Skip to content

有紧急法律疑难?请立即致电 (852) 3416 1711 与本行联系。

DOJ secures ‘joint enterprise’ enforcement (Part II)

By Michael Jackson

Hong Kong, 15 April 2021: In Part I, we outlined the recent decision of the Hong Kong Court of Appeal (“HKCA”) in Tong Wai Hung ([2021] HKCA 404) in which the court ruled that the doctrine of “joint criminal enterprise” (as it has been styled by the Court of Final Appeal – hereafter “JCE”) is of general application to all offences in Hong Kong, unless excluded by statute; secondly, and more specifically, that it applies to unlawful assembly and riot, contrary to ss.18 and 19 respectively of the Public Order Ordinance (“POO”); and thirdly, that presence at the scene is not a necessary requirement of joint enterprise liability. In this part, we offer some additional thoughts on Tong, and ask what will actually need to be proved to impose liability for unlawful assembly or riot on parties to a joint enterprise (assuming the correctness of the latter two rulings).

Presumably consideration is being given to appealing the decision. In the meantime, the following comments are offered by way of critique of the legal analysis and reasoning offered and relied on in Tong and to elaborate its effect, and not to suggest that the social instability which the HKCA addressed in its judgment and perceived by it as a motivating reason for ensuring the doctrine of joint enterprise liability is available in relation to ss.18 and 19, is anything less than a legitimate and pressing concern of the criminal law.

As a prefatory comment, the general ruling of the HKCA in Tong must surely be correct. Once it is accepted that liability based on the common law doctrine of JCE remains part of Hong Kong’s criminal law, as was affirmed by the Court of Final Appeal (“CFA”) in 2016 in Chan Kam Shing ((2016) 19 HKCFAR 640), there is no reason in policy or practice to limits its application, subject only, as the HKCA acknowledged, to statutory exclusion.

Authority also supports the third ruling regarding presence, although the point is arguably not as clearly decided as the HKCA suggests in Tong. The HKCA treats the point as largely settled by the CFA in Sze Kwan Lung ((2004) 7 HKCFAR 475). However, Bokhary JA’s observation on this point in Sze is slightly qualified (at 490, [37]): “The preponderance of authority is to the effect that presence is not always necessary for criminal liability under the doctrine of joint enterprise. And even if presence were necessary, the first appellant can, in all the circumstances, be regarded as having been present” [emph. added]. Arguably, it remains open to the courts of Hong Kong to decide that presence at the scene should be required in the absence of any other physical conduct by a party to a JCE whereby he or she actually participates in or helps or encourages the commission of the JCE offence – either in all such cases, or perhaps at least in those involving extended JCE liability. All the instances itemised by the HKCA in Tong (at [56]) of “participation” other than by acts constituting the actus reus of the JCE offence, arguably involve conduct which helps or encourages those who actually commit the JCE offence as principal offenders. If no such conduct can be proved against a party to a JCE, perhaps presence should be required, in order to evidence that party’s “mutual embarkation” on or “continuing participation” in the JCE. The concern here is usually in enabling conviction of the so-called “mastermind” who sets up the JCE and then supposedly plays no further part in its commission (Tong’s example of this ((1) at [56]) refers to a mastermind who “remotely oversees the situation and gives commands or directions to the participants on the ground”, ie. he or she thus continues to play a role). Potentially the mastermind is liable for incitement or conspiracy to commit whatever offence is intended, along with any such offence actually committed as the intended object of the JCE. As presently explained, extended JCE liability also enables the mastermind to be liable for incidental or additional offences committed in executing the JCE, on the basis he or she, having joined the JCE, “participates” or “continues to participate” in such incidental offences, without the need for any further physical conduct or encouragement. This analysis is achieved by “attributing” the “acts” constituting such incidental offences to the mastermind (see further below) – provided, of course, that the mastermind “contemplated” the commission of the incidental offence (that is, contemplated both the “acts” which are attributed to the mastermind, and also the relevant mens rea which makes those acts an offence) as a “possible incident” of carrying out the JCE. A requirement of “presence” would limit this liability for such incidental offences, unless actually “intended” by the mastermind at the time the JCE was formed (or perhaps, using the UKSC’s analysis in Jogee (([2017] AC 387), [2016] 2 WLR 681), unless “conditionally intended”). Whether this mastermind scenario is realistic, or, more importantly, whether any supposed “gap” thereby arising should dictate the nature and requirements of JCE liability (at least in extended JCE cases) arguably remains for future consideration.

Ruling that joint enterprise liability applies to ss.18 and 19: over-reaching?

When we examine the HKCA’s ruling in Tong, based on statutory construction, that the common law JCE doctrine applies to the statutory offences of unlawful assembly and riot, one aspect of the Court’s analysis seems contestable. This relates to the historical survey undertaken by the HKCA in relation to the enactment of ss.18 and 19, POO. The court relies on this survey to claim that the legislature “intended”, at the time the common law offences of unlawful assembly and riot were abolished and replaced by statutory equivalents in Hong Kong in 1967 (and then amended in 1970), to “retain” the common law doctrine of JCE liability in relation to the new statutory offences. With respect, this seems to involve considerable overreach. The Court cites (at [51]) the judgment of Sachs LJ in Caird ((1970) 54 Cr App R 499, at 505), as regards the scope of liability at common law, namely, that it attached to “any person who actively encourages or promotes an unlawful assembly or riot, whether by words, by signs or by actions, or who participates in it …” [emph. added] and asserts (at [53]) that “although Sachs LJ only spoke of a traditional accessory, there is nothing in his judgment to indicate that a party to a joint enterprise should not be held equally liable as the principal,” adding (at [53]) that “for the common law offences of unlawful assembly and riot, an accessory or a party to a joint enterprise is liable as the principal” [emph. added]. Of course, the common law recognised a doctrine of “common purpose”, and the notion of a “joint enterprise” was likewise part of the common law at that time. But “common purpose” essentially provided a basis for inferring the necessary intention for secondary liability as an accessory, as stated by the UKSC in Jogee (at [87]): “The long-standing pre-Chan Wing Siu practice of inferring intent to assist from a common criminal purpose which includes the further crime, if the occasion for it were to arise, was always a legitimate one; what was illegitimate was to treat foresight as an inevitable yardstick of common purpose”.  Similarly, “joint enterprise”, as used, for example, in Anderson & Morris ([1966] 2 QB 110), referring to the liability under English common law of those who “embark on a joint enterprise” with others, did so in discussing liability based on “tacit agreement” and, as such, is generally regarded as involving the application of accessorial liability principles. So the doctrine of JCE liability, as currently adopted in Hong Kong, derives from the “wider principle” formulated in 1985 in Chan Wing Siu. Interestingly, in Anderson & Morris, counsel for the Crown, in arguing that Morris was correctly convicted of manslaughter, relied on a series of older cases which, he submitted, were “all part and parcel of a much wider principle” justifying Morris’ conviction (although he conceded that the “present trend of authority was against” his version of such a wider principle). Lord Parker, rejecting this submission, stated (at 120), “it is in the opinion of the court quite clear that the [proposed] principle is wholly out of touch with the position today”. Instead, it was Chan Wing Siu’s formulation of a “wider principle”, and the substantive changes brought about by it, more than a decade after the enactment of ss.18 and 19, which ground today’s JCE doctrine. If confirmation of Chan Wing Siu’s seminal role is sought, reference can be made to Jogee, in which the UKSC asserted (at [62]), “From our review of the authorities, there is no doubt that the Privy Council lad down a new principle in Chan Wing Siu …” Likewise, the actual ruling of the CFA in Chan Kam Shing in response to the certified question of law (at [98]) was that: “The joint criminal enterprise doctrine based on Chan Wing Siu and the cases following it, endorsed by this Court in Sze Kwan Lung, continues to apply in Hong Kong, operating alongside traditional accessorial liability principles”.

It may well be right, as the HKCA asserts in Tong (at [53]), in discussing the maintenance of public order as a justification for construing ss.18 and 19 as it does, that “all the reasons … explaining why an accessory should be held liable [for unlawful assembly or riot at common law] apply with equal force” to a person who “participates” by way of a joint enterprise in the commission of an unlawful assembly or riot, but it is doubtful that “the legislature must have intended to retain this common law rule for the statutory offences because it plainly furthers the statutory objective of maintaining public order” [emph. added], as the HKCA asserts (at [54]).

This overreach is further evident when we recall that the central ruling of the UKSC in Jogee was to declare that Chan Wing Siu had “taken a wrong turn” in formulating a “wider principle” of common law liability, and that JCE liability should be abolished completely (ie. both basic JCE and also extended JCE) from English criminal law, save as an expression of accessorial liability. The HKCA seems almost studiously to have omitted discussion of Jogee, save for two references in elaborating the law relating to accessorial liability. But, putting it at its simplest, if JCE liability, as a distinct common law basis of participation liability, is not part of English criminal law (and, to the extent it became so, was a result of “wrongly” following Chan Wing Siu), it is difficult to embrace the suggestion that application of JCE liability to ss.18 and 19 may be justified on the basis of a supposed legislative intention back in 1970 to maintain continuity.

What must be proved?

What will need to be proved to establish joint enterprise liability in relation to unlawful assembly, contrary to s.18, POO, or riot, contrary to s.19, POO? Answering this question is much more difficult than simply confirming that JCE liability applies. The impact and implications of Tong in applying JCE liability to unlawful assembly (and riot), both of which require group activity – giving rise to the so-called “corporate nature” of these offences – still have to be worked out. Care must be taken not to simplify liability, or misunderstand or misapply the law. For example, in Tong (at [57]), the HKCA itself, after  itemising various ways in which a person may “participate” in an unlawful assembly or riot, stated: “Whatever role the above participants might have played, they have all acted in concert with the principal offenders thereby sharing both their physical acts and culpability”. This appears to be offered as a broadly worded explanation or theoretical foundation of JCE liability, for the HKCA goes on to say: “If the doctrine of joint enterprise were excluded from ss.18 and 19, they would not be held liable as such, leaving a significant lacuna in the law of unlawful assembly and riot”. But this pithy summary is highly problematic, for it reproduces the exact same language adopted by the UKSC in Jogee to explain traditional accessorial liability. Specifically, in the first paragraph of the judgment, the UKSC says: “It is a fundamental principle of the criminal law that the accessory is guilty of the same offence as the principal. The reason is not difficult to see. He shares the physical act because even if it was not his hand which struck the blow, ransacked the house, smuggled the drugs or forged the cheque, he has encouraged or assisted those physical acts. Similarly he shares the culpability precisely because he encouraged or assisted the offence.” It may be right to say that a party to a joint enterprise likewise “shares both the physical acts and culpability [of the principal offenders]”, but employing this language in this manner potentially invites misunderstanding and confusion.

Rather, as mentioned earlier, “attributing” the acts of a principal offender to all other parties to the JCE has been offered as a theoretical foundation for JCE liability as a distinct basis of liability, and it is this which purportedly enables us to say that JCE liability is not derivative (unlike accessorial liability). It purports to explain how one party can be said to have “participated” in the commission of the relevant “acts” constituting the actus reus of an offence, without himself actually physically performing either those acts or any others which help or encourage the principal offender. Similarly, it purportedly provides the basis for culpability of a party to a JCE, as summarised by Ribeiro PJ in Chan Kam Shing (at [70]): “Culpability of the secondary party may therefore be seen to be based on implied authorisation of the actual perpetrator to act ‘as the instrument of the other participants to deal with the foreseen exigencies of carrying their enterprise into effect’” [emph. added]. In turn, this is also said to explain how a party to a JCE may even be liable for a more serious offence than that committed by P, the principal offender  (eg. wounding or causing GBH with intent, contrary in Hong Kong to s.17, Offences Against the Person Ordinance, even though P is only liable for wounding or inflicting GBH, contrary to s.19, OAPO), in cases in which P lacked mens rea for the more serious offence, or successfully raises a defence to the charge (see Howe ([1987] AC 417, overruling Richards ([1974] 1 QB 776), as endorsed by the CFA in Sze Kwan Lung). In such cases, D is attributed with the “acts” of the principal offender, and may then be held liable for the more serious offence if D himself had the necessary mens rea for the more serious offence, even though P may have lacked such mens rea (thus D may be liable for murder, despite P only being liable for manslaughter, if D “intended” P’s “acts” to kill or cause GBH).

In offering the following tentative framework, we draw on a five-fold list of requirements originally proposed by the English Court of Appeal in Rajakumur ([2014] 1 Cr App R 12), prior to the abolition of joint enterprise criminal liability from English common law. This list has been adapted by directing it at unlawful assembly, but also by adding explicit references to whether the relevant offence was committed “pursuant to the JCE” and whether the offence was within the “scope of the JCE” (and using D to denote the defendant whose case is being considered). Arguably, the prosecution must prove:

(1) that D was a party to a JCE “of some kind”. Of course, this requires proof that D “agreed” with others to commit an offence “of some kind”. Commonly, the object of the JCE will be unlawful assembly itself, with riot as the incidental offence. But the JCE could also relate to the commission of any other public order offence (such as s.17B of the POO (disorder in a public place)) if it involves the assembly of persons, with unlawful assembly (and perhaps also riot) as the offence “contemplated as a possible incident” of carrying out the intended offence. When we identify “unlawful assembly” as the relevant offence, we actually mean the commission of conduct bringing the parties within the terms of s.18, including proof of any state of mind (or mens rea) required to establish liability; although this does not require proof that D “knew” such conduct with such state of mind constitutes the offence of “unlawful assembly” as such.  Significantly, if D merely “joins” others in a lawful assembly, even if D contemplates (or even perhaps desires or intends) that it may deteriorate into an unlawful assembly, D has not thereby undertake with others to carry out a “common criminal purpose”, and, consequently, JCE liability cannot get off the ground. The agreement to commit an offence “of some kind” may be made in advance, or may potentially arise more spontaneously, such as during the course of a lawful assembly. If unlawful assembly itself is said to have been the object of the JCE, then it must be proved D “agreed” with at least two others to “assemble together”, and then, in pursuit of a “common purpose” (which need not itself be unlawful) to engage in conduct of the disorderly character required by s.18, intending “their” conduct to give rise reasonably to a fear of a breach of the peace, either by themselves or by others provoked by their disorderly conduct etc. If the case is one in which it is alleged the disorderly conduct was “likely” to cause such fear of a breach of peace, then D arguably ought to be proved to have “known” this likely effect (even though it is objectively assessed for the purposes of the actual commission of the offence). The corporate nature of unlawful assembly thus adds a layer of complexity to proving JCE liability, whether unlawful assembly was the intended object of the JCE, or was merely contemplated as a possible incident of carrying out whatever other JCE was agreed, which is not obviated by charging D using JCE liability. Furthermore, the “agreement” and JCE based on it must surely be circumscribed in this context if they are not to be rendered meaningless, as would happen if the JCE were said to comprise all persons assembled at the scene and pursuing the alleged common purpose of the unlawful assembly. In this sense, therefore, the necessary “agreement” which founds the JCE should not be simply understood as being the “common purpose” required for an unlawful assembly (unless the “common purpose” is itself merely committing an unlawful assembly). Acting at the same time is not the same thing as acting pursuant to a previously formed JCE. Similarly, knowledge of the plans of others (such as may be communicated by social media etc) ought not in itself be sufficient to establish a JCE between all those who have read or even responded to the social media postings;

(2) that the commission of unlawful assembly (or likewise riot) was ‘within the scope of the JCE agreed’. This will be so if the ‘agreed’ plan was either for the parties to the JCE to commit the offence of unlawful assembly (in which case this would entail basic JCE liability, with unlawful assembly as the intended offence) or if ‘acts’ giving rise to the commission of unlawful assembly by the parties to the JCE (provided committed with the necessary mens rea of unlawful assembly) were ‘contemplated’ by D as a possible incident of carrying out their intended JCE (in which case this would entail extended JCE liability);

(3) that “acts” constituting the offence of unlawful assembly (or perhaps even riot), and done with the requisite mens rea, were committed by parties to the JCE “pursuant to” the JCE. This may necessitate proof that it must be three parties to the JCE who have “assembled together” and whose conduct is of the relevant character giving rise to a reasonable fear of a breach of peace, in order to satisfy the corporate nature of the “conduct” constituting the actus reus of unlawful assembly. Otherwise, the actus reus of unlawful assembly (and the “acts” of the parties whereby it is committed, which are “attributed” to the other parties to the JCE) will not have been committed by parties to the JCE “pursuant to” the JCE. If one of the parties to the JCE “assembles” with other persons in the absence of other parties to the JCE, or assembles with other parties to the JCE but then engages in conduct of a disorderly character with others, this ought not to suffice, since it would again entail attributing “acts” of persons who are not party to the JCE to the other parties. As stated in s.18, the commission of unlawful assembly depends upon proof that “they” – that is, the three or more persons “assembled together” – “conduct themselves” in a disorderly manner, etc, so as to cause others to fear that “the persons so assembled” will commit a breach of the peace etc. Proving this in any particular case will necessarily be a “fact-sensitive question” akin to that identified by the HKCA in Tong (at [74]) in distinguishing liability for unlawful assembly from that for riot;

(4) that D “participated” in “the” JCE, in that he “played a part in putting that plan into effect”. As discussed above in relation to “presence”, in the case of a basic JCE, D may actually physically participate in the “acts” constituting the actus reus of unlawful assembly, or physically help or encourage those who perform such acts, but it seems that nothing by way of physical conduct by D beyond the initial “agreement” to commit the offence which was the object of the plan of JCE (ie. unlawful assembly) is actually necessary. Whether this is the same for “participation” in the case of an “extended JCE” arguably remains open for further discussion. The courts speak of the need for proof that the parties to the JCE “mutually embarked” on carrying out the JCE, but it remains unclear whether this necessitates any additional physical conduct on D’s part, or perhaps might possibly justify requiring at least voluntary presence;

(5) that the JCE was still in existence at the time P committed the offence of unlawful assembly (or riot); and

(6) that D had not withdrawn from the proven JCE before P committed either the intended offence or an incidental offence. The separate inclusion of this requirement arguably indicates that “participation”, or “mutual embarkation”, as required in (4), should not simply mean, or be conflated with, the need to prove D had not withdrawn from the JCE.

What is clear, therefore, is that much is still to be determined. What should also be apparent is that reliance on JCE liability to impose liability for offences against ss.18 or 19 may not be quite as simple or straightforward as it might seem at first sight.

Potential for abuse of joint enterprise liability

Finally, it is worth noting one letter to Hong Kong Lawyer magazine in April 2017, in response to an earlier note by us (reproduced in Hong Kong Lawyer, March 2017) discussing Chan Kam Shing. What was not covered by our note at that time, wrote David Southern, are “instances in which the doctrine is susceptible to abuse in Hong Kong”. Elaborating his concerns, Southern referred to “the possibility of the doctrine being used to oppress particular groups – such as those in politically sensitive groups or those involved in social movements. Whilst the doctrine has not been used in this way in Hong Kong to date, this is a worrisome possibility, as joint enterprise liability is applicable to a wide variety of offences, not just gruesome, gang-related murders”. For some, this worry appears to have been realised by the decision of the HKCA in Tong to validate the use of the joint enterprise doctrine in relation to unlawful assembly and riot. Only time will tell.

An Associate Professor with the Faculty of Law at the University of Hong Kong, Michael Jackson has been consulting for BC&C for nearly 30 years, specialising in criminal law and procedure, cybercrime, and privacy and data protection law. He can be contacted at







  • This field is for validation purposes and should be left unchanged.

Law & More: Episode 31 – Mike Riley

Hong Kong, 28 September 2023: This time on Law & Mo […]

Read more

Fancy footwork from our finance chief

Grindelwald, Switzerland, 20 September 2023: Don your b […]

Read more

Landslide exposes prosecution risk

By Alex Liu Hong Kong, 18 September 2023: Owners of lux […]

Read more

BC&C lawyers attend wills briefing

Hong Kong, 13 September 2023: Boase Cohen & Collins […]

Read more

NSL judges given sentencing flexibility

By Colin Cohen Hong Kong, 11 September 2023: A landmark […]

Read more