1 For the last ten years, courts in Singapore have, for the most part, sentenced rape offenders in accordance with the guidelines laid down by the High Court in Public Prosecutor v NF [2006] 4 SLR(R) 849 (“PP v NF”). Under the framework expounded therein (which we shall refer to as the “NF Framework”), cases are divided into four broad categories of differing levels of seriousness, each of which attracts a different benchmark sentence. This was an approach which was first articulated by the English Court of Appeal in Regina v Billam and others [1986] 1 WLR 349 (“Billam”) and later revised by the same court in Regina v Millberry and other appeals [2003] 1 WLR 546 (“Millberry”). In his judgment in PP v NF, V K Rajah J (as he then was) explained that the English approach “of classifying rape offences into various broad categories [and assigning a benchmark sentence to each category] is both helpful and useful and may be broadly adopted and employed with appropriate adaptation” (at [23]).

2 By and large, the introduction of the NF Framework has brought a measure of consistency in the sentences imposed in rape offences. This can be seen in the very thorough analysis of the decided cases set out in the Prosecution’s submissions. That said, it has been ten years since the framework was first propounded and the present appeal has provided an opportunity for us to review it, if for no reason other than to ensure that it is still valid in light of subsequent developments in the law. This will also give us an opportunity to address several recurrent problems – many of which are raised in this appeal –in relation to the application of the NF Framework. These include the complaint that the four categories do not adequately cover the full range of circumstances under which the offence of rape could arise, thus leading to a clustering of sentencing outcomes as well as the perception that the NF Framework does not provide adequate guidance in cases of statutory rape.

3  At the first hearing of this appeal, we notified counsel of our intention to undertake a review of the sentencing framework for the offence of rape and invited further submissions on whether, and if so, how, the NF Framework should be revised. Additionally, we also invited Mr Rajaram Vikram Raja (“Mr Rajaram”) to act as amicus curiae to assist the court. We observe from the outset that while all parties agreed that reform was due, they disagreed as to the direction that the change should take. As will be clear in the course of our judgment, the approach we eventually decided on was informed by, but does not precisely resemble, the approaches urged upon us by the parties.

4 This judgment will be divided into three parts. The first part will discuss the law and the sentencing practice of the courts at the present moment. We will outline the problems with the NF Framework and set out the case for reform. The second part sets out the revised approach which will replace the NF Framework (“Revised Framework”). We will explain the considerations that we took into account in the design of the Revised Framework and will elaborate how we envisage it will apply in practice, using illustrative examples drawn from previously decided cases. The final part of the judgment will address the appeal brought by the Appellant against the sentence of 13 years’ imprisonment and 12 strokes of the cane imposed on him for the count of statutory rape under s 375(1)(b) of the Penal Code (Cap 224, 2008 Rev Ed) (“Penal Code”) which he pleaded guilty to. We will apply the Revised Framework to the case at hand and consider whether the sentence imposed by the Judge is, as argued by the Appellant, manifestly excessive.

5 In order to properly understand the NF Framework, it is necessary to go back to the earlier decision of this court in Chia Kim Heng Frederick v Public Prosecutor [1992] 1 SLR(R) 63 (“Frederick Chia”). After a review of the sentencing practice in the United Kingdom, Malaysia, and Singapore at the time, this court held that “for a rape committed without any aggravating or mitigating factors, a figure of ten years’ imprisonment should be taken as the starting point in a contested case, in addition to caning” (at [20]). Yong Pung How CJ (who delivered the judgment of the court in that case) also added that:

6 At this point, we pause to note that when Yong CJ used the expression “starting point”, we think what he had in mind was a benchmark sentence that was broadly appropriate for the offence of rape in general, absent consideration of any aggravating or mitigating factors (eg, a plea of guilt). This approach – of beginning with a single benchmark sentence and then considering whether the benchmark sentence should be reduced or enhanced based on a consideration of the aggravating and mitigating factors in the case – was described in a later decision of this court as the “conventional approach” towards sentencing (see Public Prosecutor v UI [2008] 4 SLR(R) 500 (“PP v UI”) at [22]).

7 The approach in Frederick Chia held sway for 13 years, until the decision in PP v NF. The facts of the latter case were these. The offender was the father of the victim. He returned home inebriated one day and raped the victim, who was only 15 at the time. A few months later, the victim’s teacher found out about what had happened and reported the matter to the police. The offender surrendered himself to the police and was charged with rape under s 376(1) of the Penal Code (Cap 224, 1985 Rev Ed) (“1985 Penal Code”). After setting out the facts, Rajah J observed that “a disturbing and distinct strand of cases has emerged involving vulnerable victims, where the perpetrator is either the parent of the victim, a close relative or a person occupying a position of trust and authority” (at [18]). The sentencing practice in this area, he noted, was “less consistent”. As a result, he embarked on a review of the sentencing precedents for the offence of rape and proposed the NF Framework to help trial courts to impose sentences for rapes which would have a greater degree of consistency.

8 The four categories proposed by Rajah J, and the starting points applicable to each, are as follows:

(a) “Category 1” rapes are those at the “lowest end of the spectrum” and “feature no aggravating or mitigating circumstances” (see PP v NF at [20]). The benchmark sentence is, following Frederick Chia, a term of 10 years’ imprisonment and six strokes of the cane.

(b) “Category 2” rapes are those that feature any one of the seven specific aggravating factors listed at [20] of PP v NF (“Category 2 aggravating factors”). These include, among other things, the rape of a vulnerable victim or rape committed by two or more offenders acting in concert. The “starting point for category 2 rapes” is 15 years’ imprisonment and 12 strokes of the cane (at [36]).

(c) “Category 3” rapes are those which involve the rape of the same victim on multiple occasions or the rape of multiple victims. The benchmark sentence is the same as that for Category 2, namely, 15 years’ imprisonment and 12 strokes of the cane. Rajah J explained that there was no need for a higher benchmark sentence to be set for this category because the Prosecution would, in the usual course of things, prefer multiple charges against the offender and the sentencing court could order that two or more of the sentences imposed run consecutively to reflect the overall gravity of the offending conduct (at [37]).

(d) “Category 4” rapes are those in which the offender has “manifested perverted or psychopathic tendencies or gross personality disorder, and where he is likely, if at large, to remain a danger to women for an indefinite time” (at [21], citing Billam at 50–51). Rajah J noted that unlike in England, the option of a life sentence for an offence of rape was not available in Singapore. Thus, he considered that the benchmark sentence for Category 4 rapes ought to be the maximum sentence of 20 years’ imprisonment and 24 strokes of the cane.

9  Apart from the advantages of consistency and predictability, Rajah J held that the promulgation of benchmark sentences would also serve the aim of general deterrence, by informing would-be offenders of the likely punishment facing them (at [39]). However, he stressed that the NF Framework “should never be applied mechanically, without a proper and assiduous examination and understanding of the factual matrix of the case” (at [43]). He explained that the court ought always to have regard to the presence of any “further mitigating or aggravating factors” [emphasis added] which might serve either to enhance or reduce the appropriate sentence for an offender in accordance with his legal and moral culpability (at [45]). The use of the word “further” is deliberate, for it is clear that what Rajah J had in mind were factors other than those which were already considered in his definition of the categories. Examples of such further aggravating factors would include the harm caused to the victim as well as the list of nine aggravating factors set out in Millberry (eg, the use of a weapon, the causing of particularly serious harm, or the use of a drug to overcome resistance from the victim): at [46] and [55]. In closing, he reiterated that the task of sentencing always involves the “exercise of measured discretion” and that the court had to calibrate the benchmark sentence to fit the facts of the case (at [75]).

10 Analytically, PP v NF requires the sentencing court to proceed in two steps. First, the court has to consider the category into which the particular rape offence should be placed – this will determine the starting point which is to be applied. Second, the court has to adjust this starting point based on the aggravating and mitigating factors which are disclosed on the facts. In a general sense, the manner in which the NF Framework is to be applied is not dissimilar to that used in Frederick Chia as it also adopts the “conventional approach” of beginning with a benchmark sentence and then adjusting that benchmark either upwards or downwards to account for the presence of aggravating or mitigating factors. However, it differs from Frederick Chia in the sense that it does not set a single starting point, but four different ones (one for each category).

11 The NF Framework was first considered and approved of by this court in the context of rape simpliciter in Public Prosecutor v Mohammed Liton Mohammed Syeed Mallik [2008] 1 SLR(R) 601 (“Mohammed Liton”), which was a case involving an adult victim. In PP v UI, this court endorsed its use in cases of aggravated rape (ie, cases of rape punishable under s 376(2) of the 1985 Penal Code – now s 375(3) of the Penal Code), albeit with “requisite modifications to take into account the mandatory minimum sentence prescribed by the latter subsection” (see PP v UI at [9(f)]). In Public Prosecutor v AOM [2011] 2 SLR 1057 (“PP v AOM”), the High Court applied the NF Framework in relation to a case involving the rape of a victim under the age of 14.

12 Having carefully considered the sentences which had been imposed in the cases which have come before the courts post-PP v NF, we are satisfied that the NF Framework needs revision, for essentially the following reasons:

(a) First, the categories are not sufficiently comprehensive and do not cover the full spectrum of the circumstances in which the offence of rape may be committed.

(b) Secondly, there is no conceptual coherence to the Category 2 aggravating factors. As a consequence, Category 2 embraces factual scenarios of widely differing levels of culpability which should not (but currently do) attract the same starting point.

(c) Thirdly, it is not clear as to how the statutory aggravating factors (and the statutory minimum sentence prescribed in relation to those factors) should be taken into account within the NF Framework.

13 Turning, first, to the general structure of the NF Framework, what immediately stands out for attention is the somewhat eclectic methodology used to define the categories. No single yardstick is used to define the categories under the NF Framework; instead, each category uses separate criteria to determine inclusion into its fold. Whereas Category 2 has been defined in a broad and expansive way (by reference to a list of factors the presence of any one of which would justify inclusion in the category), the boundaries of Categories 1, 3 and 4 are much narrower. Category 1 rapes were those which “feature no aggravating or mitigating circumstances” (see PP v NF at [20]). In our recent decision in Haliffie bin Mamat v Public Prosecutor and other appeals [2016] 5 SLR 636 (“Haliffie (CA)”), we explained that Category 1 acts as a kind of “residual category” that covers cases which do not fall into the other three (see Haliffie (CA) at [75]) and may not necessarily be limited only to cases at the lowest end of the spectrum. Categories 3 and 4 relate to tightly defined pockets of offending. Category 3 encompasses situations where the offender has either raped multiple victims or has raped a single victim multiple times. Category 4 covers cases where the offender has “manifested perverted or psychopathic tendencies or gross personality disorder” and is therefore likely to “remain a danger to women for an indefinite time” (see PP v NF at [21]).

14 Because of the narrow confines of Categories 1, 3, and 4, few of the some 25 post PP v NF cases which we have reviewed have come within them. The vast majority of those cases fell within Category 2; only five cases fell within Category 1 (Mohammed Liton (see [96]), Sivakumar s/o Selvarajah v Public Prosecutor [2014] 2 SLR 1142 (see [69]), Haliffie (CA) (see [90]), and Public Prosecutor v Lim Choon Beng [2016] SGHC 169 (see [28])), two cases fell within Category 3 (PP v AOM (see [19]) and Public Prosecutor v Azuar bin Ahamad [2014] SGHC 149 (“Azuar”) (see [133])), and none fell within Category 4. The result is that the present situation is little different from that under Frederick Chia, where a single starting point was adopted for all offences of rape. The only difference is that for Category 2 (which most of the cases fall within) the starting point is 15 years’ imprisonment and 12 strokes of the cane rather than the old Frederick Chia benchmark of 10 years’ imprisonment and 6 strokes of the cane. It is clear from this alone that the NF Framework does not make adequate use of the full spectrum of sentences enacted by Parliament and has instead encouraged a clustering of sentencing outcomes. This is a point which we will revisit when we discuss the problems with Category 2.

15 Furthermore, the problem of multiple offending – which Category 3 is concerned with – is adequately dealt with in practice because the Prosecution would prefer a separate charge for each count of rape disclosed on the facts. This is not just a matter of Prosecutorial practice, but a statutory requirement under the Criminal Procedure Code (Cap 68, 2012 Rev Ed) (“CPC”), which provides that there must be a separate charge for every distinct offence of which a person is accused (see s 132 of the CPC) and that a separate sentence must be passed for each charge that an accused is convicted of (see s 306(1) of the CPC). This was a point which Rajah J himself acknowledged, and it was for this reason that he said there was no need to set a benchmark for Category 3 that was any higher than that for Category 2 (see PP v NF at [37]). But if this is the case then the question may be asked: what purpose does Category 3 serve?

16 Given that Category 2 dominates the field, it is useful to discuss it at some length. In PP v NF, Rajah J explained that the “common thread running through category 2 rapes is that there has been exploitation of a particularly vulnerable victim” (at [25]). This statement, however, is not strictly correct. As the Prosecution rightly points out, not all of the Category 2 aggravating factors relate to the vulnerability of the victim. For ease of reference, the full list of Category 2 aggravating factors listed at [20] of PP v NF is reproduced here:

17 As is clear from the above, many of the Category 2 aggravating factors relate to the manner in which the offence has been committed (eg, the commission of the offence by multiple persons) or the harm to the victim (eg, where the offender commits the offence knowing he has a sexually transmitted disease). They are not confined only to factors which relate to the vulnerability of the victim. There does not appear to be any conceptual unity or discernable unifying theme (apart from the fact that they were the category defining factors used by the English Court of Appeal in Millberry) to the Category 2 aggravating factors. Indeed, it is not at all clear why these factors were singled out as category-defining factors while others – such as the use of a weapon in the commission of the offence or the covert use of a drug to overcome resistance (see PP v NF at [55]) – are not category-defining factors even though they could be equally, if not more, serious than some of the Category 2 aggravating factors.

18 Further compounding this problem is the fact that the Category 2 aggravating factors cover a wide range of situations, not all of which are of equal normative significance. Take, for example, the fourth Category 2 factor – the “rape of a child or a victim who is especially vulnerable”. Cases falling within this description can run the gamut from the violent rape of a young toddler on the one hand to the rape of a domestic helper by her employer on the other. The latter, while undoubtedly serious, cannot be compared with the former. One could also compare the second factor (rape committed by an offender “in a position of responsibility”) with the seventh (an act of rape committed “by a man who is knowingly suffering from a life-threatening sexually transmissible disease”). It is plain that the gravity of the cases which fall within these descriptions can vary greatly, and it is not clear why all of the cases which present themselves with a Category 2 aggravating factor should warrant the same starting point. This is a particular problem because there is a 50% increase in the length of the benchmark sentence (from 10 years’ imprisonment to 15 years’ imprisonment) when one moves from Category 1 to Category 2.

19 It is perhaps because of these reasons that there is a surprisingly large number of cases (eight out of 25, or about 30% of the total), where the NF Framework has not even been cited by the court, let alone applied. And even in cases where PP v NF has been cited, the benchmark sentences it sets out have not been strictly adhered to. Two examples will suffice to illustrate this.

(a) The first concerns cases of statutory rape (that is to say, the rape of a victim who is below the age of 14). The fourth Category 2 aggravating factor is the “[r]ape of a child, or a victim who is especially vulnerable …” (see PP v NF at [20(d)]). By this definition, all cases of statutory rape would, without more, be classified as a Category 2 rape because a girl under 14 is a “child” (see s 2(1) of the Children and Young Persons Act (Cap 38, 2001 Rev Ed) and would attract the Category 2 benchmark sentence of 15 years’ imprisonment and 12 strokes of the cane. However, this is not the prevailing judicial practice. Offences of statutory rape simpliciter are classified as falling “somewhere between category 1 and category 2 rapes” (see Public Prosecutor v Sim Wei Liang Benjamin [2016] SGHC 240 (“Benjamin Sim”) at [35]; see also Public Prosecutor v Lee Seow Peng [2016] SGHC 107 (“Lee Seow Peng”) at [99]) and attract sentences of between 10 and 13 years’ imprisonment. The present case is another such example (see [84] below).

(b) The second concerns cases of familial rape – such cases involve the commission of rape by a person in a position of trust or responsibility, which is the second Category 2 aggravating factor. However, in PP v UI, this court observed that the benchmark sentence for offences of familial rape committed against a minor was imprisonment for a term between 12–15 years per charge (at [23]).

20 The final difficulty concerns the treatment of the statutory aggravating factors which were first introduced in the Penal Code (Amendment) Act 1984 (Act No 23 of 1984). Section 375(3) of the Penal Code provides that a mandatory minimum sentence of eight years’ imprisonment and 12 strokes of the cane will apply if an offender either: (a) voluntarily causes hurt to a victim or any other person in order to commit or facilitate the commission of rape; (b) puts the victim in fear of death or hurt to herself or any other person in order to commit or facilitate the commission of rape; or (c) rapes a victim under 14 by having sexual intercourse with her without her consent. We shall refer to these as cases of “aggravated rape” and specifically to scenario (c) – the rape of a minor without her consent – as a case of “aggravated statutory rape”. While in PP v UI this court has said that the NF Framework can also apply in cases involving aggravated rape provided “requisite modifications” were made (see [11] above), it is not clear what these modifications are or should be.

21 The existence of these statutory aggravating factors (and the minimum sentence which they attract) cannot simply be ignored as anachronistic holdovers from a previous age. As a three judge panel of the High Court stressed in a recent judgment on corrective training, where Parliament has made specific provision for something in statute, “the court is not entitled to ignore its existence” (see Sim Yeow Kee v Public Prosecutor and another appeal [2016] 5 SLR 936 at [84]). There are at least two issues which must be considered here. The first is whether the statutory aggravating factors are themselves Category 2 aggravating factors or whether they are non-category defining aggravating factors to be taken into account to justify a further enhancement in the sentence after an appropriate starting point has been determined by reference to the Category 2 aggravating factors. The second is how the prescribed statutory minimum sentence interacts with the judicial benchmark of 10 years’ imprisonment and six strokes of the cane for the lowest category of rapes, which has been in place since Frederick Chia was decided in 1993.

22 In summary, the NF Framework was a response to the limitations of Frederick Chia, which – in prescribing a single starting point for all cases of rape – did not provide sufficient guidance for sentencing courts. While we recognise that the NF Framework has gone some way towards addressing this problem, in the light of experience gained, there are aspects in the framework where improvements could be made. Accordingly, we are of the view that it needs to be revised.

23 As argued in Saul Holt, “Appellate Sentencing Guidance in New Zealand” 3 NZPGLEJ 1 (“Appellate Sentencing Guidance”), a good guideline sentencing judgment should strive to (at 38):

(a) ensure consistency in sentencing;

(b) maintain an appropriate level of flexibility and discretion for sentencing courts;

(c) encourage transparency in reasoning; and

(d) create a “coherent picture of sentencing for a particular offence” – that is to say, it must respect the statutory context by taking into account the whole range of penalties prescribed, including the mandatory minimum punishments set out in the relevant statute.

24 In our judgment, if these goals are to be achieved, what is required is not just a “recalibrat[ion] of the benchmarks” set out in the NF Framework, as the Prosecution has submitted, but a more fundamental change to the way the sentencing framework for rape is structured. We will turn to detail these changes presently, but before that we begin with a general discussion of the basic nature and structure of sentencing guidelines.

25 In Singapore, the task of issuing sentencing guidelines falls on the judiciary, rather than an executive body specially constituted for this purpose (see Chao Hick Tin, “The Art of Sentencing – An Appellate Court’s Perspective”, Sentencing Conference 2014: Trends, Tools & http://www.supremecourt.gov.sg/docs/default-source/default-document-library/media-room/speech-on-sentencing---2-oct-14.pdf> (accessed 5 May 2017)). Some guideline judgments relate to pure points of principle, such as the manner in which the court decides which sentences to run consecutively and which concurrently (see Mohamed Shouffee bin Adam v Public Prosecutor [2014] 2 SLR 998 (“Shouffee”)) or the criteria to be applied in determining when a mentally unstable offender who has committed a serious offence and is likely to reoffend should be sentenced to life imprisonment for the sake of public protection (see Public Prosecutor v Aniza bte Essa [2009] 3 SLR(R) 327). More commonly, however, guideline judgments lay down the presumptive sentences that should be imposed for the commission of an offence in defined factual scenarios. This is what is usually referred to when the expression “guideline judgment” is used, and it is in this sense that the expression will be used in the remainder of this judgment.

26 Guideline judgments in Singapore have taken many forms and employ a host of different approaches including:

(a) The “single starting point” approach;

(b) The “multiple starting points” approach;

(c) The “benchmark” approach; and

(d) The “sentencing matrix” approach.

27 The single starting point approach was that used in Frederick Chia. As explained above, this calls for the identification of a notional starting point which will then be adjusted taking into account the aggravating and mitigating factors in the case. This is the approach which is still being used in some Canadian jurisdictions today for the offence of rape (see the decision Court of Appeal of Newfoundland and Labrador in R v Atkins (1988) 69 Nfld. & P.EI.R. 99) and, until 1994, this was also the approach favoured in New Zealand (see R v Clark [1987] 1 NZLR 380). Mr Subir Singh Grewal (“Mr Singh”), counsel for the Appellant, submitted that the approach of having a single starting point (which he contends, following Frederick Chia, should be set at 10 years’ imprisonment and 6 stroke of the cane) has the virtue of affording the sentencing court with the most extent of flexibility, and is one which “better caters for a holistic consideration of all the facts in each case.”

28 With respect, we must decline Mr Singh’s invitation to revert to the position in Frederick Chia. In our judgment, the single starting point approach would be suitable where the offence in question almost invariably manifests itself in a particular way and the range of sentencing considerations is circumscribed. This might be the case, for instance, where one is concerned with a regulatory offence. This is not the case where rape is concerned, as the range of relevant sentencing considerations is wide, and there is great variance in the manner in which the offence presents itself. In this regard, it will be helpful to recall that the impetus for the shift to the NF Framework was the inconsistency in sentencing practice that prevailed under Frederick Chia (see [7] above).

29 The multiple starting points approach involves the setting of different indicative starting points, each corresponding to a different class of the offence. Once an indicative starting point has been established by reference to the classification of the offence, it will then be adjusted in the conventional way (that is to say, by having regard to the aggravating and mitigating factors in the case). An example of a case that uses such an approach is the decision of the High Court in Vasentha d/o Joseph v Public Prosecutor [2015] 5 SLR 122, which concerned the offence of drug trafficking. In that case, the court held that different indicative starting points would apply depending on the weight of the drugs trafficked. Once a starting point has been identified based on the weight of the drugs, it would be adjusted to account for the offender’s culpability and the presence of aggravating and mitigating factors (at [47]–[48]). Another example is Yap Ah Lai v Public Prosecutor [2014] 3 SLR 180, which relates to the offence of importing uncustomed tobacco. There, different starting points were established with reference to the quantity of tobacco involved.

30 It can be seen that this approach is suitable where the offence in question is clearly targeted at a particular mischief which is measureable according to a single (usually quantitative) metric that assumes primacy in the sentencing analysis. Drug trafficking and cigarette smuggling are paradigmatic examples of such offences. In drug trafficking, for instance, the mischief which the law aims to address is the prevention of the proliferation of drugs. Prima facie, therefore, an offender who traffics in a greater quantity of drugs deserves greater punishment and the weight of the drugs provides a clear quantitative index for assessing the gravity of the offence (see Public Prosecutor v Tan Thian Earn [2016] 3 SLR 269 at [34]). However, the offence of rape is a complex one, and the seriousness of an offence of rape will depend on a multitude of factors. Because of that, any sentencing framework for this offence must accord to the sentencing court a greater degree of flexibility to take into account the multitude of variables that may come into play in determining the appropriate sentence.

31 The benchmark approach calls for the identification of an archetypal case (or a series of archetypal cases) and the sentence which should be imposed in respect of such a case. This notional case must be defined with some specificity, both in terms of the factual matrix of the case in question as well as the sentencing considerations which inform the sentence that is meted out, in order that future courts can use it as a touchstone. In Abu Syeed Chowdhury v Public Prosecutor [2002] 1 SLR(R) 182, Yong Pung How CJ explained the purpose of a “benchmark” in the following way (at [15]):

32 Like the single starting point approach, the benchmark approach is particularly suited for offences which overwhelmingly manifest in a particular way or where a particular variant or manner of offending is extremely common and is therefore singled out for special attention. One example of the latter type of case is the offence of assaulting a public transport worker. In Wong Hoi Len v Public Prosecutor [2009] 1 SLR(R) 115 (“Wong Hoi Len”), the Singapore High Court held that the benchmark sentence for an uncontested charge of assaulting a public transport worker (if prosecuted under s 323 of the Penal Code) was a sentence of four weeks’ imprisonment. Another example is the case of Public Prosecutor v Fernando Payagala Waduge Malitha Kumar [2007] 2 SLR(R) 334, which concerned offences of credit card fraud prosecuted under s 420 of the 1985 Penal Code. The High Court observed that instances of credit card fraud usually presented themselves in one of two ways: (a) as a syndicated offence involving counterfeit or forged credit cards or (b) a non-syndicated offence involving cards which were stolen or misappropriated, rather than counterfeited or forged (at [21]). However, there is no such thing as a “typical” case of rape, and each case must be assessed on its own terms. For this reason, we do not consider the benchmark approach to be suitable.

33 The sentencing matrix approach is modelled on the approach used by the United Kingdom Sentencing Council. The court first begins by considering the seriousness of an offence by reference to the “principal factual elements” of the case in order to give the case a preliminary classification (in practice, this is done by locating the position of the case in a sentencing matrix, with each cell in the matrix featuring a different indicative starting point and sentencing range: see, eg, Poh Boon Kiat v Public Prosecutor [2014] 4 SLR 892 (“Poh Boon Kiat”) at [77]–[78]). Based on this assessment, the starting point and the range of sentences will be identified. At the second stage of the analysis, the precise sentence to be imposed will be determined by having regard to any other aggravating and mitigating factors, which do not relate to the principal factual elements of the offence: see Poh Boon Kiat at [79].

34 The availability of such an approach is crucially dependent on the availability of a set of principal facts which can significantly affect the seriousness of an offence in all cases (see Koh Yong Chiah v Public Prosecutor [2017] 3 SLR 447 at [47]). For instance, in Poh Boon Kiat, the High Court held that the "principal factual elements” of vice-related offences were: (a) the manner and extent of the offender’s role in the vice syndicate (which is the primary determinant of his culpability) and (b) the treatment of the prostitute (which is the primary determinant of the harm caused by the offence): see Poh Boon Kiat at [75]–[76]. The difficulty in this context is that the offence of rape can take place in a wide variety of different circumstances and it is difficult to identify any set of “principal factual elements” which can affect the seriousness of such an offence across the board. Thus, we do not consider the sentencing matrix approach to be suitable.

35 It will be clear from the foregoing that none of the approaches enumerated above commends itself fully to us. In the 1990s, several Australian courts began adopting a more structured approach towards sentencing under which the court first determines a proportionate sentence having regard to the facts of the offence before adjusting this presumptive sentence in light of the circumstances which are peculiar to the offender (see Austin Lovegrove, “Intuition, Structure and Sentencing: An Evaluation of Guideline Judgments” (2002) 14 Current Issues in Criminal Justice 182 at 183). This approach has not taken hold in Australia, which still favours what is sometimes referred to as the “instinctive synthesis” approach (see the decision of the High Court of Australia in Wong v R (2001) 185 ALR 233 at [76]), but it has been much more warmly received in New Zealand, where it has been widely applied.

36 R v Taueki [2005] 3 NZLR 372 (“Taueki”) was a guideline judgment on offences of serious violence issued by the New Zealand Court of Appeal and it provides the clearest statement of the New Zealand approach. There are two steps to the so-called “Taueki methodology”. The first step involves the identification of a “starting point sentence” which reflects the intrinsic seriousness of the offending act. This is done by identifying the sentencing band appropriate to the offence in question (having regard only to the seriousness of the offence) and then the precise position along the prescribed sentencing band within which the offence falls. In Taueki, for instance, the court described three different “bands” of increasing levels of seriousness and set a sentencing range for each: 3–6 years’ imprisonment for band 1; 5–10 years’ imprisonment for band 2; and 9–14 years’ imprisonment for band 3. Each band was described in general qualitative terms (for instance, band one was described as being “appropriate for offending involving violence at the lower end of the spectrum”: at [36]). At the second step, this starting point sentence is then adjusted either up or down to reflect circumstances which are personal to the offender such as his offending history or the expression of remorse. In R v AM [2010] NZLR 750 (“AM”), the New Zealand Court of Appeal applied the Taueki methodology in the formulation of a guideline judgment for rape. Following the decision of the New Zealand Supreme Court in Hessell v R [2011] 1 NZLR 607 (“Hessell”), it is now seen as desirable that a “third step” be added wherein the court applies a discount to the presumptive sentence derived after the first two steps have been executed to account for the value of any guilty plea (at [73]).

37 The Taueki methodology most closely resembles the sentencing matrix approach but it differs from the latter in two important respects. First, it calls for a general holistic assessment of the seriousness of the offence by reference to all the offence-specific factors rather than just the “principal factual elements”. Second, it draws a distinction between factors which relate to the offending act and those which are personal to the offender. In our opinion, the Taueki methodology has clarity, transparency, coherence, and consistency to commend it and should be adopted. The principal advantages of this approach are as follows:

(a) First, it allows the court to clearly articulate the seriousness of the offence while allowing the sentence to be tailored according to the circumstances of each case. This promotes the communicative function of the criminal law, as it allows the court to express disapprobation for the act even if there are exceptional personal mitigating circumstances which might warrant a significant sentencing discount for the offender.

(b) Secondly, it promotes transparency and consistency in reasoning. Courts will have to openly and clearly articulate the precise weight that is being ascribed to a particular factor. This is especially important when an adjustment is made to account for the personal circumstances of the offender, where the dangers of inconsistency and arbitrariness are greater. If applied consistently over a period of time, the accumulation of transparently reasoned precedents will undoubtedly help future courts to accurately benchmark the seriousness of an offence against others of like nature.

(c) Thirdly, it will promote greater coherence. The dichotomy between offence-related factors and offender-specific factors is conceptually sound (see, generally, Jessica Jacobson and Mike Hough, Mitigation: The Role of Personal Factors in sentencing (Prison Reform Trust, 2007) at p vii) and it addresses one of the principal criticisms of the PP v NF approach, which is the lack of a principled reason for distinguishing between category-defining factors and non-category defining factors (see [17] above).

(d) Fourthly, we consider that the approach of having several sentencing “bands” which are defined in general terms has significant advantages over the NF Framework. Chiefly, these advantages are: (i) it will cover the entire range of offending acts instead of several select pockets of offending; and (ii) the use of sentencing ranges rather than fixed starting points will afford courts with greater flexibility to arrive at a proportionate sentence.

38 The only point on which we demur concerns the introduction of a “third step” for the application of a discount by reason of a plea of guilt or for the rendering of assistance to the police. In our opinion, these are offender-specific mitigating factors and can and should to be taken into account at the second stage of the analysis instead of being considered separately.

39 We turn now to describe the Revised Framework in greater detail. Under this new approach a sentencing court should proceed in two steps:

(a) First, the court should identify under which band the offence in question falls within, having regard to the factors which relate to the manner and mode by which the offence was committed as well as the harm caused to the victim (we shall refer to these as “offence-specific” factors). Once the sentencing band, which defines the range of sentences which may usually be imposed for a case with those offence-specific features, has been identified the court should then determine precisely where within that range the present offence falls in order to derive an “indicative starting point”, which reflects the intrinsic seriousness of the offending act.

(b) Secondly, the court should have regard to the aggravating and mitigating factors which are personal to the offender to calibrate the appropriate sentence for that offender. These “offender-specific” factors relate to the offender’s particular personal circumstances and, by definition, cannot be the factors which have already been taken into account in the categorisation of the offence. In exceptional circumstances, the court is entitled to move outside of the prescribed range for that band if, in its view, the case warrants such a departure.

40 Before elaborating on each of the two steps in this process, we would clarify that the benchmark sentences we are laying down apply to “contested cases” – that is to say, convictions entered following trial. There are at least two reasons for this. The first is based on sentencing theory. The mitigating value of a plea of guilt cannot be fixed, but is personal to the particular offender, and it is affected by factors such as the degree of remorse displayed and the extent to which the offender had “no choice” but to plead guilty because he had been caught in flagrante delicto (see PP v UI at [71]). We will elaborate on the proper weight to be ascribed to a plea of guilt later, but it suffices to say for now that it is clear that this makes it difficult to set a benchmark sentence by reference to uncontested cases when no uniform weight can be attached to a plea of guilt. The second is an argument based on constitutional principle. The law accords every accused person a basic right to plead not guilty and to claim trial to a charge (see Kuek Ah Lek v Public Prosecutor [1995] 2 SLR(R) 766 at [65]). If the benchmarks were set by reference to uncontested cases then it would follow that an uplift should be applied where an offender claims trial. This would lead to the “appearance” that offenders who claim trial are being penalised for exercising their constitutional right to claim trial (see, generally, the decision of the New South Wales Court of Criminal Appeal in R v Henry [1999] NSWCCA 111 at [333] per Simpson J).

41 This is not to say that it would never be appropriate to promulgate a benchmark sentence on the basis of an uncontested case. This might be suitable, for example, where the “typical case” is one where the charge is uncontested, and so fixing the benchmark sentence by reference to an uncontested case makes eminent sense. One such example is the case of Wong Hoi Len, where a benchmark sentence was laid down for a “typical case of road rage where … the accused is a first-time offender pleading guilty” (at [19]). However, the “typical” case of rape (to the extent that such a case exists) is not uncontested. Of the 25 post PP v NF cases of rape which we considered, it was almost an even split – 13 cases were contested while 12 were uncontested.

42 At the first step, the court should have regard only to the “offence-specific factors”. To recapitulate, “offence-specific” factors are those which relate to the circumstances of the offence such as the harm caused or the specific role played by an offender in the commission of a group offence. These factors “indicate the level of gravity of the crime in specific relation to the offence upon which the accused was charged” (see the decision of the High Court in Public Prosecutor v Huang Hong Si [2003] 3 SLR 57 at [8]) and are the factors that the court should consider when selecting an appropriate sentencing band.

43 It is of course impossible to provide an exhaustive list of all the offence-specific factors and what is listed here are just some factors which are often referred to in the decided cases. It should be noted that we have incorporated the statutory aggravating factors in s 375(3) of the Penal Code in this list. Parliament has singled out these factors for particular attention because it had thought that they should be visited with special disapprobation. As we shall explain at [53] below, the presence of more than one of these factors will usually place that offence within the second band of offending.

44 Many of the offence-specific factors which have been discussed in the case law go towards the aggravation of the offence. These include:

(a) Group rape: It has long been held that offences which are committed by groups of persons, even if not the product of syndicated or planned action, are more serious (see Public Prosecutor v Law Aik Meng [2007] 2 SLR(R) 814 (“Law Aik Meng”) at [25(b)]). The reason for this is that the alarm suffered by the victim is invariably enhanced and also because group offences pose a greater threat to social order. This applies with particular force to the offence of rape. When the offence is committed by multiple persons acting in concert, the trauma and sense of helplessness visited upon the victim as well as the degree of public disquiet generated increases exponentially

(b) Abuse of position and breach of trust: This concerns cases where the offender is in a position of responsibility towards the victim (eg, parents and their children, medical practitioners and patients, teachers and their pupils), or where the offender is a person in whom the victim has placed her trust by virtue of his office of employment (eg, a policeman or social worker). When such an offender commits rape, there is a dual wrong: not only has he committed a serious crime, he has also violated the trust placed in him by society and by the victim.

(c) Premeditation: The presence of planning and premeditation evinces a considered commitment towards law-breaking and therefore reflects greater criminality. Examples of premeditation include the use of drugs or soporifics to reduce the victim’s resistance, predatory behaviour (eg, the grooming of a child or young person), or the taking of deliberate steps towards the isolation of the victim (eg, by arranging to meet at a secluded area under false pretences).

(d) Violence: The actual or threatened use of violence in the course of or to facilitate the commission of rape is a statutory aggravating factor (see s 375(3) of the Penal Code). In other jurisdictions, violence must be gratuitous before it can be considered an aggravating factor (eg, in Millberry, Lord Woolf referred to “the use of violence over and above the force necessary to commit the rape” (at [32]); in AM, the New Zealand Court of Appeal referred to a level of violence which was “more than mild” (at [38])). Arguably the reason why such additional factor is alluded to is because rape is inherently a violent offence (see AM at [38]) and by making such a reference, what the courts had in mind was that only excessive force that would be an offence-specific aggravating factor. There is logic in this view and it seems to us that violence as an aggravating factor as prescribed in s 375(3) should be interpreted in that light. What would be “the use of violence over the force necessary to commit the rape” or “excessive force” is a factual inquiry to be determined by the court.

(e) Rape of a vulnerable victim: The rape of a victim who is especially vulnerable because of age, physical frailty, mental impairment or disorder, learning disability. Concerns of general deterrence weigh heavily in favour of the imposition of a more severe sentence to deter would-be offenders from preying on such victims (see Law Aik Meng at [24(b)]). Such cases would often, but not invariably, be accompanied by evidence of an abuse of position/trust and/or some element of premeditation and planning.

(f) Forcible rape of a victim below 14: During the Second Reading of the Penal Code (Amendment) Bill (Bill No 16 of 1984), the Minister said that rape was a “particularly vicious offence” that it was “particularly despicable when it involves the use or threat of violence and when the victim is a child of tender years” (see Singapore Parliamentary Debates, Official Report (26 July 1984) vol 44 at col 1868 (Chua Sian Chin, Minister for Home Affairs)). The policy of the law is that a female under 14 cannot consent to sexual activity. Thus, under s 375(1)(b) of the Penal Code, any man who uses his penis to penetrate the vagina of such a female commits rape, irrespective of whether the victim assents to the act. This mirrors the position at common law, which has long held that girls below the age of 13 cannot consent to sexual intercourse (see R v G (Secretary of State for the Home Department intervening) [2008] 1 WLR 1379 at [19] per Lord Hope). If the victim did not consent, the offence is particularly serious and Parliament has singled such a case out for special attention by making it a species of aggravated rape.

(g) Hate crime: The commission of rape as an expression of racial or religious prejudice, or when actuated by animus towards particular minority groups (eg, the disabled) is especially despicable. The former is a particular concern given the importance of community and racial harmony in Singapore (see Law Aik Meng at [24(f)]). For this reason, racially instigated rape and other cases where the victim has specifically been targeted because of her membership of a vulnerable minority group should be severely dealt with.

(h) Severe harm to victim: As Rajah J stressed in PP v NF, every act of rape invariably inflicts immeasurable harm on a victim (at [46] and [47]). It seriously violates the dignity of the victim by depriving the victim’s right to sexual autonomy and it leaves irretrievable physical, emotional, and psychological scars. Where the rape results in especially serious physical or mental effects on the victim such as pregnancy, the transmission of a serious disease, or a psychiatric illness, this is a serious aggravating factor. In many cases, the harm suffered by the victim will be set out in a victim impact statement.

(i) Deliberate infliction of special trauma: This differs from the previous factor in the sense that this relates to the intention of the offender as manifested in the manner of the offending, rather than the effect which it had on the victim. Cases in which it can be said that there has been deliberate infliction of special trauma include repeated rape in the course of one attack, where there was further degradation of the victim (eg, by forced oral sex or urination on the victim or participation in fetishistic sexual acts), or where there is a rape by a man who knows that he is suffering from a life-threatening sexually transmissible disease, whether or not he has told the victim of his condition (and whether or not the disease was actually transmitted to the victim).

45 We would caution against taking into account the following two factors, which are usually irrelevant to an assessment of the seriousness of the offence:

(a) Forgiveness by the victim: This was a subject that we dealt with at in PP v UI (at [46]–[67]). Forgiveness is a private matter between the victim and the offender, and should not affect the sentence imposed on the offender by the courts, which reflects the public interest in criminal punishment (at [56] and [67]). Two possible exceptions are: (i) where the sentence imposed on the offender would aggravate the victim’s distress and (ii) the victim’s forgiveness is relevant to a determination of the harm she has suffered as a result of the offence (at [57]).

(b) Consent by a victim under 14. Where it is a case of statutory rape, the consent of the victim is irrelevant to liability (see [44(f)] above). In PP v AOM the High Court held that the fact that a victim consented to intercourse was not a mitigating factor save in “exceptional” cases, eg, where the offender and the victim were of the same or similar age at the time the offence was committed (at [35]). We note, parenthetically, that saying that consent cannot be a mitigating factor is perfectly consistent with saying that its absence is an aggravating factor (see [44(f)] above). The reason for making consent irrelevant to liability in cases involving children is to protect them from exploitation. It does not undermine (and in fact it furthers) this objective to say the offender’s culpability is appreciably greater if it can be demonstrated that the child had actively refused to assent to sexual intercourse.

46 One particularly controversial area concerns the relevance of the existence of a prior relationship between the parties: This was discussed at length in Mohammed Liton, where it was held that the existence of a prior relationship between the offender and the victim could neither be treated automatically as an aggravating factor nor as a mitigating factor. Instead, its effect would depend on all the circumstances of the case (at [116]). An allied issue concerns evidence of consensual sexual activity shortly before the offending. In some cases, it has been held that this was something which, while not a general mitigating factor per se, was one which could go towards lessening the offender’s culpability (see, eg, AM at [59] and Mohammed Liton at [118]–[[19]). That said, we emphasise that this view should not be seen as resiling from the principle of sexual autonomy: persons have the right to choose whether or not to participate in sexual activity. They can change their minds if they wish, and their choices must be respected. However, it stands to reason that all the circumstances must be assessed in order to determine the proper weight to be given to such a factor (see AM at [60] and Mohammed Liton at [121]).

47 Once the gravity of the offence has been ascertained, the court should then place the offence within an appropriate band. These sentencing bands represent different sections along a single continuum of seriousness and no longer, unlike in PP v NF, deal only with discrete pockets of offending. Bearing in mind the prescribed maximum punishment for rape and the existing sentencing precedents, our proposed bands are:

(a) Band 1: 10–13 years’ imprisonment, 6 strokes of the cane.

(b) Band 2: 13–17 years’ imprisonment and 12 strokes of the cane.

(c) Band 3: 17–20 years’ imprisonment and 18 strokes of the cane.

48 It will be seen that there is only marginal overlap in the sentencing ranges for each band. As we have stressed on multiple occasions, sentencing is ultimately an exercise in evaluative ethical judgment. There might well be cases which will eventually be assessed to straddle the border between the bands. In the discussion that follows, we shall set out the qualitative description of the cases which fall within each band and identify some representative cases (drawn primarily from the post PP v NF case law) as illustrations of the sorts of cases which would fall within each band. Following the lead of the New Zealand Court of Appeal in AM (at [91]), we would clarify that the use of these representative cases is not to be taken as being indicative of our view of the correctness (or not) of the actual sentences passed in those cases, which were, in any event, decided under a different sentencing framework. Thus, we will not be making reference to the actual outcomes in those representative cases because we do not think it would be very helpful.

49 We would make two preliminary points. First, we would add that there may be some cases which are so unusual that a sentencing point outside the prescribed band should be adopted. In such cases, cogent reasons should be given for the departure from the prescribed sentencing range (see AM at [83]). One possible example is the case of Mohammed Liton, which involved extremely unusual facts (at [119]). As was stressed by this court in that case, it is imperative that a court considers each case on its own facts in order to arrive at a just sentence (at [121]). Second, it will be noted that the lower bound of 10 years’ imprisonment for Band 1 exceeds the minimum sentence of eight years’ imprisonment for aggravated rape (see [20] above). This – the fact that the benchmark sentence of rape of all forms exceeds the statutory minimum sentence for the aggravated rape – has been the case since our decision in Frederick Chia (see [21] above). In our judgment, there is nothing objectionable about this and it is does not have the effect of rendering the statutory minimum otiose. As the Prosecution argued, and we agree, the statutory minimum has the effect of setting an absolute floor beyond which sentences imposed for aggravated rape cannot fall, irrespective of how exceptional the personal mitigating factors. Under the Revised Framework, the statutory minimum sentence set in s 375(3) of the Penal Code still plays this role.

50 Band 1 comprises cases at the lower end of the spectrum of seriousness. These are cases which feature no offence-specific aggravating factors or where the factor(s) are only present to a very limited extent and therefore should have a limited impact on the sentence. Cases falling in the middle to upper ranges of Band 1 would include those where the offence was committed with only one of the aggravating factors listed at [44] above.

51 Cases of statutory rape in which (a) the victim consents (that is to say, the offence is punishable under s 375(2) of the Penal Code) and (b) there are no further notable aggravating factors (such as an abuse of position or evidence of particular vulnerability over and above the age of the victim) should fall in the upper end of this band. The indicative starting point for such cases ought to be 12 years’ imprisonment, though this is of course subject to the facts of each case. The inclusion of statutory rape in Band 1 is not meant to signal a softening of the courts’ stance against such offences, which are undoubtedly serious. However, in the formulation of a set of sentencing guidelines (and in the sentencing exercise more generally), the court has to make a distinction between degrees of seriousness. This sometimes entails the making of quite invidious comparisons between factors which, by any account, are serious and deserving of condemnation (see AM at [78]). When viewed against the entire spectrum of offending, we are of the view that cases of statutory rape falls within the upper end of Band 1. Such offences do not belong at the lowest end of the range, because the offence is inherently aggravated by reason of the age of the victim but at the same time they do not belong in Band 2, which comprise cases which are appreciably more serious. We consider that this accords with existing judicial practice (see [19(a)] above). However, we clarify that aggravated statutory rape (ie, without consent) is different and clearly falls within Band 2, as would the other forms of aggravated rapes listed under s 375(3).

52 A case which we consider would fall within Band 1 is Haliffie bin Mamat (CA). There, the victim was waiting for a taxi after leaving a club and she was tipsy. The offender happened to see her and he stopped his car and offered her a lift. After driving for some time, he stopped the car and raped the victim. After he had done so, he forcefully removed her from his car (which caused her to suffer mild abrasions) and stole her bag, which contained some valuables. The offender had taken advantage of the fact that the victim was inebriated (though not unconscious), but there were otherwise no other offence-specific aggravating factors of note.

53 Band 2 comprises cases of rape which are properly described as being of a higher level of seriousness. Such cases would usually contain two or more of the offence-specific aggravating factors (such as those listed at [44] above), thus underscoring the seriousness of the offence. A paradigmatic example of a Band 2 case would be the rape of a particularly vulnerable victim coupled with evidence of an abuse of position (such as where the rape took place in a familial context, as was the case in PP v NF). Cases which contain any of the statutory aggravating factors and prosecuted under s 375(3) of the Penal Code will almost invariably fall within this band. At the middle and upper reaches of this Band are offences marked by serious violence and those which take place over an extended period of time and which leave the victims with serious and long-lasting injuries physical or psychological injuries.

54 Examples of cases which might fall within this band are:

(a) Public Prosecutor v Robiul Bhoreshuddin Mondal [2010] SGHC 10 (“Robiul”): The offender broke into a house in the dead of night and raped the victim, a domestic helper employed to work on the premises, four times that night. The offender was familiar with the premises because he had done some gardening work for the neighbouring house and he waited for an opportune moment before breaking in. During the course of the rape, he threatened to kill her if she did not remain quiet.

(b) Benjamin Sim: The offender was described as a “prowler on the Internet looking out for young girls whom he could first befriend and then lure into sexual activities” (at [30]). There were multiple victims and the offences took place over a period of three months. He used a pseudonym when he contacted the victims, whom he had only befriended in order to predispose them towards sexual contact, and represented himself as being younger than he actually was. After chatting with his victims for some time, he would initiate sexual contact. In the victim impact statements, it was recorded that the victims felt “dirty and troubled” and blamed themselves for what had happened (at [12]).

(c) Public Prosecutor v Ravindran Annamalai [2013] SGHC 77 (“Ravindran”): The offender rushed at the victim, a domestic helper employed by his neighbour, as she was leaving the neighbouring flat. He forced his way inside and proceeded to rape her twice. Between the two acts of rape, the victim attempted to escape but was unable to. The offender physically assaulted the victim and also threatened to kill her. After the second act of rape, he dragged the victim into the kitchen and strangled her with a piece of string in an attempt to kill her to prevent her from identifying him. The victim lost consciousness due to oxygen deprivation as a result of the strangulation but did not die.

(d) Public Prosecutor v BNN [2014] SGHC 7 (“BNN”): The offender was the stepfather of the victim and his abuse of the victim began when she was just 11. He first touched her inappropriately and his abuses grew in intensity and perversion over a three year period. He raped her when she was 14. The sexual abuses were also accompanied by a pattern of physical abuse against the victim as well as her mother and younger sister, all of whom lived in fear of the offender.

(e) Public Prosecutor v Mohamed Fadzli bin Abdul Rahim [2008] SGHC 177 (“Fadzli”): The offender and his accomplices devised a plan to rob and then to rape commercial sex workers. He would lure each victim into his vehicle on the pretext that he wanted to engage her services. Once the victim had entered the car, she would be driven to a secluded area where he and his accomplices would set upon the victim. They physically assaulted each victim (there were six in total), raped and robbed her, and then abandoned each victim in various states of undress.

(f) PP v AOM: The offender was in a relationship with the victim’s mother and was the victim’s de facto guardian who resided with them. He began sexually abusing the victim when she was only 12 and did not know what sexual intercourse was. Over the next two years, he continued to sexually abuse the victim on multiple occasions by having unprotected sex with her. He sought to deceive her by saying that what he did was to ensure that she would not be cheated in the future. The victim contracted chlamydia as a result of the abuses. These abuses also left indelible psychological scars: the victim had recurrent thoughts about the abuse and feared contact with adult males.

(g) Public Prosecutor v AHB [2010] SGHC 138 (“AHB”): The offender had previously been convicted of outraging the modesty of the victim, his biological daughter. After his release, he did not mend his ways; instead, he continued to outrage her modesty and later raped her when she was 14. The victim became pregnant as a result and eventually delivered a daughter, who was eventually given up for adoption. After the victim first told the offender of her pregnancy, he did not show any concern and instead directed her to lie if questioned before coercing her into fellating him. As a result of the abuses, the victim suffered from flashbacks and feared that she would not be able to have healthy relationships with persons of the opposite sex in the future.

55 We will add these further comments on the cases just cited. Benjamin Sim clearly falls within Band 2 because of the following pronounced aggravating features: (a) the offences were planned and premeditated – the offender had acted in a predatory manner; (b) the offences had taken place over a longer period of time; and (c) the victims had suffered lasting harm. Robiul probably lies at the boundary between Bands 1 and 2, but what takes it to the lower end of Band 2 was the use of a threat which put the victim in fear of death (a statutory aggravating factor) and the fact that the offender broke into the house where the victim resided. It should be observed that as one moves further up the band, the number as well as the severity of the aggravating factors increase. The cases of BNN, Ravindran, and Fadzli all contained a number of aggravating factors which warrant their inclusion in the middle to upper ranges of Band 2. In particular, Fadzli was a case of gang rape which, as we have explained at [44(a)] above, is a significant aggravating factor.

56 The cases which fall at the upper end of Band 2 are PP v AOM and AHB. The former involved a sustained pattern of abuse against a young victim over an extended period of time. It was a serious case of aggravated statutory rape both because of the breach of trust and because of the serious harm occasioned to the victim, who contracted a sexually transmitted disease. AHB was a particularly serious case because of the clear abuse of trust, the vulnerability of the victim, the length of time over which the offences were committed, their callous manner (including the additional indignities inflicted), as well as the lasting effects of the offences on the victim (who became pregnant and delivered a child who had to be given up for adoption). Both PP v AOM and AHB are cases which should be placed at the high end of this band.

57 Band 3 rapes are those which, by reason of the number and intensity of the aggravating factors, present themselves as extremely serious cases of rape. They often feature victims with particularly high degrees of vulnerability and/or serious levels of violence attended with perversities. In many of these cases, the offences would have been committed as part of a “campaign of rape” (see Billam at 351E). There is often a compelling public interest in meeting out a lengthy sentence in the interest of public protection – both to specifically deter the offender in question as well as to mark society’s condemnation for the execrable nature of the offence. Lying at the very apex of this band would be cases in respect of which it might be said that “the offender has manifested perverted or psychopathic tendencies or gross personality disorder, and where he is likely, if at large, to remain a danger to women for an indefinite time” (at ibid). Such cases may properly be described as being among the “most serious instances of the offence in question” (see Sim Gek Yong v Public Prosecutor [1995] 1 SLR(R) 185 at [13]) and could attract the maximum sentence of 20 years’ imprisonment and 24 strokes of the cane.

58 Cases which we consider qualify for inclusion in Band 3 are:

(a) Public Prosecutor v ABJ [2010] 2 SLR 377 (“ABJ”): The offender sexually abused the victim, his friend’s daughter, over a continuous period of seven years beginning from the time she was eight. At points during that seven year period, the assaults took place daily and the offender inflicted a litany of indecencies on the victim: he raped the victim, penetrated her anally, forced her to fellate him, and penetrated her with various objects. As a result of the offences, the victim suffered “indelible” psychological scars and was driven to self-destructive behaviour (including self-mutilation) in an attempt to forget what the offender had done to her (at [11] and [17]).

(b) Azuar: The offender had surreptitiously administered stupefying drugs to his victims (on the proceeded charges alone there were four victims, but there were many charges taken into account) before raping them and filming the acts. The offences were carefully, methodologically, and meticulously planned – the offender was careful to avoid detection by using a pseudonym when he introduced himself and would slip the drugs into his victims’ drinks in the course of their conversations. The assaults were carried out over a 14 month period.

(c) Public Prosecutor v Bala Kuppusamy [2009] SGHC 97 (“Bala Kuppusamy”): The offender had robbed, violently assaulted, and sexually violated four different women in the space of 1.5 months. He was charged with, among other things, aggravated sexual penetration under s 376(4) of the 1985 Penal Code, which carried the same punishment as aggravated rape. This was his third conviction for a serious sexual offence (he had previously twice been convicted of aggravated rape, reoffending shortly after release each time). Against the third and fourth victims, multiple assaults were committed: he not only molested them but also digitally penetrated the victims and forced them to fellate him. He was assessed to have a high risk of recidivism and was described in the judgment as a “merciless, marauding monster for whom liberty is a licence to rob, rape or ruin the lives of innocent females” (at 29]). In his interview with the prison psychiatrist, he expressed his belief that “using physical violence is the fastest avenue in attaining submission” (at [28]).

59 ABJ resembles AHB because of the length of time over which the rapes took place, as well as the lasting harm caused to the victim’s well-being. However, the reason why we think that ABJ crosses the threshold to Band 3 is because of the extent of the degradation caused as well as the extreme youth of the victim (she was just eight when the abuses started). This court described ABJ as an “abhorrent case comprising a deliberate, systematic and remorseless pattern of sexual assaults” (at [17]). The court also denounced the offender’s conduct as “depraved and wanton” and stressed that it needed to receive the greatest censure (at [20]). For these reasons, it was stressed that there was a compelling public interest in meting out a severe sentence in the interest of general deterrence (also at [20]).

60 Azuar and Bala Kuppusamy were both cases where the interest of specific deterrence is particularly compelling, given the high risk of recidivism. The former is chilling for the methodical way in which the offender managed to perpetrate a string of abuses over a long period of time. The experts who took the stand testified that the offender posed a “risk of serious sexual harm to the public over an extended period of time and that psychiatric care or treatment would not do much to help him” (see Azuar at [126]). The latter involved an offender who embarked on what can only be described as a “campaign of rape” (see Billam at 351E). The court observed that the offender posed a “very grave danger to society” (see Bala Kuppusamy at [28]). Both cases gave rise to grave judicial disquiet and should attract very substantial sentences of imprisonment.

61 At this point, we reiterate that the identification of the appropriate band is only a half-step. After identifying the appropriate band, the court has to go on to identify precisely where along the range prescribed for the band a particular sentence falls. This involves, as we have said, an evaluative exercise in ascertaining the gravity of the aggravating factors which are present. This exercise yields the “indicative starting point” for the offence.

62 At the second step of the analysis, the court should have regard to the “offender-specific” factors. At this point what we have to consider are those aspects which relate to the personal circumstances of the offender – that is to say, matters such as his character, personal attributes, expression of remorse, or any other considerations which are particular to the offender rather than factors relating to the manner and mode of the offending or the harm caused by the offence. The court will have to decide what weight to place on these factors and the effect that they will have on the indicative starting point. It is possible that an adjustment beyond the sentencing range prescribed for the band may be called for. However, clear and coherent reasons should be set out if this is to be done. The court should explain clearly the reasons and considerations that prompted such a departure to ensure that transparency and consistency in sentencing, two of the main objectives of this revised framework, can be preserved.

63 It is not possible to give an exhaustive catalogue of all relevant factors nor do we need to, because the offender-specific factors which would generally apply in cases of rape are, by and large, similar to those that would apply in most other offences. We will confine ourselves only to the more common factors and save for the subject of guilty pleas, which the parties addressed us on at length, we do not propose to elaborate on the factors at length.

64 Some offender-specific aggravating factors include:

(a) Offences taken into consideration for the purposes of sentencing (“TIC offences”): While a court is not bound to increase a sentence merely because there are TIC offences, it will normally do so where the TIC offences are of a similar nature (see PP v UI at [38]).

(b) The presence of relevant antecedents: This is a well-established aggravating factor. If the antecedent offence(s) was the same as that of the proceeded charge, then considerations of specific deterrence may come to the fore (see, eg, AHB at [54(g)] above).

(c) Evident lack of remorse: Such a conclusion may be drawn if, for example, the offender had conducted his defence in an extravagant and unnecessary manner, and particularly where scandalous allegations are made in respect of the victim. In AHB, the offender not only failed to take responsibility, but also blamed his wife, whom he said had withheld vaginal intercourse from him, for his behaviour (at [21]).

65 Mitigating factors which are commonly considered include:

(a) Display of evident remorse: This can be demonstrated by, among other things, cooperation with the police. In Public Prosecutor v Wang Jian Bin [2011] SGHC 212 (“Wang Jian Bin”), the court gave the offender credit for apologising after the incident, though it held that the mitigatory value of this was somewhat attenuated by the medium used (he only sent a text message, and did not apologise in person): at [28]. Remorse is also an important factor to be considered when dealing with the mitigatory value of a plea of guilt, as we shall soon discuss.

(b) Youth: In certain cases, the youth of the offender, and in turn his rehabilitation, is a factor to be taken into consideration (see, eg, Wang Jian Bin at [31]; Haliffie (CA) at [90])). However, the countervailing consideration is that serious offences must be met with a condign punishment in order to satisfy the needs for deterrence and retribution (see Al-Ansari at [85]). Balancing the two opposing imperatives (ie, rehabilitation and deterrence) requires an exercise of judgment.

(c) Advanced age: As explained in PP v UI at [78], the advanced age of an offender is not generally a factor that warrants a sentencing discount. However, there is no question that the imposition of substantial custodial terms deprives the elderly of a larger fraction of their expectation of life. This – a concern for the overall proportionality of punishment – and not the age of the offender per se, is the real reason for affording leniency on account of advanced age. As was the case in relation to youthful offenders, the mitigating value of this must be balanced against the need to ensure that older offenders are still be punished appropriately, in line with the gravity of the offence committed.

66 We turn now to the subject of pleas of guilt, which formed a significant part of the parties’ submissions. In Millberry at [27] and [28], the English Court of Appeal identified three reasons why a court might reduce a sentence on account of a plea of guilt: (a) the plea of guilt can be a subjective expression of genuine remorse and contrition, which can be taken into account as a personal mitigating factor; (b) it spares the victim the ordeal of having to testify, thereby saving the victim the horror of having to re-live the incident; and (c) it saves the resources of the State which would otherwise have been expended if there were a trial: see also the decision of the New Zealand Court of Appeal in R v Strickland [1989] 3 NZLR 47 at 51 per Richardson J. Adopting the terminology used in Angliss Singapore Pte Ltd v Public Prosecutor [2006] 4 SLR(R) 653 (“Angliss”) at [53], the first affords a “remorse-based” justification for affording a sentencing discount for a plea of guilt while the latter two present “utilitarian” justifications for doing so. After an extensive discussion, Rajah J held that “only a remorse-based approach… has any currency in the context of our current jurisprudence” (at [56]). At [77], he stated the position as follows:

67 This was also the position taken in PP v NF where Rajah J considered that an early plea of guilt could be a mitigating factor, but only if it was “indicative of genuine remorse” (see PP v NF at [57]). In PP v UI¸ this court likewise held that a plea of guilt per se did not entitle an offender to a sentencing discount. Instead, whether “an early plea of guilt [was] to be given any mitigating value depends on whether it is indicative of genuine remorse” (see PP v UI at [71]). Pointing to this, Mr Francis Ng (“Mr Ng”), counsel for the Prosecution, and Mr Rajaram both argued in their written submissions that it was settled law that a plea of guilt would not per se entitle an offender to a sentencing discount unless it was motivated by genuine remorse.

68 At the time they furnished us with their submissions, Mr Ng and Mr Rajaram did not have the benefit of our decision in Chang Kar Meng v Public Prosecutor [2017] SGCA 22 (“Chang Kar Meng”). There, we held that “offenders who plead guilty to sexual offences, even in cases when the evidence against them is compelling, ought ordinarily to be given at least some credit for having spared the victim additional suffering” [emphasis in original] (at [47]). This endorsement of the utilitarian justification (at least in the context of sexual offending) is consistent with the position taken in previous decisions of this court (see Frederick Chia at [20] and Fu Foo Tong and others v Public Prosecutor [1995] 1 SLR(R) 1 (“Fu Foo Tong”) at [13]) and with at least three previous decisions of our High Court (see Public Prosecutor v Shamsul bin Sa’at [2010] SGHC 132 at [38], Wang Jian Bin at [29], and PP v AOM at [41]). It also accords with the approach taken in the United Kingdom (see R v Caley and others (Consolidated Appeals) [2013] 177 JP 111 at [5]-[6]), New Zealand (see Hessell at [45]), and in some Australian states (see Cameron v R (2002) 209 CLR 339 at [39]).

69 We think the principle of the matter is this. The criminal law exists not only to punish and deter undesirable conduct, but also to (a) help the victims of crime; (b) ensure that those suspected of crimes are dealt with fairly, justly and with a minimum of delay; and (c) to achieve its aims in as economical, efficient, and effective a manner as possible: see Chan Sek Keong, “Rethinking the Criminal Justice System of Singapore for the 21st Century” in The Singapore Conference: Leading the Law and Lawyers into the New Millennium @ 2020 (Butterworths, 2000) at p 30. The utilitarian approach properly reflects the contributions that a guilty plea makes to the attainment of these wider purposes of the law. The consideration here is not just a matter of dollars and cents. An important consideration here is the need to protect the welfare of the victims (particularly victims of sexual crimes, whose needs the law is particularly solicitous of) who must participate in the criminal justice process (see Hessell at [45]; see also [40] above). It would be consistent with the policy of the law in this regard to encourage genuine pleas of guilt to be entered (instead of encouraging a guilty accused to trying his luck by attempting to trip the victim up in her testimony), in order that the trauma suffered by victims need not be amplified by having to recount the incident in court.

70 However, we decline to follow the approach suggested by the UK Sentencing Guidelines Council in a 2007 report (see Sentencing Guidelines Council, Reduction in Sentence for a Guilty Plea Definitive Guideline, https://www.sentencingcouncil.org.uk/wp-content/uploads/Reduction_in_Se ntence_for_a_Guilty_Plea_Revised_20071.pdf> (accessed 5 May 2017)), in setting prescribed sentencing discounts based on the timeliness of the plea of guilt. We agree with the New Zealand Supreme Court that the setting of fixed sentencing discounts does not allow the court to take into account the many and varied reasons for which a plea of guilt is entered and the effects it might have on the victim and the criminal justice process as a whole (see Hessell at [62]). All of these form a complex web of inter-related considerations that should be assessed in the round, and no one factor should be singled out as being of particular significance. The point was well put by Gleeson CJ in Wong v R where he wrote at [76] that to “attribute specific numerical or proportionate value to some features, distorts the already difficult exercise which the judge must perform”.

71 In the premises, we are of the view that the plea of guilt should be assessed as one of the many offender-specific mitigating factors that should be taken into account at the second step of the sentencing analysis. In assessing the proper mitigatory weight to be given to a plea of guilt, the sentencing court should have regard to the three Millberry justifications set out at [66] above and consider the matter together with all the other offender-specific factors in calibrating the sentence to fit the facts of the case. In a case where a plea of guilt is entered timeously and is a clear indication of contrition, it has been suggested that such a plea could warrant a discount of as much as one-quarter to a third of what would otherwise be an appropriate sentence (see Frederick Chia at [20]). At [71] of Chang Kar Meng, we disapproved of such a broad proposition. We expressly observed that whether, and if so, what discount should be accorded to an accused person who pleaded guilty was a fact-sensitive matter that depended on multiple factors (see also Fu Foo Tong at [12]–[13]). Moreover, in cases that were especially grave and heinous, the sentencing considerations of retribution, general deterrence and the protection of the public would inevitably assume great importance, and these cannot be significantly displaced merely because the accused had decided to plead guilty. It is impossible to be prescriptive about this exercise and the discretion is one which must be exercised by the sentencing court, acting judiciously and in the light of the principles we have set out above. We would reiterate that, at the end of the day, the fundamental principle of sentencing is that the punishment imposed must fit both the crime and the offender.

72 The process we have described above should enable a sentencing court to derive the appropriate sentence for each individual offence of rape. Where an accused faces multiple charges, it may be necessary for the sentencing court to recalibrate the sentences imposed for each offence by reason of the totality principle (particularly since s 307(1) of the CPC mandates that a court which convicts and sentences an offender to three or more sentences of imprisonment must order the sentences for two of them to run consecutively). In such a situation, it is important for the court to proceed sequentially: it must first decide on the appropriate sentences for each offence (that is to say, absent consideration of the totality principle) before deciding on the adjustments that are required to be made to the individual sentences imposed in the light of the totality principle. This was done in Azuar (at [133]) and PP v AOM (at [47]). In our judgment, this promotes transparency and consistency in sentencing. At [66] of Shouffee, Sundaresh Menon CJ explained the point as follows:

73 To summarise, the revised framework which we are proposing is as follows:

(a) At the first step, the court should have regard to the offence-specific factors in deciding which band the offence in question falls under. Once the sentencing band, which defines the range of sentences which may usually be imposed for an offence with those features, is identified, the court has to go on to identify precisely where within that range the present offence falls in order to derive an “indicative starting point”. In exceptional cases, the court may decide on an indicative starting point which falls outside the prescribed range, although cogent reasons should be given for such a decision.

(b) The sentencing bands prescribe ranges of sentences which would be appropriate for contested cases and are as follows:

(i) Band 1 comprises cases at the lower end of the spectrum of seriousness which attract sentences of 10–13 years’ imprisonment and 6 strokes of the cane. Such cases feature no offence-specific aggravating factors or are cases where these factors are only present to a very limited extent and therefore have a limited impact on sentence.

(ii) Band 2 comprises cases of rape of a higher level of seriousness which attract sentences of 13–17 years’ imprisonment and 12 strokes of the cane. Such cases would usually contain two or more offence-specific aggravating factors (such as those listed at [44] above).

(iii) Band 3 comprises cases which, by reason of the number and intensity of the aggravating factors, present themselves as extremely serious cases of rape. They should attract sentences of between 17–20 years’ imprisonment and 18 strokes of the cane.

(c) At the second step, the court should have regard to the aggravating and mitigating factors which are personal to the offender to calibrate the sentence. These are factors which relate to the offender’s particular personal circumstances and, by definition, cannot be the same factors which have already been taken into account in determining the categorisation of the offence. One of the factors which the court should consider at this stage is the value of a plea of guilt (if any). The mitigating value of a plea of guilt should be assessed in terms of (i) the extent to which it is a signal of remorse; (ii) the savings in judicial resources; and (iii) the extent to which it spared the victim the ordeal of testifying. Thus under our proposed framework, while for the first step an uncontested case will proceed in the same way as a contested case, it is at the second step that the appropriate discount will be accorded by the court for the plea of guilt by the offender.

(d) The court should clearly articulate the factors it has taken into consideration as well as the weight which it is placing on them. This applies both at the second step of the analysis, when the court is calibrating the sentence from the indicative starting point and at the end of the sentencing process, when the court adjusts the sentence on account of the totality principle. In this regard, we would add one further caveat. In a case where the offender faces two or more charges, and the court is required to order one or more sentences to run consecutively, the court can, if it thinks it necessary, further calibrate the individual sentence to ensure that the global sentence is appropriate and not excessive. When it does so, the court should explain itself so that the individual sentence imposed will not be misunderstood.

74 In deciding on the operative date for the application of this framework, we have regard to the considerations set out in Public Prosecutor v Hue An Li [2014] 4 SLR 661 at [124], where the court discussed the doctrine of prospective overruling. After careful consideration, we are of the opinion that this is not a case in which the doctrine should apply. Our reasons are as follows. First, the Revised Framework does not effect a radical change in the sentencing benchmarks. For the most part it seeks only to rationalise existing judicial practice to promote a more systematic, coherent, consistent, and transparent approach towards sentencing in this area (particularly for cases of statutory rape: see [51] above). Secondly, as will be seen later, applying the Revised Framework to the present case would not give rise to a higher punishment to be imposed on the Appellant. In our judgment, the Revised Framework should take effect immediately. With this, we turn to the appeal in this case.

75 The Appellant pleaded guilty in the court below to two proceeded charges, both of which relate to sexual acts he engaged on two separate occasions with a female who was then under 14 years of age (“the minor”). The first charge was for rape under s 375(1)(b) of the Penal Code and punishable under s 375(2) of the same (the “statutory rape charge”); the second was for the digital penetration of the vagina of a person under 16 years of age, an offence under s 376A(1)(b) of the Penal Code (the “digital penetration charge”). The Appellant consented to having two further charges of rape committed against the same minor taken into consideration for the purposes of sentencing (“TIC charges”). The first TIC charge related to sexual intercourse which took place on the same occasion as the digital penetration charge while the second arose out of sexual activity he engaged in with the minor a month later.

76 The High Court Judge (“the Judge”) sentenced the Appellant to 13 years’ imprisonment and 12 strokes of the cane for the statutory rape charge and one year’s imprisonment and two strokes of the cane for the digital penetration charge. He ordered both sentences to run consecutively, resulting in an aggregate sentence of 14 years’ imprisonment and 14 strokes of the cane. The grounds of the Judge’s decision were published as Public Prosecutor v Ng Kean Meng Terence [2015] SGHC 164 (“the GD”). At the first hearing of this appeal on 7 July 2016, the Appellant informed us that he would neither be pursuing the appeal against the sentence imposed for the digital penetration charge nor would he be arguing that the Judge had erred in ordering the sentences for both charges be run consecutively. Thus, the only issue before us is whether the sentence imposed in respect of the statutory rape charge is manifestly excessive.

77 Before we turn to the facts, we would like to make the following observations. The Judge did not refer to precedents when imposing the sentence for the digital penetration charge. While he recognised that the law prescribed the same maximum punishment for the digital penetration charge as that for rape, he noted at [26] that “the cases have consistently regarded rape as being much more serious … than digital penetration” and that the intrusion in the present case was only “for about 15 minutes”. For this reason, he felt that one year’s imprisonment and two strokes of the cane for that charge would be reasonable. No appeal has been taken by the Prosecution. It seems to us that if the digital penetration charge were the only charge brought against the Appellant in this case, the sentence imposed would have been wholly inadequate. However, we will be dealing with the appropriate benchmark sentence for such a charge in a case which we have heard very recently (and in respect of which judgment has been reserved) so we shall say no more.

78 At the material time, the Appellant was a 42 year old cobbler who operated out of a makeshift stall in the vicinity of an MRT station in the west of Singapore. He was divorced and lived in a Housing Development Board flat (“the flat”) with his parents. The minor was 13 years old and she was a secondary 1 student. On 17 October 2013, the Appellant, having noticed that the minor had been loitering around his stall, struck up a conversation with her during which he learnt that she was afraid to return home for fear of punishment as she had run away. The Appellant invited the minor to the flat and she accepted the invitation. Nothing untoward happened on that day. After they arrived, the appellant called the minor’s mother, who was ill at the time, and was informed that the minor had run away from home on multiple occasions in the past (and stayed away for extended periods of time). Upon hearing this, the Appellant offered to take care of the minor and to act as her godfather. The minor’s mother agreed to this, as did the minor’s father and the minor herself.

79 The minor began meeting the Appellant daily at his stall. They would talk late into the evening and the Appellant would send the minor home afterwards. There were days when the minor would spend the evening at the Appellant’s flat. The Appellant knew that the victim was only 13 years old at the time but despite this, he began to be sexually attracted to her. On 29 October 2013, the Appellant returned to his flat with the minor where they found themselves alone. The minor informed the Appellant that she was experiencing stomach pain and the latter offered to apply some ointment to ease the discomfort. The minor agreed. After the Appellant had applied the ointment, he asked the minor if she would like to have sexual intercourse and she agreed. They proceeded to the toilet where they both removed their shorts and the Appellant digitally penetrated the minor’s vagina for 15 minutes. This was their first sexual encounter and it formed the subject matter of the digital penetration charge.

80 Sometime in the month of November 2013, the Appellant and the minor once again found themselves alone in the flat. The Appellant invited the minor to enter his bedroom and she did. While they were there, the Appellant “decided to determine if the [minor] wished to have sexual intercourse with him.” He did so by touching the minor’s right thigh and by unbuttoning her shorts. The minor responded by removing her shorts and lowering her panties, whereupon they were removed by the Appellant. They then had unprotected penile-vaginal intercourse for about 10 minutes. This gave rise to the statutory rape charge.

81 The minor’s mother passed away on 9 December 2013. After that, the minor’s father directed that the minor cease all contact with the Appellant. The Statement of Facts (“SOF”) did not detail the reasons for this decision save to say that the minor complied. On 7 February 2014, the minor filed a police report in which she stated that she had sexual intercourse with a Chinese male individual three times. The Appellant was arrested shortly afterwards. In the course of investigations, the Appellant was assessed by a psychiatrist from the Institute of Mental Health as being of sound mind and fit to plead in court.

82 In the court below, counsel for the Appellant (who was not Mr Singh) urged the court to impose a sentence of less than seven years’ imprisonment for the statutory rape charge (see the GD at [20]). His essential point was that there were no aggravating factors, stressing that that the minor was sexually experienced, had consented to the acts of intercourse, and did not appear to have suffered as a result of the offences. While the Appellant was identified as the minor’s “godfather”, it was submitted that this did not indicate any formal relationship of dependency or trust and that the parties’ relationship was merely platonic. Emphasis was also laid on the fact that the minor came and left as she pleased (at [18]–[19]). Pointing to all of these factors, counsel for the Appellant contended that the starting point for the statutory rape charge should be a term of 10 years’ imprisonment and 6 strokes of the cane which ought then to be reduced by a further third owing to the appellant’s plea of guilt (at [17]).

83 These arguments were not accepted by the Judge. In relation to the argument based on consent, he held that as a matter of legislative policy, the consent of a minor was not a mitigating factor save in “exceptional cases”, such as where the offender and victim were very close in age, which was not the case here (at [8]). Instead, the Judge held that the minor’s consent was relevant only insofar as it constituted the absence of an aggravating factor – that is, that intercourse was not procured through coercion (at [23]). The minor’s prior sexual experience, the Judge held, was entirely irrelevant. At [24], he stressed that “the law does not countenance that men who have sex with sexual experienced minors have committed less serious offences”.

84 However, the Judge gave the Appellant credit for pleading guilty. Further, the Judge noted that while the Appellant had a number of criminal antecedents, these related to property offences he had committed more than a decade ago, when the Appellant was much younger and were not germane to the present offences. Thus, the Judge did not take them into account (at [22]). He agreed with the Prosecution that this was a case which fell “between a Category 1 and Category 2 rape [of the NF Framework]”. Bearing in mind the two other statutory rape charges which were taken into consideration for the purposes of sentencing, the Judge held that a sentence of 13 years’ imprisonment and 12 strokes of the cane would be appropriate (at [25]).

85  Before we proceed to apply the new sentencing framework which we have canvassed above, we would clarify that we are performing this exercise de novo for illustrative purposes. This is, of course, an appeal, and the grounds for appellate intervention are limited. For this reason, the fact that an application of the new framework to the present case would yield a different result from that reached by the Judge cannot be a ground for disturbing the sentence imposed, so long as it is not manifestly excessive or otherwise wrong in principle.

86 We begin with the first step. There are two offence-specific aggravating factors which call out for attention here: (a) vulnerability of the victim and (b) the evidence that there had been an abuse of trust. From the brief recitation of the facts set out in the SOF, it is clear that the minor grew up in challenging circumstances. We say this for three reasons: first, the minor, then only 13, had already run away from home on several occasions for extended periods of time; second, it is implicit that the minor’s mother was quite ill – she passed away soon after the offences were committed – and was probably unable to exert any meaningful form of parental control over the minor; and third, the minor did not spend very much time at home, as evinced by the fact that she was able to spend extended periods of time at the Appellant’s flat. The overall picture which emerged was thus one of aching vulnerability. It was in these circumstances that the Appellant came into the minor’s life. He came putatively to act as a “godfather”. Initially, this was all that he was. For the first few days, he would allow her to accompany him at his stall and would chat with her throughout the day, before taking her back to her home afterwards. Before long, however, he developed “an attraction for her”, and that precipitated the commission of the present offences

87 Before us, Mr Singh submitted that the Judge had erred in finding that there was an abuse of trust. He pointed out that the Appellant and minor had only known each other for two weeks prior to their first sexual encounter and argued that there was nothing on the facts that suggested that the minor depended on the Appellant as a child would depend on a parent. Echoing the arguments made in the court below, he pointed out that the minor came and went as she pleased and did not depend on the Appellant for food, shelter or guidance. The true picture, he said, was that the minor viewed the Appellant as “someone more like a friend that she felt on the same level with”. While Mr Singh accepted that the Appellant had admitted to being the minor’s “godfather”, he urged us to “look past the nomenclature used”, contending that the Appellant was only a godfather in name and that their relationship was “merely platonic”.

88 With respect, we cannot accept this submission. The Appellant had been allowed unrestricted access to the minor with the express consent of her parents only because he undertook to act as her “godfather” and promised to “take care” of her (see [78] above). This was a position he secured by promising the minor’s mother that he would “take care of her and bring her for various activities”. Seen in this light, the commission of the offences (which took place after the Appellant initiated sexual contact) was not just an abuse of the trust reposed in him, but a complete abnegation of his duty to act in loco parentis. To be fair, there is no evidence that he had deliberately sought out the position of godfather in order to commit the offences (which would have made the situation even more aggravated), but even as things stand, it is plain the relationship was not “merely platonic” as Mr Singh claimed, but familial.

89 Taking this into consideration, this seems like a case which falls at the margin of Band 1 and Band 2, and we think it is more of a case at the lower end of the Band 2. We would regard the infringement in this case as serious because of the clear abuse of position and trust. However, it is not as serious as the cases of Benjamin Sim, where the offences were premeditated (see [54(b)] and [55] above), and PP v AOM, which was a serious case of familial rape where the victim suffered significant harm (see [54(e)] above). Considering matters in the round, it seems to us that this is a case in which an indicative starting sentence of 13 years’ imprisonment and 12 strokes of the cane, which is at the lowest end of the Band 2, would be appropriate.

90 We move to the second step. The most significant offender-specific aggravating factor is the fact that there are two further charges of statutory rape which have been taken into consideration for the purposes of sentencing. According to established principles, this will almost invariably result in an increase in the sentence imposed (see [64(a)] above). We agree with the Judge that the Appellant’s antecedents, being unrelated and of some vintage (those were offences he committed nearly 17 years ago, when he was a young man) should not be held against him. As for the mitigating factors, the most significant is the fact that he pleaded guilty. We are not disposed to accept Mr Singh’s submission that he had done so out of remorse. As pointed out by Mr Ng, the Appellant has consistently tried to downplay his responsibility for the offence. In his interview with the Prison psychiatrist, he attempted to push responsibility for the offence onto the victim by saying that she had initiated the sexual encounter. While he did not repeat these allegations before the Judge, he had tried to argue what had happened was the result of an “unexpected momentary loss of self-control”, even though the fact that there were two other charges of statutory rape (both of which related to sexual activity with the minor on different occasions) puts paid to this contention. However, we accept Mr Singh’s contention that a sentencing discount should be afforded to him in recognition of the fact that the Appellant’s decision to plead guilty had spared the minor the ordeal of having to testify.

91 As for the weight to be ascribed to the plea of guilt, we accept Mr Rajaram’s suggestion that a discount of about 10% would be due. When this figure was put to Mr Ng during the hearing, he accepted that this was fair. However, considering that an uplift is warranted on account of the TIC charges, we consider that the offender-specific factors cancel each other out. Thus the proper sentence that should be imposed on the Appellant would still be 13 years and 12 strokes of the cane, which was the sentence imposed by the court below.

92 For the foregoing reasons, we consider that the sentence imposed by the Judge for the statutory rape charge was entirely appropriate and that there is no basis for appellate intervention. We therefore dismiss the appeal.

93 It remains for us to record our deep appreciation to Mr Ng, Mr Singh, and to Mr Rajaram, for the invaluable assistance which they have rendered to the court in this appeal. This is the second time that Mr Rajaram has been invited to act as amicus curiae, and as was the case on the first occasion, his submissions were clear, comprehensive, and succinct.