Reforming non-consensual sexual offences in Hong Kong: How do the Law Reform Commission of Hong Kong's proposals compare with recent recommendations in other jurisdictions?

1. Introduction

‘There are … concerns’, the Law Reform Commission of Hong Kong (‘LRCHK’) observed in its 2012 Consultation Paper about rape and other non-consensual sexual offences, ‘that the existing sexual offences [in the Crimes Ordinance (Cap 200) (‘Crimes Ordinance’)] may not adequately reflect the range of non-consensual conduct which should be subject to the criminal sanction.’ ¹ In this article, we shall argue that the concerns to which the LRCHK referred were to an extent justified and that the Hong Kong legislature should give statutory effect to some of the reforms to non-consensual sexual offences that the Commission proposed in its Final Report. ² But it should not implement all of the Commission's proposals. In so arguing, we compare the LRCHK's recommendations with rape and sexual assault law reform proposals recently supported by other Law Reform Commissions – most particularly, the New South Wales Law Reform Commission ³ (‘NSWLRC’) and the Queensland Law Reform Commission ⁴ (‘QLRC’). Comparisons of this sort have the capacity to draw into focus the (un)desirability of law reform proposals, providing guidance about whether Parliament should adopt the approach that its Law Reform Commission has urged upon it; and this comparison reveals that the LRCHK's approach to the questions of consent and the withdrawal of consent is preferable to that advocated by the NSWLRC and QLRC (and therefore should be implemented in Hong Kong). Further, the LRCHK is (subject to one matter) on solid ground with its proposals about the mental element of the non-consensual offences. The same cannot be said, however, of the LRCHK's unwillingness to have the law provide much detail about the circumstances in which a person ‘freely and voluntarily agrees’ ⁵ to sexual activity; and its views about fraudulently procured sexual activity are also questionable. That said, its approach to this last matter is little different from that of the QLRC – and the NSWLRC's position also seems flawed.

This article is structured as follows.

In part 2, we briefly set out the law in Hong Kong concerning non-consensual sexual offending ⁶ and the procurement of sexual activity by threats ⁷ and false pretences. ⁸ As noted in that part, insofar as non-consensual sexual offences are concerned, there are three main issues: consent and its withdrawal; the circumstances in which a complainant does not consent; and the accused's mental state. The Crimes Ordinance says nothing about the first of these issues and very little about the second; but it does make it clear that, to be guilty of the offence of rape, a man must either actually know that the woman with whom he is having ‘unlawful sexual intercourse’ ⁹ is not consenting, or be reckless as to her consent. ¹⁰ According to the LRCHK, as well as changing the name of this offence to ‘sexual penetration without consent’, ¹¹ removing its gender-specificity ¹² and broadening its scope, ¹³ the legislature should create an ‘objective’ mental element for this crime – and for two new sexual assault offences ¹⁴ and a new offence of causing a person to engage in sexual activity without consent. ¹⁵ That is, according to the Commission, the person who can be proved unreasonably to have believed that a non-consenting complainant was consenting to the relevant sexual activity, should be convicted. ¹⁶ And while the Commission is content for the law more or less to remain in its largely reticent current form concerning the circumstances in which a person consents, ¹⁷ it has also argued that consent should be defined in the Crimes Ordinance ¹⁸ and that that statute should also expressly provide that such consent may be withdrawn at any time. ¹⁹ What should we make of these proposals?

In part 3, we argue that the LRCHK was right to advocate the adoption in Hong Kong both of a positive definition of consent and a provision stating that sexual conduct that occurs or continues after consent has been withdrawn, is non-consensual. ²⁰ It was also right to go no further than this. For while the NSWLRC has recently supported a provision that would state that consent is valid only once it is communicated, ²¹ and while both it and the QLRC believe that the law should state that ‘a person may, by words or conduct, withdraw consent’, ²² in truth neither view is sustainable. The person who is internally willing to engage in sexual activity is autonomously participating in that activity, the person who lacks such an internal willingness is not participating autonomously; and the law should recognise this.

In part 4, we deal with the circumstances in which a person does not consent to sexual activity – and, most especially, with the question of fraud. As foreshadowed above, we argue that the LRCHK's position is in one important way too broad. Contrary to that position, which is consistent in this way with that of the NSWLRC ²³ and the QLRC, ²⁴ a person who fraudulently procures ‘consent’ should not always be liable to be convicted of a serious offence. But, in another respect, it is too narrow. Contrary to the LRCHK's view ²⁵ (like that of the QLRC ²⁶ ), in fraud cases where such an accused is rightly convicted, his or her crime should always be a non-consensual sexual offence. That is because, as the NSWLRC has argued, ‘[a] person who is induced by fraud, of any kind, to participate in sexual activity cannot be said to have [consented].’ ²⁷

In part 5, we argue that the LRCHK's position concerning the mental element for non-consensual sexual offences is correct. As in NSW ²⁸ and Queensland, ²⁹ that is, a person should be guilty of rape/sexual penetration without consent if he or she lacked an honest and reasonable but mistaken belief that the complainant was consenting to sexual intercourse. ³⁰ It is true that modern Law Reform Commissions – including the NSWLRC ³¹ – have far too readily dismissed proposals to create a lesser offence of gross negligence rape, which would cover those cases where the accused had no reasonable belief in consent but offended neither knowingly nor recklessly. Such proposals are not as self-evidently unacceptable as is sometimes assumed. ³² But there are pragmatic arguments against such an approach; and to convict of sexual penetration without consent the accused whose belief in consent is unreasonable for him or her ³³ is never seriously to misstate his or her degree of criminality.

In part 6, we conclude that, while the LRCHK's proposals, if enacted, will modernise the law relating to non-consensual sexual offending, those proposals should have gone slightly further. That said, they are in some ways superior to those of the QLRC and (especially) the NSWLRC. Although those Commissions rightly rejected proposals to remove a mens rea requirement for the offences with which they were concerned, ³⁴ pressure from those who adopted such a position might have led the latter into conceptual distortion regarding the question of what consent is. ³⁵

2. The current law in Hong Kong – and the LRCHK's reform proposals

We must first briefly discuss the non-consensual sexual offences with which this article is concerned.

First, there is the offence of rape, created by s 118 of the Crimes Ordinance. If a man has penile-vaginal intercourse ³⁶ with a woman, without her consent, and knowing that she is not or might not ³⁷ be consenting, he commits this offence. Accordingly, if a jury is satisfied that a man accused of rape might have believed, however wrongly, that the complainant was consenting, it must acquit him – although, the more unreasonable the accused's belief, the less inclined a jury might be to find that he possibly held it. ³⁸

Secondly, there is the offence of non-consensual buggery, created by s 118A of the Crimes Ordinance. This, the LRCHK notes, ‘is the only non-consensual buggery offence in the Crimes Ordinance’ ³⁹ – although it elsewhere observes that assault with intent to commit buggery, created by s 118B, is another crime that involves an attack on the complainant's sexual autonomy. ⁴⁰ We can set these offences aside. Because ‘buggery’ is nothing more or less than ‘penetration of the anus’, ⁴¹ and because the LRCHK has recommended that such conduct, where it is non-consensual, be covered by a new offence of sexual penetration without consent, ⁴² the s 118A offence will become redundant if the Commission has its way. Certainly, this is what should happen: there is little justification for the gender-specific nature of the crime of rape, ⁴³ and there is every reason for criminalising non-consensual anal, vaginal and oral penetration under the same offence label. ⁴⁴ As the LRCHK has argued, ‘penile penetration of another person's anus or mouth is as severe an infringement of sexual autonomy as violation of a vagina’ and there is no reason why ‘men and women victims of non-consensual penile penetration should be treated differently.’ ⁴⁵ Indeed, as it was later persuaded to accept, the same is true of non-consensual penetration of an anus or vagina by something other than a penis. ⁴⁶ Similar remarks apply to the s 118B offence: the Commission was clearly correct to recommend its replacement with an offence that would criminalise the person who commits any crime – not just an assault – with intent to commit a sexual offence – not simply ‘buggery’. ⁴⁷

Thirdly, there is the offence of indecent assault, created by s 122(1) of the Crimes Ordinance. Section 122 does not state what the elements of this offence are, but they are clear nonetheless. If a person intentionally or recklessly (a) applies unlawful force to another person without his or her consent, or (b) causes that person to apprehend the immediate application of such non-consensual force, ⁴⁸ he or she will be guilty of indecent assault if, in addition, he or she: (i) knew of the complainant's failure to consent, or was reckless as to this matter; ⁴⁹ (ii) acted indecently (that is, in a manner that right-minded people regard as ‘an affront to the sexual modesty’ of a person ⁵⁰ ); and (iii) intended to ‘do something indecent to the [complainant].’ ⁵¹

It will be noted, then, that the complainant's non-consent is an element of both rape and indecent assault. But the Crimes Ordinance makes no attempt to define consent. Further, while it does state that it is rape if a man, by personating a married woman's husband, induces her to have penile-vaginal intercourse with him, ⁵² it provides no further detail about the circumstances in which a person does not consent to the sexual activity with which we are concerned.

Before considering the LRCHK's reform proposals, we must deal with two further matters. The first is the Crimes Ordinance's solution to what we shall describe as the Fairclough v Whipp ⁵³ problem. In other words, to adapt the facts of that case slightly, ⁵⁴ what if ‘one person invites another person’, whom s/he knows not to be consenting, ‘to touch’ him/her sexually ⁵⁵ and s/he does so? Because there has been no assault in such circumstances, there has been no indecent assault either; and yet the defendant's conduct seems clearly to warrant criminal punishment. ⁵⁶ Due to the crime created by s 119(1) of the Crimes Ordinance, in at least many such cases the accused will be guilty of a serious offence. So long as the prosecution can prove that s/he procured the ‘unlawful sexual act’ by threats or intimidation, s/he will be liable to 14 years’ imprisonment. ⁵⁷

The second thing to note is that, as well as creating this offence of procurement by threats, the Crimes Ordinance creates an offence of procuring another person ‘by false pretences or false representations to do an unlawful sexual act.’ ⁵⁸ We have just seen that a man is guilty of rape if, by personating a woman's husband, he induces her to engage in penile-vaginal intercourse with him. And, although the Crimes Ordinance does not say this, the same is true of the man who has such intercourse with a woman, knowing that she is participating because of her ‘mistake or misapprehension’ about the ‘character of the physical act that is done’ ⁵⁹ (though, seemingly, not about its purpose ⁶⁰ ) or her erroneous belief that the man is her regular sexual partner. ⁶¹ Likewise, the person who, because of any of the mistakes or states of ignorance just noted, participates in any of the conduct to which the indecent assault offence relates, has not consented. In any other case where the accused fraudulently induces the complainant to engage in sexual activity, however, the prosecution must rely on the procuring offence created by s 120(1). On its face, that provision places no restrictions on the ‘false pretences or false representations’ ⁶² to which it can apply – a point to which we shall return.

The LRCHK has recommended many reforms relating to the offences just considered, and we have already noted our approval of some of these. The buggery offences created by ss 118A and 118B should be repealed, we have contended; and the rape offence should have its scope broadened and its gendered operation removed. And there is one further proposal regarding rape that we have no hesitation in supporting. According to the LRCHK, that offence should be re-named ‘sexual penetration without consent.’ ⁶³ We agree with this because, whether or not the ‘term “rape” … carries a stigma against women’, ⁶⁴ it does tend to perpetuate the mistaken view that, when it comes to this offence, ‘[t]he use of force, rather than the absence of consent … [continues to be] the critical element.’ ⁶⁵ As the LRCHK indicates, ‘sexual penetration without consent’ better reflects ‘the substance of the offence.’ ⁶⁶ This will especially be so if the Hong Kong legislature, in accordance with the LRCHK's recommendations, has that crime apply to a far wider class of non-consensual penetrative acts than does the s 118(1) offence.

Also unobjectionable are the LRCHK's recommendations that two new sexual assault offences replace indecent assault. The first such offence would essentially involve intentional sexual touching without both the complainant's consent and a reasonable belief that s/he was consenting. ⁶⁷ The second offence would cover cases where an accused, without both consent and a reasonable belief in it, intentionally performed a sexual act that caused the complainant to apprehend ‘the use or threat of use of immediate and unlawful personal violence.’ ⁶⁸ It seems questionable to suggest, as the Commission does, that such reforms will create greater certainty (or substantially different outcomes from those produced by the indecent assault offence). ⁶⁹ Under the Commission's proposals, the jury would apply an inherently flexible objective test when assessing whether the touching, or the act, was ‘sexual’ ⁷⁰ – just as it does now when determining the ‘indecency’ question. Nevertheless, the language of ‘indecency’ is antiquated, the ‘right-minded’ people test is rather quaintly-expressed; and, as the LRCHK indicates, it is not as though the new offences would fail to criminalise any conduct that the s 122(1) offence catches. ⁷¹

This brings us to what we regard as the LRCHK's more contentious proposals. According to the Commission, the legislature should abolish the procurement by threats offence and insert into the Crimes Ordinance a new offence, modelled on that created by s 4(1) of the Sexual Offences Act 2003. A person would be guilty of this new offence if he or she intentionally caused another person to engage in non-consensual sexual activity and lacked a reasonable belief that that other person was consenting. ⁷² The issue that arises is whether this new offence would fail to catch some conduct that the s 119 offence covers. Take the accused who uses a non-violent threat to prevail upon a person to masturbate him or her. This accused is clearly guilty of the s 119 offence. But whether he or she is guilty of the LRCHK's proposed new offence depends on whether his or her threat renders the resulting sexual activity non-consensual. The LRCHK assures us that is does. ‘[I]f there is [a] threat or intimidation’, it tells us, ‘clearly there is no consent as [a] … person [who participates in sexual activity] under threat or intimidation would not be able to give consent freely and voluntarily.’ ⁷³ Because of one of the LRCHK's other proposals, however, this is not clear at all.

As we have noted above, the LRCHK believes that ‘consent’ should be defined in the Crimes Ordinance; and that that law should also state that consent may be withdrawn and that, once it is, any further sexual activity is non-consensual. ⁷⁴ It also believes that the definition of consent should resemble that which many other jurisdictions have adopted. ⁷⁵ According to such a definition, a person consents if s/he ‘freely and voluntarily agrees to the sexual activity’ and ‘has the capacity’ to do so. ⁷⁶ However, the Commission rejected the notion that such a definition is ‘too general.’ ⁷⁷ ‘[A]n elaborated definition of consent is not necessary’, it argued, before noting that ‘there are certain advantages in keeping the definition of consent broad as it is difficult to list all the possible scenarios or circumstances which constitute consent or lack of same.’ ⁷⁸ As argued below, there are difficulties with such an approach. The important point for present purposes, however, is that, under it, the Crimes Ordinance would not state that ‘consent’ given because of a threat or intimidation is no consent at all. ⁷⁹ In such circumstances, how can the LRCHK be sure that there would always be a conviction where the prosecution proves that a person used such means to cause another to engage in sexual activity? If ‘the issue of consent should be left to the decision of the judges on a case to case basis’, ⁸⁰ surely it would be open to a trier of fact in an individual case to acquit a person accused of the new offence, because, though s/he used a threat or intimidation to secure the complainant's compliance, the complainant nevertheless freely and voluntarily participated in the sexual activity. After all, many believe that a threat, say, to report a person for shoplifting, ⁸¹ or to ‘tell her fiancé that she has been a prostitute’, ⁸² or to ‘report her to the Tax Office for tax evasion’, ⁸³ is insufficient to render her participation in sexual activity other than autonomous. ⁸⁴

We must note two more of the LRCHK's proposals. The first concerns the actus reus of the offences with which we are concerned; the second concerns mens rea. We have just seen that the LRCHK has expressed the view that the Crimes Ordinance should provide little detail about the circumstances in which a person's participation in sexual activity is ‘free and voluntary.’ There are two exceptions to this. First, as we have seen, the LRCHK has proposed that the statute should provide that the person who lacks capacity to consent is not consenting. Moreover, according to the Commission, the Crimes Ordinance should provide some guidance about those persons who lack capacity – that is, those with a ‘mental condition’, those who are intoxicated and the young – and the precise circumstances in which such persons have not made a truly autonomous decision. ⁸⁵ We object to none of this. Central to modern ideas of criminal responsibility is the notion that, in some circumstances, persons in each of these categories cannot make a genuine choice. ⁸⁶ There is every reason why the criminal law should take the same approach to questions of responsibility more broadly; ⁸⁷ indeed, in certain respects, it already does so. ⁸⁸

The second exception is this: according to the LRCHK, the law should state that, in two circumstances, a complainant's false belief about a fact negates her or his apparent consent. We do have a concern about this. Somewhat inconsistently with the views that it expressed elsewhere in its Report, ⁸⁹ the Commission justifies explicitly identifying these matters, in the following way. ‘While a statutory definition of consent assists in understanding the term’, it argues, ‘it does not provide any guidance in determining whether or not consent exists in a particular case.’ ⁹⁰ Accordingly, it continues, the Crimes Ordinance should state that there is no consent, and the accused has no reasonable belief in consent, where he or she has either intentionally: ‘deceived the complainant as to the nature or purpose of the relevant sexual act’ or ‘intentionally induced [her or him] … to consent to the relevant act by impersonating [sic: personating ⁹¹ ] a person known personally to the complainant.’ ⁹² Other cases of fraud should be dealt with, the Commission thinks, by the s 120 offence, which should be retained. ⁹³ We return to this issue below. ⁹⁴

The point concerning mens rea is this. As already noted, the LRCHK believes that the mental element for sexual penetration without consent – and for the new sexual assault and causing non-consensual sexual activity offences – should no longer be ‘subjective.’ Rather, the accused who sexually penetrates another without consent, it thinks, should lack the requisite mental state only if s/he might have reasonably believed that the complainant was consenting. ⁹⁵ Further, according to the Commission, the law should state, as in certain other jurisdictions, ⁹⁶ that, when assessing the reasonableness of any belief in consent, the jury must consider ‘any steps the accused took to ascertain whether the complainant consented.’ ⁹⁷ Again, we shall return to this point. ⁹⁸

3. Consent and the withdrawal of consent

The first point that arises is whether the LRCHK was right to propose that ‘consent’ be positively defined in the Crimes Ordinance? It was. To consent to sexual activity is autonomously to participate in that sexual activity ⁹⁹ and there is every reason for the legislature to make this clear. Similarly, the LRCHK was correct to contend that the law should state the self-evident truth that the person who withdraws consent is not participating autonomously in any further sexual activity that then occurs. There should be no doubt about this matter. That said, should Hong Kong law go further than this? Specifically, should it provide that a person consents only once s/he does or says something to communicate to her or his partner her or his internal willingness to engage in the relevant sexual activity? ¹⁰⁰ And should it also state that a person's withdrawal of consent only becomes effective once s/he communicates such a withdrawal by ‘words or conduct’? ¹⁰¹

The NSWLRC would deliver an affirmative response to both these questions. ¹⁰² The QLRC would say ‘yes’ only to the second. ¹⁰³ Both are wrong. Indeed, the NSWLRC's position is not just wrong, but inconsistent. ¹⁰⁴ If, as it thinks, a person is not consenting until s/he does or says something to communicate that consent, ¹⁰⁵ surely logic requires that s/he must do exactly the same thing to withdraw consent? ‘Conduct’ would not necessarily be enough: ‘conduct’ can be an act or an omission, and if the complainant's ‘conduct’ were the latter, s/he would have done nothing to communicate her or his revocation. ¹⁰⁶

To see why both the NSWLRC and QLRC are wrong about these issues, we must remind ourselves that, as just noted, to consent is to participate autonomously. In our view, two examples – one concerning consent, the other concerning withdrawal – show that a person participates autonomously without having to communicate her or his willingness in any way, just as a person is no longer participating autonomously once s/he internally decides that s/he objects to what is taking place.

Take, first, two women lying in bed together after having engaged in sexual activity the night before. Without any communication at all, one woman begins performing cunnilingus on her drowsy companion, who considers this to be ‘the best alarm clock ever.’ ¹⁰⁷ Is she raping her? Of course she is not. For, while the other woman has neither said nor done anything to communicate her consent, her ‘autonomy is fully protected.’ ¹⁰⁸

Consider now the person who is engaging in consensual sexual activity before – unbeknownst to her or his partner – ‘freezing.’ Because this person does nothing and says nothing, and because her or his passivity fails unambiguously to communicate to her or his partner her or his changed mental attitude to the sexual activity, s/he has not withdrawn consent by ‘words or conduct.’ Yet any further sexual activity that then occurs is obviously non-consensual. This person's sexual autonomy is being violated. ¹⁰⁹

The QLRC has acknowledged the force of this latter example, ¹¹⁰ while maintaining that considerations of fairness to the accused justify its approach to the matter of withdrawal. ‘As a matter of fairness,’ it says, ‘it is necessary that the [accused] … is made aware that consent is withdrawn and given the opportunity to respond to that withdrawal by ceasing to engage in the relevant act.’ ¹¹¹ Similarly, the NSWLRC has stated that ‘[f]airness dictates that if consent has been freely and voluntarily given, its withdrawal should be communicated before a person acting on the consent that had been given could be convicted of a criminal offence.’ ¹¹² But this is to forget that, for an accused to be convicted of a non-consensual sexual offence, it is not enough for the Crown to prove that the complainant has been wronged. It must also prove that the accused was culpable. Accordingly, those who support the NSWLRC's and QLRC's approach to withdrawal must explain why it is at all unfair to convict the accused who engages in non-consensual sexual activity with a person who has withdrawn consent – and knows full well that this frozen complainant might no longer be internally willing. ¹¹³

We have just noted the NSWLRC's and QLRC's, flawed, reasons for contending that withdrawal of consent must be communicated to be effective. Why does the former think that consent, too, only becomes valid once communicated – and that, additionally, the accused must say or do something to communicate her or his consent? ‘Some submissions,’ the NSWLRC tells us, ‘argue that consent is an internal state of mind and can exist without communication.’ ¹¹⁴ But, it proceeds to say, ‘[w]e consider that consent is not just a subjective state of mind or attitude (like “willingness”), but a communicated state of mind.’ ¹¹⁵ Really? Why? The Commission provides no principled defence of its position. Rather, its stance is a purely pragmatic one. ‘It does not matter what consent really is,’ it effectively says. More important, it thinks, is that its view of consent is popular ¹¹⁶ – albeit not among legal philosophers ¹¹⁷ – and would ‘shift … the focus of the inquiry at trial.’ ¹¹⁸ ‘The question’, it continues, would become ‘whether the complainant said or did anything to communicate consent, rather than whether the complainant resisted or otherwise demonstrated an absence of consent.’ ¹¹⁹

Two things can be said in response to this. First, the law should be slow to resort to fictions in an attempt to achieve practical benefits. As Bob Williams has observed, ‘[t]he experience of the common law … is that the adoption of [such] fictions … ultimately gives rise to new problems as the[ir] internal contradictions … become apparent.’ ¹²⁰ Secondly, the law should be particularly reluctant to proceed on a fictitious basis where, as here, it seems very doubtful whether the hoped for practical benefits would be forthcoming.

To explain this latter claim, let us imagine that the NSWLRC's proposed position were in force (as is now so in NSW ¹²¹ ). In cases where the jury was satisfied that the complainant did resist, presumably it would continue to attach much significance to such resistance. After all, it would be the surest evidence that the complainant was not consenting. ¹²² Nothing would change in such trials. And nor would there be much change in cases where the jury was satisfied that the complainant did not resist.

In the latter kinds of cases, the jury would often be satisfied that the complainant, despite her or his resistance, did something at the time of the sexual activity. For example, in the case that prompted the NSWLRC's review, R v Lazarus, ¹²³ the trier of fact was satisfied that the non-resisting complainant performed acts such as getting down on her hands and knees and pointing her buttocks towards the accused. ¹²⁴ In such cases, the jury would often think it possible that, given the lack of resistance, the complainant performed such acts to communicate consent. In other words, the jury's focus on (non-)resistance, as at the moment, would lead it to acquit the accused. And if, despite the complainant's lack of resistance, the jury convicted, it would not do so because the NSWLRC's provision had changed the focus of the trial. Rather, it would do so on the basis of the same items of evidence that would prevail on it to convict in such cases at the moment. In Lazarus, for example, the trial judge found that the complainant was not consenting, because of her tearfulness immediately afterwards and her consistent complaints to friends and relatives of having had non-consensual intercourse. ¹²⁵ Under the NSWLRC's approach, her Honour would have reasoned similarly: the complainant's conduct after she left the accused's company made it clear that nothing that she did while still in his company was done to communicate her consent to sexual intercourse.

In short, by favouring the adoption of a provision that requires communication before there can be consent as a matter of law, the NSWLRC has not only effectively supported ‘criminalis[ing some factually] consensual sexual activity.’ ¹²⁶ It has also supported a provision that, however much it might appear to assist complainants, seems to have, at best, a very limited capacity to do so. The LRCHK was right not to supplement its proposed definition of ‘consent’ with a provision like that supported by the NSWLRC. And it was right, in its proposed provision about withdrawal, to refrain from providing that withdrawal must be communicated.

4. The circumstances in which a person does not consent

Was the LRCHK also right, however, to argue that, unlike the position in many other jurisdictions, ¹²⁷ the Crimes Ordinance should contain no detailed list of circumstances in which a person does not ‘freely and voluntarily’ participate in sexual activity? ¹²⁸ We believe that it was not; and, in so arguing, we agree with those ‘women affairs concern groups’ in Hong Kong who urged that, if consent were to be defined merely ¹²⁹ as ‘free and voluntary’ agreement, the Crimes Ordinance would provide insufficient guidance about what consent is. ¹³⁰ Of course, many judges who hear criminal trials have a fair understanding of when a person's conduct is and is not autonomous – and can direct juries accordingly. But what harm would be done by stating in the Crimes Ordinance that the complainant who ‘consents’, for example, because of either violence or a violent threat, ¹³¹ or because s/he is unlawfully detained, ¹³² is in fact not consenting? Provided that the list of vitiating circumstances is carefully thought through and expressly non-exhaustive, such an approach carries no danger of requiring or facilitating inflexibility. ¹³³

Moreover, there seem to be cases that should not ‘be left to the decision of the judges on a case by case basis.’ ¹³⁴ We referred above to non-violent threats. ¹³⁵ As suggested there, it is controversial which, if any, of those who ‘consent’ because of such threats are really consenting. Evidently, the LRCHK considers that none of them are: as we have seen, ¹³⁶ it contends elsewhere in its Report that there is ‘clearly’ no consent in such circumstances. ¹³⁷ But if this is so – and, certainly, the Commission's view is academically respectable ¹³⁸ – why not state this in the Crimes Ordinance? A different approach creates a real risk that cases with the same facts will be decided in different ways. One jury will find that there was no consent where, say, the complainant has had intercourse with the accused because of a threat to break off the relationship, whereas another jury will reach the opposite conclusion. ¹³⁹

Our main criticism of the LRCHK's approach, however, relates to the question of fraud. We accept that, as Rebecca Williams has recently observed, ‘[t]he question of fraudulent vitiation of consent to sexual contact is a notoriously difficult and controversial one.’ ¹⁴⁰ We also accept that some academics support an approach to this issue that is essentially the same as that adopted by the LRCHK. ¹⁴¹ Under that approach, it will be recalled, those who have induced a complainant to ‘consent’ to sexual activity by (i) deceiving her or him as to the nature or purpose of the relevant sexual act or (ii) personating someone known personally to the complainant, would be convicted of the applicable non-consensual sexual offence; other fraudulent inducements would attract s 120 liability. But the problem with this solution – which, as noted above, appeals also to the QLRC ¹⁴² – is that it is thoroughly unprincipled. It gives binding legal effect to an intuition that, however widely-held, does not withstand critical scrutiny.

The current editors of Simester & Sullivan ask us to imagine ‘V, a prostitute’ who asks D whether ‘he is an Everton supporter.’ ¹⁴³ ‘She informs him, truthfully,’ they say, ‘that she would not have sex with him if he were in that category because of her strong allegiance to Liverpool.’ ¹⁴⁴ Simester & Sullivan conclude that, if D then falsely informs V that he does not support Everton, and he and V have sexual intercourse, this episode should not be characterised as rape. ¹⁴⁵ Such a characterisation, they argue, ‘seems, in some important respects to misrepresent what is wrong about [D's conduct].’ ¹⁴⁶ In our view, however, there is no such misrepresentation. If, as Simester & Sullivan correctly observe in a footnote, ‘the essence of [rape] … is penetration without consent’, ¹⁴⁷ D has culpably committed the precise wrong that that offence targets. V's participation in the relevant sexual activity was no more ‘free and voluntary’ than is the participation of s/he who engages in such activity because of her or his mistake about the nature or purpose of the act, or the accused's identity. V consented to X, but Y is what she received. ¹⁴⁸

Now, we must acknowledge that some believe to be wrong the conclusion that we have just stated. If a person obtains sexual intercourse for him or herself by deceiving another about his or her HIV + status or intention to wear a condom, Rogers contends, s/he has not violated that other person's sexual autonomy. ¹⁴⁹ Rather, he argues, the wrong in such a case is the creation of a risk that the complainant will contract HIV or unwillingly become pregnant. ¹⁵⁰ In other words, while Rogers concedes that ‘views about what “sexual autonomy”’ means ‘may be disputed’ ¹⁵¹ , he considers that ‘the criminal law should regard [its true meaning to be] … the willingness to be used for the sexual gratification of another in a way that shows regard to one's own sexual preferences.’ ¹⁵² Under such an approach, the person who is mistaken ‘about the very fact that [s/he is] … engaging in sexual activity of any kind’ ¹⁵³ has not ‘freely and voluntarily’ agreed to participate in it. Nor has the person who, say, mistakenly believes that the penetration of her vagina is medically necessary. Neither of these individuals is willing to have sexual intercourse. But, for Rogers, a person such as V in Simester & Sullivan's example, has participated autonomously. S/he was willing to engage in the physical acts that s/he engaged in, and, according to Rogers, that is enough: it is irrelevant that V was only willing to do so on terms that D has breached.

The difficulty with Rogers's approach is that it reflects a far too narrow and formalistic view of when a person's decision-making is truly autonomous. According to it, for example, if a woman has sexual intercourse with a burglar, mistakenly thinking that he is her regular sexual partner, ¹⁵⁴ her sexual autonomy has not been violated. After all, she was willing to engage in the sexual acts that occurred. Yet, in truth, her willingness to have sexual intercourse with her partner tells us nothing about whether she has made a genuine choice to do the same with a stranger who has broken into her house. Or, to put essentially the same idea in different terms, there is no principled difference between the person who ‘consents’ to sexual activity because of a mistake as to the nature or purpose of the act, on one hand, and s/he who does so because of a mistake as to the accused's identity, on the other. ¹⁵⁵ And, once that is accepted, it follows that there is also no principled difference between any of these persons and, say, the person who ‘consents’ because s/he wrongly thinks that the accused has had a vasectomy, ¹⁵⁶ will pay her or him, ¹⁵⁷ will not deliberately ejaculate inside him or her, ¹⁵⁸ or is not HIV+. ¹⁵⁹ To use Wilde CJ's words in R v Case, in each of these cases – indeed, wherever a complainant ‘consents’ because of a mistake or misapprehension – the complainant has ‘consented to one thing’ and the accused has led her or him to do something ‘materially different, on which she [has] been prevented … from exercising her judgment and will.’ ¹⁶⁰ Because that is not a true consent, the accused has engaged in non-consensual sexual activity. ¹⁶¹

This, however, is not necessarily to say that we should invariably punish the accused for non-consensual sexual offending if s/he has the requisite mens rea. And this returns us to Simester & Sullivan. For those commentators, it seems, one reason for dealing with most such conduct as ‘procuring sexual activity by false pretences’ – and reserving rape and like offences for deceptions as to the nature or purpose of the act, or the identity of the accused – is that non-consensual sexual offences are ‘very serious offences.’ ¹⁶² That is, according to this argument, however ‘unpleasant’ ¹⁶³ D's conduct in the above example ¹⁶⁴ is, it is not serious enough to amount to rape. In our view, there is something in this argument – although, with that said, Simester & Sullivan never explicitly refer to the matter that we have in mind, and in fact never seem to get beyond the claim that, because they personally do not regard D as a rapist, neither should the law.

For us, the personal intuitions of the editors of Simester & Sullivan do not matter, but those of the public more broadly do. In other words, we believe that what is important is that the vast majority of people would agree with Simester & Sullivan that, whatever D's sins, he has not acted sufficiently culpably to be labelled a rapist. ¹⁶⁵ This is important, we think, because the law must not remain oblivious to such popular intuitions – however unprincipled they might be. For, if it does depart too radically from them, it risks losing the respect of those it governs. ¹⁶⁶

This is not to say that the law must slavishly give effect to what it perceives to be the community's values. As Williams has noted, ‘[t]he criminal law can lead public opinion as well as follow it’; ¹⁶⁷ and, to the extent possible, it should take a principled approach. ¹⁶⁸ It is instead merely to say this. If a person has engaged in non-consensual sexual activity with another, the law should say so, unless this would risk bringing it into disrepute – or unless there is some other very good reason for it to take a different tack.

The NSWLRC has accepted that, consistently with what we have just argued, a person does not consent to sexual activity if s/he has been induced by ‘fraud … of any kind’ to participate in it. ¹⁶⁹ But even it thinks that there should be no non-consensual sexual offence conviction where the accused's lie was about a ‘trivial’ ¹⁷⁰ matter, such as his or her ‘marital status, occupation or wealth.’ ¹⁷¹ In its view, a provision that stated that ‘a person does not consent to a sexual activity if [s/he] … participates in the sexual activity because of a fraudulent inducement’ ¹⁷² would be unlikely to apply to such trivial lies. ¹⁷³ ‘The criminal law’, it says, ‘has historically distinguished between fraud and “puffery” … [and o]ur view is that deceits such as lies about a person's marital status would be most likely to be viewed as puffery, and therefore not within the concept of fraud.’ ¹⁷⁴ But this seems wrong. As one of us has noted elsewhere, ¹⁷⁵ the authorities make it clear that there will be fraud wherever the accused's untruthful statement is one of fact rather than one of opinion; ¹⁷⁶ and statements such as ‘I am not married’, ¹⁷⁷ ‘I am a banker’, ‘I am a billionaire’ – or ‘I am not an Everton supporter’ – fall into the former category, not the latter. And it must be noted that, while the LRCHK does not accept the NSWLRC's view that most fraudulently induced sexual activity is rape/sexual penetration without consent, its approach (and that of the QLRC) is in one way similar to the NSWLRC's. Because, on their face, the procuring offences created by s 120 of the Crimes Ordinance and s 218(1) of the Criminal Code Act 1899 (Qld) criminalise all sexual activity that is procured by false pretences, ¹⁷⁸ it seems arguable that they criminalise those responsible for the sorts of ‘trivial’ deceptions mentioned above.

Where does this leave us? The answer to this question is that the LRCHK seems to have erred in two ways. Its first error was to regard s 120 liability as generally being the correct response where a person has participated in sexual activity because of the accused's deceit. This could only have been right if most such complainants are consenting – although, if that were so, there would be a real question about whether criminal liability should arise at all. ¹⁷⁹ The LRCHK's second error was to support a provision that would seemingly have the potential to visit serious criminal consequences on anyone who uses false pretences to induce another to participate in sexual activity. It is true that we should not overstate such dangers. Judging by the figures in the LRCHK's Consultation Paper, the Hong Kong authorities only use the procurement by false pretences offence quite sparingly. ¹⁸⁰ Nevertheless, it is submitted that criminal prohibitions should be drafted as precisely as possible, to ensure that there is not even the prospect of such prosecutions. ¹⁸¹ It should not be left to the Crown ¹⁸² – or the courts ¹⁸³ – to remove legislative harshness/overbreadth.

Moreover, it is not only in cases of ‘trivial lies’ where it seems clear that the accused who has fraudulently induced a complainant to ‘consent’ should nevertheless escape liability. There are cases, too, where a compelling privacy interest of the defendant makes it reasonable for him/her to mislead the complainant about a matter that might be material to her or his decision to have sexual relations with him or her – and where, therefore, there should be no liability. ¹⁸⁴ A classic example is the transgender accused. Herring's ¹⁸⁵ and the NSWLRC's ¹⁸⁶ view is that, if a person, post-‘transition’, induces another person to ‘consent’ to sexual activity by being dishonest about his or her gender history, s/he should be held liable for the relevant non-consensual sexual offence. ‘The law,’ the NSWLRC says, ‘must offer protection to complainants who are fraudulently induced to participate in sexual activity.’ ¹⁸⁷ The contrary view, with which we agree, is that, while sexual autonomy is a very important interest, it is not an absolute one. When account is taken of the ‘physical risks’ involved in disclosing past gender history, and ‘the psychological and emotional impact of disclosure’, ¹⁸⁸ it does seem that, in cases involving transgender defendants, the accused's privacy interest outweighs the complainant's sexual autonomy interest. ¹⁸⁹ Especially in circumstances where the risk of prosecutions in such cases is a real one in Hong Kong ¹⁹⁰ (or anywhere else with as broadly-worded a provision as s 120), the law should make it as clear as possible that such a person is guilty of no offence.

In short, while we agree with the LRCHK that, in most cases of fraudulently induced consent, the accused should be guilty of a serious offence, and while we agree with the NSWLRC that that offence should be a non-consensual one, the law should provide for exceptions. One way of achieving this would be for it to provide for a non-exhaustive list of mistakes or misapprehensions that vitiate consent. ¹⁹¹ These mistakes and misapprehensions could be those that have commonly arisen in the decided cases. As well as stating that this list is non-exhaustive, the relevant law could also explicitly state that there must nevertheless be no conviction where (i) a privacy or similar interest of the defendant made it reasonable for him or her to be dishonest about the relevant matter, or (ii) public policy considerations warrant such a course. ¹⁹² Moreover, to ensure that trial judges and juries are given adequate guidance as to exactly which cases fall within (i) or (ii), examples of both categories of case could be given in the relevant provision. ¹⁹³

5. The mental element

As HLA Hart observed, ‘[t]here seems a world of difference between punishing people for harm they unintentionally but carelessly cause, and punishing them for the harm which no exercise of reasonable care on their part could have avoided.’ ¹⁹⁴ Accordingly, while one of us has recently published ¹⁹⁵ a lengthy repudiation of an Australian proposal to ‘render the [honest and reasonable] mistake of fact excuse inapplicable to the issue of consent in rape and sexual assault cases’, ¹⁹⁶ we unhesitatingly accept that the person who lacks an honest and reasonable but mistaken belief that the complainant is consenting to sexual activity, is acting culpably enough to warrant being convicted of a serious offence. And we add three points.

First, when assessing the question of reasonableness, juries should take account of any features of the accused – such as youth, or limited intelligence, or mental illness – that prevented him or her from conforming to the standards required of the reasonable person who lacks any such features. In other words, as the LRCHK has accepted, ¹⁹⁷ the question must always be whether it might have been reasonable for the accused to believe that the complainant was consenting. For, as Hart also noted, ‘if our conditions of liability … are not adjusted to the capacities of the accused, then some individuals will be held liable for negligence though they could not have helped their failure to comply with the standard.’ ¹⁹⁸

Secondly, judges should normally (at least) direct juries that, when resolving the reasonableness question, they must ¹⁹⁹ (or at least may ²⁰⁰ ) consider whether the accused said or did anything to ascertain whether the complainant was consenting. For, while there will necessarily be some focus on the complainant's conduct when juries assess (a) whether s/he might have been consenting and, even if s/he was not, (b) whether the accused might reasonably have thought that s/he was, it is only fair that juries take account of the accused's (in)activity. ²⁰¹ That is, given ‘the ease with which the [sexual offence] defendant can ascertain the consent of his partner’, ²⁰² it is perfectly reasonable for judges to suggest to juries that a particular defendant's failure to take verbal or physical steps to ensure that the complainant was consenting, might, depending on the circumstances, count against him or her when those juries decide whether he or she had the requisite mens rea. The LRCHK has accepted this, too ²⁰³ – although, as we shall see, its proposal is deficient in one crucial way.

Thirdly, to say, as we have, that the accused who lacks an honest and reasonable but mistaken belief in consent has acted in a sufficiently blameworthy manner to justify his or her being convicted of a serious offence, is to say nothing about what that offence should be. Many think that, assuming that such an accused has performed the actus reus of rape/sexual penetration without consent, he or she should be convicted of that offence. ²⁰⁴ Others, however, think that there should be a hierarchy of sexual offences, and that such an accused should be liable, not for rape/sexual penetration without consent, but instead for a lesser offence of gross negligence rape/sexual penetration without consent. ²⁰⁵ Understandably enough, those in the latter group point to the law of homicide. ²⁰⁶ If it is anathema ²⁰⁷ to convict of murder a person who displays mere ‘objective culpability’ ²⁰⁸ in respect of a death that s/he has caused – that is, if the law is right to regard this person's offence as manslaughter – what reason could there be for convicting the grossly negligent rapist of the same offence as the person who actually knows that the complainant was not or might not have been consenting?

We shall deal with this last issue first. The argument just noted – namely, that if there is an offence hierarchy concerning homicide, there should also be an offence hierarchy concerning non-consensual sexual offences – is difficult to combat. Further, those who have sought to combat it, have often done so inadequately. ‘I am entirely opposed to the concept of “gross negligence rape”’, ²⁰⁹ announces Sir John Gillen, in his recent Northern Irish Report concerning sexual offences. But the adverbial flourish makes the ensuing reasoning no more persuasive. ‘My abiding concern remains that convicting for negligent rape’, he says, ‘does not fully acknowledge the egregious harm done to the victim who is raped even if negligently.’ ²¹⁰ As the Heilbron Report recognised, ²¹¹ however, arguments like this lead logically to one destination: absolute liability. For, if a blameless actor engages in non-consensual sexual activity with a person, that person is dealt an ‘egregious harm.’ Yet Sir John rightly accepts that, where rape is alleged, the Crown should be obliged to prove not just the actus reus but also that the accused was culpable – that is, lacked a reasonable belief in consent. ²¹² And so the question remains: if the accused who displays objective culpability has not acted culpably enough to be convicted of murder, why has s/he acted in so blameworthy a manner as rightly to be convicted of rape/sexual penetration without consent?

Monaghan and Mason argue that, in fact, such an accused might have acted as culpably as the offender who displays subjective fault. ‘[C]ulpability’, they rightly say, ‘is a function of more than whether a defendant's knowledge is actual, reckless or constructive.’ ²¹³ But those who contend that murder should be a subjective mens rea offence – and Monaghan and Mason do not dispute that claim – are saying nothing different from this. All they are arguing is that, all things being equal, the killer who displays objective fault regarding the result s/he has caused (or a similar one) is less culpable than the killer who intended to kill or cause grievous bodily harm, ²¹⁴ or foresaw the probability that his or her conduct would result in death. ²¹⁵

Nor, with respect, is it accurate to suggest, as Monaghan and Mason do, that a reasonable belief mental element constitutes no real departure from ‘the principle of mens rea.’ ²¹⁶ It is true, as they note, ²¹⁷ that when a jury determines whether a defendant should be excused on the basis of honest and reasonable mistake of fact, it must have regard to his or her actual state of mind: the accused can only hope to be acquitted if s/he might have had a positive belief in consent. ²¹⁸ But that has nothing to do with culpability. Insofar as that matter is concerned, the accused is judged, not according to his or her actual state of mind regarding the relevant circumstance, but on the basis of what s/he ought to have realised. There is no getting away from the fact that a reasonable belief threshold is not a subjective mens rea requirement – and (again, with respect) Monaghan and Mason's statement that ‘[n]othing in the principle of mens rea precludes the inclusion of an objective element alongside a focus on the defendant's subjective mental state’, ²¹⁹ also requires qualification. In the case to which they refer here, He Kaw Teh v The Queen, ²²⁰ the High Court of Australia drew a clear distinction between offences with a subjective fault requirement and those of which the accused will be convicted unless he or she was (if we ignore the onus of proof) acting under an honest and reasonable mistake of fact. The latter offences are a departure from the general rule that, where a statutory offence-creating provision is silent as to mens rea, the fault element is subjective. ²²¹

And so we pose the question once more: why should we support an objective fault requirement when it comes to the LRCHK's proposed non-consensual sexual offences? That is, why should we, in this instance, punish those who display objective blameworthiness in the same way as those who know of the wrong that they are, or might be, committing?

According to the Law Reform Commission of Ireland (‘LRCI’), ‘[a] failure to notice an obvious risk can show just as much of an insufficient concern for others as a conscious choice to disregard it.’ ²²² The word ‘can’, however, is important. We accept that, if an accused person of standard intelligence ²²³ gives no thought at all to the question of consent, s/he is at least as culpable as the accused who bothers to consider this question but then proceeds with sexual activity despite realising that there is a real risk that the complainant is not consenting. ²²⁴ We also accept that, if an accused's misogynous values lead him to form a genuine belief that a clearly distressed and/or resisting complainant is consenting, he has acted in as blameworthy a fashion as a person who knowingly or recklessly engages in non-consensual sexual activity. ²²⁵ But what about the accused who, having used neither violence nor threats, is faced with a non-consenting complainant who, despite her or his unwillingness, has ‘expressed no dissent or resistance’? ²²⁶ If this accused genuinely though unreasonably believes that the complainant is consenting, should s/he really be treated in the same way as the accused with subjective mens rea?

One major difficulty with an offence hierarchy for non-consensual sexual offences is that it has the potential to downgrade the liability, not only of the last-mentioned accused, but also of the other two accused just mentioned. ²²⁷ If we focus on the accused who believes fervently that ‘no’ means ‘yes’, the question is: should such a person's morally obtuseness ever ²²⁸ result in a downgrading of his or her liability? While normally we would say that it is better for liability sometimes to be downgraded undeservedly than it is for subjective and objective guilt to be punished as though it were the same thing, in the case of non-consensual sexual offences two other considerations seem to pull decisively in the opposite direction. These considerations, we believe – though with some misgivings – distinguish homicide from such offences sufficiently to make an offence hierarchy approach unjustified in the case of the latter.

First, whereas many people kill without intending to do so (or otherwise possessing subjective fault), though negligently, it would seem relatively rare for an accused to make a genuine mistake about whether his or her sexual partner is consenting. As has often been observed, the controversial issue at most rape trials is whether the complainant was consenting, not whether the accused had the requisite mens rea. ²²⁹ In such circumstances, it seems likely that, while an offence hierarchy approach to non-consensual sexual offences might ensure that the occasional accused is dealt with more fairly than s/he would be under the LRCHK's one-size-fits-all ‘reasonable belief’ standard, it might also lead to the far more frequent undeserved downgrading of liability. We have noted the problem of the morally obtuse accused. Another potential problem is that, as the LRCI has noted, ‘the lesser offence could render the more serious offence … obsolete in some cases.’ ²³⁰ The fear here is of compromise verdicts, as well as of over-cautious prosecutors choosing to prosecute the less serious offence when, in fact, the accused is guilty of the more serious one. ²³¹

Secondly, in the unusual case where a non-violent accused genuinely though unreasonably believes that a non-resisting complainant is consenting, an offence label such as ‘sexual penetration without consent’, or ‘sexual assault’, or ‘causing sexual activity without consent’, does not seriously misrepresent what s/he has done. Indeed, perhaps it does not misrepresent it at all. And while this accused does have a less culpable state of mind than others who are convicted of such offences, that can often be addressed on sentence. ²³²

We can now consider the LRCHK's proposed ‘steps’ provision, which we stated above was deficient in one crucial way. It will be recalled that, under that provision, juries would be required to have regard to ‘any steps the accused took to ascertain whether the complainant consented’, when determining whether the accused believed reasonably that the complainant was consenting. ²³³ But what is a ‘step’?

According to the NSW Court of Criminal Appeal in R v Lazarus, an accused person could take a ‘step’ within the meaning of what was then s 61HA(3)(d) of the Crimes Act 1900 (NSW) – which was worded similarly to the LRCHK's proposed provision – without performing any physical or verbal act. ²³⁴ Rather ²³⁵

a “step” for the purposes of the section … include[s] a person's consideration of, or reasoning in response to, things or events which he or she hears, observes or perceives.

The basic criticism of this was that such a reading seems inconsistent with s 61HA(3)(d)'s purpose, namely, to place more focus on what the accused did to establish that his or her partner was a willing participant. ²³⁶ That is, for so long as forming a positive belief in consent was a ‘step’, trial judges could imply to juries that, if the accused did form such a belief, this should be a point in his or her favour when those juries considered the reasonableness of that same belief. Under such an interpretation, people were given no encouragement to take more active measures to ascertain their sexual partners’ wishes.

It was because of criticisms such as this that the NSWLRC and QLRC recommended the enactment of provisions that, instead of using the language of ‘steps’, would refer to what the accused said or did to ascertain whether his or her partner was consenting. That said, the NSWLRC's and QLRC's respective proposals differed from one another. According to the NSWLRC, the law should provide that, when assessing whether the accused had the mens rea for the relevant non-consensual sexual offences, the trier of fact ‘must have regard to … whether the accused person said or did anything, at the time of the sexual activity or immediately before it, to ascertain whether the other person consented to the sexual activity, and if so, what the accused person said or did.’ ²³⁷ Under this approach, the jury would be obliged to consider the accused's verbal or physical steps, if any, regardless of the circumstances of a particular case. According to the QLRC, on the other hand, the law should provide that, when deciding whether the accused's belief in consent might have been reasonable, ‘regard may be had to what, if anything, the person said or did to ascertain whether the complainant gave consent.’ ²³⁸ Under this approach, it would be open to the jury to consider any verbal or physical steps the accused took, but it would not be obliged to do so. Moreover, trial judges would seemingly be entitled not to direct juries’ attention to the accused's failure to take such steps. Which of these approaches is preferable?

In our view, the answer to this question is unclear. The advantage of the NSWLRC's approach is that, under it, there would be less danger of trial judges ignoring the steps provision when charging juries about the mental element for non-consensual sexual offences. For, if they were to do so, they would fall into legal error. ²³⁹ The advantage of the QLRC's provision is that, under it, there is greater flexibility. Consider a case where the accused has an intellectual disability that has led him or her not to take verbal or physical measures to determine whether the complainant was a willing participant. In such a case, should a judge suggest to the jury that the accused's failure, verbally or by gesture, to check that the complainant was consenting, might count against him or her when that jury decides whether he or she possibly had a reasonable belief in consent? That said, perhaps the danger just identified is not as great as it might at first appear to be. After all, there is nothing in the NSWLRC's proposals that would prevent a trial judge, in this type of case, from warning the jury of the dangers of attaching too much significance to the accused's failure to take verbal and/or physical steps to ensure that the complainant was consenting. On balance, then, perhaps the Hong Kong legislature should insert into the Crimes Ordinance a provision along the lines of that proposed by the NSWLRC. ²⁴⁰ Certainly, it should re-work the LRCHK's proposal, which, while heading in the right direction, is not as clear as it could be about what a ‘step’ is – and which therefore could be interpreted in such a way as to frustrate its purpose.

6. Conclusion

The LRCHK's proposed reforms to the Hong Kong law concerning non-consensual sexual offending are generally sensible, balanced and progressive. While the Commission's proposals concerning fraudulently obtained consent are unprincipled and, in one sense, too broad; and while the Commission should have recommended that the Crimes Ordinance provide more detail about the circumstances in which a person does not consent to sexual activity; and while a comparison between the LRCHK's proposals and those of the NSWLRC and QLRC reveals a difficulty with the ‘steps’ provision that the LRCHK has recommended, the LRCHK's proposals, if enacted, will go a long way towards modernising the Hong Kong law in this area. And in one way the LRCHK's proposals are preferable to those of the NSWLRC and QLRC. Both of those Commissions were faced with submissions from those who argued either that the mental element for non-consensual sexual offences should be removed, ²⁴¹ or, alternatively, that the law should require the accused to ensure that the complainant was consenting if he or she were to benefit from honest and reasonable mistake of fact. ²⁴² Both rightly rejected such submissions. But did the views of such commentators nevertheless cause the NSWLRC to support certain reforms that, on their face, seem to require communication about consent if an accused is to be excused? We refer here to that Commission's claim that the law should state that a person only consents once s/he does or says something to communicate her or his internal willingness to another person. Whatever motivated the NSWLRC to advocate such a reform, the LRCHK and the QLRC were right not to support such a proposal. And, for like reasons, the LRCHK was right to refrain from stating in its proposed provision concerning withdrawal of consent, that withdrawal, to be effective, must be communicated in some way.