Gan Chai Bee Anne v Public Prosecutor [2019] SGHC 42

This appeal raises two issues of sentencing principle. The first is the proper approach for determining the aggregate sentence in cases involving multiple similar offences each of which, taken alone, may be a minor offence meriting only a light punishment, but when taken together, demonstrate a single course of criminal conduct. The second is whether restitution as a sentencing consideration has any significance apart from evidencing remorse.

Background

The appellant was the owner of a company called D3 Pte Ltd (“D3”), which was in the business of designing and installing store displays. In 2011, Nike Singapore Pte Ltd (“Nike”) engaged D3’s services, and as part of the engagement, three workers from D3 were attached to Nike. D3 would pay their wages and foot the expenses they incurred during the course of their work, and claim the corresponding sum from Nike by submitting an invoice for the same to Ms Joanne Cheong, who was a product presentation manager with Nike. Ms Cheong would then check the invoice to ensure that the claim was in order, before submitting it to Nike’s finance department, which would then disburse the invoiced sum to D3. Nike did not require these invoices to be accompanied by supporting documents or to particularise the expenses incurred.

Ms Cheong initiated a plan to exploit this arrangement with a view to wrongfully extracting pecuniary gain from Nike for herself and two of her colleagues. She enlisted the help of the appellant, who agreed to participate in facilitating Ms Cheong’s plan in order to maintain a good business relationship with Ms Cheong. Ms Cheong would collate receipts for expenses incurred by herself and her two colleagues which were not claimable from Nike, including personal expenses, and hand those receipts to the appellant. The appellant would then inflate the invoice to be issued by D3 to Nike by the amount of those expenses, before submitting it to Ms Cheong. Passing through Ms Cheong’s hands with approval, the inflated invoice would reach the finance department, and in this way, Nike would be misled into thinking that the invoice covered only D3’s legitimate expenses. D3 would then be paid the invoiced sum, and the appellant would transfer the illegitimate gains to Ms Cheong.

This scheme ran from 2012 to 2014, during which time the appellant issued, Ms Cheong approved and Nike paid on some 154 inflated invoices. In this way, $77,546.40, representing the total value of the unauthorised claims, was siphoned from Nike. The authorities eventually received a tip-off from someone who had heard of the scheme from one of Ms Cheong’s two colleagues, and the scheme was uncovered.

Corresponding to each of the inflated invoices, one charge under s 6(c) of the Prevention of Corruption Act (Cap 241, 1993 Rev Ed) (“the Act”) was brought against Ms Cheong for knowingly using, with intent to deceive her principal, Nike, a receipt which contained a false statement and which she knew was intended to mislead her principal. She therefore faced 154 charges in total. After making full restitution to Nike of the sum of $77,546.40, in May 2016 she pleaded guilty to 22 of those charges, on which the Prosecution had proceeded, and she consented to having the remaining charges taken into consideration for the purpose of sentencing. She was sentenced to 20 weeks’ imprisonment.

In June 2016, 154 related charges were brought against the appellant under s 6(c) of the Act, but under the limb of that provision which makes it an offence knowingly to give to an agent – Ms Cheong being Nike’s agent – any receipt which contains a false statement and which to the giver’s knowledge is intended to mislead the principal. The Prosecution proceeded on ten of the 154 charges, and in June 2018 the appellant pleaded guilty to them and consented to having the remaining charges taken into consideration for the purpose of sentencing.

The District Judge’s decision

The District Judge sentenced the appellant to 13 weeks’ imprisonment. His reasons are elaborated in his grounds of decision: see Public Prosecutor v Gan Chai Bee Anne [2018] SGDC 224 (“GD”). Four considerations weighed in his mind:

(a) First, the primary sentencing objective is general deterrence, which is the “key sentencing consideration for all offences prosecuted under [the Act]”, given that integrity in business dealings must be protected and corruption must be deterred (at [21]).

(b) Second, the custodial threshold was crossed for the following reason. The offence under s 6(c) of the Act was akin to, but even more serious than, the offence of cheating under s 417 of the Penal Code (Cap 224, 2008 Rev Ed) (“the Penal Code”). He referred to Idya Nurhazlyn bte Ahmad Khir v Public Prosecutor and another appeal [2014] 1 SLR 756 (“Idya”), where the High Court held that for an offence under s 417 of the Penal Code, a custodial sentence is generally appropriate where the offence causes a victim to part with property that has more than negligible value. Therefore, it “logically follow[s]”, thought the District Judge, that an offence under s 6(c) which causes such an outcome must also result in the custodial threshold being crossed. Here, Nike’s loss of $77,546.40 was not a negligible sum, and therefore the custodial threshold was crossed (at [25]).

(c) Third, the appellant ought to receive a lower sentence than Ms Cheong, whose culpability was higher. But the appellant did not play only a minor and passive role in the scheme, as the Defence had suggested. In the District Judge’s view, the appellant played a “pivotal role” in the scheme because the scheme would not have been possible without her involvement (at [29]). Moreover, the manner in which the appellant structured the payments made the offences harder to detect: the appellant would first transfer the moneys to the D3 workers attached to Nike before having the workers transfer the moneys to Ms Cheong and her colleagues. While the appellant received no financial gain, she stood to gain an indirect and intangible benefit in the form of continued business dealings with Nike through Ms Cheong. Accordingly, the appellant’s culpability was “not insignificant” (at [31]).

(d) Fourth, it was “appropriate that for the individual offences, the length of each sentence should bear a correlation with the amount stated on the respective invoice” (at [32]). The District Judge explained that this was “an approach taken on the broad assumption that the illegitimate claims by [Ms Cheong] and her two colleagues were distributed among the inflated invoices proportionately” [emphasis in original] (at [32]).

In the light of these four main reasons, the District Judge sentenced the appellant as follows:

Charge

Amount of unauthorised claim

Ms Cheong’s sentence

Appellant’s sentence

DAC-921755-2016

$561.29

weeks

weeks

DAC-921768-2016

$851.71

weeks

weeks

DAC-921825-2016

$1,336.53

weeks

weeks

DAC-921827-2016

$384.83

weeks

weeks

DAC-921828-2016

$391.77

weeks

weeks

DAC-921831-2016

$481.06

weeks

weeks

DAC-921832-2016

$904.03

weeks

weeks

DAC-921833-2016

$498.02

weeks

weeks

DAC-921834-2016

$370.23

weeks

weeks

DAC-921853-2016

$495.10

weeks

weeks

Indicated in the second column of this table is the amount of the unauthorised claim in each invoice to which the respective charge corresponds. As the District Judge noted, there is no information on the actual value of that amount in each invoice: GD at [32]. The figures in the second column are instead calculated on the assumption that the total excess payment is distributed proportionately across the falsely inflated invoices. The figures were put forward by the Prosecution, and the appellant does not object to them.

The third column indicates the sentences that Ms Cheong received for the charges corresponding to the invoices which now form the subject of the appellant’s charges. The fourth column indicates the sentences that the District Judge imposed on the appellant for each charge. For both Ms Cheong and the appellant, the first three sentences in their respective columns in the table were ordered to run consecutively, resulting in 20 weeks’ imprisonment and 13 weeks’ imprisonment respectively. The appellant now appeals against her sentence.

The parties’ cases on appeal

The appellant argues that her sentence of 13 weeks’ imprisonment should be substituted with a fine. This is warranted, in her view, because the District Judge erred by (a) characterising the gravity of her actions incorrectly; (b) regarding s 6(c) of the Act as more serious than the offence of cheating under s 417 of the Penal Code; (c) concluding that the custodial threshold had been crossed; (d) placing excessive weight on the total amount of unauthorised claims even though full restitution has been made; and (e) placing insufficient weight on the mitigating factors in this case.

The Prosecution, on the other hand, urges me to uphold the sentence imposed by the District Judge. The Prosecution defends that sentence on the basis that (a) the District Judge correctly held that the custodial threshold had been crossed; (b) the aggravating factors in this case justify the sentence imposed; (c) the absence of mitigating factors in this case further justify it; (d) the District Judge duly considered the difference in culpability between the appellant and Ms Cheong; and (e) the appellant’s sentence was in line with precedent.

The most significant point made in the appellant’s written case which was explored during oral argument was the idea that because Ms Cheong had made full restitution, “Nike has not suffered any loss, damage or injury and the Court must approach sentencing by balancing the nature of the Appellant’s moral culpability”. The Prosecution’s response was that Ms Cheong’s restitution was a “neutral factor” because the reason the making of restitution carried mitigating weight was that it evidenced the maker’s remorse, and since it was Ms Cheong who had made restitution, that act evidenced her remorse, not the appellant’s, and therefore should not mitigate the seriousness of the appellant’s offences. However, the Prosecution later accepted that Ms Cheong’s restitution did, to some degree, reduce the harm that Nike suffered, and that this was relevant to the appellant’s sentencing.

In addition to this, one feature of the case which troubled me, and which I put to the parties during oral argument, was that the sentences imposed for most of the proceeded charges, taken individually, appeared excessive in the light of the small amount of the unauthorised claim assumed to be involved in each charge. For example, if the only charge that the appellant had been convicted on was the first proceeded charge, which involved an unauthorised claim of only $561.29, it is possible that she would have received only a fine, and not three weeks’ imprisonment. While the District Judge opined that the custodial threshold had been crossed, he did not explain this conclusion with reference to the individual charges, but, it appears, only from an overall view of the multiple similar offences the appellant had committed.

Issues to be determined

In my judgment, therefore, there are two principal issues to be determined:

(a) What is the proper approach for determining the aggregate sentence in cases involving multiple similar offences each of which, taken alone, may be a minor offence that merits only a light punishment, but taken together, demonstrate a single course of more serious criminal conduct?

(b) Applying this approach, what is the appropriate sentence in this case? In particular, what is the significance of Ms Cheong’s restitution to the appellant’s sentence?

I shall address these issues in turn.

Issue 1: The proper approach

In cases involving financial crime, it is not unusual for the offender to face multiple charges for having committed multiple iterations of the same offence. For example, he may have received corrupt gratification on multiple occasions, and then be faced with a corresponding charge for each of them. In these circumstances, the conceptual paradigm of treating each offence individually and on its own terms – which is given effect by the statutory requirement that there must be a separate charge for every distinct offence: s 132 of the Criminal Procedure Code (Cap 68, 2012 Rev Ed) (“the CPC”) – appears to sit uneasily with the reality that in committing all of these offences, the offender had engaged in a unified course of criminal conduct. This reality makes it somewhat artificial to look at any one offence in isolation from the others. But nor does it seem principled to conclude from this that the sentencing judge should decide the total sentence based on his overall impression of the offender’s conduct, for that would be too subjective a method of analysis. To discern the right approach, it seems necessary to return to first principles.

A two-step analysis

The proper approach to determining the aggregate sentence in cases involving multiple offences was set out in my judgment in Mohamed Shouffee bin Adam v Public Prosecutor [2014] 2 SLR 998 (“Shouffee”). There, I said that the sentencing judge must begin by deciding the appropriate individual sentence in respect of each charge, and in doing so, he must consider the relevant aggravating and mitigating factors that bear upon the sentence for each charge (at [26]). Having done this, the sentencing judge must then consider which of the sentences should run consecutively (at [27]). In doing this, he must have due regard to the one-transaction rule and the totality principle (at [27] and [47]), which are discussed in the judgment and need not be elaborated here.

The Shouffee methodology implies, at a more general level, that sentencing for multiple offences comprises two analytically distinct steps which are to be taken in sequence. First, the court must determine the appropriate individual sentence in respect of each charge. Second, the court must determine the overall sentence which should be imposed. The issue of which of the individual sentences ought to be run consecutively is an important aspect of the inquiry at the second step, but it is only one aspect of it. Another aspect of that inquiry, which is particularly relevant in this kind of case, is whether the totality of the offender’s conduct justifies an adjustment in the individual sentences decided at the first step in respect of each charge.

This aspect of the inquiry relies on the totality principle, which has generally been taken to possess a limiting function, in the sense that it operates to prevent the court from imposing an excessive overall sentence. That is why it usually examines whether the aggregate sentence is “substantially above” the normal level of sentences for the most serious of the individual offences committed and whether its effect on the offender would be “crushing” and not in keeping with his past record and future prospects: Shouffee [54] and [57]. But as a matter of logic, the totality principle is equally capable of having a boosting effect on individual sentences where they would otherwise result in a manifestly inadequate overall sentence. This is because the totality principle requires not only that the overall sentence not be excessive but also that it not be inadequate. As the Court of Appeal explained in Haliffie bin Mamat v Public Prosecutor and other appeals [2016] 5 SLR 636, “the totality principle recommends a broad-brushed ‘last look’ at all the facts and circumstances to ensure the overall proportionality of the aggregate sentence” [emphasis added]. In a similar vein, in ADF v Public Prosecutor and another appeal [2010] 1 SLR 874 at [146], the Court of Appeal said, “In the ultimate analysis, the court has to assess the totality of the aggregate sentence with the totality of the criminal behaviour.” And Shouffee itself contemplates that the principle is capable of boosting individual sentences for it is stated there that the sentencing judge may consider running more than two sentences consecutively if the accused is shown to be a persistent and habitual offender, where there are extraordinary cumulative aggravating factors or where there is a particular public interest (at [81(j)]).

The scope of the totality principle is an important reason why the distinction should be observed between the first step of looking at the appropriate sentence for the individual offences and the second step of deciding the overall sentence. The principle indicates to the sentencing judge that there is a designated analytical space, namely, the second step, for accounting for considerations pertaining to the totality of the offender’s criminal conduct. In a case like the present, where multiple offences indicate a single course of criminal conduct, it will be tempting not to observe that space, and instead to determine the individual sentence for each offence based on an overall impression of the case. But this creates the risk that totality considerations might be incorrectly given less or more of their “due weight”: see Public Prosecutor v Raveen Balakrishnan [2018] 5 SLR 799 (“Raveen”) at [91]. It is also opaque and subjective. Where a string of 154 minor offences reveals a single course of criminal conduct, it is perhaps clear that what could have been a fine for each offence should be increased to a jail term. It is less clear where the number is five, ten or 20, and an impressionistic approach cannot reliably explain whether the custodial threshold has been crossed.

In such cases, the only way to ensure consistency in outcomes and transparency in reasoning is, in my judgment, to engage in the two steps in sequence. The sentencing judge must first reach a provisional view of the individual sentence for each offence. If the amount involved for each offence is small, he ought not to be concerned over saying that prima facie the offence warrants only a fine and not imprisonment. Turning to the second step, he ought then to consider whether the existence of any cumulative aggravating factors – such as the total amount of dishonest gain or the totality of the criminal enterprise – justifies calibrating the individual sentences upwards and running those calibrated sentences consecutively. If any such adjustment is thought necessary, the reason for making it should be spelt out. The observations of Chao Hick Tin JA in Ng Kean Meng Terence v Public Prosecutor [2017] 2 SLR 449 (“Terence Ng”) at [72] are pertinent in this context:

I acknowledge that the challenge of sentencing for multiple similar offences that constitute a single course of criminal conduct may, to some degree, have been ameliorated by recent amendments to the CPC which have permitted a single charge to be framed in respect of two or more incidents of the commission of the same offence which, having regard to their time, place and purpose, amount to a course of conduct: see s 32 of the Criminal Justice Reform Act 2018 (No 19 of 2018); s 124 of the CPC. If an accused person is convicted on such a charge, then the sentencing judge would likely not be faulted if he approached sentencing for that charge in a single step, focusing solely on the totality of the accused’s criminal conduct. However, the amended provisions permit the amalgamation of only property-related offences. Even in this context, there may be disagreement over whether the individual offences, when taken together, truly amount to a course of conduct for the purpose of amalgamation. If sentencing is to be approached with clarity, the need to follow the two-step analysis where multiple similar offences have been committed remains important as a matter of general principle, and therefore warrants careful consideration here.

The District Judge’s analysis

In my judgment, by not engaging the two steps in sequence to determine the appellant’s sentence, the District Judge erred in law. In particular, he did not adequately explain his reasons for imposing the individual sentences that he did for each charge. He appears to have had two main reasons for arriving at those sentences. But both, in my view, are not persuasive.

First, the District Judge appears to have taken the individual sentences imposed on Ms Cheong for her proceeded charges as the starting point, and reduced those sentences by anywhere up to five weeks to arrive at the individual sentences for the appellant’s corresponding proceeded charges: GD at [34]. Presumably, the District Judge did this because he assessed the appellant’s culpability to be lower than that of Ms Cheong. That this was his approach seems evident from the consistent difference in the individual sentences for each of the two offenders’ respective corresponding charges: see the table at [8] above.

While this approach justifies the individual sentences imposed for the appellant’s charges in relative terms, in the sense that they are consistently lower than the individual sentences imposed for Ms Cheong’s charges, it does not justify the individual sentences imposed for the appellant’s charges in objective terms, in the sense that there is no explanation for why each sentence is inherently appropriate. Indeed, viewed in isolation, the sentence imposed for each charge appears excessive in the light of the small amount of the unauthorised claim assumed to be involved in each instance. While it is important for a sentencing judge to achieve parity between accomplices, it is also important when sentencing each offender to have an objective, and not just a relative, basis for imposing a particular sentence on him. Here, it is not clear what that objective basis was.

The second reason the District Judge appears to have had in arriving at the individual sentences he did is captured in [32] of the GD, where he says:

The District Judge appears to be saying that as there was no evidence of the amount of unauthorised claim involved in each invoice, it would be assumed that the total amount which Nike was cheated of, ie, $77,546.40, was distributed over each of the 154 invoices in a manner proportionate to the total sum. On this assumption, the larger the total sum stated on any given invoice, the larger the amount Nike is assumed to have been cheated of when it paid on that invoice.

As I have mentioned at [9] above, this line of reasoning was not objected to, and does not seem to me to be unreasonable in the present circumstances where no evidence was led as to the actual amounts involved in each invoice. But like the reduction of individual sentences to achieve parity between the appellant’s and Ms Cheong’s sentences, it provides only a relative basis for the individual sentences imposed, not an objective one. The District Judge’s distributive assumption explains why, for example, the sentence for a charge in respect of an invoice assumed to involve an unauthorised claim of $561.29 ought to be lower than the sentence for a charge in respect of an invoice assumed to involve an unauthorised claim of $1,336.53. But it does not explain why the former sentence should be three weeks’ imprisonment and the latter sentence should be six weeks’ imprisonment.

If, on the other hand, he had applied the two-step analysis, it would have been clear what he considered prima facie to be the appropriate sentence for each charge, leaving aside the overall effect of the multiple offences the appellant had committed. It would then have been clear, from his analysis at the second step, why, in the light of various factors relating to the totality of the appellant’s offending conduct, he thought the individual sentences should be adjusted, and to what extent. The final sentence, then, would not have appeared, as it does now, as merely the result of reducing the figures in Ms Cheong’s sentences based largely on an overall impression of both offenders’ culpability. Indeed, if he had applied this two-step analysis, the District Judge may well have reached a different result, and my analysis below leads me to conclude that indeed he ought to have.

Issue 2: The proper approach applied

Step one: Determining the individual sentences

To determine the appropriate individual sentences for each of the proceeded charges the appellant faced, it is useful to consider a number of similar precedents, which were helpfully cited to me by the Prosecution.

In the unreported case of Public Prosecutor v Boey Mun Chong (DAC 900364/2016 and others) (“Boey Mun Chong”), two companies, A and B, agreed that A would be entitled to commission from B for contracts for sale that A referred to B. The director of A and the general managers of B then entered into a secret arrangement to enable the managers to circumvent B’s limits on claims for entertainment expenses. The managers would cause B to pay to A commission for contracts that A played no part in procuring, and in return, the director would pay for the managers’ entertainment bills. The director pleaded guilty to two charges under s 6(c) of the Act, one involving an illegitimate payment of $32,430.45 and the other, $37,402.82. Six similar charges were taken into consideration for the purpose of sentencing. The director was sentenced to one month’s imprisonment for each proceeded charge, both sentences to run concurrently. His two accomplices each pleaded guilty to three charges under s 6(c) corresponding to the payments the director received. Each charge attracted six weeks’ imprisonment, and two were ordered to run consecutively, resulting in 12 weeks’ imprisonment.

Next is the unreported case of Public Prosecutor v Koh Kian Wee (DAC 937868/2015 and others) (“Koh Kian Wee”). An assistant manager of companies C and D accepted bribes from the director of company E in exchange for awarding contracts from C and D to E. The manager would perform the awarded contracts without E’s involvement, keep the payments on the contracts for himself, and the director would assist him by falsifying invoices issued to C and D. In relation to the fraud perpetrated on D, the manager pleaded guilty to twelve charges under s 6(c) and the director, to eight charges under the same provision. The amount involved in each charge was between $2,400 and $3,600, and both of them received two weeks’ imprisonment for each charge. Their total sentences were 18 weeks’ each, but this included sentences for offences under s 6(a) of the Act for giving and receiving corrupt gratification, with which this case is not concerned.

A sampling of the individual sentences imposed for the s 6(c) offences in Boey Mun Chong and Koh Kian Wee may be compared with the individual sentences imposed in this case by the District Judge with reference to the amount of illegitimate payment involved in each offence:

Offender

Amount of illegitimate payment involved in the offence

Sentenced imposed

The director in Boey Mun Chong

$32,430.45 (in DAC-900368-2016)

month’s imprisonment

The director in Koh Kian Wee

$3,600 (in DAC-937904-2015)

weeks’ imprisonment

The appellant in the present case

$391.77 (in DAC-921828-2016)

weeks’ imprisonment

This comparison suggests that taken alone, the charge involving $391.77 which the appellant faced would have likely merited only a fine. This is principally because the amount of the unauthorised claim involved in that charge is very small. The same can be said of most of the other charges the appellant faced, which involved broadly similar amounts. A possible exception is DAC-921825-2016, which involved a sum of $1,336.53, which might perhaps be regarded as having crossed the custodial threshold, though still on the borderline, and certainly does not appear to merit six weeks’ imprisonment, which the District Judge imposed, without further justification.

I am aware that in Idya, I said at [47] that the cases “indicate that custodial sentences for terms of between four and eight months’ imprisonment have been imposed for cheating offences [meaning offences under s 417 of the Penal Code] that resulted in losses of between $1,000 and $15,000”. However, as I explain at [50]–[53] below, while s 6(c) of the Act is similar to s 417 of the Penal Code, they are qualitatively different offences with different sentencing emphases. I note that to the extent that they are similar, it may be asked whether the appropriate sentencing range for s 6(c) offences should be calibrated with reference to the sentencing range mentioned at [47] of Idya. However, as this issue was not argued before me, I shall leave it to be examined in an appropriate future case.

In any event, Idya may be distinguished because the appellant’s offending conduct differs in an important way from the cheating involved in the cases referred to in Idya: none of those cases involved an offender who was not intended to benefit from the victim’s loss. Nor is there any case, to my knowledge, in which the indicative sentencing range mentioned at [47] of Idya was applied to an offender who was not intended to benefit from the victim’s loss. The same is true of the case law on s 6(c). Public Prosecutor v Charan Singh [2013] SGHC 115 (“Charan Singh”) is an example. The offender, an officer with the Land Transport Authority, bought a motorcycle from the company he had been tasked to investigate, and gave the Authority a false receipt suggesting that he had bought the vehicle from another vendor, but his lie was eventually exposed. While the High Court considered that the offender’s actions damaged the Authority’s “institutional credibility” (at [24]), it also gave weight to the fact that the offender neither abused his position to obtain a lower price for the motorcycle nor misused any public funds (at [43]). Hence, no economic loss was caused to the Authority, and there was no such loss from which the offender could benefit. Accordingly, a stiff fine of $20,000 was considered a sufficient deterrent for the s 6(c) offence he had committed by that act (at [43]).

Contrasted with Boey Mun Chong and Koh Kian Wee, Charan Singh shows that the offender’s intention to benefit from the victim’s loss is a significant indicator that the custodial threshold is crossed. The same point was made in Idya, where I observed that the cases “indicate that custodial sentences have been imposed where the s 417 offence in question was committed for financial gain” (at [47]). Such an approach is principled, in my judgment, because offending for personal financial gain is a central aspect of the offender’s culpability. I shall return to this point below at [44]–[46] below, where the appellant’s culpability is examined in the light of the totality of her offences. For now, it is sufficient for me to observe that not even the Prosecution has suggested that the sentences in Boey Mun Chong, Koh Kian Wee and Ms Cheong’s case are manifestly inadequate in the light of Idya. There is therefore no reason to qualify the comparison made at [35] above.

At the first step of the analysis, therefore, I think it would have been difficult to justify a custodial sentence for each charge in this case given that the amount of the unauthorised claim involved in each charge was small, and further because none of the charges pertained to circumstances where the improper claim was intended to accrue to the appellant’s benefit. But when the amounts paid on the claims reflected in all the charges, both proceeded with and taken into consideration, are aggregated, it will readily be seen that that the property which the victim has parted with was of more than negligible value. What follows from this is that the conclusion that the custodial threshold was crossed is one that pertains to the totality of the appellant’s offending conduct, and it could properly be drawn only at the second step of the analysis. To that I now turn.

Step two: Determining the overall sentence

When the appellant’s offending conduct is considered in totality, a number of features emerge whose significance to her sentencing fall to be examined:

As the appellant’s offences, taken individually, would likely have each merited only a fine, the principal question now is whether the custodial threshold is crossed by reason of any of these six factors, and if so, to what extent. I will address this issue in relation to the first three factors first because they are closely related. I will then deal with the remaining three.

The substantial sum involved, Ms Cheong’s restitution and the fact that the appellant was not intended to benefit from Nike’s loss

There is a well-established general principle that in sentencing for financial and property offences, the greater the economic value involved in the offence, the heavier the sentence: see Public Prosecutor v Fernando Payagala Waduge Malitha Kumar [2007] 2 SLR(R) 334 (“Fernando”) at [47]; Lim Ying Ying Luciana v Public Prosecutor and another appeal [2016] 4 SLR 1220 (“Luciana Lim”) at [69]. The rationale for this is that economic value is a proxy for the degree of criminal benefit received by the offender and the degree of harm caused to the victim, and both are relevant sentencing considerations: Kow Keng Siong, Sentencing Principles in Singapore (Academy Publishing, 2009) (“Sentencing Principles”) at paras 17.022 and 17.090.

Naturally, this rationale builds into the general principle a number of qualifications. The first is that where one of the offenders involved in the commission of the offences was not intended to benefit from what the victim had been caused to part with, the value of the victim’s economic loss will not be an accurate proxy for that offender’s culpability. The second is that economic value is an accurate proxy for only economic harm suffered by the victim: Luciana Lim at [33]. Thus, if an item involved in the offence is of sentimental value, the loss of that value will not be reflected by the economic loss suffered by the victim. Nor will the economic value reflect other forms of intangible harm, such as harm suffered by society at large if the offence has the effect of infringing public interest. Both qualifications are implicated in this case, and I shall address their significance in sequence.

(1) Culpability

It is well established that an accomplice in a dishonest scheme who was not intended to benefit from the victim’s loss will be treated less severely that one who was so intended: see Sentencing Principles at para 18.008, citing Public Prosecutor v B R Chaandrran [2006] SGDC 301 at [109] and Public Prosecutor v Wang Xiao Hui [2004] SGDC 301 at [5] and [9]. As See Kee Oon JC (as he then was) observed in Luciana Lim, the commission of an offence for personal gain is an aggravating factor, and the absence of this factor may warrant a lower sentence (at [52]). Of course, as See JC was quick to point out, this does not override the general principle that the absence of an aggravating factor is not a mitigating factor (at [54]).

Here, it is not disputed that none of the $77,546.40 that Nike was made to part with was intended for the appellant. All payments on the unauthorised claims were intended for, and did in fact go to, Ms Cheong and her two colleagues. While the appellant admitted that she participated in the scheme “simply with the objective of building a good business relationship with Nike with [Ms Cheong]”, precisely why this was necessary for her to do is unclear because she also said, and the Prosecution did not dispute, that “D3 was already doing business with Nike since 2006 so there was no need to curry any further favours from Nike”. Therefore, neither the total value involved in the offences nor the business relationship between D3 and Nike suggests that the appellant participated in the dishonest scheme to profit from it. She may have had some other reason for doing so, but that is not in evidence, and so I can give it no weight.

Therefore, while the substantial value of the sum that Nike was caused to part with may have been an accurate proxy for Ms Cheong’s culpability, it does not in any useful way approximate that of the appellant.

(2) Harm

Next is the degree to which the value involved in the offences indicates the harm that the appellant caused. As I have mentioned, this value is an accurate proxy for the economic harm caused to the victim. However, a prior question is whether the victim’s economic interest is the principal interest s 6(c) was designed to protect. If s 6(c) in fact protects a broader kind of public interest, as it might appear to do, being part of a piece of anti-corruption legislation, then the victim’s economic loss would only partially inform the degree of harm caused by the offender, and would have to be considered in that light. Hence, it is useful first to examine the mischief that s 6(c) was designed to address.

(A) The mischief targeted by s 6(c)

Section 6 of the Act reads:

In the well-known decision of the High Court in Knight Glenn Jeyasingam v Public Prosecutor [1992] 1 SLR(R) 523 (“Knight Glenn”) at [20], L P Thean J (as he then was) contrasted s 6(c) with ss 6(a) and 6(b) in these terms:

Two main ideas emerge here. First, what characterises a s 6(c) offence is dishonesty, as opposed to corruption (which in this context denotes perversion or destruction of integrity in the discharge of duties by bribery or favour: see “corruption, n.” OED Online (Oxford University Press 2018) (accessed 8 February 2019); for judicial usage of the word to similar effect, see Ong Bee Chew v Ong Shu Lin [2017] SGHC 285 at [3], [66] and [71]). In particular, s 6(c) is different from ss 6(a) and 6(b) in terms of the requisite mental element. While those two provisions require proof of corrupt intent, as seen from their use of the word “corruptly”, s 6(c) requires proof of an intent to deceive, as seen from its use of the expression “knowingly … with intent to deceive”. The second idea is that s 6(c) is essentially a more serious version of the offence of cheating defined under s 415 and punishable under s 417 of the Penal Code, given that s 6(c) prescribed heavier penalties than s 417 did.

In brief, I agree with the first idea, but hesitate to endorse the second, in view of the substance of s 415 and the cases that followed Knight Glenn. I begin with the text of s 415:

There are a number of qualitative differences between s 415 and s 6(c). First, while s 415 requires the victim actually to have been deceived, s 6(c) does not. This was the very reason the High Court in Ong Beng Leong v Public Prosecutor [2005] 1 SLR(R) 766 (“Ong Beng Leong”) rejected the accused’s argument in that case, made in reliance on Knight Glenn, that his alleged s 6(c) offence was not made out because the victim was not actually deceived by the accused’s false representations. Yong Pung How CJ held that the accused was wrong “to draw a complete parallel between s 6(c) of the [Act] and the cheating offences under the Penal Code” (at [49]). Second, while s 415 requires that the victim must have been induced to perform an act or omission which causes or is likely to cause damage, s 6(c) does not. And third, while s 6(c) targets also the party who is one removed from the person intending to mislead the victim (specifically, the person who gives a false document to an agent knowing that it is intended to mislead his principal), s 415 does not. In the light of these differences, it is not surprising that Yong CJ in Ong Beng Leong stressed (at [51]) that Thean J in Knight Glenn did not mean to hold that s 6(c) and s 415 were identical in every respect.

From these differences flow at least two implications which are relevant to this case. First, they suggest, as the appellant argues and contrary to the District Judge’s view, that it may not be accurate to regard s 6(c) as “cheating under a different statutory provision” or as a “more serious” form of cheating: see [7(b)] and [11] above; cf Knight Glenn at [20]. Specifically, as the elements of s 6(c) of the Act are qualitatively different from the elements of s 415 of the Penal Code, it may not be fair to conclude from the difference in the statutory sentencing ranges for the two offences that one is an aggravated form of the other. While the same conduct may be capable of establishing both offences, each offence regards different aspects of that offender’s conduct as blameworthy. This leads to the second implication, which is that the emphasis in s 6(c), as Thean J alluded to in Knight Glenn at [20], is on the offender’s dishonesty, in the light of the provision’s central requirement that either the offender “knowingly” gave a false document “which to his knowledge [was] intended to deceive the principal” or the agent “knowingly” used such a document “with intent to deceive his principal”. This makes it all the more significant to the sentencing exercise here that no dishonest gain was intended to be obtained by the appellant at Nike’s expense in this case.

The idea that dishonesty characterises a s 6(c) offence was approved by the Court of Appeal in Review Publishing Co Ltd and another v Lee Hsien Loong and another appeal [2010] 1 SLR 52 (“Review Publishing”). The court had to consider s 6(c) in the rather different context of deciding whether the claimants, having been associated in the defendant’s publication with a person convicted on a s 6(c) charge, had been defamed as corrupt by the defendant. The court answered in the affirmative, opining that an ordinary reasonable person was likely to consider that s 6(c) because of its place in the Act carried an imputation of corruption (at [77]). This perception was, however, different from the intended scope of s 6(c), and in that regard, the court considered that Thean J in Knight Glenn was right to highlight that an offence under s 6(c) is distinguished by dishonest conduct (at [76]).

This is also supported by the provision’s origin and the jurisprudence on its equivalent in English law. The Act was enacted against the background of its older English counterpart, the Prevention of Corruption Act 1906 (c 34) (UK) (“the 1906 Act”): Public Prosecutor v Syed Mostofa Romel [2015] 3 SLR 1166 (“Romel”) at [12]. In pari materia with s 6(c) of the Act is the third paragraph of s 1(1) of the 1906 Act. That paragraph was examined in Sage v Eicholz [1919] 2 KB 171, where Bray J emphasised that the word “corruptly”, which is used in the first two paragraphs, is “deliberately omitted from the third” (at 175). Bray J found that the choice of words made it clear and unambiguous that “knowingly” did not necessarily involve an element of corruption, and that the type of conduct envisioned to be policed under the third paragraph was independent of whether there is an “element of corruption”, which was more difficult to prove (at 176). The meaning derived from the wording used in the provision was not to yield to the fact that the statute was entitled “An Act for the better Prevention of Corruption” nor by the context of the two preceding paragraphs: see also Colin Nicholls QC et al, Corruption and Misuse of Public Office (Oxford University Press, 2nd Ed, 2011) at para 2.71. Bray J’s view was later approved by the English Court of Appeal in R v Tweedie [1984] 2 WLR 608 at 611C per Lawton LJ.

This is not to say that s 6(c) does not serve any broader social purpose. Its enactment recognises that the agent-principal relationship, which is an important and often inevitable incident of business, has vulnerabilities that make it readily susceptible to abuse through dishonest means, and therefore should be granted penal protection to promote its integrity and utility. This was also the view of the Hong Kong Court of Appeal in HKSAR v Luk Kin Peter Joseph [2015] HKCU 2767, where McWalters JA opined that s 9(3) of the Prevention of Bribery Ordinance (Cap 201) (HK), which is in pari materia with s 6(c) of the Act, “[b]y targeting non-bribery conduct that also undermines the integrity of the principal : agent relationship … in its own distinct and separate way, plays a broader role in protecting and preserving the integrity of this special relationship” (at [160]).

That s 6(c) targets “non-bribery conduct”, in the words of McWalters JA, seems clear on further examination of the case law on bribery itself. In Romel at [26], I described three non-exhaustive ways in which bribery can manifest in the private sector:

In my judgment, therefore, the central mischief targeted by s 6(c) is the dishonest exploitation of an agent-principal relationship, whether by a person who knowingly presents a false document to the agent or by the agent himself who knowingly presents a false document to his principal. The usual result of the offence is economic harm to the principal, and therefore the degree of such harm will often be a significant indicator of the seriousness of the offence in most cases. That said, an offence under s 6(c) may have wider repercussions. Depending on the facts, it may be seen to undermine integrity in business dealings or, as was the case in Charan Singh ([37] above), to damage the credibility of a public institution. As the Court of Appeal recognised in Review Publishing ([54] above) at [75], “[i]t is not inconceivable that some [kinds] of dishonest act … falling within s 6(c) … may be regarded as corrupt conduct”. Where this has happened, the sentencing judge should take account of it as an aggravating factor.

In this case, nothing indicates the appellant has been involved in corruption in the private or public sector. Therefore, the only harm relevant to the appellant’s sentence is the economic harm her offences caused Nike. I shall therefore analyse the degree of that harm.

(B) Restitution and economic harm

As I have mentioned, the Prosecution during oral argument accepted that Ms Cheong’s making full restitution to Nike of the amount Nike had been deceived to part with reduced the degree of economic harm that Nike had suffered, and that this was relevant to the appellant’s sentence. I endorse this view, and expand on it below.

An offender’s making of timely and voluntary restitution for loss caused by his offending conduct has generally been regarded in the cases as evidence of his remorse, and therefore as a mitigating factor: see, for example, Angliss Singapore Pte Ltd v Public Prosecutor [2006] 4 SLR(R) 653 at [74]–[75]. One implication of this principle is that the making of restitution is evidence for the remorse of only the maker, and not of anyone else. Hence, in this case, it is not possible to infer from Ms Cheong’s act of restitution any remorse on the part of the appellant.

However, restitution is not necessarily limited in its significance to being evidence of remorse. In my judgment, it may also indicate that the economic harm that the victim has suffered has been reduced, and indeed, substantially reduced if full restitution was made. In such a case, it would not be completely eliminated because at the time of the offence, the victim would have been made to part with something of value, and after that, would have remained worse off until he received recompense. But the fact that what the victim lost has now been restored to him may, in my judgment, bear on the sentence imposed on the offender who caused the loss. This would be especially relevant in cases such as the present where the offender before the sentencing court was never intended to benefit personally, and where one of the principal metrics of culpability is the harm caused to the victim for the benefit of another party who in fact has substantially diminished that harm by making restitution. Moreover, giving significance to the impact of restitution, regardless of whose remorse it evidences, incentivises all offenders involved in a dishonest scheme to restore the loss suffered by the victim if they are able, which in turn promotes for the victim a form of restorative justice.

Similar thinking may be found in the literature and in the cases. Thus, in Sentencing Principles, the learned commentator, having acknowledged that restitution had mitigating value because it evidenced the offender’s remorse, suggested a further reason why it may justify a less severe sentence (at para 20.003):

Likewise, in Fernando ([42] above), Rajah J (as he then was) observed that the extent of loss or damage actually suffered as a result of the offence was relevant to sentencing. Hence, where there had been no loss or minimal loss because the offender was apprehended and the proceeds of crime recovered, that would affect the sentence to be imposed (at [49]). While he acknowledged that restitution is of mitigating value mainly because it reflects true remorse, he also opined that “restitution of any kind” would be relevant, and referred to the decision of the Supreme Court of Western Australia in R v Debra Jane Mitchell [1998] WASCA 299, where it was held that making good the loss that the victim had suffered was a mitigating factor because it reduced the harm to the victim.

Similarly, in the decision of the Hong Kong Court of Appeal in HKSAR v Tsang Pui Yu, Shirlina [2014] 5 HKC 111, McWalters J (as he then was) observed that “[t]he act of restitution will always reduce the harm to the victim and may or may not evidence genuine remorse” [emphasis added] (at [52]). Thus, the court felt unable to give full weight to the restitution made in that case because although the offender was genuinely remorseful, he had made only partial restitution. And in R v Wayne Edward Combo [2015] WASCA 34, McLure P of the Supreme Court of Western Australia observed that the voluntary repayment of a significant part of the amount defrauded is “mitigatory, in that it reduces the harm to the victim without it having to execute [a judgment containing a restitution order], even if it does not reflect contrition” [emphasis added] (at [70]).

In the light of these principles, it is relevant to have regard to the restitution made by Ms Cheong because that lessened the sting of the harm suffered by Nike when it parted with $77,546.40.

The principles I have outlined and the conclusion that I have reached do not necessarily imply that when a compensation order is made against an offender under s 359(1) of the CPC the offender should receive a less severe sentence in view of the compensation due to the victim under that order. This is for at least three reasons. First, the harm suffered by the victim should be assessed as at the time of sentencing. There is no guarantee that the offender will comply with the compensation order that is made. Second, to regard a compensation order as having a mitigating effect in the same case would invite accused persons to bargain with the court for lower sentences in exchange for making restitution, which is wrong in principle for it would effectively enable offenders to buy themselves out of prison: see R v Yip Muk Kan [1988] HKC 868 at 869F. Third and relatedly, there would also be the undesirable anomaly that impecunious offenders would receive harsher sentences than those who are able to comply with the order: see R v Chamczuk [2010] AJ No 1407 at [14].

(3) Conclusion

In the result, the substantial value involved in the offences is of largely attenuated significance to the appellant’s sentencing. It is not a proxy for the appellant’s culpability because the appellant was not intended to benefit from Nike’s loss. While the value is a proxy for the economic harm Nike sustained, that harm was substantially reduced by Ms Cheong’s making of full restitution to Nike. The value involved in the offences therefore cannot be a reason for concluding that by reason of this alone, the custodial threshold has been crossed in this case. To the extent that the District Judge held otherwise, in my judgment, he was in error: see the GD at [25].

Premeditation and planning

It is well established that committing an offence with premeditation and planning is an aggravating factor. This factor is a distinct aspect of the offender’s culpability and it is amply present in this case. The appellant and Ms Cheong took deliberate steps to minimise the risk of their scheme being detected. They arranged for the moneys to be paid first to D3’s workers before being transferred to Ms Cheong and her colleagues. This ensured the absence of any record of payment from Nike to Ms Cheong directly. The appellant and Ms Cheong also ensured that each invoice was inflated by only a small amount so as not to arouse any suspicion of their scheme. These actions were deliberated, carefully planned and designed to create a veneer of legitimacy so as to avoid detection. They constitute a significant aggravating factor, and as they demonstrate a high degree of conscious choice on the appellant’s part, they enliven the need for a sentence that deters the appellant specifically from repeating such conduct: see Public Prosecutor v Law Aik Meng [2007] 2 SLR(R) 814 at [22].

Long period of offending

It is also well established that persistent and habitual offending is an aggravating factor: see Fernando ([42] above) at [48]. This may be inferred from the commission of multiple offences over an extended period of time. It is important to distinguish this from the sheer multiplicity of offences committed, which ought not to be treated as an aggravating factor in itself because it is properly accounted for through the number of charges preferred and through the aggregation of the individual sentences for each charge: see Terence Ng ([22] above) at [15].

Here, the appellant and Ms Cheong submitted a total of 154 inflated invoices to Nike over a period of more than two years. During this time, the appellant was fully aware of what she was doing, having on each occasion to inflate and falsify an invoice. It can only be inferred that she had every intention of persisting in her offences until the scheme was unravelled. This again justifies imposing a sentence that deters the appellant specifically from future reoffending: see Fernando at [43].

Plea of guilt

Calibration and running of sentences

Having regard to the totality of the appellant’s offending conduct, I consider that the custodial threshold is crossed. This is not because of the value involved in the offences, but because of the premeditation and planning with which the appellant participated in the dishonest scheme, and because of the sustained and deliberate manner in which she committed the offences which revealed a mind consciously habitualised to crime. All of this demonstrates that the appellant participated in a calculated course of criminal conduct designed to siphon moneys from the victim. It spells the need for a sentence that will deter her specifically from reoffending in the future and deter others from engaging in similar conduct. It also evinces a high degree of culpability on her part, despite the fact that she was not the one who stood to benefit from the offences and also despite the fact that the harm to the victim in this case was significantly diminished by the restitution made. On account of this, she deserves a suitably harsh sentence which signals society’s disapproval of her criminal conduct.

In my judgment, these deterrent and retributive aims can be satisfactorily met only with a sentence of imprisonment. The totality of her criminal conduct justifies increases in the individual sentences for her charges from what would, taken alone, have been fines, to short custodial sentences.

To determine the length of the sentences, I bear in mind the precedents examined at [32], [33] and [37] above. Also of significance to me is that contrary to the Prosecution’s and the District Judge’s views, I do not think that the value involved in the offences is meaningfully treated as an aggravating factor in this case, for the reasons I have summarised at [69] above. For this reason, the calibrated sentences need not, and indeed, should not, be proportionate to the amount of the unauthorised claim involved in each charge. They should be uniform, given that the relevant aggravating factors, namely, the appellant’s premeditation, planning and habitual offending, arose from all those charges considered in totality, and was not true of any one of them more than the other. Bearing all of this in mind and having regard to all the circumstances of the case, I substitute a sentence of one week’s imprisonment for each of the appellant’s proceeded charges.

Next, in my judgment, in addition to complying with the one transaction rule and the totality principle as elaborated in Shouffee and Raveen, the way in which sentences are run should, as far as possible, reflect the substance and totality of the offending conduct. Where multiple offences are involved, the most appropriate mechanism for doing so is the consecutive running of multiple sentences. In my view, this is preferable to imposing what may, to the accused, be unexpectedly disproportionate individual sentences for relatively minor offences.

In view of the appellant’s multiple offences, I think it is appropriate for the sentences for the first five proceeded charges, as listed at [8] above, to run consecutively, and for the sentences for the remaining charges to run concurrently. This results in an aggregate sentence of five weeks’ imprisonment. In my judgment, this strikes the proper balance between communicating a deterrent and retributive message to the appellant for her active and knowing involvement in a dishonest scheme, and reflecting how the economic harm caused to the victim of that scheme, which was never meant to benefit the appellant, has in any event been all but reversed. The appeal is therefore allowed to this extent.