1 From sometime in 2020 until his arrest on 1 March 2021, the accused messaged multiple girls on Instagram, asking them whether they needed a part-time job or cash. He would then invite them to his residence for the purpose of engaging in sexual acts with them in exchange for cash. At least four girls, between 13 and 17 years old at the material time, responded and eventually went to his residence. The accused performed sexual acts on three of these girls at his residence. These incidents formed the basis of the ten charges brought against him. He pleaded guilty to four of these charges and consented to have the remaining charges taken into consideration (“TIC”) for the purposes of sentencing.

2 One of the charges faced by the accused was for attempted statutory rape committed after the coming into force of s 167 of the Criminal Law Reform Act 2019 (Act 15 of 2019) (“CLRA”), which repealed the then s 511 of the Penal Code (Cap 224, 2008 Rev Ed) (“PC”) and replaced it with new ss 511 and 512. One effect of this amendment was that the maximum term of imprisonment for an attempted offence would no longer be limited to half of the maximum imprisonment term prescribed for the completed offence, but would henceforth be the same as that prescribed for the completed offence. This case therefore raised a novel issue concerning how the approach to sentencing for attempted rape should be adjusted in the light of this legislative development.

3 The four charges that the accused pleaded guilty to concerned three different victims, whom I shall refer to in these grounds as “V1”, “V2” and “V3” respectively. Specifically, the four proceeded charges alleged that:

(a) the accused sexually penetrated, with a vibrator, the vagina of V1 who was then below 14 years of age, thereby committing an offence under s 376A(1)(b) punishable under s 376A(3) of the PC (the “Second Charge”);

(b) the accused sexually penetrated, with a vibrator, the vagina of V2 who was then below 16 years of age, thereby committing an offence under s 376A(1)(b) punishable under s 376A(2)(b) of the PC (the “Third Charge”);

(c) the accused attempted to commit rape by attempting to penetrate, with his penis, the vagina of V1 who was then below 14 years of age, thereby committing an offence under s 375(1)(b) read with s 511 of the PC, punishable under s 375(2) read with s 512 of the PC (the “Sixth Charge”); and

(d) the accused communicated on more than one occasion with V3 who was then below 16 years of age and met her for the purpose of committing the offence of sexual penetration of a minor, thereby committing an offence under s 376E(1) punishable under s 376E(4)(b) of the PC (the “Tenth Charge”).

4 The TIC charges comprise:

(a) three other charges for sexual penetration of V1 (the “First Charge”, the “Fifth Charge” and the “Seventh Charge”);

(b) one other charge for attempted statutory rape of V1 (the “Eighth Charge”);

(c) one charge for attempted sexual penetration of V2 (the “Ninth Charge”); and

(d) one charge under s 376B(1) of the PC for obtaining for consideration the sexual services of a fourth victim (“V4”) who was then below 18 years of age (the “Fourth Charge”).

5 I begin by considering the Second and Sixth Charges as they both concern V1, before turning to consider the charges relating to V2 and V3. As between the Second and Sixth Charges, I will deal first with the Sixth Charge as that offence was committed earlier in time.

6 On 21 August 2020, the accused attempted to penetrate the vagina of V1 with his penis at his residence. V1 was 13 years old at the time. This formed the basis of the Sixth Charge.

7 In the version of the PC in force prior to 1 January 2020 (hereafter “the Pre-2020 PC”), s 511 provided, among other things, that the longest term of imprisonment that may be imposed for an attempt to commit an offence shall not exceed one-half of the longest period provided for the completed offence (“the statutory one-half limit”). In the light of this statutory one-half limit, I decided in Public Prosecutor v Udhayakumar Dhashinamoorthy Criminal Case No 43 of 2018 (“Udhayakumar”) that the punishment for the offence of attempted statutory rape should be determined by adapting the framework laid down in Ng Kean Meng Terence v Public Prosecutor [2017] 2 SLR 449 (the “Terence Ng framework”) through halving the sentencing ranges for each of the three sentencing bands of the Terence Ng framework. The approach adopted in Udhayakumar was subsequently cited with approval and applied by Woo Bih Li J (as he then was) in Public Prosecutor v Ridhaudin Ridhwan bin Bakri and others [2020] 4 SLR 790 (“Ridhaudin (sentencing)”) for determining the punishment for the offence of attempted rape.

8 With effect from 1 January 2020, upon the coming into force of s 167 of the CLRA, s 511 of the Pre-2020 PC was repealed and replaced by new ss 511 and 512 PC. Under the new s 512 PC, the statutory one-half limit was abolished, with the result that the maximum imprisonment term that may be imposed for an attempt to commit an offence is now the same as the maximum that may be imposed for the completed offence. This raises the issue of the proper approach towards sentencing for attempted rape offences committed on or after 1 January 2020.

9 Noting that there are no local cases on the sentencing of attempted rape offences committed after 1 January 2020, the Prosecution proposed an approach (the “Prosecution’s Proposed Approach”) which consisted of two stages. At the first stage, the Terence Ng framework would be used to determine what the appropriate sentence would have been if the attempt were successful and the rape offence had been completed. At the second stage, the court exercises its discretion to determine if a discount should be applied on account that the offence was not completed and, if so, the extent of the discount. Factors to be considered in determining the discount to be applied may include the steps taken by the offender towards the attempt to commit the offence, the reason that the attempt was unsuccessful, and whether a long sentence is required to reflect the seriousness of the offence and the need to protect the public.

10 The Defence disagreed with the Prosecution’s Proposed Approach and proposed two alternative approaches. The first was to continue applying the framework adopted in Ridhaudin (sentencing) by halving the sentences within each band of the Terence Ng framework (the “Defence’s First Proposed Approach”). In the alternative, the Defence proposed a second approach which involved transposing the Terence Ng framework to apply to attempted rape offences in a manner similar to how the Terence Ng framework was transposed to apply to the offence of sexual assault by penetration in Pram Nair v Public Prosecutor [2017] 2 SLR 1015 (“Pram Nair”) (the “Defence’s Second Proposed Approach”).

11 Since there are, as yet, no local precedents applying s 512 PC to the offence of attempted rape, I began by considering the legislative history of s 512 PC, followed by examining any local cases which may have applied s 512 PC to other types of offences, before turning to consider the sentencing approach taken in other jurisdictions where the maximum sentences prescribed for attempted offences are the same as the maximum sentences for the completed offences.

12 The CLRA was enacted to, among other things, give effect to the recommendations of the Penal Code Review Committee (“PCRC”) co-chaired by Ms Indranee Rajah SC, Minister in Prime Minister’s Office, Second Minister for Education and Second Minister for Finance and Mr Amrin Amin, Senior Parliamentary Secretary, Ministry of Home Affairs (see Singapore Parliamentary Debates, Official Report (6 May 2019), vol 94 (Mr Amrin Amin, Senior Parliamentary Secretary, Ministry of Home Affairs)). One of the recommendations of the PCRC was that “attempts should generally be punishable with the same prescribed punishments as the primary offence, save where express provision is made by the Penal Code or any other written law” (Penal Code Review Committee Report 2018 (the “PCRC Report”), at 201).

13  In recommending the removal of the statutory one-half limit, the PCRC Report explained that while “attempts ought not to be punished as severely as the completed offence”, the precise discount to be given “is an assessment more suited for judicial discretion rather than an arbitrary reduction in the maximum term of imprisonment as s 511 of the Penal Code currently sets out” (at 202). The PCRC Report further elaborated that (at 202):

14 The PCRC Report also cited para 2.108 of the UK Law Commission’s report on Criminal Law: Attempt, and Impossibility in Relation to Attempt, Conspiracy and Incitement (Law Com No. 102, 1980) in support of the PCRC’s recommendation. That paragraph expressed agreement with a passage from the UK Law Commission’s Working Paper No 50 Inchoate Offences: Conspiracy, Attempt and Incitement (1973), at para 113. The relevant part of that passage reads:

15  The PCRC Report went on to refer (at 202) to the case of Public Prosecutor v Huang Shiyou [2010] 1 SLR 417 (“Huang Shiyou”) in which the accused pleaded guilty and was convicted on five charges including one charge of attempted rape. In that case, the victim’s evidence, in relation to the charge of attempted rape, was that she felt something poking her vagina about ten times before the offender desisted and left (at [8]). To the PCRC, Huang Shiyou illustrates the arbitrariness of the statutory one-half limit because the culpability of the offender in that case was “not, by any measure, half of the culpability of a person who had completed the offence”.

16 From the foregoing discussion, it could be discerned that the PCRC had the following considerations in mind when making its recommendation to abolish the statutory one-half limit:

(a) as a general principle, an attempt ought not to be punished as severely as the completed offence;

(b) in punishing an attempt, the precise discount to be given from the punishment for a completed offence should be in the discretion of the court; and

(c) factors that would affect the punishment of an attempt include the culpability of the offender, the reasons the attempt did not proceed to completion, and the extent that the attempt had progressed towards completion before it was stopped or called off.

17 At this point, it might be observed that the use of the term “discount” in the PCRC Report lent some support to the Prosecution’s proposal for a two-stage approach, while the factors mentioned in [16(c)] above lent support to the list of factors which the Prosecution contended to be relevant at the second stage.

18 While there were no local authorities for the sentencing of attempted rape under s 512 PC, the Prosecution drew my attention to the case of Public Prosecutor v Merina Ng Su Yi [2022] SGDC 17 (“Merina Ng”), where the offender pleaded guilty to one cheating charge and one attempted cheating charge (committed after 1 January 2020). In that case, after picking up a misplaced wallet containing a few credit cards, the offender used one of the credit cards to successfully purchase a handbag costing $11,210. Later, she attempted to purchase an iPhone valued at $2,168 with the same credit card but the transaction was declined. She then tried paying with a different credit card from the misplaced wallet but the transaction was again declined. At that point, she decided not to proceed with the purchase and left the store. The learned Senior District Judge Bala Reddy (“SDJ Reddy”) sentenced the offender to seven months’ imprisonment for the cheating charge and one month’s imprisonment for the attempted cheating charge.

19 In determining the sentence for the attempted cheating charge, SDJ Reddy began by noting that the sentencing precedents for (completed) cheating offences showed that, where the amount cheated was between $1,200 and $6,000, custodial sentences between 6 and 12 months had been imposed (at [31]). Next, he considered that the sentencing tariffs and precedents involving cheating offences were useful in determining the sentence to be imposed for the attempted cheating offence (at [34]). Then, applying considerations similar to those for the (completed) cheating offence, SDJ Reddy decided that one month’s imprisonment would be appropriate for the attempted cheating charge, bearing in mind that the amount of $2,168 involved was only potential loss and not actual loss (at [23] and [34]).

20 In this regard, I agreed that by considering the amount of potential loss if the attempt had succeeded, and by starting with the sentencing tariffs and precedents for (completed) cheating offences before calibrating downwards on account of there being no actual loss, the approach adopted in Merina Ng lent support to the two-stage approach proposed by the Prosecution.

21 The survey of local precedents need not be confined to cases decided under s 512 PC in relation to attempted offences which were committed on or after 1 January 2020. This is because, even before 1 January 2020, there were already provisions in various statutes which expressly provided that attempts to commit certain specific offences were subject to the same punishment as the completed offences. It would therefore be helpful to consider precedents on sentencing of attempted offences under such provisions.

22 One example of such provisions is s 12 of the Misuse of Drugs Act (Cap 185, 2008 Rev Ed) (“MDA”), which provides that:

In Public Prosecutor v Mas Swan bin Adnan and another appeal [2012] 3 SLR 527 (“Mas Swan”), the offender was charged with importing 21.48g of diamorphine. His defence that he believed he was importing ecstasy pills and not diamorphine was accepted by the court. Consequently, he was acquitted of the charge of importing diamorphine and convicted on the amended charge of attempting to import ecstasy pills. In deciding on the sentence to be imposed for attempted importation of ecstasy pills, the Court of Appeal began with the precedents laying down the factors to be considered when determining the sentence for a (completed) drug trafficking offence (at [56]). The Court of Appeal went on to hold that these factors were also applicable to attempted drug trafficking offences. As one of these factors is the quantity of controlled drugs involved, the Court of Appeal held (at [57]):

23 The Court of Appeal then referred to the approach taken in two English cases. The first case is Regina v Tomasz Szmyt [2010] 1 Cr App R (S) 69 (“R v Szmyt”), where the offender was found with 1,998 harmless tablets which he thought were ecstasy tablets. The English Court of Appeal held that the proper approach was to first consider what would have been the appropriate sentence if 1,998 ecstasy tablets had been imported, and then apply a discount to account for the fact that the actual substances imported were harmless tablets. The second case is R v Magdalen Genevieve Wolin [2006] 1 Cr App R (S) 133 (“R v Wolin”), where the offender imported lignocaine, a non-prohibited substance, thinking she was importing cocaine. The court there similarly applied a discount to account for the fact that the offender was charged with an attempt and not a completed offence, and the fact that the substance imported was not a prohibited substance.

24 As for the specific sentence to be imposed on the offender in Mas Swan, the Court of Appeal began by noting that the sentencing precedents on the importation of ecstasy showed that the range of imprisonment term imposed was between seven and ten years and, if precedents on the importation of Yaba were included, the range would be between five and 13 years (at [58]). (“Yaba” was the street name of a methamphetamine-based drug which had been classified together with ecstasy as belonging to the category of “amphetamine-type stimulants” by the United Nations Office on Drugs and Crime.) Nonetheless, the Court of Appeal decided not to use the sentencing range in cases of importation of ecstasy or Yaba to determine the sentencing range in Mas Swan, and further held that no discount should be given (at [59]). This was because, unlike the cases of R v Szmyt and R v Wolin where the actual items imported were not prohibited substances, the actual substance imported in Mas Swan was diamorphine, which was a more serious controlled drug than ecstasy. In the end, the Court of Appeal decided to impose a sentence of 15 years’ imprisonment and eight strokes of the cane.

25 Although the Court of Appeal ultimately did not sentence the offender in Mas Swan using the approach taken in R v Szmyt and R v Wolin, it was not because the Court of Appeal had disapproved of the approach in these two English cases. Instead, it was clear from the judgment in Mas Swan that the Court of Appeal was prepared to follow such an approach were it not for the fact that the actual substance imported in Mas Swan was something more serious than the ecstasy pills which the offender attempted to import. This much was evident from the Court of Appeal’s remark that it was “necessary for the court to formulate an acceptable means of calculating the notional quantity of ecstasy attempted to be imported for the purpose of sentencing” (at [57]). In my view, Mas Swan’s acceptance of the approach taken in R v Szmyt and R v Wolin lent support to the Prosecution’s Proposed Approach.

26 Given the dearth of local authorities, I considered that it was useful to look also at the approach adopted in jurisdictions where attempted offences are subject to the same maximum punishment as the completed offences. In this regard, the Prosecution identified as relevant the UK, as well as the Australian jurisdictions of the Australian Capital Territory and New South Wales.

27 The provisions for attempted offences in England and Wales are set out in the Criminal Attempts Act 1981 (c 47) (UK) (“UK CAA”). Section 4(1)(b) of the UK CAA provides that a person guilty under s 1 of attempting to commit an offence (not being murder or any other offence the sentence for which is fixed by law) shall “be liable on conviction on indictment to any penalty to which he would have been liable on conviction on indictment of that offence”.

28 As already noted above, in the context of attempted drug trafficking, the English courts adopt a two-stage approach similar to the Prosecution’s Proposed Approach. In the context of attempts to commit sexual offences, the Prosecution cited the English Court of Appeal case of R v Reed and another [2021] 1 WLR 5429 (“Reed”). Reed concerned five offenders whose offences involved attempting to incite fictional children (who were actually police officers pretending to be children) into engaging in sexual activity, and one offender whose offence involved attempting to incite an actual child into engaging in sexual activity. The lower courts in these six cases applied conflicting authorities in determining the sentences for the attempted offences. Some applied the sentencing approach in Attorney General’s Reference No. 94 of 2014 (R v Baker) [2016] 4 WLR 121 (“Baker”), which was to categorise the attempted offence within Category 3 (the least severe “Harm” category) of the Sexual Offences: Definitive Guideline published by the Sentencing Council for England and Wales (the “Guideline”). Others applied the sentencing approach in R v Privett & Others [2020] EWCA Crim 557 (“Privett”) (see [31] below).

29 At this juncture, it will be useful to briefly explain the Guideline. Section 125(2) of the Coroners and Justice Act 2009 (c 25) (UK) requires that every court “must, in sentencing an offender follow any sentencing guidelines which are relevant to the offender’s case”. The Guideline uses a “Harm/Culpability” matrix with three categories of “Harm” and two bands of “Culpability” defined for each sexual offence covered by the Guideline. As to Harm, Category 1 is the most serious and Category 3 the least; and as to Culpability, the “A” band is more serious than “B”. The three categories of Harm and two bands of Culpability yield a Harm/Culpability matrix with six permutations, each with its own starting point sentence and its own sentencing range (termed “category range” in the Guideline).

30 The sentencing approach in Baker was to treat an inchoate offence as involving the least severe level of harm and this would be determinative of the starting point sentence to be applied. This was perhaps justified on the facts of Baker, which involved an offender who attempted to incite (over a series of ‘Whatsapp’ messages) a 13-year-old girl to engage in sexual penetration of the girl’s mouth with his penis, however the proposed sexual activity never took place as he eventually ended the series of ‘Whatsapp’ communications. The court in Baker rationalised that where the offence of inciting a child to engage in sexual activity punishable under s 10(1) and s 10(2) of the Sexual Offences Act 2003 (c 42) (UK) (“UK SOA”) “does not lead the child to behave in the manner incited, although the culpability is likely to be identical, the harm is necessarily less: the same is so in relation to attempts” (Baker at [32]).

31 Privett concerned four appeals where the offender in each case had been in contact with an undercover police officer who had been posing as the mother of a fictitious child. Each of them believed the intended victims were real and they arranged with the ‘mother’ to engage in sexual activity with the children. They were each charged for arranging or facilitating the commission of a child sex offence contrary to s 14 of the UK SOA. The English Court of Appeal held that the proper sentencing position under the Guideline was, first, to identify the category of harm by reference to the type of sexual activity the defendant intended, and, second, to adjust the sentence in order to ensure it is “commensurate” with, or proportionate to, the applicable starting point and range if no sexual activity had occurred (including because the victim was fictional) (Privett at [67]).

32 The issue which the English Court of Appeal posed to itself in Reed was (at [1] and [5]): What is the overarching principle to be applied for sentencing in respect of all instances of offences under the UK SOA where a sexual offence has been committed in respect of a child and no sexual activity occurs? In preferring the approach in Privett, the court in Reed held that the decision in Baker was “unsustainable” because it was “in clear contravention” of s 63 of the Sentencing Act 2020 (c 17) (UK) (“UK Sentencing Act”) (Reed at [21]), which provides:

33 The court in Reed went on to hold that the approach in Privett should apply to all child sex offences “when the defendant attempts to commit these offences or incite a child to engage in certain activity but the activity does not take place” (at [23]). The court then summarised the approach in the following terms (at [23]):

34  I spent some time analysing Reed as that was the key authority cited by the Prosecution. Even though Reed was a case dealing with inchoate child sex offences generally, it was clear from the passages quoted above that the two-stage approach adopted in Reed applied also to the sentencing of attempts, which was a subset of the inchoate offences dealt with in Reed.

35 Apart from Reed, it would also be relevant to consider the case of Attorney General’s Reference No. 79 of 2013, R v C (G) [2014] EWCA Crim 1152 (“R v C(G)”), which directly concerned the offence of attempted rape. In deciding the sentence for attempted rape, the court in R v C(G) began by citing the following passage from R v Billam and others [1986] 1 WLR 349 (at 352B):

The court then referred (at [15]) to the Guideline for the starting point and sentencing range for the offence of rape, and proceeded to comment that (at [16]):

In this passage, the court in R v C(G) looked first at the likely sentence for a completed rape offence before applying a discount on account that the offence was not completed. This, again, is similar to the Prosecution’s Proposed Approach.

36 In the light of the foregoing, I was persuaded that the approach taken by the English courts provided clear support for the Prosecution’s Proposed Approach.

37 Two Australian jurisdictions, namely the Australian Capital Territory (“ACT”) and New South Wales (“NSW”), stipulate that attempts are subject to the same maximum punishment as the completed offence. Section 44(9) of the Criminal Code 2002 (ACT) (“ACT Criminal Code”) provides that “[t]he offence of attempting to commit an offence is punishable as if the offence attempted had been committed”. Section 344A(1) of the Crimes Act 1900 (NSW) (“NSW Crimes Act”) provides that “[s]ubject to this Act, any person who attempts to commit any offence for which a penalty is provided under this Act shall be liable to that penalty”.

38 The Prosecution cited R v BI (No 4) [2017] ACTSI 71 (“R v BI”), where the offender was convicted after trial for six sexual offences, including an offence of attempting to engage in sexual intercourse with a person under the age of 16 who was, to the offender’s knowledge, his lineal descendant. After reviewing authorities from across various Australian jurisdictions, the court in R v BI extracted the following ten principles concerning the sentencing of an attempt to commit an offence (at [40]):

39  Among the authorities cited in the foregoing passage, the NSW case of Taouk v R [1992] 65 A Crim R 387 (“Taouk”) stood out as one which was cited the most number of times. In fact, Taouk had been cited in many other NSW cases, including R v McQueeney [2005] NSWCCA 168 and R v Schofield [2003] NSWCCA 3 which were also cited in the foregoing passage. The relevant passage from Taouk reads (at 390–391):

40 The Australian cases did not explicitly apply the two-stage approach adopted in the English cases. However, from the reference in R v BI to the principle that “an attempt to commit an offence will likely attract a lesser sentence than would the offence had it been completed” and the references in Taouk to “the seriousness of that which was attempted” and “the seriousness of the intended consequences”, it would appear that when sentencing an attempt to commit an offence, the Australian courts would also implicitly have regard to the likely sentence which would have been imposed on a completed offence and the potential harm which would have been caused by that completed offence. To this limited extent, the Australian authorities lent some support to the Prosecution’s Proposed Approach.

41 A few observations may be made from the foregoing survey. First, the fact that the Australian cases did not explicitly apply a two-stage approach similar to that adopted in the English cases demonstrates that the two-stage approach is not necessarily the only viable or logical approach. Nevertheless, it would appear from the discussion at [37] above that, even in Australia, implicit regard would still be given to the likely sentence that a completed offence would attract and the potential harm which would have been caused by the completed offence.

42 Second, the two-stage approach adopted in England and Wales appeared to have been borne out of considerations arising from the Guideline and s 63 of the UK Sentencing Act. This raised the question of whether, in the absence of similar guidelines and statutory requirements in Singapore, it would still be open for the Singapore courts to follow the English approach. A similar issued was considered in AQW v Public Prosecutor [2015] 4 SLR 150 (“AQW”), where Sundaresh Menon CJ made the following comments (at [13]):

I was therefore of the view that the absence of similar guidelines and statutory requirements in Singapore did not prevent the Singapore courts, in the exercise of their sentencing discretion, from adopting an approach similar to that applied in the UK, if such an approach was assessed by the Singapore courts to be just and proper. In fact, it could be argued that, in the light of the acceptance of the sentencing approach in R v Szmyt and R v Wolin by the Court of Appeal in Mas Swan, the two-stage approach was already part of our law.

43 Compared to the Prosecution’s Proposed Approach, I found neither the Defence’s First Approach nor the Defence’s Second Approach attractive. The Defence’s First Approach was a non-starter as, by sticking to the Ridhaudin (sentencing) framework, it ran contrary to the reason given by the PCRC for removing the statutory one-half limit, which was that “there is no reason in principle why someone who attempts an offence is only half as blameworthy as someone who has completed the offence” (see [13] above). The Defence’s Second Approach did not fare much better. The Defence suggested that a transposition of the Terence Ng framework to apply to attempted rape should bring the sentencing ranges for each sentencing band lower than those under the Pram Nair framework (for a completed sexual assault by penetration offence), on the basis that a completed offence of sexual assault by penetration would be more serious than attempted rape because “no physical public harm was really done”. The Defence did not explain how far lower than the Pram Nair framework the proposed framework under the Defence’s Second Approach would go. Currently, the sentencing ranges for Band 1 and Band 2 of the Pram Nair framework would be two to three years lower than the corresponding sentencing ranges for the Terence Ng framework. Assuming, for the sake of argument, that the Defence’s Second Approach involved a further two-year reduction from the Pram Nair framework, this would produce a Band 1 sentencing range which was about 40% to 50% lower than that in the Terence Ng framework and a Band 2 sentencing range which was about 25% to 40% lower than that in the Terence Ng framework. In that event, we would end up simply replacing an arbitrary one-half limit with an arbitrary two-third limit or arbitrary three-quarter limit.

44 In contrast, the Prosecution’s Proposed Approach, by asking the court to consider, at the second stage, how close the offender came to completing the offence and the reason(s) the offence was not completed, better allowed the court to calibrate the sentence to, in the words of the PCRC Report, “more accurately capture the culpability of the offender notwithstanding that the prohibited outcome did not materialise” (at 202). The Prosecution’s Proposed Approach also had the virtue of simplicity in that the court is asked to apply an existing sentencing framework developed for the completed offence and then consider the amount of discount on account that the offence was not completed. This obviates the need for the court to develop an additional framework for sentencing of attempts for each and every offence for which an existing sentencing framework already exists (as would have been the case if the Defence’s Second Approach was accepted), and avoids the unnecessary proliferation of sentencing frameworks.

45 For the foregoing reasons, I held that the Prosecution’s Proposed Approach should be adopted for the sentencing of attempted rape offences committed on or after 1 January 2020. To recapitulate, this approach consists of two stages: (a) the court should first apply the Terence Ng framework towards determining the sentence for a notional completed offence; and (b) then apply a discount at the second stage, in recognition of the fact that the offence had not been completed.

46 I make two observations on the application of this approach. At the first stage, the factors which the court takes into account in determining the sentence for a notional completed offence must be based on the facts and evidence before the court, and not based on mere conjecture. At the second stage, it may be relevant to consider, inter alia, whether the offender voluntarily desisted at an early stage or whether he was only prevented at a late stage from carrying out an offence which would likely have otherwise been completed. It would be useful to keep in mind the following observation of Fulford LJ in Reed (at [24]–[26]):

47 At the first stage, the Prosecution submitted that the accused’s attempted statutory rape offence fell within the middle to lower end of Band 2 of the Terence Ng framework, assuming the statutory rape had been completed (ie, assuming the accused penetrated V1’s vagina with his penis while not wearing a condom). The Prosecution submitted that the indicative starting sentence was 13 to 15 years’ imprisonment and 12 strokes of the cane. After taking into account the offender-specific factors, the appropriate sentence was 11.5 to 13.5 years’ imprisonment and ten strokes of the cane for a notional completed offence. At the second stage, the Prosecution proposed applying a discount of around three years’ imprisonment and four strokes of the cane on account of the lesser harm occasioned and the accused’s “medium” culpability, such that the sentence for the Sixth Charge would be 8.5 years to 10.5 years’ imprisonment and six strokes of the cane.

48 I also invited the Defence to make submissions on the appropriate sentence assuming the Prosecution’s Proposed Approach was to be adopted. The Defence broadly agreed that, at the first stage, the notional completed offence would fall within the lower end of Band 2 of the Terence Ng framework, with an indicative starting sentence of 13 years’ imprisonment. The Defence submitted that this should be adjusted downwards to 11 years’ imprisonment in light of the offender-specific factors. At the second stage, the Defence submitted that a discount of 4.5 years’ imprisonment (with no position taken on the number of strokes of the cane) was appropriate, on the basis that there was no actual penile-vaginal penetration and to account for the lesser harm occasioned. The Defence thus contended that an imprisonment term of 6.5 years’ imprisonment was appropriate.

49 I first determined what the appropriate sentence would have been if the statutory rape offence had been completed. The offence in the Sixth Charge fell around the middle to lower end of Band 2 of the Terence Ng framework given the presence of two offence-specific aggravating factors. One factor was a high level of premeditation and predatory behaviour (Terence Ng at [44(c)]). The accused exploited the prevalence and reach of social media to facilitate the commission of the offence. The first time he reached out to V1 was on 20 August 2020, when he messaged her on Instagram to introduce himself and ask her whether she needed cash. He informed V1 that he would pay her S$200 if she came to his residence and allowed him to touch her body. When she agreed, he arranged for them to meet up the next day at his residence and booked a Grab ride to pick V1 up from outside a mall and send her to his residence. It was also significant that he had prepared an iPad to record his sexual acts against V1 and positioned V1 and himself in front of the iPad during the sexual encounter.

50 The second offence-specific aggravating factor was that V1 was a vulnerable victim. At the material time of the Sixth Charge, V1 was 13 years old and had just run away from a welfare home. There was a wide age gap between the accused and V1 of about 31 years, and the accused exploited her youthful naivety and need for money to satisfy his own sexual appetite. In my view, an indicative starting point of 14 years’ imprisonment and 12 strokes of the cane was appropriate.

51 At the second step of the Terence Ng framework, the most significant offender-specific aggravating factor was the fact that the accused was a serial offender, preying on at least four young victims to satisfy his sexual appetite. In particular, there was a further TIC charge for attempted statutory rape on 23 February 2021 of the same victim, V1 (ie, the Eighth Charge). The most significant personal mitigating factors were the accused’s early plea of guilt and cooperation with the authorities. I accepted that the accused had demonstrated remorse and took into account his early plea of guilt. I therefore adjusted the sentence at the end of the first stage to 12 years’ imprisonment and ten strokes of the cane.

52 As set out above at [45], the precise discount to be afforded on the basis that the offence was inchoate was a matter for judicial discretion. Undoubtedly, it should capture the fact that less harm was caused to the victim by an attempted statutory rape offence, as opposed to a completed one. I accepted the Prosecution’s submission that there was a reduced risk of V1 contracting a sexually transmitted disease or getting pregnant, and to some extent, less violation of V1’s body in the sense that the accused did not successfully insert his penis into her vagina. The discount must also accurately capture the culpability of the offender, notwithstanding that the prohibited outcome did not materialise. This was the impetus for the removal of the statutory one-half limit (see [13] above). In this case, it was significant that the accused had removed his pants, attempted to push his penis into V1’s vagina and tried doing so for a while. He was not wearing a condom. He failed not because V1 resisted or struggled, but only because he could not achieve an erection. The lesser harm caused by the attempt (as compared to if the offence was completed) thus had to be balanced against the accused’s culpability. Taking these factors into account, I agreed with the Prosecution’s submission for a discount of three years and four strokes of the cane.

53 Taking all the foregoing factors into account, I imposed the sentence of nine years’ imprisonment and six strokes of the cane for the Sixth Charge.

54 The Second Charge concerned sexual penetration of V1’s vagina with a vibrator at the accused’s residence on 23 February 2021. V1 was still 13 years old at that time.

55 In respect of the Second Charge, the Prosecution sought a sentence of ten to 12 years’ imprisonment and seven strokes of the cane, whereas the Defence submitted that a sentence of ten years’ imprisonment was appropriate.

56 The Pram Nair framework was applicable to offences sentenced under s 376A(3) PC which did not relate to penile-vaginal penetration (ABC v Public Prosecutor [2022] SGHC 244 (“ABC”) at [43] and [46]). Both parties agreed that the accused’s offence of penetrating V1’s vagina with a vibrator fell within Band 2 of the Pram Nair framework, but differed as to the precise indicative starting point for the sentence. The Prosecution placed the offence in the Second Charge as falling within the middle to lower end of Band 2, with an indicative starting point of 11 to 13 years’ imprisonment and eight strokes of the cane. On the other hand, the Defence submitted that it fell at the lower end of Band 2, with an indicative starting point of ten years’ imprisonment.

57 I accepted that the offence in the Second Charge fell within the middle to lower end of Band 2 of the Pram Nair framework given the presence of two offence-specific aggravating factors, which were those that I have already referred to in relation to the Sixth Charge (see [45] and [46] above). There was a high level of premeditation and a vulnerable victim. It was significant that, as early as 6 December 2020 (ie, some three months after the events of the Sixth Charge), the accused deliberately and persistently messaged V1 on Instagram to ask her whether she needed a “job” and to convince her to go to his residence again. He continued messaging V1 even when she did not reply to several of his messages. Again, he prepared an iPad to record his sexual encounter with V1 on 23 February 2021. In my view, an indicative starting point of 11 years’ imprisonment and eight strokes of the cane was appropriate at this stage.

58 At the second step of the Pram Nair framework, the most significant offender-specific aggravating factor was the fact that there were TIC charges of a similar nature against the same victim (ie, the First Charge, Fifth Charge and Seventh Charge). The personal mitigating factors, as above, were the accused’s early plea of guilt and cooperation with the authorities. I therefore calibrated the sentence for the Second Charge slightly downwards to ten years’ imprisonment and seven strokes of the cane (subject to further adjustment on account of the period spent by the accused in remand, as explained at [69] below).

59 Turning to the Third Charge, this was committed on 28 September 2020 against V2, who was 15 years old at the material time.

60 The Prosecution sought an imprisonment term of three to four years, applying an uplift to the sentencing starting point of three years’ imprisonment set out in Public Prosecutor v BAB [2017] 1 SLR 292 (“BAB”) to reflect the multiple aggravating factors in this case. On the other hand, the Defence contended that not more than 12 months’ imprisonment should be imposed. In the alternative, the Defence submitted that an uplift of not more than two to four months’ imprisonment from the benchmark sentence of between ten to 12 months’ imprisonment in AQW was appropriate.

61 I declined to apply the starting point of three years’ imprisonment set out in BAB as proposed by the Prosecution. The relevant passage from the Court of Appeal’s decision in that case reads (at [65]):

62 It was thus clear that the sentencing starting point of three years’ imprisonment in that case had been expressly qualified by reference to its circumstances, ie, where there was an element of abuse of trust. The Prosecution conceded that there was no such element in the present case.

63 I thus decided to start with the benchmark sentence of between ten to 12 months’ imprisonment in AQW and applied an uplift to reflect the multiple aggravating factors in this case. The Prosecution submitted that these factors were as follows: (a) premeditation in the form of predatory behaviour, (b) a degree of pressure exerted on V2 to accept his “offer” for cash, (c) the fact that the accused was a serial offender who targeted multiple minors on social media, and (d) that he had filmed the sexual encounter with V2 with an iPad. While the Defence accepted that there was premeditation involved, it submitted that the accused did not really coerce or exert pressure on V2 to participate in the sexual act after she had arrived at his residence. I would consider that some pressure was exerted by the accused to persuade V2 to accept his “offer” for cash, as he persisted and continued sending messages to her on social media even though V2 initially did not respond to his offer for some time. However, this factor should not be given too much weight as the Statement of Facts did not reveal that the accused exerted pressure on V2 at his residence.

64 I found helpful the case of Public Prosecutor v Chock Choon Seng DAC 904172/2014 and others (“Chock Choon Seng”) which was referred to in AQW (at [33]) as an instance where a “significantly higher sentence” of 24 months’ imprisonment was warranted for a charge involving fellatio under s 376A PC. The offender was a 37-year-old man who got to know the 14-year-old minor through a mobile phone application designed for homosexual people to meet one another. One day, they met in the offender’s home, whereupon the sexual act took place. There was an element of coercion and pressure involved in that the offender had overridden the minor’s expressed reluctance to engage in sexual activity and had used a measure of force in getting the minor to perform fellatio on him. While his penis was inside the minor’s mouth, the offender “held the [minor’s] head and moved the [minor’s] head in an upward and downward motion”.

65 Overall, although the actual aggravating factors were different in this case, in the light of the high level of premeditation and predatory behaviour on the part of the accused, I considered this case to be of similar seriousness as the case of Chock Choon Seng. I also considered the accused’s personal mitigating factors, which I have already mentioned in relation to the other charges. I thus imposed a sentence of two years’ imprisonment for the Third Charge.

66 Before I proceed to consider the Tenth Charge, I should observe that it may be open to question whether sentencing precedents such as BAB and AQW (both of which predated Pram Nair) should continue to apply to offences sentenced under s 376A(2) PC, in the light of subsequent developments in Pram Nair and ABC. In Pram Nair (a case concerning an offence punishable under s 376(3) PC), the Court of Appeal observed (at [161]–[162]) that the new sentencing bands could be relevant to s 376A PC because of the commonality and overlap between s 376 and s 376A PC. It was also observed (at [164]) that the starting point in general for cases sentenced under s 376A(3) PC might have to be reviewed in light of the new framework developed in Pram Nair for offences punishable under s 376(3) PC. The occasion for this review came in ABC, where Sundaresh Menon CJ held that the Pram Nair framework should apply, not only to offences that are sentenced under s 376(3) but also to all offences sentenced under s 376A(3) PC which do not involve penile-vaginal penetration (at [43] and [46]). These developments suggested that the continued applicability of existing sentencing precedents for offences punishable under s 376A(2) PC may also require review. After all, apart from the age of the victim, the elements of an offence punishable under s 376A(2) PC and an offence punishable under s 376A(3) PC are exactly the same. There was no opportunity to undertake such a review in the present case as I did not have the benefit of parties’ submissions on this issue. This was thus a matter that would have to be addressed on another occasion.

67  The Tenth Charge of sexual grooming under s 376E(4)(b) PC was committed against V3, aged 14 years at the material time. The Prosecution submitted that a sentence of nine months’ imprisonment was appropriate whereas the Defence submitted that a sentence of not more than 12 months’ imprisonment would be appropriate.

68 I considered the case of Public Prosecutor v Lee Seow Peng [2016] SGHC 107 (“Lee Seow Peng”), where a sentence of 12 months’ imprisonment was imposed for the offence of sexual grooming under s 376E(1) punishable under s 376E(4)(b) of the PC, a helpful starting point. The sexual act that the accused intended to commit in respect of V3 (which if done would have amounted to sexual penetration under s 376A PC) was less serious than the statutory rape that was intended by the offender in Lee Seow Peng. However, there was also heightened premeditation as the accused repeatedly pressured V3 into accepting his “job” for “indecent touching”, despite V3 saying no on multiple occasions. I also considered the accused’s personal mitigating factors, which I have already mentioned in relation to the other charges. Taking all these factors into account, I imposed a sentence of nine months’ imprisonment for the Tenth Charge.

69 As there were four charges involving three victims, I would run the sentences for one charge each in relation to each victim consecutively – namely, the sentences for the Second, Third and Tenth Charges would run consecutively whereas the sentence for the Sixth Charge would run concurrently with the sentence for the Second Charge.

70 Although the accused was out on bail at the time of the sentencing hearings, he had spent 87 days in remand before being granted bail. I decided to give credit to the time spent in remand by reducing the sentence for the Second Charge by three months, to nine years and nine months. There was no need to make any further adjustments on account of the totality principle as the aggregate sentence was neither disproportionate nor crushing.

71 I therefore sentenced the accused to:

(a) nine years and nine months’ imprisonment and seven strokes of the cane for the Second Charge;

(b) two years’ imprisonment for the Third Charge;

(c) nine years’ imprisonment and six strokes of the cane for the Sixth Charge; and

(d) nine months’ imprisonment for the Tenth Charge.

72 The sentences for the Second, Third and Tenth Charges were to run consecutively while the sentence for the Sixth Charge was to run concurrently with the sentence for the Second Charge. The global sentence was therefore 12 years and six months’ imprisonment and 13 strokes of the cane.