R v Carr [2025] EWCA Crim 401: Post‑offence lies as aggravation and the permissibility of ‘cusp’ starting points under the 2023 “causing or allowing a child to die” guideline

Introduction

This Court of Appeal (Criminal Division) decision concerns an appeal against sentence brought by the mother of a two‑and‑a‑half‑year‑old child, Maya Chappell, who died following a sustained period of mistreatment culminating in a fatal assault by the mother’s partner. The appellant was convicted by a jury at Teesside Crown Court of:

Allowing the death of a child, contrary to section 5 of the Domestic Violence, Crime and Victims Act 2004 (count 4), and
Child cruelty, contrary to section 1(1) of the Children and Young Persons Act 1933 (count 3).

On 21 December 2023, Bright J imposed concurrent sentences of nine years’ imprisonment (s.5) and six years (s.1(1)), with credit for 168 days spent on remand. The co‑defendant, Michael Daymond, was convicted of murder and child cruelty, and received life imprisonment with a 20‑year minimum term.

The appeal challenged the categorisation and methodology under the Sentencing Council’s definitive guideline (effective 1 April 2023) for “Causing or allowing a child or vulnerable adult to die or suffer serious physical harm,” arguing that:

The case should have been placed squarely in medium culpability (Category C), not at the cusp of high (Category B) and medium;
The judge’s “notional starting point” of 7 years 6 months was too high or mechanically derived;
Post‑offence lies were improperly double counted; and
Insufficient weight was given to compelling personal mitigation (age, traumatic background, and mental health).

The Court of Appeal dismissed the appeal. In doing so, it clarifies two important sentencing points under the 2023 guideline: first, post‑offence lies and blame‑shifting may be treated as aggravating features without double counting where they did not inform the culpability category; and second, where a case lies “on the cusp” between culpability categories, a judge may select an intermediate notional starting point without adopting a mathematical formula—“sentencing is an art and not a science.”

Summary of the Judgment

The Court of Appeal held that the overall sentence of nine years’ imprisonment was severe but neither wrong in principle nor manifestly excessive (paras 27, 30–31). The trial judge’s assessment of culpability as being on the cusp between Category B (high) and Category C (medium) under the 2023 guideline was open to him, given his command of the trial evidence and evaluative judgments involved (paras 27–28).

The Court endorsed the judge’s approach of adopting a notional starting point of 7 years 6 months before aggravating and mitigating adjustments (para 16), noting that such a figure sat within the Category 1C range in any event and that the guideline invites an adjustment where a case does not fall squarely within a single category (para 28).

Importantly, the Court rejected the double‑counting complaint: the judge was entitled to treat the appellant’s persistent lies and blame‑shifting as aggravating features because those matters are not among the culpability determinants distinguishing Category B from Category C (para 29). The Court accepted that the judge had proper regard to the appellant’s age, prior good character, and background but concluded that, given the gravity of the offending, the final sentence could not be said to merit appellate intervention (paras 16, 30).

Factual context

The abuse occurred over approximately five weeks after the appellant and her partner, Daymond, moved in together with Maya in late August 2022. The appellant worked most days; Daymond cared for Maya day‑to‑day. Multiple bruises were noticed by others, including a large and pronounced bruise over the pubic mound, which experts later concluded could only have been caused by forceful contact (para 6).

On 28 September 2022, amidst financial stressors and drug‑debt pressure on Daymond, Maya suffered unsurvivable brain trauma caused by violent shaking and severe internal abdominal injuries; her body bore numerous bruises (paras 10–11). The jury convicted Daymond of murder and child cruelty, and convicted the appellant of allowing Maya’s death and child cruelty.

Analysis

Statutory and guideline framework

Two statutory regimes frame the sentencing exercise:

Section 5 of the Domestic Violence, Crime and Victims Act 2004 (causing or allowing the death of a child or vulnerable adult). For offences committed after 28 June 2022, the maximum is life imprisonment (para 12).
Section 1(1) of the Children and Young Persons Act 1933 (child cruelty).

The Sentencing Council’s definitive guideline (effective 1 April 2023) applies to s.5. Where a child dies, harm is inevitably Category 1. Culpability is graded A–D; the relevant gradations were:

Category B (high culpability), including: prolonged and/or multiple incidents of serious cruelty; use of very significant force; deliberate disregard for welfare; failure to take any steps to protect the victim where those factors are present.
Category C (medium culpability), including: significant force; prolonged and/or multiple incidents of cruelty; limited steps to protect the victim in cases with Category B factors present; and cases falling between B and D (para 12).

The guideline provides Category 1 starting points and ranges:

1B: starting point 9 years’ custody; range 7–14 years.
1C: starting point 5 years’ custody; range 3–8 years (para 13).

Critically, the guideline (in bold) invites courts to adjust the starting point where a case does not fall squarely within a category before then applying aggravating and mitigating factors (para 13).

Precedents and authorities cited

The Court did not cite case authorities by name. It applied settled appellate principles on sentence review—namely, that intervention is warranted only if a sentence is wrong in principle or manifestly excessive—and reiterated the often‑stated proposition that “sentencing is an art and not a science” (paras 27–28). The analysis was anchored in:

The 2023 definitive guideline for s.5 offences; and
The structure and language of the guideline concerning culpability categories, starting points, ranges, and the “non‑mathematical” nature of adjustment.

The Court also relied on the statutory framework increasing the maximum penalty for s.5 to life for post‑June 2022 offences (para 12).

The Court’s legal reasoning

The appellant’s central complaint was that the judge wrongly placed the case at the cusp between Categories B and C. The defence highlighted a passage in the sentencing remarks where the judge had described pre‑28 September force used by Daymond as “significant but not serious” (para 18), arguing that this indicated medium culpability. The Court of Appeal rejected that reading, emphasising context: the judge had earlier identified a serious, non‑accidental pubic‑mound injury indicative of a kick or forceful impact on a hard surface (para 19), alongside multiple bruises and the pattern of mistreatment (paras 6, 11).

Given the evaluative nature of culpability assessment in s.5 cases—particularly where failure to protect is alleged and the abuse unfolded over weeks—the Court held it was for the trial judge to balance the indicators of high culpability (e.g., multiple incidents; serious injury signalling very significant force; failure to take steps to protect) against any features pulling towards medium. Placing the case “on the cusp” was a course reasonably open to him (para 27).

On method, the Court endorsed the judge’s adoption of a notional starting point of 7 years 6 months before adjusting for aggravation and mitigation (para 16). It stressed the guideline’s explicit invitation to calibrate the starting point where the case does not fit neatly within a category, and rejected a mathematical or “mid‑point” approach between the 9‑year (1B) and 5‑year (1C) starting points (paras 13, 28). The chosen notional figure also fell within the 1C range (3–8 years), underscoring the legitimacy of a flexible, judgment‑driven selection (para 28).

Aggravating features and the “double counting” complaint

The Court treated the appellant’s persistent lies and blame‑shifting—including attempts to blame her own mother—as aggravating features, and held this did not constitute double counting (paras 14, 21, 26, 29). The reason is structural: those post‑offence behaviours are not among the culpability determinants that mark the line between Categories B and C in the 2023 guideline. In other words, lies did not influence the category placement and, therefore, could be considered at the aggravation stage without duplication.

The Court also accepted the presence of wider non‑statutory aggravating features identified by the Crown—deliberate concealment, blaming others, and failure to respond to warnings (para 26)—all of which bear on the seriousness and persistence of the offending, particularly in a case involving visible injuries and observable fear in the victim (paras 8–11).

Mitigation: youth, background, and mental health

The appellant relied on her age (23 at the time), prior good character, a traumatic upbringing featuring parental alcohol misuse and domestic violence, and psychiatric evidence of PTSD symptoms and depression (paras 15–16, 22). The trial judge acknowledged these features and accepted their relevance, including the appellant’s emotional immaturity and susceptibility to dependent relationships (para 15). The Court of Appeal concluded that these matters were properly taken into account but, given the gravity and duration of the mistreatment and the catastrophic outcome, they could not justify reducing the sentence below nine years (para 30).

Sentencing methodology: “cusp” cases and non‑mathematical calibration

A notable feature of the judgment is its express endorsement of what might be called “cusp calibration”:

Where culpability indicators straddle Categories B and C, judges may adopt a notional starting point between the two category starting points before considering aggravation and mitigation (paras 13, 16, 28).
The exercise is not arithmetical. There is no requirement to select a strict numerical midpoint or to confine the notional starting point to one category’s nominal figure (para 28).
The chosen notional point should be justified by reference to the guideline’s descriptors and the particular facts, including the seriousness and frequency of injuries and the offender’s failure to protect (paras 12–13, 27–28).

The reaffirmation that “sentencing is an art and not a science” (para 28) underscores an evaluative, purposive application of the guideline rather than a rule‑bound computation.

Totality and concurrency

It was common ground that count 4 (s.5, allowing death) was the lead offence and that the sentences on counts 3 and 4 should be concurrent (para 2). The total sentence therefore reflected the harm and culpability associated with the s.5 offence. The Court of Appeal found no fault with this structure.

Practice points

Guideline fit: In s.5 cases, the court should first identify harm (Category 1 where a child dies) and then determine culpability. Where features straddle categories, it is permissible to place the case “on the cusp.”
Notional starting points: Selecting an intermediate notional starting point is acceptable—and encouraged by the guideline—where the case does not fit squarely within a category. Avoid mechanistic arithmetic.
Aggravation vs culpability: Keep a clear separation of stages. If a factor (e.g., persistent lies, blaming others) does not inform category selection, it can be used to aggravate without double counting.
Record the reasoning: Sentencers should explain the basis for “cusp” placement and for selecting the notional starting point, then detail aggravating and mitigating adjustments.
Mitigation: Age, background, and mental health are relevant, but their weight must be balanced against the gravity, duration, and foreseeability of harm in child protection cases.

Impact and future significance

This decision is likely to influence sentencing in three concrete ways:

Clarity on “cusp” methodology. The Court’s approval of non‑mathematical calibration between categories under the 2023 guideline gives practical latitude to sentencers dealing with complex, pattern‑of‑abuse fact patterns common in “allowing” cases.
Aggravation by post‑offence conduct. The explicit confirmation that persistent lies and blame‑shifting may aggravate without double counting (where they did not drive the culpability category) will guide sentencing remarks and sharpen submissions about the stage at which such conduct is addressed.
Firmness in child protection sentencing. By characterising nine years as “severe” yet within permissible bounds, the Court signals continuing support for robust sentences where carers fail to protect amidst obvious warning signs, particularly after the 2022 increase of the statutory maximum to life imprisonment.

More broadly, the decision promotes principled flexibility in applying structured guidelines. It underscores the importance of qualitative evaluation over arithmetic in complex, fact‑sensitive sentencing exercises.

Complex concepts simplified

Causing or allowing the death of a child (s.5 DVCVA 2004): A person can be guilty either by directly causing death or by allowing it—i.e., being aware (or ought to have been aware) of a significant risk of serious physical harm and failing to take reasonable steps to protect the child.
Harm Category 1: Under the guideline, if the victim dies, harm is automatically Category 1.
Culpability categories (A–D): These describe how blameworthy the offender is, based on factors like duration and seriousness of cruelty, degree of force, disregard for welfare, and steps (or lack of steps) taken to protect the victim. Category B = high; Category C = medium.
Starting point vs range: The starting point is the benchmark sentence for a typical case within a category. The range shows the permissible spread above and below the starting point, depending on aggravating and mitigating factors.
“Cusp” cases: Where features point both to a higher and a lower culpability category, the judge may select a notional starting point between the two before applying aggravation/mitigation.
Double counting: Using the same factor twice to increase sentence severity (e.g., once to place the case in a higher culpability category and again as an aggravating feature). The Court held that treating lies as aggravation here was not double counting because those lies did not determine culpability category.
Manifestly excessive/wrong in principle: The appellate test for interfering with a sentence. A severe sentence is not necessarily unlawful; it must be outside the proper range or based on an error of approach.
Concurrent sentences and lead offence: Where multiple offences arise from the same episode, sentences often run concurrently, with the overall term set by the most serious (the lead offence).

Conclusion

R v Carr cements two practical, principled points in the application of the 2023 “Causing or allowing a child to die” guideline. First, post‑offence lies and blame‑shifting can properly aggravate sentence without double counting when they did not inform culpability categorisation. Second, where a case lies between culpability categories, a judge may adopt a non‑mathematical, intermediate notional starting point before applying aggravating and mitigating factors.

The Court of Appeal’s refusal to disturb a severe nine‑year term highlights the gravity with which the courts will treat sustained mistreatment and failures to protect a child where warning signs are ignored. It also reinforces a deference to the trial judge’s evaluative judgments in complex sentencing exercises and the flexible—but structured—use of Sentencing Council guidelines. For practitioners and sentencers alike, Carr is an important reference point on the mechanics of guideline application in s.5 cases and the proper treatment of post‑offence conduct.