Benjamin Low

Leaving an Empty Seat: Wong Souk Yee’s answer to By-Elections in a Group Representation Constituency

By Benjamin Low Junjie

I. Introduction

In Vellama d/o Marie Muthu v Attorney-General1 [Vellama], the Court of Appeal held that Article 49 of the Constitution of Singapore,2 which deals with the filling of vacant Parliamentary seats, imposed a duty on the Prime Minister to call a by-election “to fill casual vacancies of elected MPs which may arise from time to time”.3 However, despite the supposed universality of its conclusion, the Court of Appeal was quick to confine its judgment to vacancies arising in the context of Single-Member Constituencies (“SMCs”), having observed that separate legislative provisions and rules applied to instances where vacancies arose in a Group Representation Constituency (“GRC”).4 It remained an open question then as to what the proper scope and effect of Article 49 was on GRCs.

The recent High Court decision of Wong Souk Yee v Attorney-General5 [Wong Souk Yee] has since sought to address that question by holding that by-elections for a GRC would only be held in the event that all Members of Parliament (“MPs”) of that GRC have vacated their seats. This article will analyse the reasoning and the methodology of the High Court in reaching its conclusion and will discuss the implications of the judgment in the realm of constitutional interpretation.

II. Background

In Wong Souk Yee, the Applicant sought mandatory orders requiring that the MPs for the Marsiling-Yew Tee GRC vacate their seats in Parliament and that the Prime Minister advise the President to issue a writ of election for the Marsiling-Yew Tee GRC.6 Her action was founded on the basis that a vacancy had arisen in the Marsiling-Yew Tee GRC when one of its MPs, Madam Halimah Yacob, vacated her seat in Parliament on 7 August 2017 to contest in the 2017 Presidential Election.7 Since Madam Halimah’s departure, the Marsiling-Yew Tee GRC continued to be represented by its three remaining MPs from the People’s Action Party.8

The Applicant further sought declaratory orders in the alternative that s 24(2A) of the Parliamentary Elections Act9 [PEA] be interpreted as requiring all the MPs of a GRC to vacate their seats (i) when one or more MPs of the GRC vacate their seats; or (ii) in the alternative, when the only MP in the GRC who belongs to a minority community vacates his or her seat, so as to ensure consistency between s 24(2A) of the PEA and Article 49 of the Constitution.10 On the flip side, if such an interpretation was not possible, s 24(2A) was to be declared void for inconsistency with the Constitution as per Article 4 read with Article 49.11

III. The High Court’s Decision

A. The Mandatory Orders

The Applicant justified her claim for the mandatory orders on three grounds. Firstly, she contended that Article 49(1) of the Singapore Constitution 12 requires the calling of a by- election in a GRC when the seat of an Elected MP has become vacant for any reason other than the dissolution of Parliament. Secondly, the Applicant argued that Article 39A(2) of the Constitution13 requires that a GRC be represented by an MP belonging to a minority community, until the dissolution of Parliament. Last but not least, it was argued that a by-election should be called so as to give effect to the voters’ right to be represented by an elected MP of their choice.14

Much of the judicial debate in Wong Souk Yee centred on the Applicant’s first ground, which in effect concerned the proper interpretation of Article 49. For completeness, the full text of Article 49 reads as such:

Filling of vacancies

49.—(1) Whenever the seat of a Member, not being a non-constituency Member, has become vacant for any reason other than a dissolution of Parliament, the vacancy shall be filled by election in the manner provided by or under any law relating to Parliamentary elections for the time being in force.

(2) The Legislature may by law provide for —

(a) the vacating of a seat of a non-constituency Member in circumstances other than those specified in Article 46;

(b) the filling of vacancies of the seats of non-constituency Members where such vacancies are caused otherwise than by a dissolution of Parliament.

Article 49 in its present form was left untouched by the 1988 Amendments15 to the Constitution implementing the GRC system. Prior to that, the last substantive change to Article 49 was in 1984 concerning the introduction of Non-Constituency Members of Parliament.16 No other amendments were made to Article 49 since then. Indeed, when Article 49 was first enacted in the 1980 Reprint of the Constitution, 17 the GRC system was not even in existence. As Chua Lee Ming J succinctly identified, the true question that the High Court had to address, was the correct interpretation of Article 49(1) was in relation to GRCs.18 The earlier decision of Vellama had only addressed the scope of Article 49 in the context of an SMC.

The Applicant argued, on a plain reading of Article 49(1), that a by-election had to be called whenever a Parliamentary seat was rendered vacant, regardless of whether the seat belonged to an SMC or a GRC.19 Thus, whenever a vacancy arose in a GRC, the remaining MPs would have to resign their seats in order for a writ of election to be issued for the entire GRC. This would have to occur even if these MPs had no intention to vacate their respective seats, and where the conditions for disqualification or termination under Articles 45 and 46 were not made out.20 This interpretation of Article 49 was necessary in order to give substantive effect to the requirement in s 27A of the PEA 21 that elections in GRCs had to be held between teams of candidates.

The Attorney-General submitted in response that Article 49(1) only required that a by-election in a GRC be held if all the MPs of that GRC vacated their seats mid-term or, in the alternative, that Article 49(1) did not apply to GRCs but only to SMCs.22

In accepting the Attorney-General’s submissions, Chua J noted that the Applicant’s purported interpretation of Article 49(1) was a legal improbability given that there was no provision in Articles 45 and 46 of the Constitution requiring that the remaining MPs of a GRC had to vacate their seats or were disqualified from holding their seats in the event one of their fellow MPs vacated his or her seat.23 This is correct insofar as the Constitution is silent on such a contingency occurring whereas s 24(2A) of the PEA24 specifically mentions that the writ of election would only be issued in the event all MPs in a GRC have vacated their seats. To imply another constitutional ground on which an MP’s seat (or a group of MPs in this case) would either be vacated or the MP be disqualified from becoming a Member of Parliament would be tantamount to reading-in additional provisions in the Constitution by the Judiciary in the absence of clear or express words to that effect.25

Chua J further held in favour of the Attorney-General’s interpretation of Article 49(1) on the basis that either an updating or rectifying construction could be applied to Article 49(1) to take into account the subsequent legislative amendments to the Constitution and the Parliamentary Elections Act, thereby allowing for an interpretation of Article 49(1) to mean that by-elections in a GRC would only be possible if all MPs had vacated their seats. 26

An updating construction is first and foremost a rule of statutory interpretation that is meant to provide guidance “as to the legislator’s prima facie intention regarding the legal meaning of the enactment.”27 For an updating construction to take effect, the Act in question must first have been an Act “that is intended to develop in meaning with developing circumstances”.28 Once this has been established, the updating construction operates by presuming that “Parliament intends the court to apply to an ongoing Act a construction that continuously updates its wording to allow for changes since the Act was initially framed”.29 In Singapore, the rule of updating construction was first expressly adopted in Comptroller of Income Tax v MT 30 [CIT]. This was in turn cited in Wong Souk Yee with approval,31 as part of a three-stage framework (the “CIT Framework”).32

In applying the CIT Framework to the present case, Chua J determined that since Parliament expressly intended for by-elections in a GRC to be held only in the event that all seats in that GRC had been vacated33 and had made its intentions manifest by amending the Constitution to include Article 39A, an updating construction could be applied to Article 49(1) to reflect the changes effected by Article 39A.34

B. Difficulties with Applying an “Updating Construction”

With respect, the learned Judge did not appear to have adequately considered the nature of the legislation he was required to interpret in accepting the applicability of the doctrine of updating construction. As previously stated, the presumption of updating construction is a rule of statutory interpretation. It owes its development to the peculiar characteristics of the English legal system, where there exists no written Constitution and the doctrine of Parliamentary supremacy prevails.35 Given such differences between our legal systems, the Singapore courts ought to guard against resorting to ordinary canons of statutory interpretation too quickly in order to interpret a written constitution that proclaims itself to be the supreme law of the land. To infelicitously import rules of interpretation and construction for ordinary statutes wholesale into the realm of constitutional interpretation runs the risk of treating the Constitution as being no different from an ordinary piece of statute, even if such an interpretation is justified as giving effect to the purpose of the affected legislation.36

Even if one accepts that the rule of updating construction can be applied in the context of constitutional interpretation, it is arguable that at the third stage of the CIT Framework, the application of an updating construction to Article 49(1) would itself have entailed a substantive change to the provision. With respect, a constitutional amendment may have been the more desirable solution. As the Court of Appeal in Vellama so held, Article 49(1) in its present form mandates that the Prime Minister has to call by-elections “to fill casual vacancies of elected MPs that may arise from time to time”.37 This mandatory requirement is not subject to qualifications whatsoever on a plain reading of the text. Neither is it reflected in the 1988 Constitutional Amendment that Parliament intended for by-elections in GRCs to be subject to the requirement that all MPs vacate their seats first. Not even Article 39A itself makes any reference to the requirements for by-elections to GRCs, save that they be held only on a “team” basis.38 Furthermore, the PEA’s unique requirements that must be met for a by-election in a GRC are themselves a significant amendment to the existing constitutional provision on by-elections since the departure of one or more MPs from a GRC would not trigger the by-elections mechanism unless every MP has vacated his or her seat. The result is that in practice, Parliamentary vacancies in a GRC would be left unfilled until the remaining MPs’ terms of office ends.

That being the case, the proper remedy would lie with Parliament to amend Article 49 itself to accommodate s 24(2A) of the PEA. As Menon JC observed in CIT, where a “significant legislative change has been effected in the first statute such that its importation into the second statute may have quite substantial effects, the court should ordinarily be slow to apply an updating construction to the second statute”.39

C. Rectifying Construction: A Trojan Horse of Judicial Legislation?

A similar argument may be made about the Court’s willingness to apply a rectifying construction towards Article 49(1). Rectifying construction operates on the basis that “the legislator intends the court to apply a construction which rectifies any error in the drafting of the enactment”.40 Its effect is similar to that of an updating construction, and essentially involves the addition or substitution of words into the statute to rectify the error or omission in the legislative drafting. In accepting the Attorney-General’s argument, Chua J reasoned that Parliament’s intention regarding by-elections in GRCs was manifested by the adoption of s 24(2A) of the PEA.41 That Article 49(1) was not also amended to reflect this change in the law was merely an omission on the Legislature’s part which in no way detracted from the express intention of Parliament.42

With respect, it is still difficult to reconcile Chua J’s holding with the fact that the PEA is itself a piece of ordinary legislation that must inevitably rank below the Constitution in terms of legal hierarchy.43 Constitutional supremacy as reflected in Article 4 mandates that any ordinary legislation that would at first blush be inconsistent with the Constitution ought to rendered void to the extent of the inconsistency unless the Constitution itself has been expressly amended.44 S 24(2A) arguably constitutes a marked deviation from the constitutional duty in Article 49 to fill Parliamentary vacancies through the use of elections. It follows that an amendment to Article 49(1) is necessary to reflect this change, and in the absence of such, s 24(2A) is void for inconsistency qua Article 49(1).45

Having accepted the Attorney-General’s interpretation of Article 49(1), the High Court proceeded to consider the Attorney-General’s alternative interpretation that Article 49(1) only applied to SMCs and not to GRCs. In agreeing with the Attorney-General’s alternative interpretation, Chua J reiterated his earlier holding that the legislative purpose was clear in that:

“[A]ll elections (including by-elections) in any GRC shall be held on a basis of such number of candidates as designated for that constituency by the President, and no by- election needs to be held to fill any vacancy in a GRC unless all the Members in that GRC have vacated their seats.”46

The Attorney-General’s alternative interpretation of Article 49(1) was thus adopted as the correct one that “furthers the legislative purpose”.47 Yet it seems unclear whether the legislative purpose that the High Court was referring to was the specific purpose of Article 49(1) or the general purpose of Part VI of the Constitution or even the Constitution as a whole. If Chua J was referring to the specific purpose of Article 49(1), then insufficient consideration was devoted to ascertaining the legislative purpose of Part VI of the Constitution, as per the test in Tan Cheng Bock v Attorney-General48 [Tan Cheng Bock]. If, however, the learned Judge was referring to the legislative purpose of Part VI of the Constitution, his definition of the legislative purpose would appear to be an unduly narrow one. One must consider that Part VI deals wholly with the functions, organisation as well as composition of the Legislature of Singapore in its entirety, rather than simply on by-elections alone.49

Since the Applicant’s claim could not succeed on her first ground, the High Court proceeded to consider the remaining two grounds underpinning her claim for mandatory orders. However, as the Applicant clarified in her oral submissions, her case was only that a by-election in a GRC had to be called if any seat in the GRC was vacated regardless of whether it was held by an MP belonging to a minority community or otherwise. She thus conceded her second ground that Article 39A(2) of the Constitution requires that a GRC be represented by an MP belonging to a minority community until the dissolution of Parliament.50 The High Court further rejected her third ground of argument, that there existed an implied right by voters to be represented by an elected MP of their choice until the dissolution of Parliament.51

D. The Declaratory Orders

Given that the Applicant’s alternative claim for declaratory orders was founded on similar grounds as her claim for mandatory orders, the High Court rejected the Applicant’s alternative claim for the same reasons that it gave in rejecting the Applicant’s primary claim – namely that the Applicant’s proposed interpretation of Article 49(1) was patently incorrect. 52 Nevertheless, this author contends that, given the legal impossibility of the Applicant’s interpretation of Article 49(1) and the inherent risks of applying either an updating or rectifying construction to the Constitution, the irresistible conclusion appears to be that s 24(2A) of the PEA is simply inconsistent with the Constitution and thus void. It is unfortunate that the High Court did not canvass this particular point in greater detail, being content to simply reject it as stemming from the Applicant’s incorrect interpretation of Article 49(1).

IV. Implications for Constitutional Interpretation

Wong Souk Yee appears to reinforce the proposition that in future cases, the purposive approach53 will increasingly become the dominant method of constitutional interpretation.54 Coming on the back of the earlier decision of Tan Cheng Bock, which was itself heavily cited in Wong Souk Yee, it would seem that future cases dealing with questions of constitutional interpretation will inevitably concern themselves with attempting to ascertain the purpose of a particular Constitutional provision or a part of the Constitution. Yet, two potential issues have arguably arisen from the present case that necessitate further clarification from the Singapore courts.

Firstly, the apparent willingness of the High Court in Wong Souk Yee to resort to interpretive doctrines for ordinary legislation in interpreting the Constitution intimates their general applicability across different types of legislative instruments. It is true that Tan Cheng Bock itself appeared to accept the applicability of such rules in the context of constitutional interpretation.55 However, care must be taken against bluntly importing these rules into the separate realm of constitutional interpretation, especially since they were developed by the Courts in common law, and intended for ordinary statutes. A liberal use of such rules and canons runs the risk of the courts straying into the field of constitutional amendment under the guise of interpretation, especially if such rules have the practical effect of injecting additional words and expressions into the Constitution when none previously existed. Further clarification from the Court of Appeal as to the ambit and applicability of these rules would be greatly welcomed.

Secondly, this increasing acceptance of the purposive approach as the primary method of constitutional interpretation poses a corollary question: whether the alternative rights-based approach, as laid out in the long line of cases stretching from Minister of Home Affairs v Fisher 56 [Fisher] to Ong Ah Chuan v Public Prosecutor 57 [Ong Ah Chuan] and then to Taw Cheng Kong (which cited both Fisher and Ong Ah Chuan with approval), will continue to be of any particular relevance in field of constitutional law. It will be recalled that in Ong Ah Chuan, which cited Fisher and referred to the Singapore Constitution, the Privy Council held that “the way to interpret a Constitution on the Westminster model is to treat it not as if it were an Act of Parliament but “as sui generis, calling for principles of interpretation of its own, suitable to its character...without necessary acceptance of all the presumptions that are relevant to legislation of private law””.58

It is arguable that if the so-called sui generis approach in Ong Ah Chuan were applied to the present case, a different result might have been reached by the High Court with regards to the interpretation of Article 49(1). Presumably, the Applicant’s argument that there existed an implied right to representation within the Constitution would’ve been accorded more weight in determining the proper construction of Article 49(1).

While Ong Ah Chuan was decided before the purposive approach was expressly adopted as part of Singapore law, it remains good law nevertheless and presents a possible alternative approach towards constitutional interpretation. However, it will be necessary to re-examine whether the so-called sui generis approach in Ong Ah Chuan ought to remain part of Singapore case law in light of the increasing resort to the purposive approach in constitutional interpretation.

V. Conclusion

Wong Souk Yee sought to address the latent ambiguity surrounding the scope of the by-elections clause in Article 49 of the Constitution vis-à-vis the GRC scheme by holding that the provision could be updated or rectified to incorporate changes made elsewhere to existing electoral law. While the High Court’s conclusion appears to be technically correct based on the express wording of existing legislation regarding Parliamentary elections as well as an unduly broad application of the purposive approach in interpreting Article 49(1), the reasoning of the High Court has arguably opened a can of worms regarding the proper interpretation of constitutional provisions. The decision also runs a noticeable risk of undermining the very notion of constitutional supremacy. While it remains to be seen whether an Appeal will be brought, further clarification from the Court of Appeal as to the abovementioned points of contention would be greatly welcome.

[1] [2013] 4 SLR 1; [2013] SGCA 39.

[2] Constitution of the Republic of Singapore (1999 Rev Ed), art 49.

[3] Supra note 1 at [82].

[4] Ibid at [80] and [82].

[5] [2018] SGHC 80.

[6] Ibid at [3].

[7] Royston Sim, “Halimah Yacob steps down as Speaker and MP, resigns from PAP to run for President”, The Straits Times (7 August 2017), online: <>.

[8] Valerie Koh & Kenneth Cheng, “3 remaining Marsiling-Yew Tee GRC MPs to build on Halimah’s good work”, TODAYonline (7 August 2017), online: <>.

[9] Cap 218, 2011 Rev Ed Sing.

[10] Supra note 5 at [4].

[11] Ibid.

[12] Supra note 2 at art 49(1).

[13] Ibid at art 39A(2).

[14] Supra note 5 at [12].

[15] Constitution of the Republic of Singapore (Amendment) Act, No 3 of 1988, amending Constitution of the Republic of Singapore (1985 Rev Ed).

[16] Constitution of the Republic of Singapore (Amendment) Act, No 16 of 1984, s 5, amending Constitution of the Republic of Singapore (1980 Reprint), Art 49.

[17] Constitution of the Republic of Singapore (1980 Reprint).

[18] Supra note 5 at [9].

[19] Ibid at [14].

[20] Supra note 2 at arts 45 and 46.

[21] Supra note 9 at s 27A.

[22] Supra note 5 at [16]–[18].

[23] Ibid at [23]–[26].

[24] Supra note 21 at s 24(2A).

[25] On the dangers of unduly expanding the scope of or implying into the Constitution additional interpretations, see Rajeevan Edakalavan v Public Prosecutor [1998] 1 SLR(R) 10; [1998] SGHC 2 at [19].

[26] Supra note 5 at [27].

[27] FAR Bennion & Oliver Jones, ed, Bennion on Statutory Interpretation: A Code, 6th ed (London, UK: LexisNexis, 2013) at 797.

[28] Ibid.

[29] Ibid.

[30] [2006] 3 SLR(R) 688; [2006] SGHC 120.

[31] Supra note 5 at [28]–[29].

[32] Supra note 30 at [46]–[47].

[33] Supra note 5 at [36]; see also Parliamentary Debates Singapore: Official Report, vol 50 at cols 334–335 (12 January 1988) (Mr Goh Chok Tong).

[34] Ibid at [33]–[38].

[35] As expressly acknowledged in Bennion, supra note 27 at 167.

[36] See VK Rajah, SC, “Interpreting the Singapore Constitution” in Jacyln L Neo, ed, Constitutional Interpretation in Singapore: Theory and Practice (New York: Routledge, 2017) 23 at 24.

[37] Supra note 1 at [82].

[38] Supra note 2 at art 39A(2)(c).

[39] Supra note 30 at [48].

[40] Supra note 27 at 788.

[41] Supra note 5 at [41].

[42] Ibid.

[43] Halsbury’s Laws of Singapore, vol 1 (Singapore: LexisNexis, 2017 Reissue) at 100, para 10.506.

[44] Supra note 2 at Art 4. See also Chan Hiang Leng Colin v Public Prosecutor [1994] 3 SLR(R) 209; [1994] SGHC 207 at [50] and Taw Cheng Kong v Public Prosecutor [1998] 1 SLR(R) 78; [1998] SGHC 10 [Taw Cheng Kong] at [14] for judicial recognition of the doctrine of constitutional supremacy.

[45] While no statute passed by Parliament has ultimately been held unconstitutional by the courts, there nevertheless exists judicial precedent for such an approach: see Taw Cheng Kong, ibid, at [84].

[46] Supra note 5 at [47].

[47] Ibid at [49].

[48] Tan Cheng Bock v Attorney-General [2017] 2 SLR 850; [2017] SGCA 50 at [41].

[49] This can be contrasted with Vellama, supra note 1 at [79], where the Court of Appeal interpreted Article 49(1) with reference to the broad purpose of the Constitution in establishing a Westminster model of government with a right to representation.

[50] Supra note 5 at [52]–[53].

[51] Ibid at [57].

[52] Ibid at [59]–[60].

[53] Interpretation Act (Cap 1, 2002 Rev Ed), s 9A; see also Constitution of the Republic of Singapore (1999 Rev Ed) at art 2.

[54] Some commentators have welcomed this unified approach for both constitutional and statutory interpretation: see Goh Yihan, “The Interpretation of the Singapore Constitution: Towards a unified approach to interpreting legal documents” in Jaclyn L Neo, supra note 36, 257 at 276.

[55] Supra note 48 at [38].

[56] [1980] AC 319.

[57] [1979-1980] SLR(R) 710; [1980] SGPC 6.

[58] Ibid at [23].

The PDF version of this article is available for download here.

New Sentencing Principles for Rape – An Analysis of Ng Kean Meng Terence v Public Prosecutor

By Benjamin Low


What is the most appropriate sentence that should be meted out onto an accused person who has been found guilty of the offence of rape?

The apparent simplicity of this question belies powerful, and oftentimes conflicting, sentencing policy considerations and more abstract questions of ethical principles and sentencing theory. The recent case of Ng Kean Meng Terence v Public Prosecutor1 [Ng Kean Meng Terence] is the Singapore Court of Appeal’s latest attempt at creating a new, comprehensive sentencing framework for rape in order to address the above quandaries. This article seeks to analyse the state of the law prior to Ng Kean Meng Terence and discusses the implications and effects of the case on the criminal law.


Prior to Ng Kean Meng Terence, the law on sentencing for rape offences was encapsulated in Public Prosecutor v NF2 [NF]. In NF, V K Rajah J (as he then was), formulated a new sentencing framework (the “NF Framework”) that was meant to replace the earlier sentencing guidelines in the earlier case of Chia Kim Heng Frederick v Public Prosecutor3 [Frederick Chia].

Under the NF Framework, rape cases were divided into four categories, each with a minimum sentence of imprisonment and caning of differing degree.4 The NF Framework may be briefly summarised as follows:

  1. Category 1 – Cases with no aggravating and/or mitigating factors with a starting sentence of ten years imprisonment and not less than six strokes of caning.
  2. Category 2 – Cases with any one of the seven aggravating factors listed in NF, such as where rape that is committed by two or more persons acting in concert; or where the offender is in a position of responsibility towards the victim.5 For such instances, the starting sentence would have been fifteen years imprisonment and twelve strokes of the cane.
  3. Category 3 – Cases where the victim is raped on multiple occasions, or where rape is perpetrated on multiple victims. This category shares the same sentencing benchmark as a Category 2 case. Rajah J justified this on the basis that the Prosecution would normally prefer multiple charges against the offender, thereby leading the courts to order that two or more of the sentences imposed run consecutively, thereby resulting in a higher sentence meted out.6
  4. Category 4 – Cases where 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”.7 In such cases, the maximum sentence of twenty years imprisonment and twenty-four strokes of the cane would be handed out.

Ideally, the NF Framework operates by having the Court first determine the category under which the particular rape offence should be placed, followed by the Court then adjusting the benchmark sentence upwards or downwards to take into account the additional aggravating and mitigating factors disclosed on the facts.8

The NF Framework was first considered and approved of by the Court of Appeal in the context of rape simpliciter in Public Prosecutor v Mohammed Liton Mohammed Syeed Mallik.9 It was subsequently applied by the Singapore courts to a wide range of cases involving rape committed in various permutations and factual situations.10


Notwithstanding the durability and versatility of the NF Framework over the past eleven years, the Court of Appeal in Ng Kean Meng Terence noted that problems remained with the NF Framework that, in its opinion, necessitated a revision of the sentencing law on rape. These were, firstly, that the categories laid down in NF were themselves not properly defined – to the extent that some of the categories served no clear purpose.11 Category 3, for example, was defined in NF as encompassing cases involving the repeated rape of the same victim or rapes committed by a single offender on multiple victims. However, given that the Criminal Procedure Code12 requires that (and tying in with longstanding prosecutorial practice) separate and multiple charges be pressed against an accused person based on the facts disclosed13 and separate sentences imposed for each charge which the accused person is convicted of,14 this appeared to remove the need for a separate Category 3 to deal with multiple offending under the NF Framework.

A second, more pressing problem, was the lack of conceptual coherence with which the Category 2 aggravating factors were included inside Category 2. While it is not possible to list out all the Category 2 aggravating factors here, the Court of Appeal noted with disquiet that “[T]here does not appear to be any conceptual unity or discernible unifying theme”15 underlying the inclusion of these factors and the exclusion of others. While Rajah J no doubt intended that the list of Category 2 aggravating factors themselves be complemented with the additional aggravating factors and/or mitigating factors of each specific case, there is considerable weight in the Court’s criticisms with regards to the admittedly unclear criteria on which the Category 2 aggravating factors were based on.

The final nail in the coffin was the Court of Appeal’s finding that the NF Framework did not properly account for the statutory aggravating factors already present in the Penal Code. Section 375(3) of the Penal Code16 provides for a minimum sentence of 8 years imprisonment and 12 strokes of the cane 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. None of these factors are included in the Category 2 aggravating factors within the NF Framework. In the Court’s opinion, there was a need to clarify the relationship between the two groups of aggravating factors.


Having opted to do away with the NF Framework, the Court in Ng Kean Meng Terence embarked on the difficult task of developing a new sentencing framework (the “Ng Kean Meng Terence Framework”) for the offence of rape. After considering past approaches by the courts in the preceding case law, the Court of Appeal rejected these approaches as unsuitable and opted to adopt the framework developed in the New Zealand case of R v Taueki.17

Generally speaking, the Ng Kean Meng Terence Framework operates in two steps:18

  1. Firstly, the court should identify which sentencing band the offence in question falls within by taking into account the offence-specific factors in the case. Once the sentencing band has been identified, the Court should then determine where within the aforementioned band the present offence falls in order to derive a sentence for a starting point.
  2. Secondly, the court should then consider the aggravating and mitigating factors which are personal to the offender to calibrate the appropriate sentence for the offender.

Step 1. Classification of the Offence

The Court of Appeal stressed that only “offence-specific” factors which relate to the circumstances of the offence such as the particular harm caused, the manner in which the offence was committed, or the specific role played by the offender in the commission of a offence by a group.19 This proposition is a common-sensical one that recognises that different factual permutations of rape should attract different penal consequences.

A non-exhaustive list of offence-specific factors, drawing on existing case law, was proposed by the Court:20

  • Group Rape: Offences committed by groups of persons.
  • Abuse of position and breach of trust: Where the offender is in a position of responsibility towards the victim or in whom the victim has placed her trust by virtue of his office of employment.
  • Premeditation: Where the offender engages in some form of planning or exhibits premeditation in committing the act (e.g. through the use of drugs or engaging in predatory behaviour).
  • Violence: Actual or threatened use of violence.
  • Vulnerable victim: Whether the victim was especially vulnerable by reason of age, physical frailty, mental impairment or disorder, and/or learning disability.
  • Forcible rape of a victim below the age of fourteen.
  • Hate crime: Where the offence was committed as a hate crime, such as an expression of racial and/or religious prejudice, or in other situations where the victim has been specifically targeted by reason of her membership of a vulnerable minority group.
  • Severe harm: Severe harm inflicted on the victim such as pregnancy, transmission of a serious disease or a psychiatric illness.
  • Deliberate infliction of special trauma.

The Court of Appeal also cautioned against taking into account the two factors of (i) forgiveness by the victim towards the offender and (ii) consent given by a victim under fourteen on the basis that such factors were usually irrelevant.21 Rejecting the second factor in particular, ties in with the express wording of s 375 of the Penal Code in criminalising sex with a person under the age of fourteen.

Having regard to the offence-specific factors, the Court must then place the offence within an appropriate sentencing band.22 For ease of reference, the three sentencing bands, each with different sentencing outcomes, are laid out below:

Table_Sentenncing Bands.PNG

Once the appropriate sentencing band has been identified, the Court would then identify where precisely along the range prescribed for the band a particular sentence falls.24

Step 2. Calibrating the Sentence

Once the starting-point sentence has been identified, the Court must then consider the “offender-specific” factors of the case – that is to say, those factors which relate to the personal circumstances of the offender himself,25 rather than relating to the manner and mode of the offending. In making this distinction, it seems that the Court of Appeal was clearly seeking to prevent any double-counting of factors that could lead to an imposition of more severe sentences than would usually be the case.

The possible offender-specific aggravating and mitigating factors recognised in Ng Kean Meng Terence are as follows:26

Table_Offender-specific factors.PNG

One of the more controversial aspects of Ng Kean Meng Terence was whether pleas of guilt could constitute a mitigating factor. While this contentious point will be discussed in further detail below, it will suffice to say that the Court in Ng Kean Meng Terence was prepared to treat pleas of guilt as one of the many offender-specific mitigating factors.27


From the outset, it should be noted that the Ng Kean Meng Terence Framework does not constitute a fundamental or radical revision of the sentencing law for rape. What the Ng Kean Meng Terence Framework does is to improve the present sentencing framework by better delineating the different factors that courts usually consider when determining the appropriate sentence to be meted out onto the offender.

Thus the distinction between “offence-specific” and “offender-specific” factors removes the problem of double-counting while ensuring that all relevant factors in each case are taken into consideration by the courts.28 The offence-specific factors themselves have also been modified to circumvent the conceptual difficulties of the NF Framework while addressing the third problem raised as to the statutory-aggravating factors in the Penal Code.29

A second noteworthy point is the use of sentencing bands which injects greater clarity in determining the starting-point sentence. Rather than having to work from scratch at a fixed benchmark sentence under the NF Framework, sentencing courts can now better identify which Band, and thus which part of the Band, the case in question belongs to.

It bears mentioning that since the decision of Ng Kean Meng Terence, there have been some additional developments in the local case law with regards to the Ng Kean Meng Terence Framework. While it is not possible to consider these developments in great detail, they warrant a mention in this article for completeness sake. In the later decision of Pram Nair v Public Prosecutor,30 the Court of Appeal recognised that the intoxication of a victim could constitute an additional offence-specific aggravating factor though it cautioned that this factor would not necessarily manifest itself in all cases involving an intoxicated victim.31 Much would depend on the degree of intoxication in each particular case. More importantly, the Ng Kean Meng Terence Framework was transposed onto the offence of sexual assault by penetration, with three sentencing bands imposed with varying sentencing ranges.32 As a preliminary note, these developments ought to be welcomed as promoting unity in the courts’ approaches towards both types of sexual offences.

However, notwithstanding the merits of the Ng Kean Meng Terence Framework, it is submitted that the shift in the Court of Appeal’s attitude towards pleas of guilt could carry possibly adverse implications in the sentencing law towards criminal offences. Prior to Ng Kean Meng Terence, it was settled law that a plea of guilt would not entitle an offender to a sentencing discount unless it was made out of genuine remorse.33 However, the Court of Appeal has since departed from that position by recognising in Chang Kar Meng v Public Prosecutor34 [Chang Kar Meng] that a plea of guilt could constitute a mitigating factor for utilitarian reasons (i.e. it saves victims the horror of having to recount the ordeal and it saves the resources of the State if the case were to go to trial) besides the existing remorse-based justification. The Court galvanized this proposition in Ng Kean Meng Terence by citing Chang Kar Meng with approval.35

While the utilitarian considerations admittedly cannot be discounted, the Court of Appeal’s pronouncement in Ng Kean Meng Terence has the effect of suggesting that a plea of guilt will almost certainly entitle an offender to a sentencing discount as long as he pleads guilty to a charge of rape, even if the guilty plea was not motivated by genuine remorse or contrition. Though the Court later sought to clarify that guilty pleas would only be treated as one of the many mitigating factors, it is not difficult to foresee that rape offenders would be more inclined to try their luck and enter pleas of guilt in the hopes of obtaining a lighter sentence. In a Band 1 case of rape with no further aggravating factors, a plea of guilt could potentially entitle an offender to a sentence beneath the lowest threshold in Band 1. In such situations, and where the offender does not seem to have displayed genuine remorse, can it truly be said that the public interest has been effectively safeguarded or that the overriding considerations of rehabilitation and/or deterrence have been achieved?

This leads us to a more fundamental problem with the Ng Kean Meng Terence Framework in that it does not provide any real change in the sentencing law for rape. While this was no doubt the express intention of the Court of Appeal itself,36 one cannot help but feel that a golden opportunity was missed in reviewing the present sentencing law on rape and then determining whether or not the prescribed sentencing ranges ought to be revised upwards or downwards. Indeed, the lowest sentence of ten years imprisonment and six strokes of caning in Band 1 was clearly based on the earlier decision in Frederick Chia37 but this writer submits that the sentencing ranges for the various sentencing bands should be revised upwards. A stricter sentencing regime would better accord with the longstanding deterrent aspect of the sentencing jurisprudence of the Singapore courts,38 especially given a disturbing rise in sexual crimes over the past few years.39


The process of determining and laying down suitable and comprehensive sentencing guidelines for various offences is a continuous and challenging effort that often requires the courts to continuously take into account broader societal and legal changes in an effort to ensure that such guidelines remain relevant and up-to-date while ensuring that the public interest is safeguarded. Since the NF Framework, the criminal law in Singapore has evolved immensely, and the Courts must continue to refine and redevelop the existing case law to account for the new legal climate.

In light of the intrinsic seriousness of the offence of rape in particular, and the difficulties that courts sometimes encounter in threshing out the most appropriate sentence, the Court of Appeal has boldly decided to revise the sentencing law for rape by devising a new sentencing framework that meets these requirements. Although it remains to be seen whether the Ng Khean Meng Terence framework is sufficiently capable of withstanding future challenges in the form of hard cases, there is much to commend the Court for its efforts in tackling what has been a most complex area of criminal law.

[1] [2017] 2 SLR 449.

[2] [2006] 4 SLR(R) 849.

[3] [1992] 1 SLR(R) 63.

[4] Supra note 2 at [20]–[21].

[5] Ibid.

[6] Ibid, at [37].

[7] Ibid, at [21]; see also R v Keith Billam (1986) 8 Cr App R (S) 48 at pp 50–51.

[8] Ibid, at [23]; see also Ng Kean Meng Terence at [10].

[9] [2008] 1 SLR(R) 601.

[10] See PP v UI [2008] 4 SLR(R) 500 and PP v AOM [2011] 2 SLR 1057.

[11] Supra note 1 at [13]–[15].

[12] (Cap 68, 2012 Rev Ed).

[13] Ibid, at s 132.

[14] Ibid, at s 306(1).

[15] Supra note 1 at [17].

[16] (Cap 224, 2008 Rev Ed).

[17] [2005] 3 NZLR 372.

[18] Supra note 1 at [39].

[19] Ibid, at [43].

[20] Ibid, at [44].

[21] Ibid, at [45].

[22] Supra note 1 at [47].

[23] See PP v Bala Kuppusamy [2009] SGHC 97 at [28]–[29].

[24] Supra note 1 at [61].

[25] Ibid, at [62].

[26] Ibid, at [64]–[65].

[27] Ibid, at [71].

[28] Ibid, at [73].

[29] Supra note 16 at s 375(3).

[30] [2017] 2 SLR 1015.

[31]Ibid, at [126]–[132].

[32] Ibid, at [158]–[159]; although, strangely, the Court held that the respective sentencing ranges for sexual assault by penetration were to be lower than that for rape.

[33] Supra note 1 at [67]; see also Rajah J’s holding in Angliss Singapore Pte Ltd v PP [2006] 4 SLR(R) 653 at [56] and [77].

[34] [2017] 2 SLR 68 at [47].

[35] Supra note 1 at [68]–[69]; see also Chia Kim Heng Frederick v PP [1992] 1 SLR(R) 63 at [20].

[36] Ibid, at [74].

[37] Supra note 3 at [20].

[38] See PP v Law Aik Meng [2007] 2 SLR(R) 814 at [18]–[19].

[39] Seow Bei Yi, “Rise in sex crimes over past 5 years: State Courts”, The Straits Times (3 April 2017), online: <> ; see also “Violence: Rape Victims” (24 February 2017), online: Ministry of Social and Family Development <> .

A PDF version of this entry is available for download here.