Kamis bin Basir v Public Prosecutor [2023] SGHC 348

Introduction

The appellant is Mr Kamis bin Basir. He is 54 years old. He has been in and out of prison, for the past three decades, for a whole slew of offences. In the wake of his latest offending spree, the appellant pleaded guilty to one charge of snatch theft and one charge of drug consumption, and consented to have two other charges, one for drug possession and the other for assisting an illegal moneylender, taken into consideration for the purpose of sentencing. The District Judge (“DJ”) in the court below sentenced him to ten years of preventive detention (“PD”) but declined to backdate the sentence. The DJ’s grounds of decision can be found in Public Prosecutor v Kamis Bin Basir [2022] SGDC 297 (“GD”).

The appellant filed an appeal. He contested the DJ’s decision on two grounds. First, that he should not be sentenced to PD. Second, that if he were sentenced to PD, that sentence should have been backdated.

We observed that the lower courts have taken inconsistent positions on whether a PD sentence can be backdated, and the principles guiding the court’s power to backdate. In the present case, the DJ relied on the decision of the Court of Appeal in Public Prosecutor v Rosli bin Yassin [2013] 2 SLR 831 (“Rosli”) to find that a PD sentence could only be backdated in exceptional cases (GD at [42]). In Public Prosecutor v Png Gek Kwee [2022] SGDC 179, a district judge backdated a PD sentence to the date of first remand, citing s 318 of the Criminal Procedure Code 2010 (2020 Rev Ed) (“CPC”). In backdating the PD sentence, the district judge did not cite Rosli or suggest that backdating should only be ordered in exceptional cases. In Public Prosecutor v Ow Gan Wee [2023] SGDC 16, a district judge backdated a PD sentence to the date of first remand (at [50]). Curiously, however, the district judge did so on the purported basis that there is no qualitative difference between a PD sentence and a regular imprisonment sentence. Conversely, the district judge did not cite or rely on s 318(1) of the CPC to backdate the PD sentence.

We thus appointed Ms Estad Amber Joy (“Ms Estad”) as young independent counsel to assist us with the question of when a sentence of PD should be backdated. This question of when a PD sentence can be backdated is of general importance and is likely to arise whenever the court decides to impose a PD sentence. A decision is likely to provide helpful guidance to the lower courts. In particular, we asked Ms Estad to consider the principles which ought to guide the Court’s discretion to backdate a PD sentence under ss 318(1) and 318(3) of the CPC. In answering this question, Ms Estad was asked to consider the decision in Rosli and whether the power to backdate a PD sentence under ss 318(1) and 318(3) of the CPC should only be exercised in exceptional cases.

We heard the appeal on 26 October 2023 and allowed it in part, ordering that the appellant’s sentence of ten years’ PD be backdated to his date of arrest. We now set out the reasons for our decision.

The facts and the decision below

On the morning of 20 January 2022, the appellant was in the vicinity of Jurong West Street 91. He was looking for something to steal as he had no money but wanted to purchase heroin as well as pay off his debts. He spotted the victim, an elderly woman, walking along the street near Blk 966 Jurong West Street 93. He noticed that she was wearing a gold chain. Having decided to steal her gold chain, the appellant tailed the victim back to her residence. He saw her enter her residence. In a bid to get her attention and to lure her closer to the gate, where she would be within striking range, the appellant rang the doorbell and shouted “David”. As the victim approached the gate, the appellant reached through the rails of the gate and grabbed the gold chain violently, breaking it.

These facts formed the subject of the first charge of using criminal force to commit theft under s 356 of the Penal Code 1871 (2020 Rev Ed).

After robbing the victim of her gold chain, the appellant took a bus to Boon Lay Interchange. He then went to his friend’s housing block where he disposed of his grey shirt to escape detection by the police who would be hot on his trail. Thereafter, he took a taxi to Jalan Bukit Merah where he pawned the gold chain for $870.

The appellant then took the train to Bukit Batok and went to a block at Bukit Batok where he had previously purchased heroin. After making this purchase, he took a taxi back to Boon Lay where he was arrested. Upon his arrest, he was taken to Woodlands Police Divisional Headquarters. Urine samples were taken. They were found to contain monoacetylmorphine, a known metabolite of heroin and a specified drug listed in the Fourth Schedule to the Misuse of Drugs Act 1973 (2020 Rev Ed) (“MDA”). The appellant stated that he had last consumed heroin on the morning of 20 January 2022 at around 9am.

These facts formed the subject of the second charge. The appellant was liable for enhanced punishment for drug consumption under the MDA because he had a previous conviction for heroin consumption.

The DJ considered that a PD sentence was amply justified in the present case. He considered that there was a real need to protect the public against future reoffending given the appellant’s history of offending, the circumstances of the present offences that the appellant was charged with and the risk of future reoffending. Examining the appellant’s lengthy antecedents, the DJ noticed three highly disturbing patterns. First, the appellant’s offences were all drug or property-related. Second, since 2000, the appellant was unable to refrain from crime for any substantial period whenever he was released from incarceration and re-integrated into society. Third, all of the earlier punishments had a limited rehabilitative or deterrent effect. Notwithstanding the fact that the appellant had been incarcerated for increasingly lengthy periods, undergone corrective training, and even been caned, this did not deter him from reoffending (GD at [25]–[29]).

Continuing in this vein, the DJ noted that the circumstances of the appellant’s present offences reinforced the view that he was a recalcitrant offender. Again, the offences which the appellant had been charged with were drug and property-related (GD at [30]).

It was also clear, from the facts underlying the first charge, that the offence was not a trivial one. The offence was premeditated. The appellant had formed an intention to steal from the victim. He stalked her to her flat and had no qualms in attacking the victim. He had also deliberately targeted the victim because she was an elderly person. He had also sought to evade arrest – and this cast doubt on the extent of his remorse (GD at [30]).

The DJ then considered the appellant’s prospects of reoffending. He noted that the pre-sentencing report (“PSR”) highlighted several risk factors which further lent weight to the view that the appellant should be incarcerated for a substantial period to protect the public. For one, the PSR categorised the appellant as being in the “high risk/need” level of criminal reoffending. There was a 70.2% probability of recidivism within two years of release. There were a number of risk factors which supported the view that it was highly probable that the appellant would reoffend. As the appellant himself conceded, his social circle consisted of peers who were negative influences. Further, his own history of drug abuse could also lead him to commit property offences to fuel his habit. The appellant’s behavioural and thinking patterns also suggested that he thought crime was a useful way of resolving his immediate problems. Although capable of introspection and self-reflection, the appellant could not resist the temptation to resort to crime to resolve his problems (GD at [31]–[33]).

While the appellant’s family and his choice of recreational activities did not present issues, this did not mean that they would stop him from reoffending. In the round, the picture that emerged was that the appellant was highly likely to continue to commit drug and property-related offences which could harm innocent members of the public. There was thus a need to sentence the appellant to PD to protect the public from future reoffending by the appellant (GD at [34]–[35]).

As to the length of PD, the DJ considered that ten years was warranted, taking into account the appellant’s history of habitual offending, the circumstances of the present offences and the risk factors that strongly suggested that he was likely to reoffend if not incarcerated. In particular, the DJ noted that the likely sentence the appellant would have received had he been sentenced to a term of regular imprisonment was nine years. This suggested that the appropriate length of the PD sentence should not, in any event, be less than nine years. After all, PD operated on a different penological basis from regular imprisonment, and the principle of proportionality would not apply as rigorously (GD at [38]–[42])

Finally, the DJ concluded, citing the decision of the Court of Appeal in Rosli at [20], that there was no reason to backdate the sentence meted out. The case was not an exceptional one where backdating was possibly justified. Here, we note that the DJ applied Rosli without considering s 318(1) of the CPC (GD at [43]–[44]).

Issues

There were two issues raised in this appeal, and we will deal with them in turn. First, was the DJ correct in sentencing the appellant to PD? Second, if a sentence of PD was justified, should that sentence be backdated?

Our decision

Issue 1: Whether the DJ erred in sentencing the appellant to PD

The technical requirements that must be satisfied before the court can order a sentence of PD are set out in s 304(2) of the CPC, which states:

It could not be seriously contended that the appellant did not satisfy the technical requirements in s 304(2)(a) of the CPC. The present offences that he had been convicted of were both punishable with a term of imprisonment of two years or more. It was also clear from the appellant’s lengthy antecedents that he had been convicted in Singapore at least three times for offences punishable with such a sentence and had been sentenced on at least two of those occasions to imprisonment or corrective training. In particular:

(a) On 9 March 2000, the appellant was convicted of an offence of consumption of a specified drug under s 8(b)(ii) punishable under s 33A(1) of the Misuse of Drugs Act (Cap 185, 1998 Rev Ed). He was sentenced to five years and six months’ imprisonment with three strokes of the cane.

(b) On 13 May 2005, the appellant was convicted of three counts of robbery, with one count of fraudulent possession of property, 16 counts of snatch theft and four counts of robbery taken into consideration for the purposes of sentencing. He was sentenced to seven years of corrective training and 18 strokes of the cane.

(c) On 11 July 2013, the appellant was convicted of three counts of snatch theft, with four other counts of the same offence taken into consideration for the purposes of sentencing. He was sentenced to eight years of corrective training.

Having satisfied these technical requirements, the next step was to consider whether a sentence of PD was justified. In this vein, it must be emphasised that PD is a sentencing option that is only exercised in the appropriate case by reference to a range of considerations directed at whether there is material to persuade the court that the focus in all the circumstances in relation to the offender at hand should no longer be on rehabilitation but on prevention (ie, the offender’s antecedents and ability to stay out of trouble). Thereafter, having regard to the nature of the offences and other relevant circumstances, the court considers what the appropriate term of PD should be.

In other words, the court must be satisfied that a PD sentence is expedient for the protection of the public: Ravindran s/o Kumarasamy v Public Prosecutor [2023] 3 SLR 1343 at [45]–[46] (“Ravindran”) citing Re Salwant Singh s/o Amer Singh [2019] 5 SLR 1037 (“Salwant Singh”) at [52], and Rosli at [11]. The court must also take into account the totality of the offender’s previous convictions, as well as the circumstances of the offender’s present offending: Ravindran at [47]. These considerations must be borne in mind, otherwise there may be a tendency to resort to long PD sentences without appreciating the nuances of whether such sentences would be appropriate in the case at hand.

In so far as the role of an appellate court is concerned, it should be remembered that appellate intervention in reappraising sentences imposed by a court of first instance is limited. Sentencing is, after all, a matter of discretion and requires a delicate balancing of myriad considerations that are often conflicting: Rosli at [8]. Appellate intervention in the sentencing judge’s exercise of discretion is only warranted in cases where the judge below had failed to appreciate the facts before him, or where the exercise of the sentencing discretion was contrary to principle and/or law. It does not suffice to merely show that the appellate court would or could have awarded a higher or lower sentence: Rosli at [9] citing Public Prosecutor v Mohammed Liton Mohammed Syeed Mallik [2008] 1 SLR(R) 601 at [84].

In the event, we found that a sentence of PD was amply justified. The appellant has a long list of both property and drug-related antecedents. It was also particularly noteworthy that despite having been sentenced to corrective training on two previous occasions, that did not seem to have any rehabilitative effect on him. As the DJ observed, less than six months after completing his last corrective training regime, the appellant went out and committed the present offences. The conclusion we drew was that the appellant was indeed a habitual offender who was unwilling to stay crime-free for any significant period. Furthermore, it also appeared that the appellant remained undeterred despite his earlier punishments which also included caning.

This view was reinforced by the PSR which categorised the appellant as being at a high risk of criminal reoffending. The report stated that the appellant belonged to a group of prisoners with a 70.2% probability of recidivism within two years of release.

The appellant challenged this. He questioned the basis on which the report derived the finding that he belonged to a group of prisoners who had a 70.2% probability of reoffending. There was no merit to this argument. While the report does not explain how it had arrived at this conclusion, making it difficult to assess whether the appellant did indeed fall within this group of offenders (see Liat Levanon, “Statistical Evidence, Assertions and Responsibility” (2019) 82(2) Modern Law Review 269) the DJ did go on to analyse the risk factors specific to the appellant and did not base his decision to sentence the appellant to PD entirely on the statistical probability of him reoffending. As the DJ had noted, this was not merely a matter of statistics because of the risk factors stated in the PSR (GD at [33]).

It was clear to us that the risk factors presented in the PSR demonstrated that the appellant was very likely to reoffend. The PSR stated that one reason the appellant had reoffended in the present case was the bad company he kept. He had agreed to act as a guarantor for his friend who had obtained a loan from an illegal moneylender. When his friend defaulted on repayments and became uncontactable, the illegal moneylenders went after the appellant who resorted to crime to pay off his debts.

This leads us to the second point stated in the PSR: the appellant’s solution to his predicament was to resort to crime. Despite having been in and out of jail for the past three decades, it was apparent that the appellant had not learnt to solve his immediate problems without resorting to crime. When confronted with financial troubles and demands for repayment by illegal moneylenders, the appellant chose to resort to theft instead of seeking out other avenues of help. In this vein, we also noted that the appellant has a tendency to trivialise his offences. As noted in the PSR, the appellant was of the view that his present offence was not as severe as what he had done in the past (for which he was sentenced to corrective training). Taken together, one would conclude that the appellant somehow not only failed to appreciate the gravity of his offences, but it also spoke to his attempt to rationalise and justify his actions.

We thus agreed with the DJ’s finding that the appellant was very likely to reoffend. That much was clear from the corrosive effect of bad peers which the appellant chose to mingle with, as well as his own attitudes towards offending as a means of resolving his problems.

It is also stated in the PSR that the appellant has a supportive family, including a wife who has steadfastly stood by him, and that his recreational activities did not pose a concern. While these factors posed no concern, as the DJ rightly noted, that did not mean that they were “protective”. In other words, they did not reduce the likelihood that the appellant would reoffend.

We were thus satisfied that a PD sentence was warranted. It was clear that the appellant was not only a habitual offender, but there were also very clear risk factors which made it very likely that he would reoffend in future.

Issue 2: Should the sentence of PD be backdated?

We now turn to the key issue in this appeal, that is, whether the PD sentence should be backdated. In Rosli, the Court of Appeal made the following observations:

Crucially, post-Rosli, s 318(1) of the CPC was amended in 2018 by s 90 of the Criminal Justice Reform Act 2018 (Act 19 of 2018) (“CJRA”) to provide that “a sentence of imprisonment, reformative training, corrective training or preventive detention shall take effect from the date it was passed, unless [the court] otherwise directs” [emphasis added]. (Section 318(1) was again amended in 2019 by the Criminal Procedure Code (Amendment) Act 2019 (Act 14 of 2019) to parcel out reformative training, which was then addressed separately under a new s 318(1A).) The crucial point is that prior to these amendments, s 318(1) referred only to a sentence of imprisonment, and Rosli was decided on this statutory footing. The CJRA also introduced s 318(3), which provides that “[t]o avoid doubt, a court may under subsection (1) direct that a sentence of imprisonment … corrective training or preventive detention is to take effect on a date earlier than the date the sentence is passed”. The relevant parts of s 318 now read:

The effect of ss 318(1) and 318(3) is this. The starting point is that the sentence of PD takes effect on the date it was passed. The court, however, has the discretion to backdate the sentence of PD. In exercising its discretion, the court must take into account the considerations spelt out by ss 318(4) and 318(5), which state:

The question before us, therefore, was whether what had been held by the Court of Appeal in Rosli, specifically, that the power to backdate a PD sentence should only be exercised in exceptional cases, should continue to apply in the context of s 318 of the CPC.

The Prosecution disagreed with Ms Estad’s submission that this slew of legislative amendments to the CPC (above at [43]) had created a new legislative framework governing the court’s approach to backdating a sentence of PD and represent a marked departure from the position set out in Rosli. The Prosecution argued that the CJRA amendments did not alter the fundamental nature or principles concerning PD and were not intended to displace or overrule the principles articulated in Rosli. In support of its argument, the Prosecution pointed to the Explanatory Statement to the Criminal Justice Reform Bill (Bill No 14/2018) (“Explanatory Statement”) which stated:

We could not accept the Prosecution’s argument. First, a key premise of the decision in Rosli was the absence of an express statutory power to backdate a PD sentence. Second, the decision in Rosli was also premised on the fact that a PD sentence is a “prospective” sentence and distinct from a regular imprisonment sentence (Rosli at [17]). In Public Prosecutor v Ng Kim Hong [2014] 2 SLR 245 at [32], the High Court relied on similar reasoning to find that there was no basis to backdate a sentence of corrective training. In Sim Yeow Kee v Public Prosecutor and another appeal [2016] 5 SLR 936 (“Sim Yeow Kee”), however, a 3-judge coram of the High Court held (at [120] and [121]) that a sentence of corrective training could be backdated, finding there was no justification for not backdating a corrective training sentence since there was no longer any qualitative difference between a corrective training sentence and a regular imprisonment sentence. While Sim Yeow Kee did not consider whether a PD sentence can be backdated or whether a PD sentence is qualitatively different from a regular imprisonment sentence, the court recognised that a discretion to backdate existed notwithstanding the absence of an express statutory provision.

It was clear to us that the amendments to the CPC (see [43] above) had indeed wrought a sea-change in so far as the backdating of a sentence of PD was concerned. It is hornbook law that in interpreting a statute, the first port of call must necessarily be the text of the relevant provision: Tan Cheng Bock v Attorney-General [2017] 2 SLR 850 at [37] and [43]. Here, s 318 not only expressly gives the court the power to backdate a sentence of PD, but also expressly spells out the analytical approach that the court must take in determining the date on which the sentence of PD should commence. Notably, s 318 of the CPC does not mandate that a sentence of PD can only be backdated in exceptional cases. On the contrary, the words of the provision suggest there is to be no qualitative difference, in this context of whether backdating is available, between PD and a normal sentence of imprisonment. Holding, therefore, that a sentence of PD can only be backdated in exceptional cases would be adding a gloss to the text of s 318.

Finally, as Ms Estad pointed out, under s 318, the court now has the same express statutory discretion under ss 318(1) and 318(3) of the CPC to backdate a sentence of regular imprisonment, corrective training and PD – but this discretion does not extend to a sentence of reformative training (see s 318(1A) of the CPC). As the Senior Minister of State for Law explained in the Second Reading of the Criminal Procedure Code (Amendment) Bill (Singapore Parliamentary Debates, Official Report (8 March 2019), vol 94 (Mr Edwin Tong Chun Fai, Senior Minister of State for Law)):

If one considers the legislative framework enshrined in s 318, Parliament had clearly intended that sentences of regular imprisonment, corrective training and PD be treated in the same manner, with the only exception being that of reformative training. It was therefore insufficient to argue, as the Prosecution did, that because a sentence of PD rests on a different penological basis from that of regular imprisonment or corrective training, such a sentence of PD should only be backdated in exceptional cases.

There is, in our view, no reason why the considerations in s 318 should apply differently for a term of regular imprisonment and for PD. After all, PD is meant to protect the public by keeping the offender out of society – remand also serves that purpose. It is true that the length of PD sentences are determined prospectively, in that the court looks ahead and considers how long the offender needs to be removed from society to protect the public (see Public Prosecutor v Rahim bin Basron [2010] 3 SLR 278 (“Rahim bin Basron”) at [56]). But, as Ms Estad pointed out, and with which we agree, the court should also be able to consider the period spent in remand because the offender has had his liberty curtailed during that period, and the public has been protected from the risk of him offending: see Rahim bin Basron at [57]. The CA in Rosli, which had been decided before s 318 of the CPC was amended, had made a similar observation (Rosli at [20]). The CA noted that the remand period could possibly be a factor to be taken into account when considering the length of the PD. This was, in essence, another way of effectively backdating the sentence of PD, by reducing its length by the amount of time spent in remand.

As a final point, we would note that backdating a sentence of PD would not deprive the offender of the benefits of the PD regime, for the period of backdating. A sentence of PD is broken down into three stages, and at the first stage, the offender serving a sentence of PD is treated no differently from an offender serving a term of regular imprisonment: see reg 16 of the Criminal Procedure Code (Corrective Training and Preventive Detention) Regulations 2010.

In summary, the amendment of s 318 has changed the landscape in so far as the backdating of a sentence of PD is concerned. The court does have the power to backdate a sentence of PD, and this power to backdate a sentence of PD is not limited to exceptional cases. We would also emphasise that the amendment of s 318 was not intended to have retrospective effect.

With this in mind, we now turn to consider the principles that should guide the court in deciding when to backdate such a sentence. It bears noting that the power to backdate a sentence under s 318 is a discretionary one, and it does not follow that the court must backdate a sentence of PD: Kow Keng Siong, Sentencing Principles in Singapore (Academy Publishing, 2nd Ed, 2019) (“Sentencing Principles in Singapore”) at paras 27.224–27.225.

In determining the length of a sentence of PD, the primary consideration before the court is the need to protect the public. The court will, in its assessment, consider the totality of the offender’s previous convictions as well as the circumstances of the offender’s present offending: Ravindran at [45]–[47] citing Salwant Singh at [52] and Rosli at [11]; Kuah Teck Hin v Public Prosecutor [2022] 5 SLR 720 at [3]–[4].

Having said that, considerations of proportionality, whilst attenuated in the context of PD, should not be wholly ignored. To cite an example, it has been noted that the sentence of eight years’ PD, which was meted out to the offender in Tan Ngin Hai v Public Prosecutor [2001] 2 SLR(R) 152 who had stolen $1.10 from a van, was wholly disproportionate to the offence: Public Prosecutor v Low Ji Qing [2019] 5 SLR 769 at [77] citing Sim Yeow Kee at [97].

We consider that these two considerations, viz, the need to protect the public, as well as proportionality, should guide the court in deciding whether to backdate the sentence of PD. As to how backdating should work in the context of PD, this should be done in two stages.

First, the court should decide on the appropriate length of PD, bearing in mind the overarching principle which is the need to protect the public (see above at [45]).

Second, having arrived at a landing on the appropriate length of PD, the court must then take a step back and consider the effect of backdating the sentence. Specifically, the court should consider whether the total effective period of incarceration, in the event the sentence is backdated, would still give effect to the overarching principle of the need to protect the public. This is because the practical effect of backdating the intended PD sentence would be to shorten the period for which an offender would otherwise be incarcerated: Sentencing Principles in Singapore at para 27.223, citing Mani Nedumaran v Public Prosecutor [1997] 3 SLR(R) 717 at [9]. It will often be the case that even with a backdated sentence the overall effective period of incarceration will be sufficient, and that not backdating may result in unfairness.

It is perhaps useful at this juncture to consider the scenario where the court deems that the maximum of 20 years’ PD should be imposed (see, eg, Public Prosecutor v Raffi Bin Jelan and another [2004] SGHC 120; Public Prosecutor v Syed Hamid bin A Kadir Alhamid [2002] 2 SLR(R) 1018; Heng Jong Cheng v Public Prosecutor [1999] 1 SLR(R) 769; Public Prosecutor v Wong Wing Hung [1999] 3 SLR(R) 304). In such a case, it would be wrong in principle for the court to order the maximum PD sentence of 20 years but refuse to backdate it on the sole basis that it considers that the offender ought to be kept out of society for longer than 20 years, provided that backdating would otherwise be justified. This would run counter to the statutory limit which Parliament had enacted in respect of PD sentences (see, eg, Public Prosecutor v Louis Pius Gilbert [2003] 3 SLR(R) 418 at [28]).

Ultimately, the decision to backdate a sentence lies within the discretion of the court. Sentencing is, after all, an art and not a science (see Kwan Weiguang v Public Prosecutor [2022] 5 SLR 766 at [38], citing ADF v Public Prosecutor and another appeal [2010] 1 SLR 874 (“ADF”) at [197]) – and in this vein, the discretion to backdate a sentence is one of the tools at the court’s disposal which allows it to shape the appropriate sentence, taking into account all the relevant factors. We emphasise that the views which we have set out above should be read in that light and not as a set of rules to be rigidly and mechanistically applied in every case (see ADF at [218]).

Turning back to the present case, we were of the view that the appellant’s sentence should be backdated to the date of his arrest: 20 January 2022. The appellant had spent about ten months in remand by the time he was sentenced by the DJ. If the sentence was not backdated, the appellant would effectively be incarcerated for almost 11 years.

Conclusion

For the reasons above, we allowed the appellant’s appeal in part in that the sentence of PD was backdated to the date of his arrest: 20 January 2022.

Finally, we wish to record our deep appreciation to Ms Amber Estad for her comprehensive research and submissions which assisted us considerably.