1 Section 377BB Penal Code (Cap 224, 2008 Rev Ed) (“PC”) came into force on 1 January 2020 and sets out, for the first time under Singapore law, a series of provisions specifically targeting the act of voyeurism. The present case engages the offence under s 377BB(4) PC, which is punishable under s 377BB(7) PC. The appellant was a 24-year-old student in his final year of undergraduate studies at the Nanyang Technological University (“NTU”) when he committed two counts of the offence. ¹ He pleaded guilty before a District Judge (“DJ”), who imposed an aggregate imprisonment term of seven weeks. Dissatisfied, he appealed. Central to his case on appeal is that he has shown an extremely strong propensity for reform as evinced by the active steps he took post-offence to seek psychiatric intervention.

2 This case thus puts into issue the relevance of rehabilitation as a sentencing consideration where the offender commits a s 377BB(4) PC offence punishable under s 377BB(7) PC. It also provides us with an opportunity to consider the appropriate sentencing framework for this new offence which would guide future sentencing courts and achieve broad parity and consistency in sentencing. A Young Independent Counsel (“YIC”), Ms Eden Li Yiling, was appointed to address us on these legal issues.

3 Before addressing these two points of law, we first set out the legal, factual and procedural background which provide context to our subsequent analysis.

4 Prior to the enactment of s 377BB PC, the courts relied on a patchwork of laws, including insult of modesty under the now-repealed s 509 PC and possession of obscene films under s 30 of the Films Act (Cap 107, 1998 Rev Ed), to punish voyeurs. This patchwork of laws, however, did not adequately cover the range of circumstances in which voyeurism offences are committed. Against this backdrop, the Criminal Law Reform Act 2019 (Act 15 of 2019) (“CLRA”) introduced s 377BB PC with the view to define the offending behaviour of voyeurism and provide adequate punishments (Singapore Parliamentary Debates, Official Report (6 May 2019) vol 94 (Mr K Shanmugam, Minister for Home Affairs)).

5 Section 377BB PC, as introduced by the CLRA, reads:

Subsequently, s 30 of the Criminal Law (Miscellaneous Amendments) Act 2021 (Act 23 of 2021) (“CLMAA”) replaced the word “genitals” in ss 377BB(4)(a), 5(a) and (9) PC with the words “genital region” with effect from 1 March 2022.

6 There are six offence-creating provisions in ss 377BB(1)–(6) PC, which can be distinguished with the help of the following table prepared by the YIC:

7 In so far as the ambit of s 377BB(4) PC is concerned, two key points must be borne in mind. First, s 377BB(4) PC is broad enough to encompass a scenario where the victim’s private region had been placed under observation, as well as a scenario where the victim’s private region had not been sighted even though the offender had operated equipment with the intention of enabling himself or another person to observe the victim’s private region. Secondly, the s 377BB(4) PC offence can be committed regardless of whether the offender had made a record of the victim’s image, be it by way of a photograph or a video.

8 The first incident (the “First Incident”) took place on 19 October 2020. At around 6pm that day, the appellant was in his girlfriend’s room on level 6 of Block 24 Tamarind Hall, a student residential hall at NTU, when he decided to walk down to level 1 to meet his girlfriend. Upon reaching level 1, he spotted the victim (“V1”), a 20-year-old female NTU student who was walking back to her room in Tamarind Hall. ² The appellant decided to follow her as she walked up to the lift lobby on level 3. When V1 was waiting for the lift at level 3, the appellant took out his mobile phone with the intention to take an upskirt video of V1. He switched his phone camera on to video mode, squatted down, placed his phone under V1’s dress with the camera aimed up her dress, and recorded an upskirt video despite knowing that V1 did not consent to this. ³ As V1 felt someone moving closer to her from behind, she turned and saw the appellant. Shocked, and without communicating with the appellant, V1 left the level 3 lift lobby as she was worried for her safety. In particular, she was afraid that the appellant would enter the lift with her if she had taken the lift. ⁴

9 After the First Incident, the appellant deleted the video from his phone. The First Incident subsequently came to light as V1 reported the matter to campus security and the appellant’s identity was established through CCTV footage. ⁵ An investigation officer at NTU approached the appellant to seek his assistance with investigations and thereafter called the police. ⁶ The appellant was arrested on 20 October 2020 but was released on police bail the next day. ⁷

10 Another incident (the “Second Incident”) involving a 17-year-old female victim (“V2”) took place on 25 February 2021 while the appellant was still on police bail. The appellant, who was returning home to 300 Canberra Road at around 4.30pm that day, had just parked his car at a multistorey car park when he spotted V2, who was returning home from school in her school uniform. The appellant noticed that V2 was wearing a skirt and felt the urge to take an upskirt video of her. He then followed V2 from the entrance of the car park to the lift lobby at 306 Canberra Road and stood next to her at that lift lobby. When the lift arrived, he followed V2 into the lift. V2 selected the 12th floor and the appellant selected the 15th floor. ⁸ While the lift was going up, the appellant set his mobile phone to video recording mode. When V2 was about to head out of the lift at the 12th floor, the appellant squatted down and stretched out his arm with his phone camera pointing towards V2’s thigh area with the intention to take an upskirt video of V2 despite knowing that V2 did not consent to this. ⁹

11 V2 felt the appellant touch her thigh and turned back and shouted “oi”. V2 was out of the lift at this point. The appellant did not respond but raised his hands in apology. ¹⁰

12 When the appellant reached the 15th floor, he took the stairs down to the ground floor and went home. V2 called her father about the incident and started crying. Her parents tried to search for the appellant but to no avail. V2’s mother then brought V2 to lodge a police report that afternoon. ¹¹ The appellant had intended to view the video and delete it thereafter, but he deleted the video without viewing it when he was caught in the act by V2. ¹² The appellant’s identity was later established through CCTV footage and he was arrested on 1 March 2021. ¹³

13 On 16 December 2021, the appellant pleaded guilty to two charges under s 377BB(4) PC in relation to the First and Second Incidents respectively. ¹⁴ The appellant also consented to one charge of criminal trespass under s 447 PC being taken into consideration for the purpose of sentencing (the “TIC Charge”). The TIC Charge accuses the appellant of remaining at Tamarind Hall with the intent to annoy V1. ¹⁵

14 In light of a medical report from the appellant’s psychiatrist opining that the appellant suffered from voyeuristic disorder, along with many other reports detailing the appellant’s progress after multiple psychiatric reviews and psychotherapy sessions, the DJ called for a Mandatory Treatment Order (“MTO”) suitability report on the same day the appellant’s guilty plea was taken. In the MTO suitability report dated 22 February 2022, the appointed psychiatrist stated that she was not recommending an MTO because she did not find any evidence of a psychiatric disorder. ¹⁶ The DJ then scheduled a Newton hearing on 30 and 31 May 2022 to determine if the appellant suffered from voyeuristic disorder, and if so, whether it caused or contributed to the commission of the offences. However, these issues were subsequently rendered moot when the appellant confirmed on 30 May 2022 that his mitigation would not rely on the alleged voyeuristic disorder causing or contributing to the offences. In the circumstances, the parties agreed that a Newton hearing was not required. ¹⁷ The DJ then proceeded to hear their submissions on sentence without a Newton hearing.

15 The Prosecution sought an aggregate sentence of eight to twelve weeks’ imprisonment in reliance on sentencing precedents for similar upskirt offences previously prosecuted under the now-repealed s 509 PC, and a reported case involving an offence under s 377BB(4) PC and punishable under s 377BB(7) PC. ¹⁸ Prior to the calling of the MTO suitability report by the DJ (see [14] above), counsel for the appellant initially submitted that it was appropriate to call for concurrent probation and MTO suitability reports and alternatively, an aggregate custodial sentence of not more than four weeks’ imprisonment. ¹⁹ Subsequently, the appellant urged the court to accept that there was no need for an MTO nor a custodial sentence, as he had made significant progress in his psychiatric treatment. ²⁰

16 On 24 June 2022, the DJ sentenced the appellant to an aggregate of seven weeks’ imprisonment (Public Prosecutor v Nicholas Tan Siew Chye [2022] SGMC 40 (“GD”) at [23]).

17 In arriving at this aggregate sentence, the DJ first identified deterrence as the dominant sentencing principle for upskirt video cases (GD at [12]). He also found that some measure of specific deterrence was called for in this case as the appellant had reoffended on bail in the Second Incident (GD at [14]). The DJ then held that there was no reason for rehabilitation to displace deterrence as the primary sentencing consideration. The appellant was not a youthful offender, but a sufficiently mature yet recalcitrant offender whose alleged voyeuristic disorder did not cause or contribute to his offending (GD at [15]). The appellant’s rehabilitation efforts also did not displace the need for deterrence since the purported disorder he was being treated for did not cause or contribute to his offending (GD at [20]).

18 Next, the DJ analysed the aggravating factors relating to each incident. In respect of the First Incident, the DJ noted the following factors (GD at [16]):

(a) There was “an element of stalking, planning and pre-meditation” – the appellant spotted V1 and followed her from level 1 to level 3.

(b) The offence occurred at V1’s place of residence where she had the right to feel safe.

(c) In so offending, the appellant committed criminal trespass which was the subject of the TIC Charge.

The aggravating factors relating to the Second Incident were more severe. They were (GD at [17] and [22]):

(a) The appellant reoffended while on police bail.

(b) There was again “an element of stalking, planning and pre-meditation” – he spotted and targeted V2 and followed her from the carpark to her block.

(c) The offence was committed in a lift, even though it was when V2 was about to head out of it.

(d) V2 was distressed and cried when she spoke to her father.

19 Against these, the DJ considered the following mitigating factors (GD [18]–[19]):

(a) The appellant had no antecedents.

(b) The appellant was genuinely remorseful, as demonstrated by his early plea of guilt and efforts at rehabilitation through multiple sessions with his psychiatrist and psychotherapist. The appellant’s efforts in seeking psychiatric treatment also evinced his desire to change.

(c) The appellant had good family support.

The DJ did not place mitigating weight on the appellant’s alleged voyeuristic disorder since it did not cause or contribute to his offending (GD at [19]).

20 As to the type of punishment, the DJ held that the custodial threshold has been crossed given the need for deterrence and the aggravating factors. He rejected the appellant’s submission for a fine, holding that it would be an inadequate punishment (GD at [20]). Although full weight must be placed on the appellant’s rehabilitation efforts, the DJ found that this did not mean a fine ought to be imposed since the appellant’s purported voyeuristic disorder did not cause or contribute to his offending (GD at [20]), and rehabilitation could continue in prison and even after the appellant was released (GD at [24]).

21 Instead, the DJ placed full weight on the appellant’s rehabilitation efforts by factoring a “substantial sentencing discount” into the imprisonment term (GD at [25]). For the offence arising from the First Incident, the DJ calibrated the starting point of four weeks’ imprisonment to three weeks’ imprisonment on account of the appellant’s remorse, as demonstrated by his plea of guilt and attempts at rehabilitation. As for the offence arising from the Second Incident, the DJ used a starting point of six weeks’ imprisonment but calibrated it downwards to four weeks’ imprisonment after having regard to the mitigating factors and the totality of the sentence (GD at [21]–[22]).

22 The two sentences were ordered to run consecutively, resulting in an aggregate sentence of seven weeks’ imprisonment (GD at [23]). The DJ was satisfied that this aggregate sentence was not crushing, and noted that it was much lighter than the sentence meted out in Public Prosecutor v Mark Fritz Tanel [2022] SGMC 26 (“Mark Fritz Tanel”) for a similar offence (GD at [25]).

23 On appeal, the accused sought a non-custodial sentence on the basis that rehabilitation rather than deterrence should be the primary sentencing consideration. ²¹ Meanwhile, the Prosecution submitted that the DJ’s decision should be upheld as he had correctly identified deterrence as the dominant sentencing consideration and the custodial sentence imposed was not manifestly excessive.

24 This appeal raises two main issues of law:

(a) In what circumstances, and to what extent, should rehabilitation be a relevant sentencing consideration for voyeurism offences under s 377BB(4) PC and punishable under s 377BB(7) PC?

(b) What would be an appropriate sentencing framework for voyeurism offences under s 377BB(4) PC and punishable under s 377BB(7) PC, and when is the custodial threshold crossed?

We clarify that our decision on these legal issues applies to s 377BB(4) PC offences regardless of whether they took place before or after the amendment brought about by s 30 of the CLMAA (see above at [5]).

25 Once these legal issues are answered, two factual issues fall to be considered:

(a) Did the DJ err in finding that deterrence instead of rehabilitation should be the dominant sentencing principle?

(b) Did the DJ err in imposing a seven weeks’ imprisonment term?

26 The YIC proposed that rehabilitation will be the dominant sentencing consideration for a s 377BB(4) PC offence if it is presumptively the dominant sentencing consideration, and this presumption is not displaced by another sentencing consideration such as deterrence. Rehabilitation is presumptively the dominant sentencing consideration where the offender (a) is a youthful offender (ie, below age 21), (b) is an adult offender with an extremely strong propensity for reform (as determined using the framework in Public Prosecutor v Siow Kai Yuan Terence [2020] 4 SLR 1412 (“Terence Siow”)), or (c) has a mental condition that is causally linked to the commission of the s 377BB(4) PC offence. The presumption that rehabilitation is the dominant sentencing consideration may be displaced by deterrence where (a) the offence is serious or grave, (b) the harm caused is severe, (c) the offender is hardened and recalcitrant, or (d) conditions do not exist to make rehabilitative sentencing options viable. ²² In all other s 377BB(4) PC cases where rehabilitation is not the dominant sentencing consideration, the YIC submitted that rehabilitation will at best feature as a subsidiary sentencing consideration, with deterrence as the dominant sentencing consideration. ²³

27 Having regard to the inherent severity of s 377BB(4) PC offences which may be amplified by the specific circumstances of a particular case, as well as Parliament’s intention and case law, the YIC submitted that deterrence will “almost invariably” displace rehabilitation as the dominant sentencing consideration. ²⁴

28 The appellant accepted that deterrence is generally the dominant sentencing consideration for an adult offender who commits a s 377BB(4) PC offence. ²⁵ However, he argued that rehabilitation should displace deterrence as the primary sentencing consideration where the adult offender has demonstrated an extremely strong propensity for reform, especially by taking active steps post-offence to leave his errant ways behind. ²⁶

29 The Prosecution agreed with the YIC that deterrence is the dominant sentencing principle for s 377BB(4) PC offences, ²⁷ and that courts generally give more weight to rehabilitation as a sentencing objective in the three circumstances identified by the YIC (see above at [26]). ²⁸ In response to the appellant’s argument, the Prosecution submitted that the fact that an offender has shown a strong rehabilitative potential by voluntarily seeking treatment for his voyeuristic disorder cannot by itself shift the focus from deterrence to rehabilitation. That said, it accepted that the act of seeking treatment can still be considered when calibrating the sentence in so far as it is indicative of the offender’s remorse and rehabilitative potential. ²⁹

30 We make three preliminary points. First, the YIC had, for the purpose of determining the relevance of rehabilitation as a sentencing consideration, developed the same analytical approach for youthful offenders, adult offenders with an extremely strong propensity for reform, and offenders who have a mental condition that is causally linked to the offending conduct. For all three categories of offenders, the YIC submitted that there should be a presumption that rehabilitation is the dominant sentencing consideration, and this presumption may be displaced by the need for deterrence in certain circumstances. We are, however, disinclined to adopt a standardised approach across all three categories. We prefer to think of these categories as factual circumstances in which the court has the difficult yet important task of striking a fine balance between deterrence (as well as retribution and prevention, as the case may be) on the one hand and rehabilitation on the other, where each of these sentencing considerations may very well pull the court towards different sentencing outcomes. The determination of how this balance ought to be struck involves a highly fact-centric inquiry, which is in turn shaped by distinct policy considerations relating to the offender’s personal attribute (eg, his youth, extremely strong propensity for reform, or mental condition). A standardised approach should therefore not be adopted.

31 To put things into further perspective, it is trite that rehabilitation is the presumptive dominant sentencing consideration where youthful offenders are concerned. However, the basis for this is grounded in a retrospective rationale and a prospective rationale, both of which are heavily influenced by the unique policy considerations relating to the youth of the offender. The retrospective rationale justifies giving the youthful offender a second chance by excusing his offending behaviour on the grounds of his youthful folly and inexperience. The prospective rationale justifies rehabilitation as the preferred tool to discourage future offending on the premises that the youthful offender will be more receptive towards a sentencing regime aimed at altering his values and guiding him on the right path, society will stand to benefit considerably from the rehabilitation of the youthful offender who has many potentially productive and constructive years ahead of him, and the youthful offender will appear to suffer disproportionately as compared to adult offenders if typical punitive options were to be imposed: A Karthik v Public Prosecutor [2018] 5 SLR 1289 (“A Karthik”) at [37]. Evidently, these considerations are not directly applicable to an adult offender with an extremely strong propensity for reform, or an offender who has a mental condition that is causally linked to the offending conduct.

32 Secondly, it is possible for rehabilitation to be the dominant sentencing consideration outside of the three categories identified by the YIC. Much will depend on the interplay of the various sentencing considerations in a given set of facts.

33 Thirdly, this judgment will only consider the relevance of rehabilitation as a sentencing consideration for s 377BB(4) PC offences (punishable under s 377BB(7) PC) committed by adult offenders with no mental condition contributing to their offending conduct. We are cautious of adjusting the body of case law concerning the treatment of youthful offenders (see Public Prosecutor v Mohammad Al-Ansari bin Basri [2008] 1 SLR(R) 449; Public Prosecutor v Koh Wen Jie Boaz [2016] 1 SLR 334; A Karthik) when the facts of this case do not concern one. Further, given that the appellant had unequivocally disclaimed any reliance on his alleged voyeuristic disorder, it is not necessary for us to consider the appropriate approach to take where the s 377BB(4) PC offence (punishable under s 377BB(7) PC) is committed by an offender who suffers from a psychiatric condition, such as voyeuristic disorder, that is causally linked to the commission of the offence. We prefer to leave the consideration of this issue to a future case where it directly arises on the facts.

34 We now turn to elaborate on the relevance of rehabilitation as a sentencing consideration for s 377BB(4) PC offences (punishable under s 377BB(7) PC) committed by adult offenders with no mental condition contributing to their offending conduct.

35 Generally, neither the prospective nor the retrospective rationale set out at [31] above will apply to an adult offender. As such, the starting point is that rehabilitation is not the presumptive dominant sentencing consideration where an adult offender is concerned: A Karthik at [34] and [44]. That is not to say that rehabilitation can never be the operative sentencing consideration for an adult offender. Where it is shown that the particular adult offender in question has demonstrated an extremely strong propensity for reform, rehabilitative sentencing options may still be an effective means of discouraging prospective offending despite the age of the offender. To this extent, the prospective rationale described above is engaged. It is for this reason that the law takes the view that rehabilitation may be the operative sentencing consideration where the particular adult offender in question demonstrates an extremely strong propensity for reform, provided that the significance of rehabilitation as the dominant sentencing consideration is not displaced by another sentencing consideration such as deterrence: see Terence Siow at [42], [45], [52] and [60]. These principles are well settled in case law and are equally applicable to a s 377BB(4) PC adult offender who is not suffering from any mental condition that is causally connected to the offending conduct.

36 In this connection, the following three-limbed framework developed in Terence Siow is useful in evaluating whether a s 377BB(4) PC adult offender has demonstrated an extremely strong propensity for reform (Terence Siow at [55]):

(a) First, the court should consider whether the offender has demonstrated a positive desire to change since the commission of the offence(s) (“the first limb”).

(b) Second, the court should consider whether there are conditions in the offender’s life that are conducive to helping him turn over a new leaf (“the second limb”).

(c) If, after considering the first two limbs, the court comes to a provisional view that the offender has demonstrated an extremely strong propensity for reform, the court should then consider, in light of the risk factors presented, whether there are reasons to revisit the finding of such a high capacity for reform (“the third limb”).

This framework, which weighs the factors in favour of reform against the risk factors that may counteract and so compromise the efficacy of the reformative efforts, provides a systematic approach towards a fact-sensitive inquiry: Terence Siow at [60].

37 Under the first limb, the court examines the offender’s own resolve to change, as inferred from evidence of the offender’s remorse and the trajectory of his rehabilitative progress between the time of offending and sentencing. The non-exhaustive factors in this regard are (Terence Siow at [56]):

The first of these factors, namely genuine remorse, can be evinced by an early plea of guilt and a full and frank disclosure of criminal activities beyond the offences for which the offender is presently charged, amongst other indicators. As regards the last of these factors, the offender’s hitherto clean record and otherwise unexceptional conduct and temperament can be relevant in showing that the offences in question were likely an aberration.

38 Next, the second limb focuses on whether the offender’s environment presents conditions that are conducive in helping him turn over a new leaf. This may be discerned from the following non-exhaustive factors (Terence Siow at [57]):

39 If, after considering the first and second limbs, the court comes to a provisional view that the offender has demonstrated a sufficiently strong propensity for reform, the inquiry will shift, at the third limb, to the risk factors that are present in order to determine whether, in all the circumstances, the offender can indeed be said to have an “extremely strong propensity for reform”. Risk factors include the offender’s association with negative peers, or the presence of bad habits such as an offender’s habitual drug use or dependence: Terence Siow at [58].

40 If the court is satisfied that the adult s 377BB(4) PC offender concerned demonstrates an extremely strong propensity for reform after applying the three-limbed Terence Siow framework, it remains to be considered whether it is appropriate in all the circumstances to retain the emphasis on deterrence despite the offender’s extremely strong propensity for reform: Terence Siow at [45], [52] and [60]. Where, for instance, the offence is serious or the harm caused is severe, deterrence may displace rehabilitation as the dominant sentencing consideration even though the adult offender has demonstrated an extremely strong propensity for reform: see Terence Siow at [52]–[53], citing Boaz Koh at [30] and GCO v Public Prosecutor [2019] 3 SLR 1402. In this regard, both general and specific deterrence are relevant. General deterrence aims to educate and deter other like-minded members of the general public by making an example of a particular offender, whereas specific deterrence seeks to instil in a particular offender the fear of reoffending through the potential threat of re-experiencing the same sanction previously imposed: Public Prosecutor v Law Aik Meng [2007] 2 SLR(R) 814 (“Law Aik Meng”) at [21] and [24]. General deterrence assumes significance for certain types of offences (eg, offences affecting public safety and offences against vulnerable victims) and where the circumstances of the offence demands so (eg, the offence is prevalent or causes public disquiet): Law Aik Meng at [24]–[25]. In particular, it is appropriate to place emphasis on general and specific deterrence where the crime is premeditated because deterrence works best where there is a conscious choice to commit crimes: Law Aik Meng at [22].

41 This inquiry mentioned at [40] above is an important one which balances the s 377BB(4) PC offender’s extremely strong propensity for reform against aspects of the s 377BB(4) PC offence which point towards the need for general and specific deterrence. Whilst the offender’s extremely strong propensity for reform may signal a shift to a focus on rehabilitation at first instance, this may still give way to society’s need for general and specific deterrence in light of offence-related considerations. This balancing exercise is a highly nuanced one that will turn on the facts of each case.

42 Nevertheless, for the reasons ably articulated by the YIC, we take the view that deterrence is generally the dominant sentencing consideration for this offence, and it would rarely be the case that emphasis would shift away from deterrence to rehabilitation even where the s 377BB(4) PC adult offender has demonstrated an extremely strong propensity for reform. We elaborate.

43 First, in every case, the impact of a s 377BB(4) PC offence extends beyond the particular victim concerned because it offends the sensibilities of the general public and triggers unease. Bearing in mind that the offence entails the operation of an equipment with the intention of observing the victim’s private region in circumstances where that region would not otherwise be visible and where the victim did not give his or her consent, it goes without saying that the commission of a s 377BB(4) PC offence involves an appalling attempt to invade the victim’s privacy. At its core, a s 377BB(4) PC offence is an affront of our society’s fundamental value that no woman or man should have to suffer the indignity of having his or her modesty outraged or insulted (see Singapore Parliamentary Debates, Official Report (5 March 2021) vol 94 (Mr K Shanmugam, Minister for Home Affairs); ³⁰ Singapore Parliamentary Debates, Official Report (6 May 2019) vol 94 (Mr K Shanmugam, Minister for Home Affairs)). ³¹ Adding to the public disquiet occasioned by the transgression of our society’s fundamental belief, the discovery of a surreptitiously committed s 377BB(4) PC offence also affects the extent to which members of the public feel safe as they go about their daily life. This is the case regardless of whether the offence was committed in a private location, such as the victim’s own home, or a public location, such as an escalator in a shopping mall. In light of these consequences, it behoves the sentencing court to send a stern and unequivocal signal, on behalf of society, that s 377BB(4) PC offences will not be tolerated.

44 Second, a deterrent sentence would usually be warranted as s 377BB(4) PC offences often inflict significant emotional harm on the victim. The archetypal situation envisaged by s 377BB(4) PC is one where the offender had attempted to make a record of the victim’s image but was caught before any recording could be made (ie, a failed recording situation). This can be gleaned from the fact that ss 377BB(4) and 377BB(5) PC cover substantially the same act, save for the fact that the recording of the victim’s image is an ingredient of the s 377BB(5) PC offence whereas a s 377BB(4) PC offence is made out regardless of whether the victim’s image was recorded. In Mark Fritz Tanel, for example, the s 377BB(4) PC offender bent down to take an upskirt video of the victim on the escalator but was caught by the victim before he could record a video (at [6]–[7]). ³² In a failed recording situation such as in Mark Fritz Tanel, the victim would more often than not be aware that he or she had been a victim of voyeurism and would, as a result, suffer significant emotional distress. Such harm brings both general and specific deterrence to the fore.

45 Even where the s 377BB(4) PC offender had surreptitiously captured a record of the victim’s image without the victim’s knowledge, there remains a danger of considerable emotional harm being inflicted upon the victim post-offence given the ease with which such material can be disseminated through the Internet. The degree of emotional harm that could potentially be caused would be aggravated if the victim was identifiable from the record.

46 Third, as the s 377BB(4) PC offence entails the operation of an equipment with the intention to observe the victim’s otherwise not visible private region without the victim’s consent, the commission of the offence generally involves a degree of furtiveness, planning and premeditation on the part of the offender. This is yet another feature of the offence which warrants the imposition of deterrent sentences.

47 Fourth, the need to censure s 377BB(4) PC with deterrent sentences is augmented by the increasing prevalence of voyeurism offences, a trend noted by the Minister of Home Affairs during the Second Reading of the Criminal Law Reform Bill which introduced s 377BB PC (Singapore Parliamentary Debates, Official Report (6 May 2019) vol 94 (Mr K Shanmugam, Minister for Home Affairs)). ³³ The rise in the number of voyeurism offences is in part attributable to technological advancements that have facilitated the ease with which such offences can be stealthily committed. For instance, the production of cameras and lenses which are small enough to be placed discreetly below women’s dresses and skirts has made it much easier for a s 377BB(4) PC offender to observe the victim’s private parts without being caught. In these circumstances, the imposition of stiff sentences would be both timely and necessary in curbing the rising number of voyeurism offences.

48 Finally, keeping the emphasis on deterrence in most s 377BB(4) PC cases, even where the adult offender has demonstrated an extremely strong propensity for reform, accords with Parliament’s intention to place deterrence at the fore for such offences. The clearest indication of this intention is s 377BB(7) PC, which provides for a maximum imprisonment term of two years. This is double the maximum imprisonment term under s 509 PC, the earlier provision criminalising the insult of modesty of a women and under which voyeurism offences were previously prosecuted prior to the introduction of s 377BB PC. Although Parliament has recognised that an adult offender’s strong propensity for reform may justify placing rehabilitation as the primary sentencing consideration, it has indicated in no uncertain terms that the general sentencing position in respect of adult offenders who commit sexual offences, including voyeurism, is to prioritise deterrence over rehabilitation, and that only exceptional circumstances may justify deviation from this general position (Singapore Parliamentary Debates, Official Report (5 March 2021) vol 94 (Mr K Shanmugam, Minister for Home Affairs)): ³⁴

This supports our view that it would be rare for the emphasis to shift from deterrence to rehabilitation on the ground of the adult s 377BB(4) PC offender’s strong propensity for reform. That is not to say that this will never be the case. It will, however, take a truly exceptional case to warrant this.

49 Thus, it would be appropriate in most s 377BB(4) PC cases to retain the emphasis on deterrence despite the adult offender’s extremely strong propensity for reform, with the result that the sentences imposed for s 377BB(4) PC offences would likely include an imprisonment term. Even then, rehabilitation would remain a relevant but subsidiary sentencing consideration which can be given effect to by calibrating the overall imprisonment term downwards upon the application of the totality principle. This would avoid a crushing sentence that would destroy all prospects of the offender’s rehabilitation and reintegration (see Ang Zhu Ci Joshua v Public Prosecutor [2016] 4 SLR 1059 at [5] and [8]; Public Prosecutor v Raveen Balakrishnan [2018] 5 SLR 799 (“Raveen”) at [78]).

50 At this juncture, we make two preliminary points on the form of the sentencing framework and its applicability to other provisions under s 377BB PC. For the avoidance of doubt, as alluded to at [24] above, these points are equally applicable to s 377BB of the Penal Code 1871 (2020 Rev Ed), which is substantially the same as s 377BB PC discussed in this judgment.

51 In her written submissions, the YIC proposed the following sentencing framework: ³⁵

(a) Determine if the victim’s image was recorded.

(i) If not, the indicative sentencing range is a term of imprisonment ranging from four to 16 weeks per charge.

(ii) If so, the indicative sentencing range is a term of imprisonment ranging from six to 24 weeks per charge.

(b) Next, determine the appropriate sentence within the indicative sentencing range in light of the aggravating and mitigating factors.

(i) If there are no aggravating factors and only mitigating factors, the appropriate sentence to be imposed is a fine only.

(ii) If there are many aggravating factors, the court should consider whether lengthening the imprisonment term would result in a disproportionately long period of incarceration.

(A) If not, the imprisonment term should be lengthened.

(B) If so, the court should go on to consider if the case at hand is a serious case involving violence or significant disruption to public disorder and safety.

(I) If so, the court should impose caning in addition to the term of imprisonment.

(II) If not, the court should impose a fine in addition to the term of imprisonment.

The YIC submitted that it is only in the most egregious scenario that the Court should impose a sentence combining all three types of punishment on the convicted offender, as it represents the maximum end of the sentencing range.

52 The Prosecution submitted that the sentencing framework should be based on the two-stage, five-step sentencing framework set out in Logachev Vladislav v Public Prosecutor [2018] 4 SLR 609 (“Logachev”):

(a) Stage 1: Arrive at an indicative starting point sentence for the offender upon considering the intrinsic seriousness of the offending act. This involves three steps:

(i) Step 1: Identify the level of harm caused by the offence and the level of the offender’s culpability.

(ii) Step 2: Identify the applicable indicative sentencing range in a three-by-three matrix by reference to the level of harm caused by the offence (in terms of low, moderate and high) and the level of the offender’s culpability (in terms of low, moderate and high).

(iii) Step 3: Identify the appropriate starting point within the indicative starting range having regard to the level of harm caused by the offence and the level of the offender’s culpability.

(b) Stage 2: Make adjustments to the starting point sentence identified at stage 1. This stage involves two steps:

(i) Step 4: Adjust the starting point sentence having regard to offender-specific aggravating and mitigating factors.

(ii) Step 5: Where an offender has been convicted of multiple charges, make further adjustments, if necessary, to the sentence for the individual charges in the light of the totality principle.

53 On the other hand, the appellant advocated a sentencing bands approach. This approach requires the court to choose one out of three sentencing bands with reference to offence-specific factors before determining the precise sentence in the light of the offender-specific factors. ³⁶

54 In the hearing before us, the YIC indicated that she had reconsidered her views and was inclined towards the Logachev framework.

55 In our judgment, the appropriate sentencing framework for s 377BB(4) PC offences punishable under s 377BB(7) PC should follow the Logachev two-stage five-step framework.

56 The Logachev framework is gaining ground as the preferred sentencing framework for offences which admit of a wide variety of typical presentations, and voyeurism is one such offence. Judicial precedents of voyeurism offences previously prosecuted under s 509 PC show that one extremely common scenario is the recording of “up-skirt” videos by an offender who stands closely behind the victim on an escalator in a public facility, such as a shopping mall or an MRT station. ³⁷ Voyeurism cases also frequently involve an offender recording or peeping at the victim while the latter is using the toilet or bathroom. ³⁸ Recording “down-blouse” pictures and videos of victims in public areas is yet another typical instance of voyeurism. Moreover, as technology gradually advances, it is likely that the typical manifestations of voyeurism offences will incrementally diversify. The Logachev framework, which provides for a methodical evaluation of offender-specific and offence-specific factors without overemphasising any particular factor, is particularly apt for voyeurism offences such as s 377BB(4) PC, as it assists the sentencing court in systematically navigating the broad range of common situations in which the offence may manifest whilst giving the sentencing court sufficient latitude to respond to the distinctive features of a particular case.

57 We now turn to compare the Logachev framework against the sentencing bands approach proposed by the appellant. The sentencing bands approach is not too far apart from the Logachev framework. Both involve the selection of an indicative sentencing range in view of the offence-specific factors, followed by the identification of a precise sentence having regard to offender-specific factors. However, there are significant differences between the two which lead us to the view that the Logachev framework is to be preferred in so far as s 377BB(4) PC offences punishable under s 377BB(7) PC are concerned.

58 First, unlike the sentencing bands approach, the Logachev framework categorises the offence-specific factors into those relevant to the level of harm caused by the offence and those relevant to the level of the offender’s culpability. By requiring the sentencing court to reason along the lines of harm and culpability respectively, the Logachev framework facilitates a clearer and more systematic evaluation of the seriousness of an offence, and this in turn promotes the development of consistent and coherent sentencing precedents.

59 Second, the three-by-three matrix in the Logachev framework breaks down the overall sentencing range prescribed by legislation into five distinct indicative sentencing ranges, whereas the sentencing bands approach breaks down the same overall sentencing range prescribed by legislation into three distinct indicative sentencing ranges. As such, for the same offence, each indicative sentencing range in the sentencing bands approach will be broader than each indicative sentencing range under the Logachev framework. Generally, narrower indicative sentencing ranges promote consistency in methodology whereas broader indicative sentencing ranges potentially heighten the risk of inconsistency. Thus, broader indicative sentencing ranges should only be used when necessary, such as where the offence in question manifests itself in a “broader than usual spectrum” of factual circumstances and greater flexibility is required to calibrate the precise sentencing accordingly: see Goh Ngak Eng v Public Prosecutor [2022] SGHC 254 (“Goh Ngak Eng”) at [99].

60 In light of these differences, the choice between the two forms of sentencing framework turns on whether the offence-specific factors of the offence in question lend themselves to being categorised by reference to harm and culpability, and whether the circumstances in which the offence manifests are so diverse that there is a greater need for flexibility in sentencing. Both these considerations lead us to the view that the Logachev framework is to be preferred in so far as s 377BB(4) PC offences punishable under s 377BB(7) PC are concerned. As will be seen, the offence-specific factors of this offence can be meaningfully categorised into harm and culpability respectively. The circumstances in which this offence presents itself are also relatively circumscribed even though they cover a wide variety of typical presentations. Each sub-section from ss 377BB(1) to (6) PC, including s 377BB(4) PC, covers a specific set of circumstances in which the act of voyeurism can be committed (see above at [6]). Since s 377BB(4) PC covers a range of factual situations which engage offence-specific factors amenable to being categorised into harm and culpability respectively, the sentencing framework for s 377BB(4) PC offences punishable under s 377BB(7) PC should be modelled using the Logachev framework instead of the sentencing bands approach.

61 Finally, we observe that the YIC had rightly refrained from pressing for her proposed sentencing framework during the oral hearing. In our view, this proposed sentencing framework, which takes the form of a decision tree, is likely to be difficult for sentencing courts to apply. Further, as rightly pointed out by the Prosecution, the YIC’s proposed sentencing framework does not expressly account for the full breadth of the custodial term prescribed under s 377BB(7) PC. ³⁹ Ideally, to give effect to the full range of possible sentences intended by Parliament, the prescribed sentencing framework should take into account the whole range of penalties prescribed so that sentencing courts can determine precisely where the offender’s conduct falls within that range and avoid meting out sentences which are arbitrarily clustered in a particular segment of the full range: see Ong Chee Eng v Public Prosecutor [2012] 3 SLR 776 at [24]; Huang Ying-Chun v Public Prosecutor [2019] 3 SLR 606 (“Huang Ying-Chun”) at [79].

62 Given the facts of this case, the sentencing framework in this judgment only concerns s 377BB(4) PC offences punishable under s 377BB(7) PC. Though we see no reason why the same sentencing framework should not apply to the other offences in ss 377BB(1), (2), (3), (5) and (6) PC should they also be punishable under s 377BB(7) PC, we prefer to leave this point to be considered in an appropriate future case.

63 We now turn to set out the two-stage, five-step sentencing framework applicable to the offence under s 377BB(4) punishable under s 377BB(7) PC.

64 At the first step, the court has to consider the offence-specific factors and identify the level of harm caused by the offence as well as the level of the offender’s culpability (Logachev at [76]). In this regard, we are generally in agreement with the list of non-exhaustive factors proposed by the Prosecution and YIC but have structured the harm factors in a different manner.

65 The harm caused by a s 377BB(4) PC offence can be categorised into three types, and each type may arise only under certain circumstances:

(a) Invasion of the victim’s privacy, in cases where there was unwanted observation of the victim’s private region, or the retention or dissemination of a record of the victim’s image.

(b) Violation of the victim’s bodily integrity, in cases where the offender made physical contact with the victim in the course of committing the s 377BB(4) PC offence.

(c) Humiliation, alarm or distress, in cases where the victim was made aware of the offending conduct.

66 We have organised the concept of harm into these categories to stress that the victim’s lack of knowledge of having been subjected to a s 377BB(4) PC offence does not necessarily mean that no harm was caused – whilst such a victim would not have subjectively suffered emotional harm, harm in the form of a loss of privacy, the degree of which is to be determined by reference to objective indicia (see below at [67]–[71]), might still result from the offence. On the flipside, if the victim was aware of the offending conduct, the humiliation, alarm or distress suffered would be affected, in part, by the extent to which his or her privacy was invaded. Bearing in mind the rule against double counting, the court should only take into account the emotional harm subjectively experienced by the victim in so far as it exceeds that which is objectively inferred from the extent of the invasion of privacy.

67 We begin with the first type of harm, viz, the invasion of the victim’s privacy. Harm in the form of actual or potential invasions of privacy may arise where the offender observes the victim, retains a record of the victim’s image or disseminates a record of the victim’s image. We will elaborate on each of these in turn.

68 First, the more intrusive the observation, the greater the actual loss of privacy. The intrusiveness of the observation is in turn a function of several objective factors, such as the extent of the body parts under observation, how exposed those body parts were, the duration of observation and the number of other persons who were enabled by the offender to observe the victim’s private regions. The last of these factors concerns persons who observed the victim’s private regions while the offence was being committed, as opposed to persons who viewed a record of the victim’s image following the offender’s post-offence dissemination of the same. The latter will be addressed at [71] below.

69 Second, the offender’s retention of a record of the victim’s image gives rise to the potential for repeated invasions of the victim’s privacy long after the s 377BB(4) PC offence had taken place – the offender is able to view the record repeatedly for his or her own perverted pleasure, allow others to view the same on his or her device, and even circulate it to others through the Internet. This potential harm is further amplified if the victim is identifiable from the record (eg, the photo or video reveals the victim’s face).

70 Where the offender had made a record of the victim’s image but subsequently deleted it, the potential for harm might be curtailed. That said, actual harm might have already been caused to the victim post-offence if the offender had viewed or circulated the record prior to its deletion. We add that the offender’s act of deletion may not have mitigating value, and may even be an aggravating factor, depending on the offender’s motivation for deleting the record. If, for instance, the act of deletion was an attempt to destroy evidence, the deletion would be an aggravating factor going towards the offender’s culpability. However, if the deletion was a result of the offender realising the reprehensibility of his actions, the deletion may be indicative of the offender’s remorse.

71 Third, where the offender not only retained a record of the victim’s image but disseminated it, the act of dissemination represents a significant incursion into the victim’s privacy over and above observing and recording the victim. If, however, the offender is already facing a separate proceeded charge under s 377BE(1) PC for distributing an intimate image or recording of the victim, the same act of distribution cannot be regarded as an aggravating factor for the s 377BB(4) PC offence in the light of the rule against double counting: Kow Keng Siong, Sentencing Principles in Singapore (Academy Publishing, 2nd Ed, 2019) (“Sentencing Principles in Singapore”) at para 08.082. In cases where the s 377BE(1) PC charge is stood down, the courts may consider the act of distribution either as an aggravating factor for the sentencing of the s 377BB(4) offence, or take into consideration the s 377BE(1) PC charge for the purpose of sentencing. Where the act of dissemination can be properly regarded as an aggravating factor for the s 377BB(4) PC offence, the extent of the invasion of privacy arising from the dissemination can be evaluated by reference to the objective indicia set out in Public Prosecutor v GED [2022] SGHC 301 (“GED”) at [51]–[53], such as how widely the record was distributed, whether it was distributed to certain recipients known to the victim, and the degree of identifiability of the victim.

72 We note that the s 377BB(4) PC offence may be committed even in circumstances where the offender had not observed, recorded or disseminated recordings of the victim. Though there would be an absence of the aforementioned harm in such cases, the intrusiveness of the offender’s intended observation of the victim (eg, whether the offender intended to observe a fully nude or partially dressed victim), and whether the offender intended to record or disseminate a record of the victim’s image, may be relevant in evaluating the culpability of the offender. All other things being equal, an offender who intended to record the victim (but failed to do so) is more culpable than an offender who merely intended to observe the victim, given that the retention of such a record harbours the potential for prolonged incursions into the victim’s privacy long after the offending conduct had taken place. For completeness, we add that the offender’s omission to record or disseminate such recording should not be accorded mitigatory weight as the absence of an aggravating factor cannot, as a matter of logic, be called in aid as a mitigating factor: see Edwin s/o Suse Nathen v Public Prosecutor [2013] 4 SLR 1139 at [24].

73 Next, the victim’s bodily integrity may be violated where the offender made unwanted physical contact with the victim whilst committing the s 377BB(4) PC offence. This may happen, for instance, where the offender accidentally bumped the victim on the back of her knee while using a device to take an “up-skirt” photo or video. The offender will be placed higher up on the spectrum of harm if, for example, he or she made prolonged physical contact with the victim’s private parts or visited violence upon the victim in the course of committing the s 377BB(4) PC offence.

74 Victims of s 377BB(4) PC offences may suffer emotional harm contemporaneous to the offence as well as subsequent psychological harm (eg, post-traumatic stress disorder). The existence and extent of the emotional harm suffered may either be disclosed by the victim or inferred from the circumstances and may, in appropriate cases, need to be corroborated by evidence (including expert evidence). In particular, where it is alleged that the victim had developed conditions such as depression or post-traumatic stress disorder as a result of the offending conduct, such allegations should be supported with evidence in the form of victim impact statements and medical reports: see GED at [56].

75 Generally, the subjective emotional trauma experienced by the victim would be contingent on the extent to which the victim’s privacy was invaded and other victim-specific factors such as the vulnerability of the victim, which may be a relevant consideration where the victim is relatively young, has pre-existing mental conditions or has a relationship with the offender that renders the victim susceptible to being manipulated or taken advantage of. Where the victim was made aware of the offending act and the objective indicia at [67]–[71] above have been accounted for in the evaluation of the extent of the invasion of privacy, the court must avoid double counting the same factors when assessing the subjective emotional harm experienced by the victim, and only take into account the emotional harm suffered in so far as it exceeds that which is objectively inferred from the extent of the invasion of privacy.

76 The factors going towards culpability are as follows:

When evaluating the offender’s level of culpability, the sentencing court must guard against the risk of double counting. For instance, the assessment of the degree of pre-meditation should exclude factors relating to the use of equipment if considerations relating to the type and sophistication of the equipment used, as well as whether it was concealed, are also taken into account as standalone factors going toward culpability. With this in mind, we now turn to elaborate on a few of these factors.

77 As regards the factor at [76(a)] above, an offender who had actual knowledge that the victim did not consent to the offending conduct and an offender who merely had reason to believe that the victim had not consented to the offending conduct would both be caught by s 377BB(4)(b) PC. However, the culpability of the former would be higher. As explained by this court in Huang Ying-Chun at [74]:

78 In relation to the factor at [76(d)] above, the operation of equipment with recording capability (eg, a mobile phone with an in-built camera) is strongly indicative of the offender’s intent to make a record of the victim. This, as noted earlier at [72] above, is a factor raising the culpability of the offender in situations where the offender intended to make a record of the victim’s image but had not in fact done so.

79 Turning now to the factor at [76(e)], a breach of a relationship of trust may arise where the offender and the victim are husband and wife (or vice versa) (see Public Prosecutor v GEZ [2022] SGMC 59 at [104]), ⁴⁰ landlord and tenant (or vice versa) (see Tan Pin Seng v Public Prosecutor [1997] 3 SLR(R) 494 at [34]), ⁴¹ or work colleagues who share an office space (see Public Prosecutor v Lau Zongming [2021] SGMC 71 at [56]). ⁴²

80 Regarding the factor at [76(g)] above, the offender’s motivation for the offence refers to why the offender committed the offence. Depending on the offender’s precise motive(s), the offender’s motive may either heighten or reduce the offender’s culpability: see Ye Lin Myint v Public Prosecutor [2019] 5 SLR 1005 at [53]. The offender’s culpability may be higher in instances where, for example, the s 377BB(4) PC offence was committed out of spite or for the purpose of commercial exploitation. The case of Prasanth s/o Mogan [2022] SGDC 209 is one instance in which the s 377BB(4) PC offence was committed out of malice. There, the offender bore a personal grudge against the victim and had deliberately humiliated the victim by stripping him naked and instructing him to dance while he recorded a video.

81 In relation to the factor at [76(h)] above, this factor ought to be distinguished from the offender’s prolonged pattern of offending which extends beyond the subject of the charge. It relates to the persistence of the offending conduct in question and may be a relevant consideration in situations where the offender continued to observe the victim even after the victim had made efforts to prevent or warn the offender against continuing the offending behaviour, such as by shouting at the offender.

82 The second and third steps of the sentencing framework requires the court to identify the applicable indicative sentencing range before proceeding to identify the appropriate starting point within that range.

83 The sentencing regime under s 377BB(7) PC gives the court the discretion to impose a term of imprisonment of up to two years, a fine, caning, or any combination of such punishments. Bearing in mind the need to make full use of the available statutory sentencing range, and balancing that against the sentencing objectives of each type of punishment, we adopt the following sentencing matrix which is promulgated on the basis of a first offender who is convicted after trial:

84 This sentencing matrix is similar to the Prosecution’s proposed sentencing matrix, save that we have introduced the possibility of caning in the central three diagonal cells (high harm-low culpability, moderate harm-moderate culpability and low harm-high culpability). In our judgment, the offences which involve either high harm or high culpability, as well as the offences on the more egregious end of the moderate harm-moderate culpability scale, may undermine social safety to such an extent that necessitates the extremely strong deterrent effect which is secured through the imposition of caning: see Sentencing Principles in Singapore at para 29.008. Caning will be further warranted if the moderate to high harm in a particular case flows from an act of violence against the victim, in which case the additional sentencing objective of retribution is engaged: see Public Prosecutor v BDB [2018] 1 SLR 127 at [76]; ⁴³ Sentencing Principles in Singapore at para 29.005.

85 Separately, regardless of where the offender is placed in the sentencing matrix, the option of imposing fines in addition to an imprisonment term should be considered if the offender had procured financial benefits from his offending conduct (eg, the offender was paid to procure videos of unsuspecting victims), such that it is necessary to disgorge the offender of his unlawful gains (see Goh Ngak Eng at [100]). Where, however, the profit has already been surrendered, confiscated or it has been established that the profits made has already been squandered and the offender has no means to pay any fine imposed, the imposition of a fine may not be necessary (Tan Gek Young v Public Prosecutor and another appeal [2017] 5 SLR 820 at [75]). ⁴⁴

86 Nevertheless, notwithstanding that the offender had not profited from his offending conduct, it may be appropriate to impose fines without an accompanying imprisonment term provided that the offence lies at the less severe end of the low harm-low culpability spectrum. The level of deterrence warranted by such an offence may not justify a custodial sentence and may be sufficiently met with the imposition of a fine. That said, we are of the view that s 377BB(4) PC cases will typically cross the custodial threshold given the intrinsic seriousness of the offence (see above at [43]–[46]), and we stress that it will only be in the less severe of the low harm-low culpability cases that a fine may sufficiently advance the sentencing objective of deterrence.

87 At the fourth step of the sentencing framework, the court must adjust the starting point sentence having regard to offender-specific aggravating and mitigating factors. We set out a non-exhaustive list of these factors below:

As these offender-specific factors are generally applicable across all criminal offences and are well settled in our criminal jurisprudence, we will not elaborate further on this point.

88 The fifth step of the sentencing framework is engaged where the offender has been convicted of multiple charges, in which case the court has to make further adjustments to the sentence for the individual charges in keeping with the totality principle.

89 The totality principle ensures that the aggregate sentence is sufficient and proportionate to the offender’s overall criminality through a two-limbed analysis: the first limb examines whether the aggregate sentence is substantially above the normal level of sentences for the most serious of the individual offences committed, whereas the second limb considers whether the effect of the sentence on the offender is crushing and not in keeping with his past record and his future prospects (see Mohamed Shouffee bin Adam v Public Prosecutor [2014] 2 SLR 998 (“Shouffee”) at [54] and [57]; Raveen at [73]). Further, when considering the totality principle, the sentencing court must ensure that the total term of imprisonment for sentences that are ordered to run consecutively exceeds the term of imprisonment that is imposed for the highest individual sentence: Shouffee at [77].

90 Where the aggregate sentence is considered excessive upon the application of the totality principle, the sentencing judge may opt for a different combination of sentences to run consecutively or adjust the individual sentences: Shouffee at [59]; Raveen at [73]. Conversely, if the overall sentence would otherwise be inadequate in reflecting the offender’s overall criminality, the court can order more sentences to run consecutively or make upward adjustments to the individual sentences: Shouffee at [80]; see also Gan Chai Bee Anne v Public Prosecutor [2019] 4 SLR 838 at [20]. We add that when choosing which of the multiple sentences of imprisonment should run consecutively, the sentencing judge must guard against double counting and refrain from choosing a combination of longer sentences on account of aggravating factors that were already considered when calibrating the sentences for each individual offence: Shouffee at [79].

91 The first factual issue that arises for our consideration is whether the DJ erred in finding that deterrence, instead of rehabilitation, should be the dominant sentencing principle. As we have reiterated above, it would take an exceptional case to warrant a shift of emphasis from deterrence to rehabilitation. In our judgment, this case does not come close to being such an exceptional case.

92 In the first place, we are not persuaded that the appellant had displayed an extremely strong propensity for reform under the three-limbed framework in Terence Siow such as to take him out of the space where deterrence is the key consideration. The first limb requires a consideration of whether the appellant has demonstrated a positive desire to change since the commission of the offences. In this regard, we note that the appellant had reoffended while on bail for the first offence. Though he subsequently took active steps to address his voyeuristic urges by regularly attending psychiatric reviews and psychotherapy sessions after the second offence, we are particularly troubled by the fact that he fought hard to prove that that his offence was a result of a voyeuristic disorder, only to drop it on the day of the Newton hearing (see above at [14]). This, in our view, is an attempt to evade the due consequences under the law and suggests the appellant’s lack of willingness to accept responsibility for what he did. On balance, we find that the appellant has only demonstrated only some positive desire to change, as evinced by his efforts to seek psychiatric help. At the second limb of the Terence Siow framework, we accept that there are no evidently strong risk factors, and the appellant may have a positive external support system from his family and romantic partner. Nevertheless, given our view under the first limb, we do not take the view that the appellant has demonstrated an extremely strong propensity for reform, though we are willing to accept that the appellant has shown some propensity for reform.

93 In any event, even if the appellant had displayed an extremely strong propensity for reform, the emphasis would remain on deterrence for the reasons set out at [43]–[48] above, as well as the specific circumstances of the present case. In particular, the need for deterrence is buttressed by the fact that the second offence was committed against a relatively vulnerable victim, who experienced considerable distress upon catching the appellant red-handed, and the fact that both offences involved an element of premeditation as the appellant had stalked his victims for a short while before committing the offences.

94 If it were proven that the appellant had suffered from voyeuristic disorder at the time of the offences, and that disorder contributed to his offending conduct, we might have taken a different view on the appellant’s propensity for reform and the dominant sentencing consideration at play. However, we say no more on this point given that the appellant is no longer alleging that he suffered from voyeuristic disorder at the time of his offending.

95 We now turn to consider whether the DJ erred in imposing a seven weeks’ imprisonment term. As a preliminary point, the sentence imposed by the DJ may be set aside even if it is not manifestly excessive or inadequate, as the introduction of a new sentencing framework warrants a re-evaluation of the sentence in principle: GED at [119], citing Public Prosecutor v Manta Equipment (S) Pte Ltd [2022] SGHC 157 at [48].

96 In our judgment, both offences committed by the appellant are in the low harm category. Although the appellant invaded the victims’ privacy by observing and recording videos of them, the degree of the invasion of privacy is limited – the appellant observed only the victims’ covered genitals for a brief period of time, and the video recordings were deleted shortly after the offences were committed. There is no indication that the appellant had viewed or disseminated the videos prior to deleting them. While both offences are in the low harm category, the harm caused by the second offence is higher for two reasons. First, the appellant touched V2’s thigh but did not make any physical contact with V1. Second, V2 was considerably more traumatised, as can be gathered from the fact that she called her father and cried shortly after she caught the appellant in the act. In our view, V2’s relative youth had exacerbated the level of distress she experienced from the invasion of privacy and the violation of her bodily integrity.

97 Next, we take the view that the appellant’s culpability for both offences falls within the low category. Though the appellant followed both victims, an act which suggests a degree of determination, the appellant only did so for a brief duration leading up to the commission of the offences.

98 As both offences involve low harm and low culpability, the applicable indicative sentencing range is a fine or up to four months’ imprisonment. The offence-specific factors considered above lead us to the view that a starting point of three weeks’ and four weeks’ imprisonment for the first and second offences respectively would be appropriate. The circumstances of both offences are such that the custodial threshold is crossed, and there is nothing about the case that suggests that the interests of deterrence would be adequately met by the imposition of a fine.

99 We now turn to adjust these starting points with reference to offender-specific factors, starting with the first offence. After taking into account the TIC Charge, and balancing that against the appellant’s guilty plea, cooperation with authorities (the NTU investigation officer and the police), as well as the fact that the appellant has demonstrated some (though not extremely strong) propensity for reform, we calibrate the starting point of three weeks’ imprisonment to one week’s imprisonment. The first offence is a fairly standard iteration of the offence, and we therefore think that an imprisonment term of one week would be appropriate in all the circumstances.

100 The sentencing discount for the second offence, however, will be much less given that the appellant had the audacity to reoffend on bail slightly more than four months after he was arrested for his first offence. As the appellant is no longer relying on his alleged voyeuristic disorder in his mitigation, the appellant must be treated as a normal 24-year-old adult who retained the mental ability and capacity to control himself at the time of the offences. The fact that he chose not to exercise self-control and reoffend while on police bail speaks volumes about the appellant’s lack of remorse and blatant disregard of the law. Taking these alongside the appellant’s guilty plea, cooperation with authorities and his prospect for reform, a slight downward calibration from four to three weeks’ imprisonment would be justified.

101 Both sentences should run consecutively for an aggregate term of four weeks’ imprisonment.

102 For completeness, we add that it is unclear whether the appellant’s deletion of the videos shortly after both offences were motivated by his sense of guilt or remorse, or constituted an attempt to destroy evidence because both victims were alerted to his wrongdoing. In these circumstances, the appellant’s deletion of the videos is at best a neutral factor.

103 We accordingly set aside the sentences that were imposed by the DJ and substituted in their place a sentence of one week and three weeks’ imprisonment respectively, for an aggregate imprisonment of four weeks.

104 Finally, we would like to express our deep gratitude to the YIC, Ms Li, for her thorough research and comprehensive submissions on the legal issues raised in this appeal.