By Boo Kok Chuon

The recent sentencing appeal involving former actor Edmund Chen attracted considerable public attention. Unsurprisingly, most of the discussion centred on whether five days’ imprisonment was too lenient, and whether the High Court was right to cut it down to one day.

Public discourse naturally gravitates towards outcomes. Legal minds, however, tend to find themselves more interested in the reasoning that produces those outcomes.

Reading the District Court’s brief reasons on sentence, one sentence stands out. It appears near the very beginning: “Both the Prosecution and Defence sought a fine. I did not agree” (at [2]).

It is a simple statement, but it raises a genuinely interesting question. Our criminal justice system is adversarial. The Prosecution presents its case, the Defence advances its own submissions, and the court hears both sides before deciding. But what happens when there is no disagreement between the parties at all? The Prosecution sought a fine of $3,000 to $5,000 (at [9]); the Defence sought a fine of not more than $3,000 (at [10]). If Prosecution and Defence, having examined the same facts and the same law, independently arrive at the same recommendation, why is a court entitled to reject both and impose imprisonment instead?

At first glance the result looks counterintuitive. If two adversaries reach the same conclusion, that consensus would ordinarily carry real weight. One might even think a court should only depart from an agreed sentencing position in exceptional cases.

The answer has less to do with sentencing policy than with a basic feature of how criminal courts are constituted. To see why, it helps to first understand why the District Court concluded that imprisonment, rather than a fine, was called for on these facts.

The Chen Song Framework

The District Court did not simply ask what punishment felt right. Its brief reasons follow a structured analysis laid down by the High Court in Chen Song v Public Prosecutor and other appeals [2025] 3 SLR 509 (“Chen Song“) (at [11]).

Chen Song matters because, like other sentencing guideline judgments from our appellate courts, it exists to promote consistency and principled reasoning. Rather than leaving sentencing to broad instinct, it sets out a methodology sentencing courts are expected to follow, in two stages.

The first stage assesses harm: the seriousness of the victim’s injuries, whether surgery was needed, the length of hospitalisation and medical leave, and whether any disability is permanent.

The second stage assesses culpability, which is a different question entirely. Culpability is not about what happened to the victim, but about the quality of the offender’s own driving. Was this a momentary lapse of attention, or a deliberate disregard of traffic rules? Was the dangerous conduct fleeting, or sustained?

The distinction is important because harm and culpability can pull in different directions. Two drivers might cause identical injuries, yet one merely lost concentration for a second while the other knowingly ignored several rules in a row. If sentencing tracked injury alone, both would be punished the same, despite an obvious difference in blameworthiness. Equally, two drivers might behave with the same recklessness, but only one is unlucky enough to cause serious harm. Ignore culpability altogether, and sentencing becomes a matter of luck rather than conduct.

Chen Song requires both concepts to be assessed separately before they are brought together to determine the sentencing band. It is, in short, an attempt to make punishment reflect both what happened and how it happened.

Applying the Framework to Chen’s Case

On harm, the Court noted that although the victim suffered multiple fractures requiring surgery, two days of hospitalisation and ninety-nine days of hospitalisation leave, there was no permanent disability (at [13] to [17]). Two primary harm factors were engaged, but on a holistic assessment the Court found that lesser harm had been caused (at [18] to [20]), consistent with Chen Song‘s recognition that even where two or more primary harm factors apply, a holistic assessment may still yield a finding of lesser harm where they present themselves to a limited degree (Chen Song at [127]).

Culpability was the more significant discussion. The Court ultimately assessed overall culpability as falling within the lower category (at [28]), though that should not be misread as an absence of aggravating features. It reflects an overall assessment, not a denial that anything went wrong.

Within that lower category, the Court identified two offence-specific culpability factors it regarded as significant (at [23]). First, Chen crossed chevron markings while changing lanes (at [24], applying Chen Song at [131(b)(iv)]). Chevrons are not decorative; they separate traffic flows and mark zones that motorists should not cross, and disregarding them raises collision risk, particularly at expressway speeds. Second, the Court found a high degree of carelessness in how the manoeuvre unfolded (at [25], applying Chen Song at [131(c)]). It did not treat this as a split-second lapse. Having let one car in the adjacent lane pass, Chen continued filtering sharply across the third lane regardless, which the Court read as a sustained period of inattention rather than a momentary lapse, describing the conduct as “deliberately cavalier about certain mitigatable risks” (at [25]).

Whether one agrees with that characterisation is a separate question. What matters for present purposes is the reasoning structure: the Court was not merely describing facts. It was identifying culpability factors that Chen Song itself treats as significant.

Crossing the Custodial Threshold

Given the lesser harm and lower culpability, the case fell within Band 1 of Chen Song (at [29]). The next question was whether a non-custodial sentence remained appropriate.

Band 1 is often assumed to be a “fine category.” Strictly speaking, it is not. It is a sentencing range within which either a fine or imprisonment may be appropriate, depending on the overall circumstances. The framework frames the question. It does not dictate the answer.

Chen Song itself offers guidance here: where two or more offence-specific harm or culpability factors are present, the custodial threshold will typically be crossed (at [30], applying Chen Song at [137]). The word “typically” is doing real work. It is not automatic, and it is not mandatory. Judicial discretion remains, and an exceptional case may still justify a fine despite two factors being present. But as a general signal, two factors point the court towards imprisonment as the starting point.

Having found two or more offence-specific factors present, the District Court concluded the custodial threshold was crossed and adopted an indicative starting point of about one week’s imprisonment (at [31]).

From Starting Point to Sentence

A starting point is not a sentence. It is the beginning of an exercise, and the judgment shows that exercise being worked through with some care.

Two of Chen’s antecedents were given no weight at all. His compounded offences predated 1 November 2019 and were treated as irrelevant for sentencing purposes (at [43]), and a conviction for failing to give driver’s particulars was set aside as too dated to matter, having occurred three decades earlier, in 1996 (at [44]). A personal mitigation letter, in which Chen expressed willingness to support community initiatives on road safety, was noted (at [45]). The Court also gave due weight to Chen’s cooperation with the authorities, applying the approach in Public Prosecutor v Siew Boon Loong [2005] 1 SLR(R) 611 at [16] to [18] (at [46]).

The more consequential adjustment came from the guilty plea itself. Under the Sentencing Advisory Panel’s Guidelines for Guilty Pleas, discussed in Ng En You Jeremiah v Public Prosecutor [2025] SGHC 135 at [110] to [111] and [115] to [117] (at [47]), an early plea can attract a substantial discount precisely because it spares the criminal justice system the cost of a full trial, a principle traced to Angliss Singapore Pte Ltd v Public Prosecutor [2006] 4 SLR(R) 653 at [77] (at [48]). Chen’s plea was about as early as a plea can be. His then counsel, Mr Josephus Tan, had indicated at the very first mention that Chen was ready to plead guilty that day, and the Court accordingly gave him the full thirty percent discount available at the first stage of the guidelines (at [49]).

Applied to a starting point of one week, that discount produces five days almost exactly. The Court’s own summary confirms the arithmetic: “the starting point for the sentence was about 1 week’s imprisonment. After considering the mitigating factors and the PG discount, in the round, the sentence would be 5 days’ imprisonment” (at [50]). The mitigation exercise did not soften the underlying finding that this was a custodial case. It simply calibrated, transparently and by reference to established guidelines, how much credit Chen was owed for pleading guilty at the earliest opportunity. That transparency is itself worth noticing. A reader can trace the sentence from starting point to outcome without having to take the final figure on faith.

Why a Joint Position Does Not Bind the Court

This brings us back to the opening question. If both Prosecution and Defence asked for a fine, why was the Court entitled to impose imprisonment instead?

The answer lies in how our criminal justice system distributes responsibility. The Prosecution decides whether to charge, what to charge, and what sentence to seek. The Defence advances every legitimate argument available to the accused. But the responsibility for deciding the sentence itself belongs to the court, and to the court alone.

This is neither novel nor controversial. Singapore’s appellate courts have repeatedly emphasised that sentencing frameworks and guidelines exist to guide judicial discretion, not to replace it, and the same logic applies with even more force to the parties’ own submissions. A prosecutor may decide leniency is warranted and ask the court for a lighter sentence. Defence counsel will naturally support that position. But once those submissions are made, the question before the court is not whether a merciful outcome is desirable. It is whether the sentence proposed is one that the law permits on the facts as found.

Seen this way, the District Court’s decision is not really about disagreeing with counsel. It is about the sentencing function belonging to the court rather than to whichever position the parties happen to share. A joint sentencing recommendation by two experienced advocates deserves careful consideration, and clearly received it here. But consensus between the parties cannot substitute for the court’s own assessment of what the law demands, and a court that simply adopted a joint position without that independent assessment would not be exercising judicial discretion at all.

The High Court’s Calibration

The subsequent appeal is equally instructive. The High Court reduced the sentence from five days to one day, but did not go so far as to substitute a fine. Although the High Court’s own written grounds have yet to be released, this suggests the appellate court took a different view of the calibration within the custodial range, without necessarily disturbing the finding that the custodial threshold itself had been crossed.

That distinction is worth sitting with. A one-day sentence is not a token gesture. It preserves the legal conclusion that the offence warranted custody, while recognising that the mitigating circumstances justified the shortest term available. The difference between one day and five is not simply a matter of degree. It reflects the court’s calibration of culpability after accepting that imprisonment, rather than a fine, remained the correct sentencing category.

Conclusion

The District Court’s brief reasons expressly reserve full grounds for later release, should the need arise (at [4]). When and if they come, they will be studied closely by practitioners and academics, particularly for how they draw the line between a momentary lapse and sustained poor driving under the Chen Song framework.

But the broader lesson from this case does not depend on those grounds. It is that sentencing in Singapore is not decided by whoever happens to agree in the courtroom. Prosecutors, defence counsel and courts each exercise discretion within their own roles, but the final responsibility for declaring and applying the law rests with the court alone. Edmund Chen’s case is a reminder of that principle, not an exception to it.