from facts: an association with an object, a piece of real evidence (a handset, a car), the location of the object at a crime scene at the time when the crime was committed. From these, second level inferences are drawn – presence at the scene and/or association with others, who are themselves charged with the same crime, often on similar evidence. Second level inferences depend on an accumulation of first level inferences related by associative links, or similarity. Third level inferences – the conclusions which enable decision makers to be ‘sure’ that particular defendants are guilty – are that, as associates or members of the group, defendants were aware of the violent propensities and intentions of another associate or member of the group (who may not have been identified), and that they encouraged or assisted that person to kill or cause serious injury. The UKSC oversimplifies this complex process:

Juries frequently have to decide questions of intent (…) by a process of inference from the facts and circumstances proved. The same applies when the question is whether [a defendant], who joined with others in a venture to commit crime A, shared a common purpose or common intent (…) which included, if things came to it, the commission of crime B, the offence or type of offence with which he is charged, and which was physically committed by [another defendant]. (93)

I will argue that to require tenuous evidence of the kind I have described to prove in addition that the encouragement or assistance was intentional, as is now required following the UKSC judgement in Jogee, is a small step and one which is in any case already routinely taken in joint enterprise cases.


  • subjectivity


The tactic of prosecuting successfully by the use of minimal evidence has been made possible by the courts’ focus on the importance of subjective tests of mens rea (P. Roberts, ‘Science, Experts, and Criminal Justice’ in M. McConville and G. Wilson (eds), The Handbook of the Criminal Justice Process (Oxford University Press: Oxford, 2002) 260). Sir Robin Cooke said in Chan Wing-Siu (7):

The test of mens rea here is subjective. It is what the individual accused in fact contemplated that matters.

We should examine in particular the changing role of foresight. Until the mid 20th century, the test for intent on the part of secondary participants in joint enterprises was an objective one: whether a secondary participant ‘ought to have known [it] was not improbable’ that the principal would commit a more serious crime when engaged in a joint criminal enterprise. Then a more subjective test was substituted of whether a secondary participant foresaw that a more serious crime might be committed by the principal (Dr Turner, Russell on Crime, 150, 1964, cited by J. C. Smith, ‘Criminal liability of accessories,’ L.Q.R. 1997, 113(Jul), 456-457; see also Krebs 578-9). Lord Steyn, in Powell and English (13), not only regarded foresight as necessary for proof of guilt, as well as sufficient.

The UKSC (Jogee 56) was clearly aware when it rejected the rule that foresight was sufficient that foresight was also formerly a necessary condition for conviction of a secondary participant, but when it restated the ‘principles’ of joint enterprise it chose to omit any mention of foresight. The unstated crime which the group as a whole or any unspecified member of it might commit is that which a defendant may be held to have intentionally assisted or encouraged by joining and remaining part of that group. The commission of a crime by a group member serves to expose the shared criminal culpability already present in its members.

The move towards proof by chain of inference of the guilty state of mind of a defendant – the adoption of the normative position of a notional group prepared to engage in criminal activity – permits tenuous evidence to serve as proof of that change. The change appears to facilitate the solving of serious crimes and the maintenance of order (although arguably it does neither).


  • Has the UKSC decision in Jogee affected the policy or its implementation?


Every element of the full joint enterprise doctrinal supplement is restated in the Jogee judgement.

  1. association with a group (‘supportive presence’) is sufficient to attract secondary liability. 78
  2. An individual who joins the group thereby authorises a range of possible offences committed by others and not specifically foreseen. Doing ‘an act capable of encouraging or assisting the commission of an offence’ does not relate to a specific offence. 14
  3. Both presence at the scene of a crime and association with principals (who may not have been identified) can in themselves [and perhaps almost certainly will, when combined] constitute evidence of intentional participation.

It may be assumed that those who become members of a group change their normative position to that of the group:

  1. Paragraph 11
    ‘Most people are bolder when supported or fortified by others than when they are alone…’
  2. The individual who joins a group which may engage in criminal activity may have no deliberate, conscious desire to encourage or assist criminal activity, but that individual may, by intentional support for the group, also intentionally encourage or assist its criminal activity. 90-92
  3. Following on from (E), the group is likely to be of one mind:
    The group of young men which faces down a rival group may hope that the rivals will slink quietly away, but it may well be a perfectly proper inference that all were intending that if resistance were to be met, grievous bodily harm at least should be done.’  92
  4. Such a group needs no overt agreement. 92

The consequence of specifying in one supreme court judgement all seven of these supplements which together constitute the elements of the joint enterprise doctrine, together with the removal of a requirement of foresight of what the principal might do, constitute a reinforcement of that doctrine.

How radical is the UKSC decision in Jogee?
Was foresight ever important as sole proof of intent?
We examined trial judges’ instructions to juries concerning the role of secondary participants in 14 murder cases in which we had this information to hand, involving 62 defendants who were found guilty. In only four cases, involving 20 defendants, were juries instructed to rely on foresight rather than intent of any other kind. In the other 10 cases, juries were told that in order to reach guilty verdicts they must find that each secondary participant had assisted or encouraged the principal with intent, or that they shared the principals’ intent. [In discussion with Lord Toulson, one of the two authors of the judgement, at the Conference, he said he was not surprised that many judges had continued to instruct juries to they must find that defendants had intent to encourage or assist. He also agreed that the requirement that this instruction should be given would not need additional evidence proving intent.]

But almost all the defendants were convicted on the kinds of evidence we characterised as tenuous, and inferences which we believed to be unjustified.

This means that the tenuous evidence which permits inferences that defendants assisted or encouraged, has already been used to support inferences of intent. All that the Jogee judgement adds is a confirmation that common practice is good practice and should invariably be observed in future.


  • Why? the structural context of the development of joint enterprise law


Appellate court decisions on secondary participation are not themselves policy, but implementations of policy. The policy itself is developed by prosecutors and investigators of crimes – the CPS and the police. Thus new types of evidence, often based on technological and epistemological developments, are admitted by courts and approved in appellate court decisions. So too are the evidential structures. Second level inferences of association, the concept of groups, gangs, and the ‘wolfpacks’ of feral youths which we described can and do originate in police constructions.

Joint enterprise increasingly permeates policing. The Code of Practice A which governs the police use of the power to stop and search individuals in public places, requires the police to have reasonable suspicion that a person is engaging in criminal activity before that person can be searched. In the 2015 edition, section 9A extends the meaning of ‘reasonable suspicion’ that those stopped and searched might be engaged in criminal activity to suspicion that their targets might be associated with a group another member of which might commit a crime. That provision opens the possibility of a major extension of stop and search usage (which can in itself lead to major public order disturbances), and the creation of apparent criminal association – a second order inference – on the street, at the lowest and least visible level of police activity.

For those who, when stopped and searched on suspicion that they are members of joint criminal enterprise, cannot provide an innocent explanation for their presence near a crime scene, s.37 of the Criminal Justice and Public Order Act 1994 enables inferences to be drawn about their involvement in a crime. Of course such evidence and inferences are not sufficient on their own to support findings of guilt, but if those concerned have an association with others in similar circumstances, a criminal gang may be revealed through the use of such powers.  

It should also be noted that these policing tactics – stop and search and joint enterprise prosecutions – appear to be permeated by racism, directed as they have been disproportionately against black and other ethnic minorities.

In the Wolfpacks article, we observed that the so-called gangs targeted by joint enterprise prosecutions were frequently characterised in the right wing media as ‘wolfpacks’, their members considered less than human – ‘scum’, ‘thugs’, ‘feral rats’, ‘an army of ants.’ The feral beings that individuals are thought to become when they join a wolfpack are not open to reason, have ceased to be specimens of homo economicus (whose behaviour is rational) and have become instead homo vulpus. This description of alleged joint enterprise participants now takes on racist overtones.


Families of people convicted using joint enterprise law


James Richardson QC, editor of Archbold, in Criminal Law Week six days after the judgement was handed down, claimed that the UKSC had responded to the facts that ‘joint enterprise liability’ had become controversial and caused difficulties in the appeal system by adopting the ‘pragmatic solution’ of ‘realign[ing] the law a little so that there will be slightly fewer murder convictions, but correspondingly more manslaughter convictions.’The general welcome given to the reinstatement of intentionality must be tempered by the removal of foresight as a necessary condition for the conviction of secondary participants.

Richardson blamed ‘pressure groups’ including the intervener in the application to the UKSC, Joint Enterprise Not Guilty by Association (JENGbA) for encouraging ‘misleading media coverage’ which caused the controversy over joint enterprise. I agree with JENGBA’s principal argument, implied by the name it has adopted, that the central problem with joint enterprise law is the assumption and proof of guilt through association. For a small campaigning organisation to achieve a positive change in the law (albeit a minor one) is impressive. But the injustices caused by joint enterprise law and the dangers created by its continued use remain. No doubt this aptly named organisation will continue to protest, and so it should.

Future work
Joint enterprise prosecutions will no doubt continue to provide a fruitful source of legal commentary and political and social analysis.

The reliance of such prosecutions on minimal, tenuous evidence which can bear the weight of so much meaning demands a reconsideration of what evidence in criminal prosecutions can be.

How far will these experiments in effective and useful knowledge production be copied and integrated in other areas of interaction between state and subject?

August 2016


Student members of Sheffield University Miscarriages of Justice Review Centre at work under the supervision of Dr Andrew Green

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: