Logachev strikes again: A “new” sentencing framework for evasion of income tax in Tan Song Cheng v PP [2021] SGHC 138

*By: Keith Low

I.         Introduction

Consistency in sentencing promotes fairness by ensuring that like offenders are treated similarly by our criminal justice system. This can be achieved by applying well-constructed sentencing frameworks.

II.         Material facts

The present decision involved two separate appeals against the sentence imposed for s 96(1) offences. The material facts of each appeal are as follows:

A. Tan Song Cheng v PP ^[4]

The First Appellant (“Tan”) was the director of two companies which he ran together with one Lim.^[5] Tan agreed to Lim’s false reduction of one of the company’s net profits in the years 2009 and 2011.^[6] In addition, Tan failed to declare holiday reimbursements he received from the other company as a “performance reward”.^[7] These two acts resulted in the under-reporting of his trade and employment incomes, respectively.^[8]

The resulting amount of tax undercharged was $34,992.26 in 2009 and $34,444.18 in 2011.^[9] These became the subject matter of the two s 96(1) charges faced by Tan. Six other similar s 96(1) charges were taken into consideration for the purposes of sentencing.^[10]

The lower court imposed a sentence of six weeks’ imprisonment for each s 96(1) charge and ordered them to run consecutively, leading to an aggregate sentence of 12 weeks’ imprisonment.^[11]

B. Lin Shaohua v PP ^[12]

The Second Appellant (“Lin”) was the partner of two partnerships.^[13] The partnerships’ accountant (“Lucy”) prepared the accounts for both of these partnerships.^[14] Lin instructed Lucy to reduce the reported sales to below a $1 million threshold whenever the actual sales figures for either partnership exceeded that threshold.^[15] This led to a corresponding reduction in Lin’s reported partnership income.^[16]

By under-reporting her personal income, Lin evaded $79,142.13 of tax.^[17] This was the subject matter of her s 96(1) charge. She faced another charge under the Goods and Services Tax Act,^[18] with two similar charges taken into consideration under the ITA and GSTA respectively.^[19] Through these offences, she evaded $536,379 of tax.^[20]

The lower court imposed a sentence of 10 weeks’ imprisonment for the s 96(1) charge.^[21]

In addition to their sentences, both offenders incurred the mandatory treble penalty under s 96(1) of the ITA, which requires an offender to pay a sum three times the value of tax evaded.^[22]

III.         The Decision

The Court’s review of precedent cases involving s 96(1) offences did not reflect any consistent sentencing trend.^[23] Furthermore, the prior decision of Chng Gim Huat v PP^[24] had an anchoring effect on sentences for this offence: sentences in subsequent decisions clustered around the lower end of the permissible custodial sentencing range.^[25] This was problematic because the courts generally aim to utilise the full range of possible sentences.^[26] As such, the Court was convinced of the need to establish a new sentencing framework for s 96(1) offences so as to resolve these issues and achieve greater consistency in sentencing.^[27]

A.    Finding the appropriate framework

The Court first considered the following formulations by the Court of Appeal and the High Court in prior decisions:^[28]

• The court in Ng Kean Meng Terence v PP^[29] established a two-step sentencing framework for rape, which distinguishes between offence-specific and offender-specific factors.^[30] Under this framework, the Court first considers the manner and mode in which the offence was committed, and thereafter turns to any aggravating or mitigating factors personal to the offender.^[31]
  
• The court in PP v Koh Thiam Huat^[32] held that in evaluating the seriousness of a crime, the court has to consider the harm caused by the offence and the culpability of the offender.^[33]
  
• Subsequently, the court in Logachev clarified that these considerations relating to the harm caused by the offence and the offender’s culpability should go under the offence-specific factors (i.e., the first step of the two-step sentencing framework outlined in Terence Ng).^[34] Offender-specific factors considered in the second step of the framework comprise aggravating and mitigating factors that are not directly related to the commission of the offence, and which are generally applicable across all criminal offences.^[35]

The Court found that the categorisation in Logachev neatly delineated the various sentencing considerations.^[36] The Court further noted that the offence-specific factors listed in Logachev are relevant to offences involving tax evasion.^[37] As such, the Court endorsed the lower court’s decision to transpose the Logachev framework to s 96(1) offences, and incorporated the aforesaid offence-specific factors.^[38] This transposition is also in line with the fact that the Logachev framework had already been adapted for offences concerning financial and commercial crimes, like money-laundering and corruption.^[39]

B.    The adapted five-step Logachev framework for s 96(1) offences (“the Framework”)^[40]

1.    First, the Court identifies the level of harm (based on the quantum of tax evaded) and the level of culpability (based on the offence-specific factors which affect culpability).

2. The applicable indicative sentencing range is then selected from the table below.

Harm Culpability	Level 1 Harm (<$75,000 tax evaded)	Level 2 Harm ($75,000 – $150,000 tax evaded)	Level 3 Harm (>$150,000 tax evaded)
Low Culpability	Fine or up to 6 months’ imprisonment	6 to 12 months’ imprisonment	12 to 18 months’ imprisonment
Moderate Culpability	6 to 12 months’ imprisonment	12 to 18 months’ imprisonment	18 to 24 months’ imprisonment
High Culpability	12 to 18 months’ imprisonment	18 to 24 months’ imprisonment	24 to 36 months’ imprisonment

3. The Court identifies the appropriate starting point within the indicative sentencing range. In the present case, the Court did so with reference to all of the relevant offence-specific factors.^[41] This is similar to the approach taken by the High Court in Logachev.^[42]

4. The Court adjusts the starting point sentence based on any offender-specific aggravating and mitigating factors.^[43]

5. Finally, the Court will make further adjustments to account for the totality principle,^[44] if necessary.

The various offence-specific factors and offender-specific factors are elaborated on below.

C.    Offence-specific and offender-specific factors

(1)   Offence-specific factors (Steps 1 to 3 of the Framework)

The following table lists the offence-specific factors adopted from Logachev and their relevance vis-à-vis s 96(1) offences.

Offence-specific factor (Relating to the harm caused)	Relevance to s 96(1) offences
Amount of income tax evaded	Section 96(1) of the ITA is targeted specifically at tax evasion, so the harm caused to the State by such an offence is measurable the amount of tax evaded.[45] Hence, the quantum of tax evaded determines the level of harm in Step 1 of the Framework.[46] This level is later calibrated based on other offence-specific factors relating to harm in Step 3, if necessary.[47]
Involvement of a syndicate[48]	Activities of criminal syndicates are often multi-faceted and involve serious crimes. Hence, a tough stance against criminal syndication is warranted, and the involvement of a syndicate operates as an independent aggravating factor.
Involvement of a transnational element[49]	Cross-border transactions or offshore companies and trusts can be used to obfuscate the purpose of fund transfers, and conceal the beneficial ownership of taxable assets. The increased difficulty in detecting tax evasion in these contexts makes the involvement of a transnational element an aggravating factor.
Undermining government schemes or harm done to the confidence in public administration[50]	There is a risk of double counting in relation to this harm factor, because the evasion of tax itself undermines the government scheme of taxation and may have the impact of damaging confidence in public administration. But the factor may still be applicable where a s 96(1) offence has an effect of undermining governmental schemes and confidence in public administration that is additional to the ordinary effects of such an offence.
Offence-specific factor (Relating to the culpability of the offender)	Relevance
The degree of planning and premeditation[51]	The presence of planning and premeditation is an aggravating factor because it reflects greater criminality.
Sophistication of the systems and methods used to evade income tax or to avoid detection[52]	Committing an offence through the use of sophisticated methods is considered an aggravating factor. However, the Court warned of the possibility of double counting given the potential overlap between this factor and the aggravating factor of planning and premeditation.[53]
Evidence of a sustained period of offending[54]	Although this is an aggravating factor which evinces a need for specific deterrence, the Court noted that it may have a limited impact on the overall sentence, given that there may be double counting when courts further take into consideration charges that concern similar offending behaviour on other occasions. Further, the nature of s 96(1) offences is such that they can only be committed on a yearly basis when tax is payable.
The offender’s role[55]	This aggravating factor applies mainly to offenders who play an influential role in a criminal syndicate.
Abuse of position and breach of trust[56]	Such abuse or breach can operate as an aggravating factor. An example of this would be using a client’s account to evade taxes.

(2)   Offender-specific factors (Step 4 of the Framework)

Offender-specific factors	Relevance to sentence
Aggravating factors: having offences taken into consideration for sentencing purposes; having relevant antecedents; and an evident lack of remorse[57]	These aggravating factors are well-established in case law (see Logachev at [63]–[66]; Terence Ng at [64]).
Mitigating factor: a guilty plea[58]	A guilty plea is a clear mitigating factor because it saves the resources of the criminal justice system, and may represent genuine remorse on the part of the offender. However, where there is incontrovertible documentary evidence for a s 96(1) offence, there is likely to be little or no mitigating value to such an inevitable guilty plea.
Mitigating factor: voluntary restitution[59]	An offender’s actions in making restitution without the party cheated having to resort to attempts at recovery are indicative of his remorse.
Mitigating factor: co-operation with the authorities[60]	The weight to be given to this factor depends on the specific circumstances of the case. Importantly, there is often difficulty in detecting such offences and income tax returns are heavily dependent on self-reporting.

D.    Fine as a sentencing option

Even though the ITA provides the sentencing option of a fine of up to $10,000 for s 96(1) offences, the Court held that a fine should only be imposed where its deterrent effect is not eclipsed by the imposition of the mandatory treble penalty.^[61] For instance, the sentencing option of a fine can remain relevant where the amount of tax evaded is in the lower range, leading to a lower mandatory treble penalty that does not render the deterrent effect of the maximum $10,000 fine nugatory.^[62]

E.    Application of framework to the present appeals

A fine was not an appropriate sentence for either appellant.^[63] The Court noted that the amount of income tax evaded for each of Tan’s charges was in itself already more than three times the maximum fine that can be imposed.^[64]

(1)   Tan Song Cheng

Based on the amount of tax evaded for each of Tan’s two charges (which were both less than $75,000), the Court found that both charges can be classified as “Level 1 Harm”.^[65] Further, Tan’s culpability for the offences were low.^[66] There was some planning and premeditation involved in under-reporting the firm’s revenue, but no sophisticated means were employed to evade income tax.^[67]

The Court agreed with the lower court that a starting point sentence of ten weeks’ imprisonment on each charge was appropriate.^[68] This sentence was then reduced to six weeks’ imprisonment after considering Tan’s early plea of guilt, his cooperation with IRAS, and the fact that he made full restitution for all of his charges.^[69] The Court also noted that he had six other s 96(1) charges to be taken into consideration.^[70] The tax evaded under these charges totalled $119,186.21.^[71]

Additionally, the Court agreed that the two sentences should run consecutively^[72] because the offences were committed two years apart and were therefore not part of the same transaction.^[73]

(2)   Lin Shaohua

Lin evaded more than $75,000 of tax. Thus, the harm caused was classified as “Level 2 Harm” at Step 1 of the Framework.^[74] Her culpability was low: her scheme was unsophisticated and did not extend beyond instructing Lucy to reduce the actual sales figures below the $1 million threshold.^[75]

Since there were no other harm factors involved and the amount of tax evaded lay just between the two levels of harm, the Court found that the lower court was not wrong in exercising its discretion to adjust the indicative starting sentence to 16 weeks’ imprisonment (when it would have ordinarily landed in the range of 6 to 12 months’ imprisonment).^[76]

The Court further considered the additional charge taken into consideration which involved the evasion of $7,654.46 in tax; Lin’s early plea of guilt and cooperation during the investigations; and the fact that full restitution had been made. In these circumstances, the High Court found it unnecessary to adjust the sentence of ten weeks’ imprisonment, even though it appeared to have been slightly lenient.^[77]

IV.         Discussion

The present decision is a remarkable one. First, the Court neatly transposed the Logachev framework and objectively evaluated the applicability of the offence-specific and offender-specific factors to s 96(1) offences. Its utilisation of the quantum of income tax evaded as the single quantitative metric also reduces any potential arbitrariness in determining the level of harm occasioned by such an offence.^[78]

Second, the Court was careful to provide both sentencing options of a fine and a custodial sentence under the lowest end of the sentencing matrix. The Logachev matrix provides for a mere fine for cheating at play offences in that category, but it would have been inappropriate to adopt this for income tax evasion. As stated earlier, even the maximum fine of $10,000 for s 96(1) offences would often be rendered insignificant by the mandatory treble penalty. Further, that maximum fine pales in comparison to the $150,000 maximum fine for cheating at play in Logachev.^[79] Hence, the Court rightly left open the sentencing option of imprisonment under the lowest end of the sentencing matrix, which may be the more appropriate sentencing option for most offences even in that category.

However, some concerns arise from the application of the Framework in the two appeals. The Court endorsed the lower court’s decision to adjust Lin’s indicative starting sentence outside the range of 6 to 12 months’ imprisonment for Level 2 Harm under the Framework.^[80] This was said to be justified by the absence of other harm factors (besides the quantity of tax evaded): there was no syndicate involvement or transnational element.^[81] It was further stated that the sentencing categories in the Framework are not to be regarded as rigid or impermeable.^[82] But the degree to which courts in subsequent decisions may exercise such discretion remains undiscussed. Indeed, it is uncertain if such discretion was exercised in respect of Tan’s appeal. While his offences did not involve the additional harm factors, there was no express mention of any downward adjustment in his indicative starting sentence.^[83]

It is apposite to note that none of the case precedents for s 96(1) offences referred to in the current decision involved the additional harm factors. If these factors are indeed often absent, greater consistency in sentencing may be achieved by using indicative sentencing ranges that account for their absence. In this regard, the current ranges in the Framework can be adjusted downwards, leaving some room for an uplift in sentence where these uncommon harm factors are present. In other words, an offence at the high end of “Level 3 Harm” involving the additional harm factors and an offender of the highest culpability should just attract the maximum custodial term of 36 months. 

Alternatively, consistency may be achieved by fixing a benchmark sentencing discount for the absence of these harm factors.^[84] For example, the reduction of Lin’s starting point sentence from 6 months’ imprisonment (since her offence was on the low end of Level 2 Harm) to 16 weeks can be interpreted either as a one-third sentencing discount or a downward adjustment of about 2 months. The latter method should be preferred: fixing a discount by proportion may result in a disproportionately large discount for offences that attract a higher starting point sentence under the Framework.

V.         Conclusion

The adaptation of the Logachev framework will aid the courts in utilising the full range of sentences provided for s 96(1) offences. It is also a relatively intuitive framework for identifying and analysing the relevant sentencing considerations. However, the Court in Logachev itself noted that the list of sentencing considerations is non-exhaustive and that the various categories outlined are not watertight.^[85] Section 96(1) offences also have the potential to vary widely.^[86] This means that factors deemed less significant in this decision may be of paramount importance within another factual matrix. In this regard, the court in each subsequent decision shoulders the task of attributing the appropriate weight to these various sentencing considerations and then balancing them.

For these reasons, the Framework may not, by itself, provide consistency in sentencing for s 96(1) offences. To achieve this, each subsequent application of the Framework has to be done with diligence so that future decisions can be guided by clear and cogent reasoning.

---

– – – For a PDF version of this article, click here – – –

*Written by: Keith Low, 3rd-year LL.B. student, Singapore Management University Yong Pung How School of Law.

[1] Cap 134, 2014 Rev Ed (“ITA”).

[2] Tan Song Cheng, infra n 4, at [23]; see Chng, infra n 24, at [106]–[114].

[3] [2018] 4 SLR 0609 (“Logachev”).

[4] [2021] SGHC 138 (“Tan Song Cheng”) at [4]–[7].

[5] Id, at [4].

[6] Id, at [5].

[7] Ibid.

[8] Ibid.

[9] Ibid.

[10] Id, at [6].

[11] Id, at [7].

[12] Id, at [8]–[11].

[13] Id, at [8].

[14] Ibid.

[15] Ibid.

[16] Id, at [9].

[17] Ibid.

[18] Cap 117A, 2005 Rev Ed (“GSTA”).

[19] Tan Song Cheng, supra n 4, at [10].

[20] Ibid.

[21] Id, at [11].

[22] ITA, supra n 1, s 96(1).

[23] Tan Song Cheng, supra n 4, at [27].

[24] [2000] 2 SLR(R) 360 (“Chng”). In Chng, the offender evaded a total of $354,645.65 in taxes and was sentenced to an aggregate of two months’ imprisonment. However, that case involved strong mitigating factors: the offender had voluntarily and swiftly informed the tax authorities about payments made to him that he had failed to declare; and he had made reparation in excess of the actual taxes payable.

[25] Tan Song Cheng, supra n 4, at [25]–[26]. The Court observed that the sentences in unreported cases ranged from one to six weeks’ imprisonment.

[26] Ibid, citing Suventher Shanmugam v PP [2017] 2 SLR 115 at [29]; Vasentha d/o Joseph v PP [2015] 5 SLR 122 at [45]–[46].

[27] Id, at [33].

[28] Tan Song Cheng, supra n 4, at [37].

[29] [2017] 2 SLR 0449 (“Terence Ng”).

[30] Id, at [39].

[31] Ibid.

[32] [2017] 4 SLR 1099.

[33] Id, at [41].

[34] Logachev, supra n 3, at [36].

[35] Ibid.

[36] Tan Song Cheng, supra n 4, at [38].

[37] Id, at [39].

[38] Id, at [58].

[39] Ibid.

[40] Id, at [59]–[78].

[41] See Tan Song Cheng, supra n 4, at [74].

[42] See Logachev, supra n 3, at [79].

[43] Tan Song Cheng, supra n 4, at [75].

[44] The totality principle requires the court to examine whether the aggregate sentence is substantially above sentences normally meted out for the most serious of the individual offences committed, then further consider whether the effect of the aggregate sentence on the offender is crushing and not in keeping with his past record and future prospects: Mohamed Shouffee bin Adam v PP [2014] 2 SLR 998 at [54], [57]; cited at [78] of Tan Song Cheng.

[45] Tan Song Cheng, supra n 4, at [68]. The Court rejected suggestions of using the proportion of unreported income (vis-à-vis the value of reported income) as the main indicia of harm caused by tax evasion offences, because this would allow individuals with a higher overall income to obtain a more lenient sentence for evading tax of the same amount as an individual with a lower overall income: Tan Song Cheng at [61]. It further rejected the suggestion to consider the extent to which restitution has been made as the main indicia of harm occasioned by the offence. The Court found that this would allow an offender to escape the full consequences of his act no matter how egregious his actions were simply by paying back the tax evaded: Tan Song Cheng at [62].

[46] Id, at [70].

[47] Ibid.

[48] Id, at [43]–[45].

[49] Id, at [46]–[47].

[50] Id, at [48].

[51] Id, at [49].

[52] Id, at [50].

[53] While there is indeed a potential overlap, sophistication of methods used to commit the offence should not be subsumed under the aggravating factor of planning and premeditation because sophisticated methods do not necessarily require a high degree of planning or evince the presence of premeditation. For example, skilled offenders may utilise their expertise to commit the same offence, and will be dealt with more harshly because they are more culpable given that they exploited their expert knowledge to commit the offence in question.

[54] Tan Song Cheng, supra n 4, at [51].

[55] Id, at [52].

[56] Id, at [53].

[57] Id, at [54].

[58] Id, at [55].

[59] Id, at [56].

[60] Id, at [57].

[61] Id, at [72]–[73].

[62] Ibid.

[63] Id, at [82], [90]–[93].

[64] Id, at [82]. Though not expressly stated, this argument applies with greater force in respect of Lin’s offence, since the quantum of tax evaded in her s 96(1) charge was more than double that of Tan’s.

[65] Id, at [80].

[66] Id, at [82].

[67] Id, at [80].

[68] Id, at [82].

[69] Id, at [84].

[70] Id, at [83].

[71] Ibid.

[72] Id, at [85].

[73] Ibid.

[74] Id, at [87].

[75] Id, at [88].

[76] Id, at [90].

[77] Id, at [93].

[78] Id, at [65].

[79] See s 172A(2) of the Casino Control Act (Cap 33A, 2007 Rev Ed).

[80] Tan Song Cheng, supra n 4, at [90].

[81] Ibid.

[82] Ibid; citing Edwin s/o Suse Nathan v PP [2013] 4 SLR 1139 at [22].

[83] If it is the case that “Level 1 Harm” offences (like Tan’s) preclude the involvement of the additional harm factors, such clarification would be helpful as well.

[84] The Courts may choose to depart from this benchmark where there are clear and cogent reasons for doing so.

[85] Logachev, supra n 3, at [38].

[86] Even though they do tend to manifest in a few particular forms.

---

---