This judgment text has undergone conversion so that it is mobile and web-friendly. This may have created formatting or alignment issues. Please refer to the PDF copy for a print-friendly version.
Public Prosecutor v Aniza bte Essa
[2009] SGCA 16
| Case Number | : | |
| Decision Date | : | 20 April 2009 |
| Tribunal/Court | : | |
| Coram | : | Chan Sek Keong CJ; Andrew Phang Boon Leong JA; V K Rajah JA |
| Counsel Name(s) | : | Walter Woon, Tan Kiat Pheng, Gillian Koh-Tan and Samuel Chua (Attorney-General's Chambers) for the appellant; Noor Mohamed Marican (Marican & Associates) for the respondent |
| Parties | : | Public Prosecutor — Aniza bte Essa |
Criminal Procedure and Sentencing – Sentencing – Mentally disordered offenders – Whether Hodgson criteria appropriate in Singapore to determine when life imprisonment justified – Whether Hodgson criteria cumulative
Criminal Procedure and Sentencing – Sentencing – Statement of facts referring to psychiatric reports containing statements made by accused to psychiatrists – Whether judge entitled to treat such statements as fact
Criminal Procedure and Sentencing – Sentencing – Wife abetting young lover to kill her husband – Wife suffering from depression caused by husband's abuse – Whether nine years' imprisonment appropriate sentence
|
20 April 2009 |
Judgment reserved. |
Chan Sek Keong CJ (delivering the judgment of the court):
Introduction
1 This is an appeal against sentence by the Public Prosecutor (“the PP”). He has asked this court to set aside the sentence of nine years’ imprisonment imposed on Aniza binte Essa (“Aniza”) by the trial judge (“the Judge”) and to sentence her to life imprisonment.
2 This appeal raises an important issue of sentencing practice relating to the principles the court should apply in punishing a young female offender with two young children who has committed a grave offence (which requires a substantial term of imprisonment) but who was found to have diminished responsibility when she committed the offence. The Judge applied what are known as the Hodgson criteria (see [9] and [11] below) in determining the appropriate sentence for Aniza. The rationale of the Hodgson criteria is that mentally unstable offenders who have committed serious offences should only be kept in prison for as long as is necessary for the protection of society (from the likelihood of recidivism), but no longer. In this appeal, the PP has urged this court to sentence Aniza to life imprisonment on the principle of retribution and/or deterrence, the rationale of each, respectively, being that the punishment should fit the crime and that the punishment should have the effect of deterring others and the offender herself from committing similar offences in future.
3 The sentence of nine years’ imprisonment (which was backdated to the date of remand on 3 July 2007) was imposed on Aniza for the offence of abetment of culpable homicide not amounting to murder under s 304(a) read with s 109 of the Penal Code (Cap 224, 1985 Rev Ed) (“the abetment offence”). As the maximum sentence for the abetment offence is life imprisonment (which is the same for an offence under s 304(a) of the Penal Code (“s 304(a) offence”)), the PP argues that the sentence of nine years’ imprisonment is too lenient, particularly having regard to: (a) the apparently callous manner in which Aniza committed the abetment offence; and (b) the fact that the offender in the principal offence has to serve, effectively (as submitted by the PP), a sentence of life imprisonment.
4 Aniza had pleaded guilty to, and was convicted on, the charge of abetting Muhammed Nasir bin Abdul Aziz (“Nasir”), her 16-year-old lover, to kill her husband (“the deceased”). As the facts of the case are set out fully in PP v Aniza bte Essa
Material facts
5 Aniza had married the deceased (in September 2001) when she was 19 years old and he was 24. They had two sons, aged two and five as at August 2007. Aniza was 24 years old when she committed the abetment offence. She was diagnosed by two psychiatrists from the Institute of Mental Health (“IMH”) as suffering from an abnormality of the mind at the time of the offence, which substantially impaired her responsibility for her acts and qualified her for the defence of diminished responsibility to a charge of murder. The impairment of her mental faculties arose as a result of frequent quarrels with the deceased who had also physically and verbally abused her during a substantial part of their marriage. The quarrels and abuse arose because Aniza alone had to support the family as the deceased could not find a steady job. He was sent to prison on 15 June 2005 after being convicted of the offence of being absent without leave as a national serviceman and was released on 15 August 2006.
6 While the deceased was in prison, Aniza found work as a waitress in a pub in order to support the family. There she met Nasir, a patron of the pub, in November or December 2006. They struck up a friendship and she began confiding in Nasir about her marital problems. He advised her to divorce the deceased but she told him the deceased would not divorce her. Soon after, in February 2007, they became lovers. In June 2007, they agreed on a plan to kill the deceased. There is no evidence on record as to who suggested that the deceased be killed. However, the agreed statement of facts (“the SOF”) tendered by the Prosecution to the court for the purpose of sentencing shows that Aniza had obsessively urged Nasir to kill her husband. He made three attempts to kill him, and only succeeded at the third attempt, the first two attempts having been carried out in a rather half-hearted fashion. He finally killed the deceased by repeatedly stabbing him with a knife. The autopsy showed that nine stab wounds were inflicted on the deceased, the fatal wound being the stab to his chest.
Psychiatric assessments of Aniza
7 After her arrest, Aniza was examined by two psychiatrists from IMH. It is necessary to note at the outset that their reports on Aniza’s psychiatric condition were annexed to the SOF without any qualification by the Prosecution. The first report dated 6 August 2007 was by Dr Tommy Tan, then a senior consultant. He stated that Aniza had “moderate depressive episode … characterised by a depressed mood, disturbed sleep and suicidal thoughts”, and that she had “a depressed affect, depressed mood and psychomotor retardation”.[note: 1] He stated that Aniza was not of unsound mind at the time of the offence and would have known what she was doing and that what she had done was wrong, but because of her episodic depressive mood, she had “an abnormality of the mind, which would have substantially impaired her mental responsibility for her acts or omissions in causing the death or being a party to causing the death”[note: 2] of the deceased.
8 The second report dated 22 August 2007 was by Dr Kenneth Koh, a consultant forensic psychiatrist. He gave a detailed report of Aniza’s marital history, her account of spousal abuse, the development and worsening of her depression, her relationship with Nasir and her description of her role in the offence. Dr Koh stated that she “gave a convincing account of being the victim of protracted spousal abuse” and that she suffered from chronic depression as a result of it.[note: 3] He found that “[s]he suffered from the vegetative symptoms of depression, including disturbed sleep and appetite, impaired concentration and decreased libido. She had also made a serious suicide attempt in April 2007.”[note: 4] Concurrent with an increase in the frequency and severity of her beatings from about March 2007, her depression worsened to be of moderate severity until the date of the offence. Dr Koh stated that she showed some features of post-traumatic stress disorder, although these were insufficient to warrant a clinical diagnosis of the disorder. He also held the view that she demonstrated “learned helplessness” as a result of the repeated abuse by the deceased.[note: 5] He added, “There is a complex interplay between her chronic depression, and her constantly being abused.”[note: 6] This resulted in her failure to act positively to prevent Nasir from following through with the plan to kill the deceased. Like Dr Tan, Dr Koh concluded that, at the time of the offence, she was not of unsound mind but suffered from an abnormality of the mind that impaired her judgment and responsibility.
9 In submitting on sentence, the Deputy Public Prosecutor (“DPP”) referred the Judge to the Hodgson criteria (see [11] below) and argued that Aniza should be sentenced to life imprisonment for three reasons: (a) the gravity of the offence; (b) the circumstances in which Aniza had committed the offence; and (c) the need for a deterrent sentence to protect life and the sanctity of marriage, and because homicide involving spouses was becoming common (see the Judgment at [36]). The Hodgson criteria consist of three conditions which have to be satisfied before a court is justified in sentencing a mentally unstable offender to life imprisonment. The Judge rejected the DPP’s submission on the ground that the Prosecution’s case did not satisfy the second Hodgson criterion, and therefore he could not justify sentencing Aniza to life imprisonment.
10 In this appeal, the PP went beyond the Prosecution’s position before the Judge. His principal argument is that the Judge should not have applied the Hodgson criteria at all (to determine whether Aniza should be sentenced to life imprisonment) as they are not appropriate in our penal regime. Instead, the Judge should have sentenced Aniza to life imprisonment on the principle of deterrence, having regard to the way she committed the offence. Before we examine the PP’s arguments, it is necessary that we examine first what the Hodgson criteria are and why the courts in England, Australia, Hong Kong and Singapore have used them as a guide in determining the proper punishment for mentally unstable offenders who have committed serious offences.
The Hodgson criteria
The purpose and rationale of the Hodgson criteria
11 The Hodgson criteria were first enunciated in 1968 by the English Court of Appeal in R v Rowland Jack Forster Hodgson (1968) 52 Cr App R 113 (“Hodgson”). In that case, the appellant was 23 years of age. He was convicted of a series of violent offences. He had previously been convicted twice of various assaults on women, one for wounding with intent to cause grievous bodily harm and the other for assault occasioning actual bodily harm. The trial judge sentenced him to: (a) three terms of life imprisonment for two acts of rape and one of buggery committed against two women; (b) two years’ imprisonment for assault on one of the two women occasioning bodily harm; and (c) four years’ imprisonment for an assault on a third woman with intent to rob. The trial judge held that the public, in particular, women and girls, had to be protected against him. The appellant appealed against his sentence. The Court of Appeal dismissed the appeal. At 114 of Hodgson, the court (per MacKenna J) said:
When the following conditions are satisfied, a sentence of life imprisonment is … justified: (1) where the offence or offences are in themselves grave enough to require a very long sentence; (2) where it appears from the nature of the offences or from the defendant’s history that he is a person of unstable character likely to commit such offences in the future; and (3) where if the offences are committed the consequences to others may be specially injurious, as in the case of sexual offences or crimes of violence. [emphasis added]
It should be noted that in Hodgson the appellant was not certified as having a mental disorder, but he was treated by the court as having an unstable personality by reason of the nature and the number of offences he had committed.
12 As can be seen from the above quotation, the Hodgson criteria were enunciated as a guide to justify the sentence of life imprisonment for mentally unstable offenders who met those criteria. Under English law, life imprisonment meant imprisonment for the rest of the offender’s natural life, and in 1968, life imprisonment was the most severe judicial punishment in the English statute book. On established sentencing principles, life imprisonment would be imposed only for cases falling within the worst category of cases prescribed by that offence. However, in the case of mentally unstable offenders, English judges were not disposed to sentence such offenders to life imprisonment unless it was necessary for the protection of the public. England then had (and still has) in place a statutory scheme under which the Home Secretary had the power to release such prisoners on licence if he was satisfied that they had recovered from their mental condition and would not be likely to re-offend in a similar manner. For this reason, life imprisonment was an indeterminate sentence as it could not be determined in advance of the sentence when the prisoner would be released. In reality, because life imprisonment was an indeterminate sentence, it was being used as a form of preventive detention for mentally unstable offenders until it was safe to release them under licence back to society. It was against this penal background that the English courts applied the Hodgson criteria to determine when to and when not to sentence a mentally unstable offender to life imprisonment.
13 In an article published in 1963, Glanville Williams, “The Courts and Persistent Offenders” [1963] Crim LR 730, Prof Williams explained the rationale of this policy at 742 as follows:
If an offender is convicted of a serious offence, and appears from the evidence to be mentally unbalanced, the courts will readily (if they are able so to do) award a sentence of life imprisonment, rather than a term of imprisonment or preventive detention. The reason is that any term of imprisonment or of preventive detention must come to an end on a date previously known, and so does not give such good protection to the public as an indeterminate sentence, when the authorities [namely, the Home Secretary under s 27 of the Prison Act 1952 (c 52) (UK)] can satisfy themselves as to the mental state of the prisoner before releasing him. This means that if an element of mental abnormality is injected into the case the court casts aside the ordinary notion of just retribution, and the sentence proceeds entirely on the notion of social defence. [emphasis added]
We do not think that when Prof Williams said that “the sentence proceeds entirely on the notion of social defence”, he meant to imply that no degree of retributive punishment should be imposed on such offenders, as mental abnormality does not absolve the offender from all responsibility for his criminal wrongdoing. Depending on the state of his mental abnormality and the degree of culpability associated with it, the offender could or would still be punished on a retributive basis.
14 In Attorney-General’s Reference No 32 of 1996 (Steven Alan Whittaker) [1997] 1 Cr App R (S) 261, Lord Bingham CJ explained the relationship between the elements of the Hodgson criteria (at 264) as follows:
It appears to this Court that the conditions [in which a discretionary life sentence can be imposed] may be put under two heads. The first is that the offender should have been convicted of a very serious offence. If he (or she) has not, then there can be no question of imposing a life sentence. But the second condition is that there should be good grounds for believing that the offender may remain a serious danger to the public for a period which cannot be reliably estimated at the date of sentence. By “serious danger” the Court has in mind particularly serious offences of violence and serious offences of a sexual nature. The grounds which may found such a belief will often relate to the mental condition of the offender. So much is made plain by Wilkinson (1983) 5 Cr.App.R.(S.) 105, in particular in the passage at 108 where Lord Lane C.J. cites the judgment of Lawton L.J. in Pither (1979) 1 Cr.App.R.(S.) 209 and continues:
“It seems to us that the sentence of life imprisonment, other than for an offence where the sentence is obligatory, is really appropriate and must only be passed in the most exceptional circumstances. With a few exceptions, of which this case is not one, it is reserved, broadly speaking, as Lawton L.J. pointed out, for offenders who for one reason or another cannot be dealt with under the provisions of the Mental Health Act, yet who are in a mental state which makes them dangerous to the life or limb of members of the public. It is sometimes impossible to say when that danger will subside, and therefore an indeterminate sentence is required, so that the prisoner’s progress may be mentioned by those who have him under their supervision in prison, and so that he will be kept in custody only so long as public safety may be jeopardised by his being let loose at large.”
It is therefore plain that evidence of an offender’s mental state is often highly relevant, but the crucial question is whether on all the facts it appears that an offender is likely to represent a serious danger to the public for an indeterminate time.
[emphasis added]
In short, the underlying principle of punishment in relation to mentally unstable offenders is that, as long as they pose a serious danger to the public at large, they must not remain at large and must be kept in prison or a mental hospital until they are safe to be released back to society.
The Hodgson criteria in Australia
15 The Hodgson criteria were approved by the High Court of Australia in an appeal from New South Wales in Veen v The Queen (1979) 143 CLR 458 (“Veen (No 1)”). In that case, Mason J explained (in a minority judgment) the relevance of the Hodgson criteria (at 470) as follows:
Certainly this position has been reached in England against a background of different statutory provisions commencing with s. 57 of the Criminal Justice Act, 1948 (U.K.) to which reference was made in Reg. v. Cunningham, where life imprisonment was imposed for the purpose of preventive detention. The section conferred power on the Home Secretary to release on licence a person sentenced to life imprisonment. That provision has been replaced by s. 61 of the Criminal Justice Act, 1967. In addition, in England there is the power to make hospital orders under s. 60 and restriction orders under s. 65 of the Mental Health Act, 1959. By reason of the statutory power of the Home Secretary, life imprisonment, when imposed in England as a means of protecting the community from an offender who has a propensity to crimes of violence, has somewhat different consequences. No doubt the existence of this power induced the English courts more readily to adopt the approach which has been taken there. But, notwithstanding the absence in New South Wales of a corresponding statutory provision and of the system for psychiatric treatment and assessment which obtains in England, I am of opinion that life imprisonment should be imposed when it is necessary to protect the community from an offender who has a disposition to commit violent crimes and that the conditions for the imposition of that sentence are as stated in [Hodgson]. [emphasis added]
16 As can be seen from the above passage, Mason J did not consider that the usefulness of the Hodgson criteria in determining when the sentence of life imprisonment should be imposed on mentally unstable offenders depended on whether there was a statutory system in force that enabled the prison authority to calibrate the time for the offenders’ release, if ever. This was the case, since the protection of society was the dominant purpose of life imprisonment as a form of preventive detention. But, in a later decision in Veen v The Queen (No 2) (1988) 164 CLR 465 (“Veen (No 2)”) (see [29] below), the High Court of Australia held that the Hodgson criteria, even if satisfied, did not compel the sentence of life imprisonment, and that the central principle of sentencing was still that of proportionality. This approach was followed by the Court of Criminal Appeal of Queensland in R v Chivers [1993] 1 Qd R 432 (“Chivers”) (see [29] below), a case which has been approvingly cited in R v Hester
The application of the Hodgson criteria in Singapore
17 Singapore also does not have a statutory release scheme similar to that in England. In Neo Man Lee v PP
We were of the opinion that the conditions for sentence to imprisonment for life were clearly satisfied ... The appellant is clearly a continuing danger not only to himself but also to the public. The trial judge was of the view, which we shared, that he should be detained as long as it was permissible under the law. We might add that, with remissions, life imprisonment in Singapore may be reduced in practice to no more than 14 years, and the appellant may in fact be out of prison in another seven years.
18 This passage shows that the court was aware of the rationale of the Hodgson criteria and that the life sentence in Singapore was a determinate sentence. There is no evidence as to whether the court was aware that the Hodgson criteria originated in a jurisdiction in which the life sentence was an indeterminate sentence. The court found the Hodgson criteria appropriate as a guide to determine whether the appellant should be sentenced to life imprisonment, which was the highest sentence for the offence committed by him. The court found that the Hodgson criteria had been satisfied and accordingly sentenced him to life imprisonment. It is not clear whether the court was concerned that the life sentence was a determinate sentence of not more than 20 years (or about 14 years with remission). But if it was, the determinate duration of the sentence was a policy matter over which it had no control. In Neo Man Lee, the court could hardly have sentenced the appellant to the lesser sentence of ten years’ imprisonment as such a sentence (or six years eight months with remission) was too short for the protection of society since the appellant was clearly found to be a continuing danger to society.
19 For the next eight years, there were no reported decisions where the Hodgson criteria were invoked. In PP v Loh Ah Nu
That logic [as set out in Neo Man Lee (see quotation at [17] above)] is now obviously negated in light of our decision in Abdul Nasir, and all the cases in which the life sentence was imposed must be viewed in that context.
The reference to Abdul Nasir was to the decision of this court in Abdul Nasir bin Amer Hamsah v PP
20 Lai J (in Allan Tan at [36]–[40]) went on to explain the new logic, consequent upon the ruling in Abdul Nasir, as follows:
36 On the question whether a sentence of life imprisonment was appropriate, we were naturally impressed by the implications of our decision in Abdul Nasir. Certainly, even with [reg 119A, the then equivalent of reg 125 of the new Prisons Regulations], a sentence of life is now much harsher than it was before our ruling in Abdul Nasir. Whereas an accused person previously would serve a maximum sentence of 20 years, with a potential remission commuting his sentence to one of 13 years and 4 months, he must now serve a minimum of 20 years’ imprisonment, at which point his release would be within the discretion of a Life Imprisonment Review Board. So, the minimum period of incarceration is now six years and eight months longer, whilst the maximum period of incarceration, previously 20 years, is now the remainder of the prisoner’s natural life. In this context it is equally important to note that under the old position, his release after 20 years would have been guaranteed, but a prisoner sentenced for life in respect of a crime committed after Abdul Nasir has no such peace of mind.
37 In that respect, we are of the view that the courts must now exercise caution before committing a young offender to life imprisonment. Contrary to traditional reasoning, in similar cases involving a youthful offender on the one hand and an older offender in the other, the youthful offender sentenced to life imprisonment would now be subject to a longer period of incarceration than an older offender, assuming they both lived to the same age.
38 This case highlights one consequence of our decision in Abdul Nasir where youthful offenders are concerned. With the life sentence now being a sentence for the remainder of the convicted person’s natural life, the range of sentencing options are very limited. If the trial judge does not wish to impose a sentence of life imprisonment (which carries a minimum of 20 years, but which, as in the present case, may extend to over 50 years), he must impose a sentence of up to ten years’ imprisonment (which, with remission, would amount to a sentence of up to seven years’ imprisonment).
39 In serious cases the court must choose between the two options for a weighty sentence: ten years or life imprisonment. Under the old position, the effective choices would be up to seven years’ imprisonment (after remission) or about 13 years’ imprisonment for a ‘life sentence’ (after remission), a gap of about six years. Without remission, the gap would be ten years. Now, the gap is very much wider. Even assuming a positive outcome after review by the Life Imprisonment Review Board, the gap between the sentencing options is between 7 and 20 years, more than double the old position. Assuming a negative outcome by the Review Board, or that the sentence was not commuted, the gap widens. In the present case, the gap is 44 years (the difference between 10 years and 54 years). There is no discretion for the court to impose a sentence of more than ten years, but less than life imprisonment. This compares to the position in England, where, in respect of manslaughter (murder without intent), the court has a discretion to impose a sentence up to and including a sentence for life (see the English Offences Against the Person Act 1861, s 5, as amended by the Criminal Justice Act 1948).
40 In a situation in which the court is desirous of a sentence greater than ten years, but feels that a sentence of life imprisonment is excessive, we have no choice but to come down, however reluctantly, on the side of leniency. Otherwise, the punishment imposed would significantly exceed the offender’s culpability. It would, in our view, be wrong to adopt an approach in which the court would prefer an excessive sentence to an inadequate one.
[emphasis added]
21 What concerned the court in Allan Tan was that the new Prisons Regulations had widened even more the gap between the lowest and the highest imprisonment term for a s 304(a) offence. Before Abdul Nasir, the gap was only six years eight months (the difference between imprisonment of six years eight months (ten years with remission) and 13 years four months (20 years with remission)). After Abdul Nasir, the gap was widened considerably: the minimum being 13 years four months (the difference between six years eight months and 20 years), assuming a positive review by the Life Imprisonment Review Board, and the maximum being the difference between six years eight months and the rest of the natural life of the offender, assuming a negative review by the review board. In the present case, our calculation shows that the maximum gap could reach 50 years four months (ie, based on Aniza’s age of 25 and a statistical life expectancy of 82 years, life imprisonment for her may mean imprisonment for 57 years). In Allan Tan, the court calculated the maximum gap to be as long as 44 years. This sentencing structure prescribed by the law made it impossible for the court in Allan Tan to apply the proportionality principle, and thus compelled it to elect the lower sentence where it considered life imprisonment to be excessive, even though the lower sentence was inadequate, as there was no middle ground. This was also the dilemma faced by the Judge in sentencing Aniza; on the facts before him, he decided to impose the lower sentence in line with what this court said in Allan Tan at [40] (see [20] above).
22 Neo Man Lee was followed by the trial judge in PP v Ong Wee Teck
[B]ut if I do not provide any sound basis why a person whose mental capacity (for the commission of the offence charged) was diminished by a chronic and serious mental illness should be punished to the maximum limit provided under s 304(a) it is because there is none save the utilitarian one that I have adopted for this case.
PP v Dolah bin Omar shows that, even though life imprisonment was not then an indeterminate sentence, the judge sentenced the mentally unstable offender to life imprisonment because he needed psychiatric care and treatment for an indefinite period. The judge’s pragmatic (and utilitarian) approach in that case would fit in well with the current penal regime (see [23] and [32] below).
23 Since then, the Penal Code (Amendment) Act 2007 (Act 51 of 2007) (which came into effect on 1 February 2008) has increased the possible maximum fixed term of imprisonment from ten years to 20 years. This increase has reduced the wide gap between the post-Abdul Nasir regime and the new Prisons Regulations, and has made it possible for the court to apply the proportionality principle within the new sentencing range. However, this amendment does not provide an intermediate position for mentally unstable offenders serving life sentences to be released earlier, ie, before they have served 20 years of their sentence, even if they have fully recovered from their mental conditions. But for those offenders who are likely to re-offend in a similar manner if released or who need long-term psychiatric care, the new Prisons Regulations are structured such as to permit their indefinite detention.
24 In the next two cases, viz, PP v Ng Kwok Soon
[T]he accused here was nothing less than a cold-blooded would-be murderer. …
His acts were not the uncontrolled reaction of a person who had been suddenly assaulted verbally or physically. He was cunning and calculative.
In other words, the Hodgson criteria are not an exclusive judicial guide to justify the sentence of life imprisonment. These two cases show that the courts have no hesitation in sentencing offenders to life imprisonment in appropriate cases based on the principles of retribution and deterrence.
25 In the next group of seven cases on s 304(a) offences, viz, PP v Lim Hock Hin
26 During this period there was another group of four cases on s 304(a) offences, viz, PP v Rohana
27 These cases show that our courts were fully aware of the limitations of applying the Hodgson criteria in imposing the highest sentence (life imprisonment) for a s 304(a) offence, whether or not the offender was mentally normal or mentally unstable. If the sentence of life imprisonment was not capable of protecting society from dangerous or unstable offenders being released after a minimum of 13 years four months in prison, this was not caused by any flaw in the Hodgson criteria but in the accepted meaning of life imprisonment during the relevant period. On the other hand, the Hodgson criteria had the merit of allowing the court to sentence a mentally unstable offender to a term of imprisonment that it considered appropriate to his or her culpability in the commission of the offence. In other words, the court was able to resort to the principle of retribution (just deserts) or the principle of proportionality in order to opt for the lower of the two prescribed sentences, ie, ten years’ imprisonment.
Proportionate sentences and the Hodgson criteria
28 There is an inherent conflict between life imprisonment as a form of preventive detention for the protection of the public and the principle of proportionality in sentencing offenders according to their culpability which cannot be resolved logically. In D A Thomas, “Sentencing the Mentally Disturbed Offender” [1965] Crim LR 685, the author examined the tension between the two sentencing objectives (at 694):
The attraction of life imprisonment as opposed to long fixed-term sentences is that life imprisonment is in effect an indeterminate sentence. Under section 27 of the Prison Act, 1952, the Home Secretary may release on licence a person sentenced to life imprisonment at any time, and subsequently recall him if this becomes necessary. … The Court deliberately uses the sentence of life imprisonment as an indeterminate sentence, and considers it more favourable to an appellant than a long fixed-term sentence. However, the court is not entirely happy to dispense completely with the concept of proportion in this context, and has said that the sentence should be one which a mentally-normal person might properly be given for a similar offence. The equation is necessarily a rough one, however, as sentences of life imprisonment are rarely awarded to mentally-normal persons (except on conviction for non-capital murder); the object of an extremely long-term sentence in the case of a mentally-normal person is usually deterrence, and for this object a fixed-term sentence is considered more appropriate; there is, therefore, no real standard of comparison. Also, the responsibility of a mentally-disordered offender can hardly be equated with that of a mentally-normal offender; if the only issue is to allocate punishment according to culpability, the mentally-disordered offender would presumably be treated more leniently. In practice, while the court will generally uphold sentences of life imprisonment for offences in the most serious category – manslaughter by reason of diminished responsibility, rape, serious woundings, and serious robbery are the most usual – there have been other cases of a less serious nature where a sentence of life imprisonment was legally available, but the Court has not chosen to make use of its power to impose one, although recognising the need for a long-term preventive sentence. Presumably this reluctance is in part due to a desire to avoid the appearance of a harsh sentence, although as the court has itself pointed out, an indefinite life sentence may well be more favourable to the accused than a long fixed term. A more logical policy would, it is submitted, result from the exclusion altogether from this class of case of the concept of proportion and the use of life imprisonment as an indeterminate sentence without regard to the seriousness of the most recent offence.
The Australian practice
29 The Australian courts have not accepted the “more logical policy” advocated by D A Thomas (at [28] above). In R v Murdock [1980] Qd R 504 (a rape case), the Queensland High Court adopted the Hodgson criteria without qualification, but in Chivers ([16] supra), the Queensland Court of Criminal Appeal (by a majority) held that R v Murdock had to be read subject to the decisions of the High Court of Australia in Veen (No 1) ([15] supra) and Veen (No 2) ([16] supra), which had modified the application of the Hodgson criteria to take into account the principle of proportionality. In Chivers, the court sentenced the offender to life imprisonment for attempted murder after he shot at a social worker who was bringing him his medication. At the time of the murder attempt, the offender was 53 years old. He had for some time shown signs of mental instability and had only been recently discharged from a psychiatric unit.
30 Thomas J (in his majority judgment in Chivers) held at 436 that, as a consequence of Veen (No 1) and Veen (No 2), the following propositions were justified:
1. The punishment to be inflicted must be proportionate to the crime. ([Veen (No 1)] at 468, 482, 483, 495; [Veen (No 2)] at 472). This process is popularly described as making the punishment fit the crime.
2. The protection of the community from violent crime is a very important factor to be taken into account in sentence. ([Veen (No 1)] at 467; [Veen (No 2)] at 473-475). For example there are cases in which the mental condition of the convicted person would render him dangerous if at large and in some such cases sentences of life imprisonment may have to be imposed to ensure that society is protected. This is a direct quotation from the statement of Gibbs J. in R. v. Pedder (C.A. 16/1964; Court of Criminal Appeal, 29 May 1964, unreported) expressly approved by the majority in [Veen (No 2)]. In other words there will be cases where this factor (protection of the community from danger) will be the express factor that will make a finite sentence unsuitable and a life sentence appropriate.
“When a man has deserved punishment we shall very properly look to our protection in devising it.”
(C. S. Lewis — On Punishment: A Reply, quoted with approval by the majority in [Veen (No 2)] at 473-474.)
The point was encapsulated by the majority in its statement that “the courts of this country ... (have) regard to the protection of society as a factor in determining a proportionate sentence” ([Veen (No 2)] at 474). But at what level of seriousness of offence, or of propensity to re-offend, may this factor convert a finite sentence into a life sentence? This question is answered, at least in part, in paras 3 and 4 below.
3. The antecedent criminal history of an offender is a factor which may be taken into account in determining the sentence to be imposed, subject of course to its not being disproportionate to the gravity of the particular offence ([Veen (No 2)] at 477). Such a history may show whether the offence is an uncharacteristic aberration or that the present offence is merely a continuing attitude of disobedience of the law. In the latter case a more severe penalty may be warranted. “It is legitimate to take account of the antecedent criminal history when it illuminates the moral culpability of the offender in the instant case, or shows his dangerous propensity or shows a need to impose condign punishment to deter the offender and other offenders from committing further offences of a like kind.” (Ibid. at 477).
4. The maximum prescribed penalty for an offence is intended for cases falling within the worst category of cases for which that penalty is prescribed ([Veen (No 2)] at 478). This does not mean that one must envisage the worst possible example of the crime in order to impose the maximum penalty, but rather that it must be seen as qualifying for the worst category of cases of that kind (ibid at 478). It may be difficult if not impossible to envisage “categories” of cases of attempted murder, and I would interpret this statement as requiring that the circumstances reveal conduct towards the top end of the scale when one is looking at the range from the least serious to the most serious examples of such offences.
At 440, Thomas J said:
The psychiatric evidence now suggests that there is a grave danger of repetition. There may be a temptation to impose a life sentence in those circumstances for the laudable object of protecting the community. But that is not the function of a criminal court. To do so would be to punish the offender, not for the offence of which he was convicted, but for the potential offences which he might commit in the future. It would be to order that he be held in prison, not simply for the punishment of the offence which he has committed against the law, but for another purpose entirely. To do so would infringe the principles of [Veen (No 2)], because it would be a naked example of preventive detention.
31 Cooper J gave a concurring judgment and the court set aside the sentence of life imprisonment and imposed a sentence of 16 years’ imprisonment, which, according to the judge, was the highest punishment the court was entitled to impose (see Chivers at 440).
The practice in Singapore
32 In Singapore, the huge gap between the maximum and the minimum custodial sentences created by the new Prisons Regulations after the decision in Abdul Nasir ([19] supra) had made it practically impossible for the courts to apply the proportionality principle in sentencing mentally unstable offenders whose illnesses may be transient or treatable. As this court pointed out in Allan Tan ([19] supra), the prescribed custodial sentence was either up to ten years (or six years eight months with remission) or a minimum of 20 years, with a potentially indefinite term for the rest of the natural life of the offender. This gap was reduced considerably when the lower imprisonment term for a s 304(a) offence was increased from ten years to 20 years in 2007 by the Penal Code (Amendment) Act 2007 (see [21] and [23] above). This was the problem confronting the Judge in the present case. He considered the sentence of life imprisonment for Aniza to be excessive and unwarranted. He had no choice but to opt for the next lower custodial sentence, which was up to ten years’ imprisonment, in accordance with the approach enunciated by this court in Allan Tan (see [20] above).
Are the Hodgson criteria appropriate in our sentencing regime?
33 We have earlier mentioned that the English sentencing regime provides a statutory scheme for the release of offenders sentenced to life imprisonment under the Hodgson criteria, and that such a scheme is absent in Singapore. Our courts have nevertheless expressly applied the criteria in many cases since 1991, although we have not been able to find in those cases any express reference to the statutory scheme underlying the criteria in England or to life imprisonment being an indeterminate sentence there. We now have to consider whether the Hodgson criteria are appropriate in Singapore in the light of these two features of the English penal system which are absent here. In this appeal, the PP has argued that the Hodgson criteria are not appropriate in Singapore and should be buried, but his argument is based on entirely different reasons which, as will be seen, are unpersuasive (see [43]–[46] below).
34 We have mentioned that the Australian courts have endorsed the Hodgson criteria even though there was then no equivalent statutory scheme to the English one there (at least in New South Wales) (see [15]–[16] above). However, they have subsequently subjected the application of the criteria to the proportionality principle (see [29]–[30] above). In our view, the answer to whether the Hodgson criteria are appropriate in Singapore is, similarly, in the affirmative. The Hodgson criteria fulfil two important functions. The first is that they provide an alternative (ie, alternative to the principle that the highest punishment should only be reserved for the worst types of cases), and equally principled, justification to impose life imprisonment when it is the highest punishment prescribed for an offence on an offender (ie, for the protection of society). The second is that the Hodgson criteria also provide a useful guide to differentiate between when it is appropriate to sentence to life imprisonment dangerous mentally unstable offenders who are a long-term threat to society and when lesser sentences may be meted out to those who suffer from a transient illness who can be rehabilitated and reintegrated into society. They reflect a humane approach to the punishment of mentally unstable offenders, without sacrificing the dominant objective of protecting the community from the likelihood of similar re-offending. Mental illnesses come in many forms and affect cognition and judgment in different degrees – some illnesses are genetic in origin, and some are situation-specific; some need long-term treatment or are not susceptible to treatment, whereas others are treatable. We need to consider, from a penal point of view, what public policy considerations justify keeping mentally unstable offenders in prison for life if it is not necessary to do so in order to protect the public. The only other justification is the need to punish them on the basis of retribution and/or deterrence. But, retribution involves the consideration of the proportionality principle which conflicts with the principle of public protection, and, as the Judge has correctly recognised, deterrence is not apposite for mentally unstable offenders (see [73] of the Judgment and [36] below). In the present case, the Judge’s choice of punishment for Aniza was between a maximum custodial sentence of ten years and a minimum of 20 years (and potentially for life).
The Judge’s sentence
35 In the present case, the Judge rejected as fanciful the Prosecution’s argument that there was a risk of Aniza re-offending in a similar manner if she were sentenced to a term of imprisonment less than life. On the basis of the reports of the two psychiatrists, he was satisfied that Aniza’s psychiatric illness was transient, that its dominant cause was prolonged spousal abuse by the deceased towards her, that it was a treatable condition, and that with the removal of this cause, Aniza would be able to recover and lead a normal life without any tendency to criminality within a short time, certainly less than the minimum of 20 years’ imprisonment. She would not be a danger to society as there should be no risk of her re-offending in a similar manner. At [39]–[41] of the Judgment, he explained:
39 … For a relatively young offender of 24 years of age, a sentence of life imprisonment, which is of indeterminate duration, is therefore a very severe and crushing sentence. For this reason, care must be taken to scrutinise and evaluate the relevant facts and circumstances to determine whether a life sentence is indeed justified. As a sentence of life is now much harsher than it was before Abdul Nasir, the Court of Appeal in [Allan Tan ([19] supra)] opined at [37] that:
[T]he courts must now exercise caution before committing a young offender to life imprisonment. Contrary to traditional reasoning, in similar cases involving a youthful offender on the one hand and an older offender in the other, the youthful offender sentenced to life imprisonment would now be subject to a longer period of incarceration than an older offender, assuming they both lived to the same age.
40 The Court of Appeal in [Purwanti ([25] supra)] further cautioned at [25] that:
[E]ven if the Hodgson conditions were satisfied, the court must exercise caution before committing a young offender to life imprisonment, especially since life imprisonment now means imprisonment for the rest of the prisoner’s natural life.
41 Hence, the maximum sentence of life imprisonment should not be imposed unless it is clearly necessary and appropriate to do so. It must be one where there is sufficiently persuasive evidence to support those facts that allegedly make it out to be a necessary and appropriate case for a life sentence. The burden is on the Prosecution to establish that ... life imprisonment is an appropriate and necessary sentence and I am not persuaded in this case that it is.
36 The Judge ruled out deterrence as a relevant objective in punishing an offender whose cognitive understanding had been substantially impaired at the time of the offence (see his references at [73] of the Judgment to Yong Pung How CJ in Ng So Kuen Connie v PP
37 The Judge accepted (at [51] of the Judgment) the Prosecution’s submission that Aniza had “deviously and psychologically manipulated a young gullible 16-year-old boy into killing the deceased”, that she had pressured him repeatedly into doing the act, and that “there was premeditation and planning by both of them showing that the killing was cold-blooded”. However, on the basis of the reports of the psychiatrists, the Judge attributed her conspiratorial and manipulative acts to an abnormality of mind which had substantially impaired her judgment at that time. Her illness was caused by the deceased’s protracted physical and emotional abuse which lasted six years. The abuse included frequent beatings (two or three times a week, according to Dr Koh’s report) and incessant demands for money. The Judge was satisfied from the psychiatric reports that the deceased was the very reason which drove Aniza “to conspire with Nasir to kill him to end her suffering” (at [52] of the Judgment). Whilst emphasising that each case on sentence must be decided on its own set of facts and circumstances, the Judge referred to comparable cases (where the offenders were mentally unstable) as being supportive of his decision. Reference was made to Purwanti ([25] supra) (life imprisonment because of risk of re-offending), Constance Chee ([26] supra) (ten years’ imprisonment; little risk of re-offending), PP v Juminem
38 The Judge concluded his examination of the relevant factors in sentencing Aniza (at [74] of the Judgment) as follows:
Having careful regard to all the facts and circumstances of the case, the aggravating as well as the mitigating factors, and the guidelines laid down by the Court of Appeal appropriate to cases under s 304(a), I decided that a sentence of nine years’ imprisonment (backdated to the date of her remand) is a fair and appropriate sentence, which I believe is adequate on the whole to address the need for retributive punishment, deterrence, and to safeguard the public interest (including that of public safety) whilst, at the same time, balancing them against those other mitigating factors in Aniza’s favour and her need for treatment and rehabilitation.
39 For those reasons, the Judge decided that sentencing Aniza to life imprisonment was not justified. Given the severe limitations of the sentencing options open to the Judge, we find it difficult to fault him for imposing what the PP regards as a lenient sentence, and we do not do so. The Judge made a principled decision on the sentence on the agreed facts placed before him.
40 It should be noted that the Judge’s assessment, that Aniza’s psychiatric condition is treatable and that she has no predisposition to re-offend in a similar manner, is based on the reports of the two psychiatrists’ diagnoses. The Hodgson criteria require the Judge to make such an assessment, even though there is no certainty as to whether it will turn out to be correct. As Dr Tan himself has acknowledged in another case (quoting another expert psychiatrist), “Nothing is certain in psychiatry” (see PP v Barokah
41 In Constance Chee, V K Rajah J was faced with a similar difficulty. In that case, the offender was involved in the kidnapping of a young girl and had caused the girl to fall to her death from a block of flats. The offender was diagnosed as suffering from schizophrenia which, if not treated, might create a risk of danger to herself (with a 70% chance of suicide) as well as to others. The psychiatrist recommended that she be kept in conditions of security where she might continually receive psychiatric treatment, and be reviewed on a regular basis. In spite of this report, Rajah J decided to impose a sentence of ten years’ imprisonment as neither retribution nor deterrence was a proper consideration in a case where the offence was committed as a result of a serious mental disorder. He was particularly persuaded by the fact that the offender’s family had given concrete assurances that she would be provided with a secure environment for treatment and rehabilitation upon her release (contrast his later decision in PP v Lim Ah Liang ([25] supra) at [36] and [39]). However, Rajah J was able to increase the imprisonment sentence by another three years by ordering that the sentence for the kidnapping offence run consecutively from the date of remand. He concluded his judgment with a plea at [29] as follows:
The current position, where the courts are neither empowered nor endowed with any discretion whatsoever to customise or tailor their sentences in a manner that would be consistent with either the possible recovery or decline of the medical condition of an offender who is unwell, is far from satisfactory. Judges often have to choose between a rock and a hard place when resolving their colliding instincts in determining the appropriate sentence. Should the offender’s medical condition stabilise without any real risk of a relapse it would be quite unjust for him or her to continue to be incarcerated after rehabilitation through medical attention when he or she no longer poses any further risk to the public upon a return to the community. It is apodeictic that in such an instance the underlying rationale for the second of the Hodgson criteria … no longer prevails. In order to properly and fairly sentence offenders whose medical condition might potentially be reversed through medical attention and/or with the passage of time, the courts should be conferred the discretion to impose a sentence band with appropriate minimum and maximum sentences tied to periodical medical assessments and reviews. This will minimise the rather unscientific and imprecise conjecture that is now inevitably prevalent when determining appropriate sentences for such offenders. The proposed approach, while fairer to offenders, will also concomitantly serve to address and assuage public interest concerns on adequate sentencing as well as protection from mentally ill offenders with a propensity for violence. It is my hope that Parliament will review the present position and, upon taking into account the views of all relevant stakeholders in the sentencing and rehabilitation framework, endow the courts with more comprehensive and pragmatic sentencing powers. Effectiveness need not be divorced from fairness and reality. It is a fundamental tenet of criminal jurisprudence that whenever liberty is subtracted, justice must be added. Sentencing in cases such as this requires a rapier-like rather than a blunderbuss approach. [emphasis added]
42 Parliament partially answered the plea in 2007 by amending the punishment for all offences, where the highest punishment was the life sentence, to provide for a term of imprisonment for up to 20 years and retaining life imprisonment as the highest punishment for those offences (see [23] above). This amendment has enlarged the power of the court to punish such offenders from one day to 20 years, and has thus narrowed the gap considerably between the previously existing prescribed custodial sentences. But, the sentence of life imprisonment is still determined by reg 125 of the new Prisons Regulations (which allows the release, with or without conditions, of the prisoner only after he or she has served 20 years of imprisonment). Accordingly, the life sentence remains a determinate sentence for the first 20 years, but reverts to being an indeterminate sentence thereafter, as whether the remaining term is remitted or not is determined by the Life Imprisonment Review Board.
The PP’s appeal
Whether the Hodgson criteria are appropriate
43 We will now consider the PP’s arguments that the Hodgson criteria are inappropriate in our sentencing regime, which arguments are of a somewhat academic and technical nature. As set out at paras 36–38 of his written submissions, they are as follows:
36. However, in neither of these two Court of Appeal cases [Neo Man Lee and Purwanti] was the background to the Hodgson criteria ever considered. [Hodgson] was decided on 26 September 1967. The last execution in the United Kingdom took place in 1964. The death penalty was de facto abolished in the UK in 1965 by the Murder (Abolition of the Death Penalty) Act 1965.
37. Thus, when MacKenna J pronounced his celebrated dicta in 1967, life imprisonment was the highest punishment that an English court could impose. MacKenna J’s dicta may have been apposite when speaking of the ultimate punishment; they are not apposite when the punishment in question is not the most serious that a court may impose, which is the case with a sentence of life imprisonment in Singapore.
38. To make this clear, consider that there are several sections in the Penal Code where the Court is given the discretion to impose either the death penalty or a sentence of life imprisonment: eg, section 396 (gang robbery with murder), section 364A (kidnapping in order to compel the Government to do or abstain from doing a certain act). If a person is charged with an offence under section 396 or 364A and the criteria for imposition of a sentence of life imprisonment are those promulgated in [Hodgson] (a fortiori if the criteria are treated as cumulative), in what circumstances could a death sentence be imposed? Can it be plausibly said that Parliament intended that the threshold for life imprisonment would be set so high that it is impossible to impose capital punishment?
44 In our view, the first point (that the Hodgson criteria are not appropriate in Singapore because they were enunciated for a penal regime where the highest punishment under English law was life imprisonment, and not the death penalty, as in Singapore) is difficult to follow. As we have explained earlier, the Hodgson criteria were intended as a guide to justify sentencing mentally unstable offenders to life imprisonment for the protection of society. The Hodgson criteria were not intended to address, and had nothing to do with, offences punishable with the death penalty. Therefore, the circumstance that the highest sentence in our penal regime is the death penalty does not affect in any way the appropriateness or otherwise of the criteria with respect to offences punishable with life imprisonment. Their appropriateness or otherwise must be considered on the basis of whether they are effective to achieve their objective and whether the objective is appropriate to the values of our penal regime.
45 In any event, the PP’s first point, although historically correct, is unsound in the context of English sentencing practice. Similar principles were already being applied prior to the abolition of the death penalty, and these found subsequent expression as the Hodgson criteria (see [12]–[14] above). The court in Hodgson was merely restating for the benefit of trial judges the matters they should consider before sentencing mentally unstable offenders to life imprisonment.
46 The PP’s second point (in para 38 of his written submissions), that the Hodgson criteria are unable to assist the court in determining when or whether the death sentence should be imposed with respect to offences under s 364A(i) or s 396 of the Penal Code (which offences are punishable with life imprisonment or death), is also, with respect, off the mark for the same reason, ie, the Hodgson criteria were not intended as a guide to when to sentence an offender to death. The Hodgson criteria cannot be criticised for failing to guide the courts with respect to offences under s 364A(i) or s 396 of the Penal Code. The appropriate sentence in such a case would have to be determined by reference to the offender’s culpability, applying the proportionality principle or some other appropriate principle, such as deterrence. But even here, the PP’s argument may be misdirected for the reason that the rationale of the Hodgson criteria takes into account the culpability of a mentally unstable offender. In sentencing a mentally unstable offender to a s 364A(i) or s 396 offence, his or her culpability would be relevant as a sentencing factor. In such a case, such an offender is unlikely to have the degree of culpability that would justify him or her being sentenced to death.
Whether the Hodgson criteria are cumulative
47 The PP’s third point, that the Hodgson criteria are not cumulative (and that the Judge was wrong in law in so holding) is made in relation to a s 304(a) offence and also a s 364A(i) or s 396 offence. The first reason is that, if the criteria are cumulative, the threshold for the life sentence would be set so high that there might be no circumstances in which the courts could impose the death penalty. We have already dealt with this point at [46] above.
48 The second reason is that the Judge contradicted himself when he expressed the view that, if the offender killed his victim in a cruel and sadistic manner, that by itself would justify the sentence of life imprisonment. The PP’s argument is that, if the Hodgson criteria are cumulative, then it would not be appropriate in such a case to impose a life sentence (for the purpose of denunciation or deterrence) if there was no likelihood of repetition. Once it is accepted that a life sentence might be justified if the crime was heinous enough, it must also be accepted that it is not necessary to satisfy all the three Hodgson criteria. Hence, they are not cumulative.
49 The PP’s argument is based on the Judge’s analysis of the range of custodial sentences prescribed for a s 304(a) offence (from one day’s imprisonment to life imprisonment). Given this range, he reasoned that there must be a special circumstance that justifies the court to impose the most severe sentence. He held that this special circumstance was the second Hodgson criterion, viz, the likelihood of re-offending with “specially injurious” consequences to the victims (at [46] of the Judgment). In such a case, life imprisonment would be justified as a preventive measure to protect the public. He went further to identify another special circumstance, viz, where “the offence is so cruel and inhumane that the defendant does not deserve any leniency whatsoever and that the only just sentence is the maximum of life imprisonment and any other sentence is simply too lenient” (at [47]). Such a sentence would be justified on retributive as well as deterrent grounds. However, he held that Aniza’s role in the killing of her husband was not cruel or sadistic.
50 In our view, the PP’s second reason is also off the mark. We do not know if the Judge would have sentenced Aniza to life imprisonment if he had found her role in the killing of her husband to be cruel or sadistic. There would be no contradiction if he had indeed done so (but again, we do not know). But, for the sake of argument, even if the Judge contradicted himself, it does not render the Hodgson criteria any less cumulative in their application in order to justify the sentence of life imprisonment. What the second reason proves (assuming it is valid) is that the Hodgson criteria are not an exclusive guide for sentencing an unstable offender to life imprisonment, and not that they are not cumulative.
51 The PP’s third reason is that the Hong Kong courts have held that the Hodgson criteria need not be satisfied cumulatively before an offender can be sentenced to life imprisonment. The PP referred to R v Cheung Hing-biu
52 In our view, Cheng Wui Yiu is only an illustration of the two points we have made earlier, that the Hodgson criteria are not applicable to mentally normal offenders and that the criteria are not an exclusive justification for the imposition of the life sentence (see [24] and [50] above). Accordingly, the Hong Kong position does not assist the PP’s argument that the Hodgson criteria are not cumulative.
53 We have not been able to agree with the PP’s arguments that the Hodgson criteria are not appropriate to our penal regime. But having considered these arguments, it seems to us that the PP’s real objection to the Hodgson criteria is not because they preclude or prevent mentally unstable offenders from being sentenced to life imprisonment, the sentence which the PP is seeking in this appeal. It is not possible for such an argument to be made because, in intent, the Hodgson criteria are meant to justify the imposition of a life sentence when the criteria are satisfied. Furthermore, our courts have in fact sentenced many mentally unstable offenders to life imprisonment, using the Hodgson criteria as a guide (see [17]–[22] and [25] above). Thus, it seems to us that the substance of the PP’s argument is that the Hodgson criteria were not applicable on the facts of the present case and should not have been applied to Aniza because her mental faculty was not impaired or not impaired to such an extent that she was entitled to the defence of diminished responsibility (which the psychiatrists had opined to be the case). This is borne out by his submission that the psychiatrists should not have accepted at face value the self-serving statements given by Aniza to the psychiatrists, on the basis of which, inter alia, they had made their assessment of her condition. This issue is considered below.
Mitigation pleas
Unsubstantiated allegations and facts
54 We now turn to some other matters raised in this appeal by the PP in connection with the mitigation process. In particular, he referred to the burden of proof with regard to the statement of facts in a mitigation plea made by the defendant or defence counsel. In this connection, he argued that the Judge, in assessing the proper sentence for Aniza, had erroneously accepted as facts many unsubstantiated and self-serving statements made by Aniza to the two psychiatrists in her mitigation plea. He referred to the following instances where the Judge went beyond the SOF and instead relied on the psychiatric reports for the following statements:
(a) As a result of the deceased’s frequent assaults and beatings which increased in frequency and severity, Aniza began to suffer from depression of moderate severity from March 2007 to the date of the offence (see the Judgment at [51]).
(b) Aniza harboured suicidal thoughts (ibid).
(c) Aniza exhibited features of post-traumatic stress disorder (ibid).
(d) Aniza’s mental illness was caused by her husband’s protracted physical and emotional abuse and his incessant demands for money from her during their marriage (ibid).
(e) Aniza was subjected to prolonged spousal abuse by the deceased since the tender age of 19 and she had endured it without retaliation for five years (ibid).
(f) The deceased was the source of Aniza’s trepidation and mental depression. He had assaulted and traumatised her emotionally and was the very reason which drove Aniza “to conspire with Nasir to kill him to end her suffering” (at [52] of the Judgment).
(g) Aniza was trapped in a “largely abusive and loveless marriage, filled with seemingly unending spousal violence” which impaired her state of mind (at [75] of the Judgment).
55 The PP submitted that these were merely unsubstantiated self-serving statements made by Aniza to the psychiatrists, that the psychiatrists reached their conclusions on the basis of such assertions, and that the Judge should not have accepted these assertions as facts without proof. He referred to PP v Lim Ah Seng
56 In our view, the short answer to the PP’s argument is that the Prosecution had accepted the reports of the two psychiatrists for the purpose of mitigation. The reports of the psychiatrists were part of the SOF which the Prosecution had tendered to the court (without any qualification as to their contents) for the purpose of sentencing. The two psychiatrists had formed their professional opinions on the basis of the truth of Aniza’s statements. Since the Prosecution had accepted the opinions of the psychiatrists, they would also have accepted, by implication, the basis on which the psychiatrists had formed their opinions. The Judge was fully justified to proceed with the sentencing on the basis of the Prosecution’s case as presented to him. In our view, it is now too late for the PP to move away from this position and contend that the psychiatrists should not have accepted Aniza’s statements without proof. Accordingly, we are unable to accept the PP’s submission on this point.
Burden of proof in mitigation pleas
57 The PP also submitted that, as a matter of law, where an accused has pleaded guilty, the burden of proving any extraneous fact alleged by way of mitigation is on the accused and he must prove them on a balance of probabilities. The PP referred to the English decisions of R v Michael Joseph Kerrigan (1993) 14 Cr App R (S) 179 and R v Darius Nicholas Lechmere Guppy and Benedict Justin Marsh (1995) 16 Cr App R (S) 25 for this proposition. He also referred to the seemingly uncertain and unsettled practice of our courts in this regard and sought guidance from us on the proper procedure in a mitigation hearing. He contended, on the analogy of R v Kevin John Underwood [2005] 1 Cr App R (S) 90, that, in future cases, the Prosecution should be given the Defence’s mitigation speech in advance so that it would have an opportunity to decide which statements to accept and which to call for proof. In this regard, the PP referred to certain paragraphs in the mitigation statement of the Defence which had put the blame entirely on Nasir as the initiator and the executor of the scheme to kill the deceased in contradiction to the SOF.
58 In his reply, counsel for Aniza made two points. First, he submitted that it was inappropriate for the Prosecution to challenge the mitigation plea on appeal when it had not done so before the Judge below. Second, he disclosed that the Prosecution had in fact perused the text of his mitigation speech and had agreed to it after deleting certain objectionable portions in it. The significant omissions from the text included the paragraph which stated that Nasir had planned and executed the murder, while Aniza’s role was only to provide information on his request, and the paragraphs which attributed the idea of killing the deceased to Nasir.[note: 7] Counsel argued that the Judge could not be faulted for relying on statements that both parties had agreed to.
59 The reply of defence counsel, which was not contradicted by the PP, showed that there was no live issue on the burden of proof in mitigation or any perceived difficulties with the mitigation process. There was no lack of clarity in that regard. The sentencing process included an agreed statement of facts and an agreed mitigation speech. The Judge, in sentencing Aniza, considered the SOF, the mitigation speech and also the oral submissions of the Prosecution. This has been the established procedure for mitigation pleas in our courts for many years. It is a simple and workable system. It saves a lot of time and resources for all the parties involved in the process, including the trial judge.
60 The High Court has not found it necessary, in the past, to spell out the precise procedure with respect to the sentencing process, including how mitigation pleas should be conducted. The reason is that the procedure is well established and is known to all defence counsel. The courts are, of course, aware that issues of unfairness to the Prosecution may arise if the Defence is given too much leeway to assert all kinds of unsubstantiated mitigating factors in order to influence the sentence of the court. In PP v Chan Yoke Ling Catherine
a. Firstly, the sentencing process and a trial are materially different in terms of their objectives. The reasons for requiring strict proof by admissible evidence of all relevant facts – eg the presumption of innocence – do not apply during sentencing.
b. Secondly, the usual limitations on evidentiary sources and standard of proof could potentially limit the information available to the Judge, information which is necessary for ensuring that a sentence will adequately and effectively protect, deter and rehabilitate: PP v Tan Fook Sum
c. Finally, a heightened burden of proof may also add to the time and resources spent in the sentencing process, and risk turning it into a second trial. Such a spectre is clearly undesirable, as it would result in an inefficient criminal justice process.
We endorse these views. But we wish to point out that the sentencing courts are not unaware that defendants and/or defence counsel are prone to assert unsubstantiated statements by way of mitigation, and have thus been careful to distinguish the wheat from the chaff in giving them the requisite weight (Chan Yoke Ling at [40]).
61 The existing practice of the Prosecution in tendering to the court an agreed statement of facts and of defence counsel in providing the Prosecution with a copy of the mitigation speech before the sentencing hearing has made the mitigation process simple. This practice minimises any dispute between the Prosecution and the Defence on what mitigation statements are to be admitted without proof. This practice has rendered a Newton hearing (see R v Robert John Newton (1982) 4 Cr App R (S) 388) a rarity in our courts. Where the Prosecution objects to any unsubstantiated assertions in the mitigation speech, the Defence will either have to withdraw the statements, provide proof acceptable to the Prosecution or call evidence. This may be regarded as an analogous form of the Newton hearing, which is also rare in our sentencing practice. If the Prosecution does not object to the assertions made by the Defence, the court is entitled to accept them and give such weight to them as it thinks fit.
62 We do not think that it is desirable that we lay down too many rules to micro-regulate this area of criminal practice as they may create unnecessary satellite litigation on whether there has been due compliance with the rules. The important point is that the Prosecution should be given the opportunity to object to self-serving and unsubstantiated allegations that may influence the judge to impose a lower sentence than the offender deserves. Beyond that, the role of the court is to ensure that the sentencing process is fair to both the Prosecution and the Defence, and some degree of flexibility is called for. As observed in Christopher Emmins, Emmins on Sentencing (Oxford University Press, 4th Ed, 2001) (Martin Wasik ed) at p 74:
The procedure between conviction and sentence is markedly different from that which pertains to the trial itself. The role of the judge or bench of magistrates changes from that of an umpire to one of a collector of information about the offence and the offender. Rules relating to the admissibility of evidence are somewhat relaxed, and the combative or adversarial style of the opposing lawyers is less marked. The judge takes a more central and active role in the gathering of information, which comes from a variety of sources, in reaching the sentencing decision. In fact there are relatively few legal rules governing the procedure between conviction and sentence …
Was the sentence of nine years’ imprisonment inadequate?
63 The issue of the adequacy of the sentence of nine years’ imprisonment for Aniza lies at the heart of the PP’s case. He has submitted that the sentence of nine years’ imprisonment for an offender as cunning and manipulative as Aniza is manifestly inadequate. The PP compared the present case with s 304(a) offences committed by domestic workers (such as Purwanti ([25] supra) and PP v Juminem ([37] supra)) and contended that the punishment must fit the crime and that like cases should be treated alike, citing this court’s decision in PP v Kwong Kok Hing
64 We would not disagree with the PP if all the court is concerned with are the cunning and manipulative ways of Aniza, a fact which the Judge has highlighted in the Judgment. However, as we have mentioned earlier, the PP’s submission is based on a misapprehension of the facts relating to Aniza’s psychiatric condition which the Judge was entitled to accept. Although the PP is now not prepared to accept the factual basis of the psychiatric reports, in our view, he is not entitled to disavow them at the appellate stage of the proceedings. Once Aniza’s diminished responsibility was factored into her actions, her culpability was not, in the opinion of the Judge, so high that she deserved the sentence of life imprisonment (which means, initially, a fixed term of 20 years, and thereafter an indeterminate term of imprisonment which might be for the rest of her natural life).
65 The Judge held that some degree of leniency was called for, having regard to all the circumstances of the case. At [75] of the Judgment, he stated as follows:
I have taken the view that some judicial leniency is warranted instead of imposing the maximum punishment of life imprisonment, which is generally to be reserved for the category of very bad cases or the worst kind of cases that can be typically found under s 304(a). This is hardly a case that calls for the severest and harshest sentence that overrides or outweighs any leniency consideration on account of [Aniza’s] clean record, her guilty plea, her relatively young age (which necessarily entails an extremely long imprisonment term as she is now only 24 years of age), and the fact that she is also the mother of two very young children, who will be denied her motherly care, love and affection were her sentence to be for her natural life. Consideration must also be given to the fact that Aniza unfortunately found herself trapped within a largely abusive and loveless marriage, filled with seeming unending spousal violence which led to moderate chronic depression and her eventual impaired state of mind.
66 Having regard to the reasons given by the Judge and the various factors he took into account in sentencing Aniza to nine years’ imprisonment, we find it difficult to disagree with him or conclude that he has wasted his empathy or sympathy for Aniza. On the other hand, as reminded by the PP, we are not unconcerned about sending the wrong signal to the public that potential offenders in loveless marriages can get away with only a few years in jail if they got rid of their husbands. However, we see no real concern in affirming the Judge’s decision as this case relates specifically to an offender whose judgment was substantially impaired and who was not in full control of her mental faculties when she committed the offence. Each case must depend on its own facts, and future offenders will have to face the prospect of a longer sentence (up to 20 years’ imprisonment), rather than ten years’ imprisonment, under the current law.
Parity in sentencing
67 The PP has also argued that the Judge’s sentence is unfair as Aniza could be released in about six years’ time whereas Nasir (who has been detained at the President’s pleasure) could effectively be imprisoned for life. The sentence is therefore wrong on the parity principle as there is no parity in the sentence imposed on her and the sentence imposed on Nasir.
68 In our view, the parity principle has no application in relation to the offence committed by Nasir. The argument does not help the PP’s case, and may even be turned around to support the Judge’s position. Nasir was convicted for the offence of murder and Aniza for the abetment of the offence of culpable homicide not amounting to murder. Nasir could not be sentenced to death for murder because of his age and not because of his lack of understanding of what he did. Aniza, on the other hand, was suffering from an impairment of her cognitive appreciation of what she was doing, although she might have been unusually cunning and manipulative in her actions. Their culpabilities were not comparable. Since Nasir is (according to the PP) effectively serving a sentence of life imprisonment for the more serious offence of murder, the parity principle, if applied, would justify Aniza being sentenced to a term of imprisonment shorter than life imprisonment.
69 For the reasons we have given above, this is not a case for the intervention of this court. We are unable to disagree with the principled approach of the Judge in sentencing Aniza and to the sentence of nine years’ imprisonment. Another court might well have sentenced her to ten years’ imprisonment, but this possibility would not render the sentence manifestly inadequate. As this court said in Kwong Kok Hing ([63] supra), at [13]–[16]:
13 It is trite law that an appellate court has only a limited scope for appellate intervention apropos sentences meted out by a lower court. This is because sentencing is very much a matter of discretion and requires a delicate balancing of myriad considerations which are often plainly conflicting: Angliss Singapore Pte Ltd v PP
14 This limited scope was defined in PP v Cheong Hock Lai
(a) the trial judge had made the wrong decision as to the proper factual matrix for sentence;
(b) the trial judge had erred in appreciating the material before him;
(c) the sentence was wrong in principle; or
(d) the sentence imposed was manifestly excessive, or manifestly inadequate.
15 As to what was meant by “manifestly inadequate”, Yong CJ elaborated further in [PP v Siew Boon Loong
When a sentence is said to be manifestly inadequate, … it means that the sentence is unjustly lenient … and requires substantial alterations rather than minute corrections to remedy the injustice … [emphasis added]
16 Because sentencing is a complex discretionary process, the actual grounds on which an appellate court can intervene in sentencing are relatively circumscribed. The prerogative to correct sentences should be tempered by a significant degree of deference to the sentencing judge’s discretion. As the High Court emphasised in Angliss at [14]:
The mere fact that an appellate court would have awarded a higher or lower sentence than the trial judge is not sufficient to compel the exercise of its appellate powers, unless it is coupled with a failure by the trial judge to appreciate the facts placed before him or where the trial judge’s exercise of his sentencing discretion was contrary to principle and/or law. [emphasis added]
[emphasis in original]
70 We would like to conclude our examination of the principles of sentencing by referring to a passage from the majority judgment in Veen (No 2) ([16] supra) on the nature of sentencing. At 476–477, the majority judges said:
[S]entencing is not a purely logical exercise, and the troublesome nature of the sentencing discretion arises in large measure from unavoidable difficulty in giving weight to each of the purposes of punishment. The purposes of criminal punishment are various: protection of society, deterrence of the offender and of others who might be tempted to offend, retribution and reform. The purposes overlap and none of them can be considered in isolation from the others when determining what is an appropriate sentence in a particular case. They are guideposts to the appropriate sentence but sometimes they point in different directions. And so a mental abnormality which makes an offender a danger to society when he is at large but which diminishes his moral culpability for a particular crime is a factor which has two countervailing effects: one which tends towards a longer custodial sentence, the other towards a shorter. [emphasis added]
Conclusions
71 Our conclusions on the issues raised by the PP may be summarised as follows:
(a) The Hodgson criteria are appropriate in Singapore as guidelines for sentencing mentally unstable offenders who have committed offences punishable with life imprisonment. They fulfil two important functions:
(i) they provide an alternative (ie, an alternative to the principle that the highest punishment should only be reserved for the worst types of cases), and an equally principled, justification to impose life imprisonment on an offender (where it is the highest punishment prescribed for the offence) for the protection of society; and
(ii) they also provide a useful guide to the courts to determine when it is appropriate, and when it is not, to sentence mentally unstable offenders to life imprisonment: those who pose a long-term threat to society will be detained in prison for as long as possible under the law, and those whose illness may be temporary or transient should be rehabilitated and reintegrated into society when it is safe to do so.
(b) The Judge applied the Hodgson criteria correctly in the circumstances of this case. The sentence of nine years’ imprisonment which he imposed on Aniza was not wrong in principle or manifestly inadequate. It was a principled decision, having regard to Aniza’s diminished responsibility for the offence committed by her, and the limited sentencing options available to him.
(c) As the term of imprisonment for s 304(a) offences has been increased from ten years to 20 years, the court is now able to calibrate the appropriate custodial sentence for mentally unstable offenders within this range of punishment in cases where life imprisonment cannot be justified by reference to the Hodgson criteria.
(d) Until the current sentencing regime (where life imprisonment is initially a determinate sentence for the first 20 years but becomes an indeterminate sentence thereafter) is modified, the courts will have to apply the Hodgson criteria to mentally unstable offenders in one of two ways:
(i) where the Hodgson criteria are satisfied, the sentence of life imprisonment will be justified for the protection of society; and
(ii) where any one of the three Hodgson criteria is not met (ie, the offence in question is not grave enough to require a very long sentence; or the offender is not of so unstable a character that he is likely to re-offend in the future; or the consequences of the offence are not specially injurious to others) the offender will be sentenced in accordance with the established sentencing principles, having regard to the facts of each case and bearing in mind that deterrence has little or no effect on mentally unstable offenders.
(e) Where the offender is mentally normal, the Hodgson criteria are irrelevant, and the offender will be sentenced in accordance with the established sentencing principles, having regard to the facts of each case.
(f) In relation to the reports of the psychiatrists, the Prosecution was not entitled to question the factual basis of the psychiatrists’ opinions on appeal, after it had agreed to the reports without qualification by tendering them in court as part of the SOF for the purposes of sentencing.
(g) With reference to the mitigation process following a guilty plea, the established practice of: (i) the Prosecution and defence counsel or the offender agreeing to a statement of facts; and (ii) defence counsel making available to the Prosecution a copy of the mitigation speech prior to the sentencing hearing, has provided a simple, workable and efficient procedure which facilitates and speeds up the sentencing process. This practice should continue.
72 We fully appreciate the PP’s concerns in bringing this appeal based on his assessment of Aniza’s culpability in abetting the offence of murder committed by Nasir and the apparent disparity in the punishments imposed on them. However, since the sentence of nine years’ imprisonment passed on Aniza by the Judge is not wrong in principle or manifestly inadequate and is consistent with established sentencing principles, we are unable to accept the PP’s arguments that the sentence should be set aside and be substituted with a sentence of life imprisonment.
73 For all the reasons given above, we dismiss the Prosecution’s appeal.
[note: 1]Record of Proceedings, vol 2 at p 22.
[note: 2]Id, at p 23.
[note: 3]Id, at p 28.
[note: 4]Ibid.
[note: 5]Id, at p 29.
[note: 6]Ibid.
[note: 7]See Mitigation Plea, Record of Proceedings vol 2 at p 30 (para 7), pp 50–51 (paras 58–61).
Copyright © Government of Singapore.
This judgment text has undergone conversion so that it is mobile and web-friendly. This may have created formatting or alignment issues. Please refer to the PDF copy for a print-friendly version.
Version No 0: 20 Apr 2009 (00:00 hrs)