1 This is an appeal against the sentence imposed on the appellant, M Raveendran (“Raveendran”), for the offence of driving under the influence of drink pursuant to s 67(1)(b) of the Road Traffic Act (Cap 276, 2004 Rev Ed) (“RTA”) (which for convenience, I refer to as “drink driving”). Raveendran also consented to one charge of driving without due care and attention under s 65(1)(a) of the RTA being taken into consideration for sentencing. In the court below, Raveendran was sentenced by the District Judge (“DJ”) to one week’s imprisonment and disqualification from driving all classes of vehicles for a period of 24 months: see Public Prosecutor v M Raveendran [2020] SGDC 289 (“GD”). Raveendran did not contest the disqualification order before me. He only sought to persuade me that a custodial sentence should not be imposed because of the potential impact this would have on his entitlement to receive emoluments from the Singapore Armed Forces (“SAF”) upon his retirement.

2 At the first hearing of the appeal on 21 April 2021, I directed that Raveendran file a statutory declaration setting out, in precise terms, the probable consequences of the sentencing decision in the present appeal on his entitlement to emoluments upon his retirement from the SAF. I then directed the parties to file further submissions addressing the question of whether these probable or potential consequences, which stemmed from the terms of his employment, could properly be taken into consideration as a relevant factor by the sentencing court, and if so, the basis on which this could be done. I also appointed Mr See Kwang Guan (“Mr See”) as young amicus curiae to assist me. I was greatly assisted by Mr See’s submissions, which were carefully researched and thoughtfully presented.

3 After hearing the submissions of the parties, I held that the learned DJ erred in not considering some of the mitigating factors in the present case, specifically those evidencing Raveendran’s remorse. Having regard to sentences imposed in other cases involving broadly similar circumstances, I allowed the appeal to that extent and reduced the imprisonment sentence to five days’ imprisonment. With the benefit of Mr See’s and the parties’ submissions, however, I was also satisfied that it was not appropriate for me to have regard to Raveendran’s potential loss of employment benefits and emoluments from the SAF as a factor that was relevant to sentencing, and I therefore disregarded it. I now explain my reasons for coming to this view.

4 On 8 September 2018, from about 8pm, Raveendran consumed some alcohol while he was with his friends at Newton Food Centre. Thereafter, on 9 September 2018, at about 12.35am, while driving a car along Thomson Road towards Upper Thomson Road on the way home, he lost control of the car and veered right; this caused the car to mount the centre divider and collide into twelve pieces of the centre guard railings.

5 A police officer came across the accident while patrolling along Thomson Road. He interviewed Raveendran and conducted a preliminary breath test, which Raveendran failed. Raveendran was then arrested and escorted to the Tanglin Police Division Headquarters for a Breath Analysing Device (“BAD”) test to be administered. The BAD test showed that the proportion of alcohol in his breath was 91 microgrammes of alcohol in every 100 millilitres of breath, well in excess of the prescribed limit of 35 microgrammes of alcohol per 100 millilitres of breath. Raveendran had therefore committed an offence under s 67(1)(b) of the RTA.

6 The cost of repairing the guard railings that were damaged due to the collision amounted to $1,438.50. Raveendran duly compensated the Land Transport Authority (“LTA”) by paying the full sum.

7 The DJ first considered the indicative sentencing ranges for drink driving offences set out in Stansilas Fabian Kester v Public Prosecutor [2017] 5 SLR 755 (“Stansilas”). The present case was held to be one falling within the category of slight harm and medium culpability, taking into account the damage caused to the railings, the fact that the alcohol level found in Raveendran’s breath was high, and that Raveendran had lost control of his vehicle (GD at [17]–[24]).

8 The DJ then determined that the indicative starting sentence was one week’s imprisonment. In reaching this conclusion, the DJ compared the present case with three other cases: Stansilas, Public Prosecutor v Vilashini d/o Nallan Rajanderan [2018] SGDC 142 (“Vilashini”), and Public Prosecutor v Solomon Seah [2018] SGDC 106 (“Solomon Seah”) (GD at [25]–[29]). The DJ then considered the offender-specific factors and held that Raveendran’s positive record of public service and contributions, as well as the fact that a substantial amount of his bonus and gratuity payments from the SAF might be forfeited if a custodial sentence were imposed, did not justify a reduction in his sentence. In respect of the benefits Raveendran was entitled to receive from the SAF, the DJ cited Stansilas, where I had held that the financial consequences that an offender may face were not relevant mitigating factors (GD at [40]–[41]). Finally, the DJ considered that whilst Raveendran’s plea of guilt and the restitution he had made to the LTA for the damage caused to the railings demonstrated remorse, this was not sufficiently exceptional to justify any reduction in the sentence (GD at [42]).

9 In this appeal, Raveendran submitted that the imprisonment term imposed was manifestly excessive and that the appropriate sentence should be the maximum fine of $4,000. He submitted that the DJ had erred in finding that the offender-specific factors raised in mitigation did not warrant a reduction in sentence from the indicative starting point of one week’s imprisonment. Specifically, the DJ should have given weight to the following factors: (a) he had pleaded guilty and made full restitution to the LTA; (b) he had remained at the scene and rendered all possible assistance to the attending police officer; (c) he had a strong propensity for reform, as evidenced in his professional record as an army officer and his contributions to the nation; and (d) he was at risk of losing the emoluments which he would otherwise have received from the SAF if a custodial sentence were imposed.

10 In respect of point (d) above, Raveendran averred in his statutory declaration that he had retired as a 1st Warrant Officer on 21 November 2020, after serving for around 38 years with the SAF. He stated that he would have been entitled to emoluments amounting to $273,694.02 upon retirement, and that these had been withheld from him as a result of his conviction. Raveendran further averred that these might potentially be forfeited if he were sentenced to a term of imprisonment. However, he was unable to provide any confirmation from MINDEF as to how his entitlements would be impacted by his sentence.

11 Raveendran submitted that the potential consequences he faced in connection with the potential loss of his employment benefits were relevant to sentencing in two ways. Relying on Chew Soo Chun v Public Prosecutor and another appeal [2016] 2 SLR 78 (“Chew Soo Chun”), he argued first, that this could be a basis for the exercise of judicial mercy; and second, that this could be viewed as a mitigating factor in sentencing in the sense that because he stood to suffer more than other offenders who committed the same offence, his sentence should be adjusted on grounds of proportionality.

12 The primary issue that arises for determination in this appeal is whether Raveendran’s potential loss of emoluments, in the event a custodial sentence is imposed on him, is a factor that should be considered by the sentencing court. In these grounds, I provide my reasons for having concluded that this is not a relevant factor in sentencing.

13 The starting position is that an individual who breaches the criminal law generally can and should expect that the law will take its course and that he will have to face the consequences of his actions (see Stansilas at [111]). This will only be displaced in exceptional circumstances. In this regard, it is important to note that any exceptional circumstances must be identified and applied in a principled and transparent manner. Here, it is apposite to heed the caution sounded in Chew Soo Chun at [26] in a slightly different, albeit analogous, context:

14 Though stated in the context of judicial mercy, that passage underscores the importance of identifying the conceptual basis for a court’s sentencing decisions in a principled way. The cases reveal four possible bases upon which the reduction of a sentence on account of the potential loss of emoluments could conceivably be justified, namely:

(a) the principle of equal impact;

(b) the principle of parsimony;

(c) judicial mercy; and

(d) pursuant to the express terms of an applicable statute.

I address each of these in turn.

15 The equal impact principle rests on the notion that if an offender suffers from some condition that would render the sentence significantly more onerous for him than for other offenders, a sentencing adjustment may be made so as to avoid such an “undue differential impact” upon him. Such an adjustment serves to “eliminate [the] increment in severity” that would otherwise arise as a result of the offender’s condition: see Andrew von Hirsch and Andrew Ashworth, Proportionate Sentencing: Exploring the Principles (Oxford University Press, 2005) at p 172–173. The equal impact principle is an aspect of the principle of proportionality that, in essence, provides that an offender’s sentence should be “in line with what the offence he had committed deserves, and no more”: see Public Prosecutor v Saiful Rizam bin Assim and other appeals [2014] 2 SLR 495 at [29]. In my view, the equal impact principle is applicable only to factors that are intrinsic or inherent to an offender, for the reasons that follow. To be clear, intrinsic factors refer to those that inhere in the offender and are part of his person. This would include circumstances such as his mental and physical condition, and his age. These factors exist regardless of the offence or the sentence. In contrast, factors such as the offender’s employment, wealth, or level of education would not be regarded as intrinsic. They may be part of his wider circumstances, but certainly cannot be said to be part of his personal attributes. Further, to the extent they are raised in the context of sentencing, it is because of the consequences that the particular sentence would have on the extrinsic factors in question.

16 In determining the appropriate sentence, a sentencing court will have regard to, among other things, the circumstances of the offence, its impact and consequences on the victim or others, the offender’s culpability and criminal record, and matters of mitigation that are personal to the offender: Sentencing Practice in the Subordinate Courts (3rd Ed, LexisNexis 2013) (“Sentencing Practice”) at p 127. It is therefore clear that a sentencing court does take into account factors that go toward the seriousness of the offence committed, as well as aggravating and mitigating factors that relate to the particular offender. In the overall analysis, the court will also have regard to the relevant sentencing principles and interests: Public Prosecutor v ASR [2019] 1 SLR 941 at [130]; see also Public Prosecutor v Tan Fook Sum [1999] 1 SLR(R) 1022 at [15].

17 Factors that are intrinsic to an offender typically feature as part of the matrix of sentencing factors when considering the offender-specific mitigating factors. Offender-specific factors are aspects that are relevant to sentencing and that “relate to the personal circumstances of the offender”, such as his “character, personal attributes, expression of remorse, or any other considerations which are particular to the offender rather than factors relating to the manner and mode of the offending or the harm caused by the offence” [emphasis added]: Ng Kean Meng Terence v Public Prosecutor [2017] 2 SLR 449 (“Terence Ng”) at [62]. The court sentences an offender as he is by taking into consideration these attributes.

18 Factors intrinsic to an offender broadly fall into two categories. The first category of factors is linked to the commission of the offence itself, including the impact of these factors on the culpability of an offender. Examples of factors belonging in this category include an offender’s mental condition and intellectual ability. These factors go towards determining what the appropriate punishment should be in the light of matters that shed light on the commission of the offence. If the relevant factors are applied correctly, offenders who are similarly situated in terms of what led to the commission of the offence in broadly similar circumstances, should receive broadly similar sentences. Although this has nothing to do with the question of the impact of a sentence on offenders, I refer to this to note that the factors considered in this context are intrinsic to the offender: see further my observations at [45] below.

19 The second category of factors relates to the effects or impact that a sentence would have on an offender. It is this category of factors that triggers the application of the equal impact principle. I consider how the equal impact principle has been recognised in case law in the context of these factors below. It will be seen that this has always been in respect of conditions or factors that are intrinsic or inherent to the offender such as the offender’s ill-health or age; and, in my judgment, this is so for good reason.

20 In Chew Soo Chun, the court applied the equal impact principle when sentencing an offender who was ill. The court considered that ill health could cause imprisonment to result in disproportionate suffering for the offender, such that it would amount to a crushing sentence for him. The applicability of such a consideration will obviously depend on how serious the illness is. The question in each case is whether what would otherwise have been appropriate with regard to the offence committed, could become “out of line on the ground of proportionality”. This is because “other things being equal, offenders ought to be subject to the same impact”. Where there is an underlying condition of the requisite seriousness, the sentence to be imposed should be reduced so that it would not be disproportionate to the offender’s culpability and physical condition (at [33], [34] and [38]).

21 Age has also been regarded as a factor that may bear on the principle of equal impact. In A Karthik v Public Prosecutor [2018] 5 SLR 1289 (“Karthik”) at [37], the court noted that one of the reasons why rehabilitation is regarded as the controlling principle when sentencing young offenders is that they “appear to suffer disproportionately when exposed to typical punitive options, such as imprisonment, as compared to adult offenders” [emphasis added]. This should be seen in context because, as I discussed in Karthik, there are a number of factors that result in the sentencing of young offenders usually being assessed in a different manner than is the case with other offenders. The position is somewhat clearer with elderly offenders, where the court will consider whether the imposition of a long custodial sentence would result in a disproportionate sentence, applying a fact-sensitive enquiry. In Public Prosecutor v UI [2008] 4 SLR(R) 500 (“PP v UI”), the court held that an offender’s advanced age did not generally warrant a reduction in sentence, citing Krishan Chand v Public Prosecutor [1995] 1 SLR(R) 737. However, the court considered that a sentence imposed on an elderly offender that virtually amounted to a life sentence could be crushing and in breach of the totality principle (at [78]).

22 In Yap Ah Lai v Public Prosecutor [2014] 3 SLR 180, the court referenced the decision in PP v UI, and held that the court’s key consideration was to “assess the impact of [a substantial period of imprisonment] on the offender having regard to his past record and his future life expectation and consider whether this would be disproportionate and crushing because of the offender’s particular circumstances” [emphasis in original] (at [88]). The equal impact principle applies in this context because a long custodial term in some circumstances could mean that the offender will have to spend most of the rest of his life in jail. This could be mitigating “not because the court is extending mercy to the offender in view of his advanced age, but because the court is unwilling to make such offenders suffer more than others who are similarly situated” (at [91]). The suffering in question here is the effective denial of the hope of regaining one’s liberty. However, this is subject to the limitation that the impact of the sentence must be “so severe as to be disproportionate or crushing”; in particular, there is no general principle that the advanced age of an offender would always be mitigating (at [93]).

23 As explained in Terence Ng at [65(c)], the imposition of a custodial term will necessarily deprive an elderly offender of a larger fraction of their expectation of life. Hence, a “concern for the overall proportionality of punishment – and not the age of the offender per se, is the real reason for affording leniency on account of advanced age”.

24 What of extrinsic factors that might be said to have some impact on the offender as a result of the particular sentence that the court is considering? Prof Andrew Ashworth (“Prof Ashworth”), in Justifying the Grounds of Mitigation (1994) 13 Criminal Justice Ethics 5 (“Justifying the Grounds of Mitigation”), describes extrinsic factors as consequences of the offence, whereas intrinsic factors are “pre-existing or unrelated elements in the offender’s situation”, and notes that it is to the latter that the equal impact principle usually applies (at 6). The issue before me is whether the equal impact principle should also apply to consequences flowing from the imposition of a particular sentence because of the terms of the offender’s employment, which is an extrinsic factor. In the present context, the potential consequences depend on Raveendran’s terms of employment. This is not a factor that is intrinsic or inherent to Raveendran, but turns on the fact that he is employed and that the particular consequence may happen to be a term of his employment.

25 In my judgment, the equal impact principle does not extend to factors that are extrinsic to an offender, such as financial consequences that would befall him as a consequence of his sentence. I begin by contrasting intrinsic and extrinsic factors, which lead to the reasons for why this must be the case.

26 In respect of factors extrinsic to an offender it is not possible to compare their impact either between offenders, or in relation to how the desired effect of punishment can otherwise be achieved. In short, the court is not equipped with manageable standards to take this into consideration in sentencing.

27 In respect of factors that are intrinsic to an offender, their effects on an offender are limited in scope, and the court can evaluate whether the sentence would cause that offender to suffer disproportionately as compared to other offenders on whom the same sentence might be imposed. As I explain below, the impact of ill health and advanced age are specifically defined and circumscribed. The court is therefore able to determine whether these factors should result in a reduction in sentence and if so, the extent to which the sentence should be attenuated.

28 In respect of ill health, the court in Chew Soo Chun noted that the extent to which a reduction in sentence is warranted in a given case would depend on whether the evidence revealed a “real likelihood” of disproportionate impact on the offender and the magnitude of such impact (at [36]). Specifically, this is assessed by reference to the “risk of significant deterioration in health or a significant exacerbation of pain and suffering” (at [34]). In Chew Soo Chun (at [39]–[40], the court observed as follows:

29 This has subsequently been applied in other cases:

(a) In Public Prosecutor v Tan Kok Ming Michael and other appeals [2019] 5 SLR 926, one of the offenders was in remission for cancer. She appealed against the District Judge’s decision that her illness did not warrant the invocation of judicial mercy or attract mitigating weight. The High Court held that this did not constitute evidence of the exacerbation or likely recurrence of the offender’s medical condition. Based on the evidence canvassed at a Newton Hearing, there was evidence that the Singapore Prisons Service (“Prisons”) was able to provide the requisite care and treatment, and that the offender therefore had as good a chance of detecting any recurrence of her illness in prison as she did outside it. The Judge therefore rejected the offender’s argument (at [147]–[148]).

(b) In Goh Chin Soon v Public Prosecutor [2021] 4 SLR 401 (“Goh Chin Soon”), the offender adduced various medical reports stating that he was at increased risk of heart attack and sudden cardiac death. The High Court considered that the burden was on the offender to present at least some evidence from medical professionals “directed towards suggesting specifically that imprisonment would have a significantly adverse impact on his health”. If the offender were able to do so, the burden would shift to the Prosecution to adduce evidence to the contrary, such as by showing that Prisons would be capable of addressing the offender’s medical issues in prison. The court held that the offender had not shown that imprisonment would have such a significantly adverse impact on his health, compared to his situation if he was not imprisoned, such that a custodial term would cause disproportionate suffering (at [165]–[166]).

(c) In Chua Siew Peng v Public Prosecutor and another appeal [2017] 4 SLR 1247, the offender was diagnosed with paranoid schizophrenia. The High Court agreed with the Prosecution that Prisons was able to manage the offender’s condition and that the offender was unable to prove that prison life would have a significantly adverse impact on him. The remark in the offender’s psychiatric report that her health condition might deteriorate further in prison was thought to be equivocal and therefore insufficient to meet the threshold of showing that he would suffer disproportionately from a custodial term (at [103]–[104]).

30 It is evident that the court in these cases is concerned with the question of whether an offender’s illness meets the threshold of exposing him to a “risk of significant deterioration in health or a significant exacerbation of pain and suffering” if a custodial term were imposed on him and it assesses this by having regard to the medical evidence adduced on his behalf. The court in Goh Chin Soon noted that this would “involve some articulation on the part of medical professionals who are familiar with the offender’s medical conditions as to the basis they may have for believing or fearing that imprisonment would adversely affect the offender’s health”: Goh Chin Soon at [165]. The court’s determination of whether the situation is sufficiently exceptional to warrant a reduction in sentence is therefore grounded in medical evidence that directly addresses the legal question.

31 As for advanced age, the court in Yap Ah Lai described the “critical point” as being whether an offender would suffer disproportionately by reason of his age. This would be “particularly pertinent” where the effect of the custodial sentence is that the offender would spend much of the rest of his life in prison (at [91] and [94]). On the facts in Yap Ah Lai, the court considered that the sentence of 15 months’ imprisonment was not of such a long duration that the sentence had to be moderated on account of the offender’s advanced age (at [94]). In contrast, in PP v UI, the Court of Appeal took account of the offender’s advanced age in imposing a sentence of 12 years’ imprisonment per charge for each offence of rape, resulting in an aggregate sentence of 24 years’ imprisonment. The court held that, as the offender was 55 years of age, a sentence at the higher end of the sentencing benchmark, which was 15 years’ imprisonment per charge (resulting in an aggregate sentence of 30 years’ imprisonment with both sentences running consecutively) would be crushing on the offender.

32 In contrast, as I have noted above, extrinsic factors are downstream consequences that might or would befall the offender because of the imposition of a sentence. If the courts were to take such consequences into account, there would be no logical limits as to when or how these consequences should be factored into sentencing. In my judgment, these consequences cannot meaningfully be considered in this context for at least three reasons. First, it is impossible for the court to place a value on such downstream consequences and to translate the potential financial losses into an appropriate reduction in sentence. The two are simply incommensurable.

33 Second, the different potential financial losses that could be faced by offenders cannot meaningfully be compared, given the varied consequences that one could suffer flowing from a particular sentence. Some might face these consequences upon any conviction; while others only upon the imposition of any term of imprisonment or one of a certain length. Moreover, every offender who is imprisoned for some length of time will likely lose his employment altogether. That plainly cannot be a reason for not meting out a sentence of imprisonment. This is true also of the potential losses of pensions, bonuses or prospects of promotion. On top of all this, if this was a factor that could be taken into account, it would unfairly work to the detriment of the unemployed, and possibly also those employed on a daily rate, or gig workers and other free-lancers.

34 Third, the consequences that would arise as a result of the sentence will often be indeterminate at the point when the case is heard before the court. It is helpful to recall what was said in this context in Public Prosecutor v Lim Yee Hua and another appeal [2018] 3 SLR 1106 at [71]:

35 These seem to me to be compelling reasons for concluding that the equal impact principle cannot apply in respect of extrinsic factors.

36 Despite this, the courts in other jurisdictions have not spoken with one voice. In R v Rees (1982) 4 Cr App Rep (S) 71 (“Rees”), the appellant, who was a serving soldier, pleaded guilty to assault with intent to resist arrest and was sentenced to one month’s imprisonment and ordered to pay £500 in compensation. Counsel for Rees argued that the court should substitute the imprisonment term with a fine, because the appellant’s future in the army would otherwise be “at an end”. The court considered a letter from the Ministry of Defence, which stated that a soldier sentenced to imprisonment in a civil court would be discharged from service absent exceptional reasons. The court took this into account and varied the sentence to a fine of £500 (see Current Sentencing Practice (Lyndon Harris ed) (Sweet & Maxwell, Looseleaf Ed, 2021 release) (“Current Sentencing Practice”) at para A1-2250).

37 In contrast, in R v Ranu [1996] 2 Cr App R (S) 334 (“Ranu”), the appellant, also a serving soldier, pleaded guilty to assault occasioning actual bodily harm and was sentenced to six month’s imprisonment. It was submitted before the court that he would be dismissed from service if a custodial sentence were imposed on him. Stuart-Smith LJ held that the armed forces did have some flexibility with regard to the retention of its officers. He further considered that “[if] the Army are anxious to keep this soldier, and [the court] can well understand why they should be, [the court] can see no reason why they should not do so, but … it is not a reason why, in the circumstances of this case, this Court should interfere with this sentence”, which he regarded as a “perfectly proper sentence” (see Current Sentencing Practice at para A1-2275).

38 I accept the facts in these cases may be distinguishable, but I cannot see a basis in logic and principle for these different outcomes. In my judgment, these cases illustrate the difficulty with taking such extrinsic consequences into account when sentencing an offender. It seems difficult to justify an outcome where an offender, such as the appellant in Rees, could get away with not serving an imprisonment term by virtue of the possibility that he might lose his job in the military, even though a custodial sentence would have otherwise reflected his culpability and the seriousness of the offence that he committed. Does this apply to any and all employees or is it an exception for armed forces personnel only? Does it apply to any offence where a fine is a sentencing option? Is it relevant that the impact of the reduced sentence on the offender may not achieve the applicable sentencing objectives?

39 In contrast, our jurisprudence is more settled, and our courts have generally rejected the proposition that additional hardship suffered by an offender due to the potential consequences of his sentence on his employment may be considered by the sentencing court. This was my holding in Stansilas, and this is also reflected in other precedents. In Public Prosecutor v Yue Mun Yew Gary [2013] 1 SLR 39, Quentin Loh J (as he then was) considered that the adverse impact on an offender’s career or job prospects as a result of his conviction is “but a natural consequence of his own acts and ought to be given little or no weight in mitigation” (at [67]). In Chow Dih v Public Prosecutor [1990] 1 SLR(R) 53, the offender was a medical doctor who was convicted of cheating his patients of sums of money. Chao Hick Tin JC (as he then was) held that the fact that the offender would also be dealt with professionally by the Medical Council was not a mitigating factor. The offender “must expect to be dealt with according to law as well as the disciplinary rules of his profession” (at [59]).

40 In my judgment, the position we have taken leads to principled outcomes. There are other considerations that fortify this conclusion. For one, taking account of factors that are extrinsic to an offender have nothing to do with the sentencing objectives that the court is obliged to consider in this context, and in fact, is likely to undermine the functioning of the criminal justice system.

41 Where factors intrinsic to an offender are concerned, such as his age or ill health, these are considered in terms of how the sentence imposed would be experienced by the particular offender, and whether as a result, the sentence would be disproportionate. The sentencing objectives of deterrence, rehabilitation, prevention and retribution continue to apply, but the operative interests are capable of being achieved through a reduced sentence because its impact on the offender would nonetheless be the same as a heavier sentence would have on a typical offender. Thus, the interest of deterrence, for instance, is not displaced if a reduced sentence were imposed as long as the reduction serves only to equalise the impact of the sentence on the particular offender before the court. In this context, the observations of Prof Ashworth and Prof Elaine Player are germane (see Andrew Ashworth and Elaine Player, “Sentencing, Equal Treatment and Sanctions” in Fundamentals of Sentencing Theory (Andrew Ashworth and Martin Wasik eds) (Clarendon Press Oxford, 1998) at pp 271–272):

42 Further, such intrinsic factors do not involve a wider social accounting and do not give rise to discrimination against some groups of offenders over others. They apply only in the exceptional situations where a specific offender’s impediments cause a substantial imprisonment term to be disproportionate or crushing on him.

43 In contrast, taking account of extrinsic factors could result in a fundamental assault on the criminal justice system. If the court were to place weight on factors such as the financial consequences of a particular sentence, and reduce an offender’s sentence on that basis, it would result in the more favourable treatment of certain individuals. As noted by Profs Susan Easton and Christine Piper in Sentencing and Punishment: The Quest for Justice (Oxford University Press, 2nd Ed, 2008) (at para 7.2.4):

44 In the same vein, Prof Ashworth argues that it would be “wrong to allow a source of mitigation that is only available to certain offenders who have an advantage that others lack”: see Andrew Ashworth, “Re-evaluating the Justifications for Aggravation and Mitigation at Sentencing” in Mitigation and Aggravation at Sentencing (Julian V Roberts ed) (Cambridge University Press, 2011) at p 31. I also find persuasive Prof Ashworth’s observations as follows (see Justifying the Grounds of Mitigation at 7):

45 It is a matter of fundamental importance that the criminal justice system be designed to work for all people in all circumstances. As I held in Public Prosecutor v Siow Kai Yuan Terence [2020] 4 SLR 1412 at [1]:

46 If this fundamental principle of equality were to be displaced in favour of some offenders, it would undermine and dilute the deterrent effect of the entire system of criminal justice. One of the core aims of the criminal justice system is to deter potential offenders from committing crimes. As noted by Prof Andrew von Hirsch in Censure and Sanctions (Clarendon Press Oxford, 2003), the criminal law “seems to have preventive features in its very design”. The State in criminalising conduct “issues a legal threat: such conduct is proscribed, and violation will result in the imposition of specified sanctions” (at p 12). General deterrence is “intended to create an awareness in the public and more particularly among potential offenders that punishment will be certain and unrelenting for certain offences and offenders”: Tan Kay Beng v Public Prosecutor [2006] 4 SLR(R) 10 at [31]. The criminal justice system works because the specified sanctions are imposed on any offender who violates the law and offenders are only treated differently if and where there is a principled basis for doing so. The balance that a sentencing court should assiduously strive to achieve is that between the need to mete out a just sentence on the facts of a particular case while seeking to achieve broad consistency among broadly similar cases.

47 This too strengthens the argument for keeping separate sentencing under the criminal law and whatever other consequences may flow from one’s conduct, these being factors extrinsic to both the offence and the offender. I reiterate my reasons in Stansilas at [110]–[111] for rejecting the argument that additional hardship in the form of loss of employment or disciplinary proceedings should be taken into account by the sentencing court, much, if not all, of which I have already developed above.

48 These considerations continue to apply even though, unlike the accused person in Stansilas, Raveendran is at the end of his career. Counsel for Raveendran, Mr Markus Kng (“Mr Kng”), submitted that this weighed in favour of my taking a more generous approach since Raveendran may not have the time to carve out a new career for himself and the loss of emoluments would therefore weigh especially heavily on him. I make two observations. First, that is a point that Raveendran can direct to his employers. Second, that illustrates the precise difficulty with taking into account such extrinsic factors. Where is one to draw the line? Would it make a difference if Raveendran had been five years away from retirement at the time of the offence instead of two? Or ten? This highlights the lack of manageable standards by which the court can take such considerations into account in a principled way. Therefore, I am satisfied that such extrinsic factors will generally have no mitigating weight in sentencing.

49 The next possible basis is the principle of parsimony, which postulates that offenders should only be punished to the minimum required to achieve the aims of punishment: see Bagaric et al, “Excessive Criminal Punishment Amounts to Punishing the Innocent: An Argument for Taking the Parsimony Principle Seriously” (2016) 57(1) South Texas Law Review 1 at 6. This has also been formulated as calling for the “least intrusive punishment” (Morris & Tonry, Between Prison and Probation: Intermediate Punishments in a Rational Sentencing System (Oxford University Press, 1990) (“Rational Sentencing System”) or sentences that are “no more severe, intrusive, or damaging to an offender’s ability later to live a law-abiding life” (Tonry, “Remodeling American Sentencing: A Ten-Step Blueprint for Moving Past Mass Incarceration” (2014) 13(4) Criminology & Public Policy 503). While these definitions may vary, the essential point is that the punishment should accord with the ultimate sentencing aims that the criminal justice system seeks to achieve.

50 The point on parsimony can be dealt with briefly. First, this argument was quite simply not one advanced by Raveendran. While it was helpfully brought to my attention for consideration by Mr See, it does not directly form part of the appeal by Raveendran.

51 Second, the current case law suggests that it is not applicable as a general principle in our courts. In Than Stenly Granida Purwanto v Public Prosecutor [2003] 3 SLR(R) 576 (“Than Stenly”), Yong Pung How CJ rejected the general applicability of the parsimony principle. There, the appellant had pleaded guilty to one count of conspiracy to possess forged credit cards with intent to use them as genuine, and five counts of conspiracy to cheat using credit cards. He was sentenced to a total of 6½ years’ imprisonment. On appeal, counsel for the appellant argued that the district judge had erred, among other things, on the ground that a “sentencing court, being cognisant of the gravity of the particular offence committed, should have regard to the ‘trite common law principle of parsimony’ and select the least severe sentencing option that is commensurate with the gravity of that specific offence” (at [10]). Yong CJ observed as follows (at [12]):

52 Third, in any case, the parsimony principle cannot apply to extrinsic factors for reasons that have already been set out in relation to the equal impact principle. In particular, equality before the law is a fundamental principle that cannot be easily displaced. If the parsimony principle were applied as a general rule, it could result in an outcome where an offender who is better socially situated as a result of his circumstances would be sentenced more leniently, so as to achieve the utilitarian aim of imposing the minimum possible sentence. Morris and Tonry in Rational Sentencing System argue that (at 89–90):

53 However, taking into account the “social threats and personal conditions” of an offender to determine a sentence introduces inconsistency and violates the principle of equality in sentencing. In my view, that is neither a desirable nor an acceptable outcome. Where an offender’s crime and criminal record would have warranted an imprisonment sentence, that sentence should generally be imposed save for exceptional circumstances which must be principled, transparent and articulated by the courts. As argued by Prof Ashworth in Sentencing and Criminal Justice (Cambridge University Press, 6th Edn, 2015) (at p 269):

54 I therefore regard the view in Than Stenly as sound, and do not consider the principle of parsimony to be generally applicable in the context of sentencing in our jurisdiction.

55 For completeness, I should state that the rejection of the parsimony principle as a general sentencing principle does not mean that it should not, in limited and appropriate circumstances be applied. An example is where an accused is a young offender as was observed in Public Prosecutor v ATW (A Minor) [2011] SGJC 2 at [3]:

56 It must, nevertheless, be reiterated that the principle does not apply whenever the accused person is a young offender. Due regard must be had to whether the circumstances personal to the accused in question call for a less intrusive punishment to be imposed. I respectfully consider that this was correctly articulated in Public Prosecutor v GCB (A Minor) [2019] SGYC 1 as follows:

57 I emphasise, however, that the applicability of the parsimony principle to young offenders is an exception to the general rule that it is not an applicable sentencing consideration in our jurisdiction. This needs to be restated because the point appears to have been extended incorrectly in a clutch of cases decided by the same District Judge: Public Prosecutor v Law Kok Leong [2009] SGDC 504 at [25]–[26], Public Prosecutor v Toh Beng Hua [2009] SGDC 506 at [25]–[26], Public Prosecutor v Hamidah Binte Hanif [2010] SGDC 331 at [26]–[27], Public Prosecutor v Kulandaivelu Padmanaban [2010] SGDC 407 at [20]–[21] and Public Prosecutor v Ezmiwardi Bin Kanan [2011] SGDC 152 at [49]. In each of these cases, the District Judge’s decision included the following reasoning:

58 In my judgment, the parsimony principle is not applicable by virtue of the fact that an offender happens not to have criminal antecedents. That fact will typically be considered as an offender-specific mitigating factor. The weight, relevance and impact that this fact will have on sentencing will inevitably depend on the entirety of the factual matrix before the court. For example, it will typically be less significant where the offender is facing a series of charges, since the absence of antecedents would, in relation to all but the first of the offences forming that series simply indicate that the offender had not been caught earlier. It will also have little weight if the evidence before the court similarly indicates that the offender had committed other offences but just had not been apprehended for those. On the other hand, if the evidence indicates that the offence in question is an aberration and out of step with an otherwise law-abiding record, then more weight may be accorded to the lack of antecedents. Further, the gravity of the offence and the circumstances in which it was committed will obviously be relevant considerations. These are trite observations, and they reflect the approach taken each and every day by sentencing judges in our jurisdiction. But they bear reiterating as a reminder of the complexity that inheres in sentencing, which is something we should not lose sight of by resorting to reductionist generalities.

59 The third possible basis for having regard to Raveendran’s possible loss of emoluments is the principle of judicial mercy, which was relied on in this case. The principle was carefully examined in Chew Soo Chun. Several points should be emphasised for present purposes. First, judicial mercy is founded on a humanitarian concern; it is this conceptual basis which allows a court to temper the punishment in the light of an offender’s personal circumstances: Chew Soo Chun at [21]–[22], citing John Tasioulas, “Mercy” (2003) 103 Proceedings of the Aristotelian Society 101 at p 117. Second, the grant of judicial mercy entails a holistic review in which the relative interests are weighed. As noted in Chew Soo Chun at [25], such interests include the public interest in punishing crimes in order to denounce them so as to safeguard society, and the concern to avoid imposing punishment that is unduly harsh given the particular circumstances of the offender. Finally, judicial mercy is an exceptional jurisdiction. The ultimate effect of such judicial mercy is that the culpability of the offender is displaced by considerations of humanity in the court’s determination of the appropriate sentence, such that benchmark sentences “effectively play no part”: Chew Soo Chun at [23]. It should therefore come as no surprise that such jurisdiction is exercised sparingly.

60 A survey of the case law reveals two situations in which judicial mercy has been invoked, and both are founded in concerns relating to ill health (see Chew Soo Chun at [22], citing Lim Kay Han Irene v Public Prosecutor [2010] 3 SLR 240 (“Irene Lim”) at [46]). First, the typical situation is where the offender is suffering from some terminal illness: Lim Teck Chye v Public Prosecutor [2004] 2 SLR(R) 525 at [82]. The second is where the offender is so ill that an imprisonment term carries a high risk of endangering his life: Public Prosecutor v Tang Wee Sung [2008] SGDC 262.

61 Mr Kng urged me to extend the principle of judicial mercy to cover the present situation of possible financial consequences. This was founded on the observation in Chew Soo Chun at [22] that:

62 While I agree that the situations in which judicial mercy will be extended are not closed, I am equally convinced that the court should always remind itself that this is an exceptional jurisdiction. Having done so, it is clear to me that it cannot be invoked to ameliorate the possible financial consequences of a condign sentence. The threshold to warrant the exercise of judicial mercy is an exceedingly high one. For instance, in Irene Lim, the offender suffered from a medical condition of involuntary tremors that were not controllable and hampered her fine finger activities; she also suffered from severe anaphylactic reactions to unknown food substances that would cause cardiovascular collapses, and Morton’s metatarsalgia that resulted in pain while walking and required special footwear. Yet, the court found that the cumulative effect of these conditions did not merit the exercise of judicial mercy.

63 Other examples were helpfully compiled in Chng Yew Chin v Public Prosecutor [2006] 4 SLR(R) 124 at [52] as follows:

64 These cases demonstrate that even when faced with very trying medical circumstances, the courts will not easily exercise judicial mercy, thus respecting its exceptional nature. In my judgment, the loss of possible emoluments falls far short of warranting the exercise of judicial mercy in this case.

65 Mr See raised a final possible ground under which the loss of emoluments might fall to be considered, namely, pursuant to the provisions of any applicable statute. Put simply, there may be statutes that require a sentencing court to have regard to certain consequences when determining the appropriate sentence. One such example, that is also relevant for present purposes, is s 108(2) of the Singapore Armed Forces Act (Cap 295, 2000 Rev Ed) (“SAF Act”), that provides as follows:

66 Section 108(2) of the SAF Act was considered in Ong Beng Leong v Public Prosecutor [2005] 1 SLR(R) 766 (“Ong Beng Leong”). There, the appellant, a former commanding officer of the SAF Training Resource Management Centre, was charged and convicted of ten charges of using false documents with intent to deceive his principal, an offence under s 6(c) of the Prevention of Corruption Act (Cap 241, 1993 Rev Ed) (“PCA”). After the irregularities were discovered, the SAF conducted a summary trial, and the appellant was fined a total of $2,250. For the PCA offences, a District Judge also sentenced him to a total of six months’ imprisonment. On appeal, the appellant argued that this sentence was manifestly excessive because, among other things, the District Judge had failed to consider the military punishment that had been meted out to him. In relation to s 108(2) of the SAF Act, Yong Pung How CJ observed at [57]–[58] as follows:

67 Yong CJ ultimately held in Ong Beng Leong that a custodial sentence was unavoidable given the aggravating factors in this case. The appellant’s sentence, however, was reduced particularly on account of the mitigating circumstances which included the military punishment already imposed and the fact that he had been “suspended from the SAF since the commencement of the trial, and is likely to lose considerable amounts in pension and other benefits as a result of his convictions” [emphasis added]: Ong Beng Leong at [59]–[61]. Although Yong CJ appeared to take cognisance of the overall suffering the offender would have to endure, it is plain that he was not considering or suggesting that this was an extrinsic factor that a court should consider as a matter of principle. For the reasons I have already explained, I do not regard it appropriate to do so.

68 I agree with Yong CJ’s observations that any military punishment may be taken into consideration as a mitigating circumstance. This accords with the terms of s 108(2) of the SAF Act, which provides that a civil court shall have regard to such military punishment. I also agree with Yong CJ that this should not fetter the civil court’s discretion. The pertinent consideration remains whether the sentence imposed is fair and just, having regard to all relevant circumstances. I note that in this case, no military punishment has as yet been imposed on Raveendran. The only argument available to him is that, as in Ong Beng Leong, it is likely that he will lose considerable amounts in emoluments. This, however, does not in and of itself mean that the court should avoid imposing a custodial sentence if that is appropriate.

69 For all these reasons, I was satisfied that I should not have regard to the possible consequences of the sentence on Raveendran’s entitlement to his retirement benefits and emoluments. I turn to briefly explain my decision as to the sentence that should be imposed in the circumstances. The DJ assessed Raveendran’s culpability as medium, primarily on account of his high alcohol level. He also assessed the harm engendered as slight. In the circumstances, he held that the starting point for sentencing in this case was one week’s imprisonment. I had no difficulty with the approach taken by the learned DJ up to this point.

70 The learned DJ then took into account the fact that Raveendran had pleaded guilty, and that he had made restitution. He also observed that this suggested that Raveendran was remorseful but then concluded that it did not justify any adjustment in the starting point, seemingly because of the high alcohol content in his blood. In my judgment, the learned DJ erred at this stage of the analysis. Whilst it is true that regard should be had to Raveendran’s higher level of alcohol intake in the overall sentencing analysis, this has already been reflected in the classification of the present case as one of medium culpability. That is why this case crossed the custodial threshold. However, other mitigating factors that are present, such as remorse, are not consequently nullified. The court remains obliged to consider such factors in calibrating Raveendran’s sentence from the indicative starting point.

71 In my judgment, the facts in Vilashini and Solomon Seah were relevant in this context. In Vilashini, the offender consumed alcohol and then went on a joyride in the course of which she lost control of the motor car, causing chain collisions involving seven vehicles in an open-air carpark. The offender was found to have 53 microgrammes of alcohol in every 100 millilitres of breath and was sentenced to a week’s imprisonment and a disqualification period for the drink driving offence. In Solomon Seah, the offender fell asleep at the wheel and crashed his car into a traffic light, uprooting it. The cost of repairs for the damaged traffic light was estimated to be about $1071.26 and restitution was made by the offender to the LTA. The offender was found to have 59 microgrammes of alcohol in every 100 millilitres of breath. He was sentenced to three days’ imprisonment and a disqualification period for the drink driving offence.

72 Comparing the present facts with those in Vilashini and Solomon Seah, I was satisfied that an adjustment of the starting sentence in this case to five days’ imprisonment was warranted and I allowed the appeal to that extent. I took this view because the offender in Vilashini caused far more harm than in the present case. It seemed to me, therefore, that Raveendran’s sentence should be lower. On the other hand, while the harm and surrounding circumstances here were comparable to that of the offender in Solomon Seah, the blood alcohol level of the offender there was significantly lower. Hence, I decided that Raveendran’s sentence should be higher than that imposed in that case.

73 For these reasons, the appeal was partially allowed and Raveendran’s sentence was reduced from a week to five days’ imprisonment.

74 I once again record my appreciation to Mr See for his extremely helpful submissions.