Sentencing Guidelines Are Not “Watertight Compartments”: Holistic Categorisation, Totality, and the Protective Purpose of Extended Sentences

1) Introduction

In R v Eniola Aminu ([2026] EWCA Crim 637), the Court of Appeal (Criminal Division) dismissed an appeal against an Extended Determinate Sentence imposed at Woolwich Crown Court. The appellant had pleaded guilty to robbery and three counts of assault occasioning actual bodily harm (ABH), arising from two separate episodes of serious violence four months apart.

The appeal raised four issues: (i) whether the sentencing judge miscategorised the robbery under the Definitive Guideline of the Sentencing Council on Robbery - Street and Less Sophisticated Commercial; (ii) whether aggravating/mitigating factors were misweighted; (iii) whether the overall term breached the totality principle; and (iv) whether a 5-year extended licence period was excessive, particularly in light of the appellant’s need for programmes and mental health management.

2) Summary of the Judgment

  • The Court accepted the sentencing judge had mis-categorised the robbery when viewed in isolation. Properly analysed, culpability was B and harm 3, i.e. B3 (starting point 2 years; range 1–4 years), rather than the category used by the judge.
  • Despite that error, the Court held the overall outcome was unchanged: the judge was entitled to select the robbery as the lead offence and to enhance its sentence to reflect the other offences, provided totality was respected. The resulting overall sentence was not manifestly excessive.
  • The Court rejected challenges to totality and to the balancing of aggravating/mitigating factors.
  • The Court upheld the 5-year extended licence period, emphasising that the purpose of an extended sentence is public protection, not to mirror the anticipated duration of treatment or offender programmes.

3) Analysis

3.1 Precedents Cited

The judgment does not turn on prior appellate authorities by name. Instead, it is structured around the correct approach to:

  • The Definitive Guideline of the Sentencing Council on Robbery - Street and Less Sophisticated Commercial (categorisation by culpability and harm; weighing factors across categories).
  • The totality guideline (ensuring a just and proportionate overall sentence across multiple offences).
  • The statutory framework for dangerousness and extended sentences under section 279 of the Sentencing Act 2020.

The Court’s contribution is therefore best understood as an appellate clarification of method in applying these instruments, rather than an extension of case-law lines.

3.2 Legal Reasoning

(a) Guideline categorisation must be holistic, not artificially segmented

A central theme is the Court’s warning against “artificial analysis” designed to “lever” facts into a category. Although the robbery guideline states the category should be determined “with reference only to the factors in the tables”, it also directs the court to weigh factors where more than one category is engaged. The Court treated that instruction as demanding a realistic, whole-event evaluation.

Applying that approach, the Court broke the incident into a sequence (snatch; scuffle; attempted punch; demand for return; threat to use an item from a bag) and held that the post-taking threat was part of the seriousness of the overall event, not something to be excluded because it occurred after the initial taking. The Court concluded:

  • Culpability fell between A and C and therefore into B.
  • Harm was Category 3 because there was no physical injury, despite clear distress and deliberate damage to the phone.
  • Overall: B3, with significant upward movement justified by aggravation (including relevant previous convictions and the fact the appellant was subject to a community order at the time).

(b) A categorisation error may be immaterial if totality and structure justify the same overall term

The Court accepted the robbery was miscategorised “standing alone” but treated that as non-dispositive. The key was the sentencing judge’s lawful discretion over structure:

  • A judge may adopt a lead-offence approach and “enhance” the lead sentence to reflect other offending, rather than (or instead of) consecutive terms, provided totality is kept “front and centre”.
  • The Court evaluated the seriousness of the accompanying ABH offences (including prolonged, unprovoked violence against two vulnerable middle-aged victims in one episode; and a further ABH shortly after the robbery involving a bag containing a metal bar), and held that the overall pre-plea sentence of 7½ years after trial was within the proper range.
  • After one-third credit for guilty pleas, 5 years’ custody was justified, and therefore the appeal failed even though the Court would have placed the robbery in a different guideline category.

In short: the Court treated guideline categorisation as important, but not an end in itself; the appellate question remained whether the overall sentence was manifestly excessive.

(c) Extended sentence: the extension period is about protection, not programme-length

On dangerousness, it was conceded that the statutory test was met. The dispute was the exercise of discretion to impose (and fix the length of) the extended licence period.

The Court emphasised the “whole point” of an extended sentence: enhancing public protection against the risk of serious harm. It expressly rejected the framing that the licence period should be calibrated to how long it might take the offender to complete treatment or offender programmes. On the facts—previous relevant convictions (including robbery with a knife), escalating violence against multiple victims, and evidence of erratic engagement with mental health services and drug misuse—the Court held a 5-year extension was “just and proportionate”.

3.3 Impact

  • Guidelines application: The decision reinforces that guideline categories are tools for structured judgment, not rigid “compartments”. Sentencers (and appellate courts) should resist overly technical slicing of fast-moving incidents where the overall seriousness emerges from the full interaction.
  • Appeals based on categorisation errors: A proven guideline categorisation error will not necessarily yield a reduced sentence if the sentencing court’s overall structuring (lead offence, enhancement, concurrency/consecutivity) achieves a proportionate total outcome.
  • Extended sentences and rehabilitation arguments: The judgment signals that “programme duration” arguments have limited traction against an extended licence period where the court is focused on prospective risk management and public protection.

4) Complex Concepts Simplified

  • “Culpability” and “harm” categories (robbery guideline): the court places the offence into a box that best matches (i) how blameworthy the conduct was (culpability) and (ii) the injury/loss/impact intended or caused (harm). Where features point in different directions, the court must weigh them and pick the closest fit.
  • Totality: when sentencing for multiple offences, the final overall sentence must be proportionate to the total criminality. Courts may use concurrent or consecutive sentences, or a lead-offence uplift, but must avoid a mechanistic “add up”.
  • Extended Determinate Sentence / extended licence: a custodial term plus extra time on licence designed to protect the public where the offender is “dangerous”. The extension is not a therapeutic timetable; it is a risk-management period.
  • “Manifestly excessive”: the appellate test is not whether the appeal court would have sentenced differently, but whether the sentence is outside the range of reasonable sentences open to the judge.

5) Conclusion

[2026] EWCA Crim 637 confirms three practical propositions: (1) sentencing guidelines must be applied with realistic, holistic judgment rather than technical compartmentalisation; (2) a guideline categorisation error may be immaterial where a lead-offence structure and totality analysis produce a proportionate overall sentence; and (3) extended licence periods under section 279 of the Sentencing Act 2020 are justified by public protection, not by estimates of how long rehabilitation interventions may take.