Delay-Enabled Rehabilitation as Mitigation and the Limits of Unduly Lenient Sentence Review in Controlling or Coercive Behaviour Cases

Introduction

In Wilson, R. v [2026] EWCA Crim 427, the Court of Appeal (Criminal Division) (Lord Justice Stuart-Smith, Sir Robin Spencer, and His Honour Judge Marks KC) determined a reference by the Solicitor General under s.36 Criminal Justice Act 1988 alleging that a sentence for controlling or coercive behaviour was unduly lenient.

The respondent (the offender) had been convicted by a jury of a single count of controlling or coercive behaviour in an intimate relationship. He was acquitted of rape counts (one directed acquittal; two jury acquittals). The Crown Court imposed 18 months’ imprisonment suspended for 18 months, with 150 hours unpaid work and up to 35 rehabilitation activity requirement (RAR) days.

Two features dominate the appeal: (1) extreme delay (roughly 5 years to institute proceedings and over 6 years from end of alleged conduct to trial), accepted as unexplained and inexcusable; and (2) inadequate sentencing remarks, which left the appellate court unable to identify the factual basis for sentence or the Crown Court’s approach to suspension under the Imposition of Community and Custodial Sentences Guideline (effective 1 September 2025).

The court granted leave but ultimately declined to interfere, emphasising that its outcome was driven by the “particularly unfortunate history” and stating it should not be seen as setting any form of precedent.

Summary of the Judgment

  • The Court of Appeal criticised the Crown Court’s sentencing process as fundamentally inadequate: no clear findings of fact were made despite a non-specific jury verdict, and no explicit application of the suspension guideline was given.
  • Despite that criticism, the court held it could not be sure that the sentence should have exceeded 2 years (the practical threshold beyond which suspension becomes unavailable), because the lack of findings made the seriousness assessment uncertain.
  • The court accepted that (on the material it could safely rely upon) the maximum proper mitigation was about 9 months, leading it to view the imposed reduction as lenient and potentially at the outer edge of discretion.
  • Considering suspension afresh, the court held that suspension was nevertheless justified, largely because the prolonged delay coincided with substantial evidence of rehabilitation and lasting change, and because immediate custody would significantly harm the offender’s family.
  • The reference was therefore dismissed: the sentence was not declared unduly lenient.

Analysis

1) Precedents Cited

R v Barrett [2025] EWCA Crim 840

The court relied on R v Barrett [2025] EWCA Crim 840 for the orthodox explanation of when delay mitigates sentence. At paragraph [32], it quoted Barrett’s description of the “detrimental effect” envisaged by the guideline—typically where an offender suffers genuine psychological/physical harm from the delay, or has had to “put his life on hold” awaiting a prosecution decision.

Importantly, the court used Barrett as both a baseline and a contrast. It accepted that the respondent did not suffer the classic Barrett-type detriment. But it then identified a different mitigation mechanism: delay may be materially relevant because it creates the conditions for demonstrated rehabilitation such that immediate custody years later may be particularly counterproductive and arguably disproportionate in practical penal effect.

R v Timpson [2023] EWCA Crim 453

The court also cited R v Timpson [2023] EWCA Crim 453 as a caution against excessive percentage reductions for delay in serious cases, noting that a reduction “would be most unlikely to be as great as 25%” where offences are serious, though “some reduction” may still follow.

Timpson functioned as a constraint on the Crown Court’s broad-brush approach: the Court of Appeal regarded the implied reduction in this case (from a 2 years 6 months starting point down to 18 months, i.e. about 40%) as very difficult to justify on conventional grounds, especially absent evidence that the respondent’s mental health diagnosis reduced culpability in a sentencing-significant way.

R v Hussain [2019] EWCA Crim 1542

On suspension, the court relied on R v Hussain [2019] EWCA Crim 1542 for the proposition that even if a judge considers that “appropriate punishment would only be achievable by immediate custody”, there remains a discretion to suspend where there are sufficient countervailing factors.

Hussain therefore supported the analytical structure the court adopted: (a) identify the seriousness and the custody imperative; (b) then still ask whether the guideline factors (rehabilitation prospects, impact on others, compliance, etc.) justify suspension.

2) Legal Reasoning

(a) The Attorney General reference test and why the court could not intervene

Under s.36 Criminal Justice Act 1988, the Court of Appeal intervenes only where a sentence is “unduly lenient” (not merely lenient). A central obstacle here was epistemic: the appellate court could not safely determine the Crown Court’s factual findings because the judge:

  • made no explicit findings of fact despite a verdict that did not specify the precise conduct proved;
  • did not identify which prosecution allegations in the PSR were accepted, rejected, or treated with doubt;
  • gave sentencing remarks that were “extremely short” and did not explain the sentence’s derivation;
  • failed to articulate any reasoning under the Imposition Guideline when suspending custody.

This mattered because the unduly lenient inquiry is fact-sensitive: the more serious the established conduct, the more likely an uplift from the guideline starting point and the less likely suspension becomes. The absence of findings meant the court could not confidently say the judge had to impose more than 2 years’ custody (and therefore could not confidently say suspension was unavailable).

(b) Guideline categorisation: what could be inferred—and what could not

The parties treated the case as Category A1 under the controlling or coercive behaviour guideline, with a starting point of 2 years 6 months and a range of 1 to 4 years. The Court of Appeal accepted that the A1 categorisation was justified, but stressed that categorisation alone could not supply the missing detail:

  • Category A culpability may be met by one or more of: intent to maximise fear/distress; persistent/prolonged conduct; multiple methods. The court noted that defence counsel’s concession that the conduct was “persistent” likely aligned with “persistent action over a prolonged period”, but did not specify the broader course of conduct.
  • Category 1 harm may be met by various factors (fear of violence on many occasions; very serious alarm/distress with substantial adverse effect; significant psychological harm). The court could not tell which factor(s) the sentencing judge relied upon.

The result was a narrow set of certainties: the guideline bracket; a recognition of seriousness; and the existence of substantial delay. That was insufficient to prove the sentence necessarily crossed the “unduly lenient” threshold.

(c) Delay as mitigation: not “detriment”, but “evidence of rehabilitation”

A key analytical move appears at [33]. The court accepted that the respondent did not suffer the conventional kinds of delay detriment described in Barrett. Nonetheless, it treated delay as materially mitigating because it enabled compelling evidence that the respondent had turned his life around:

  • no further relevant offending over roughly 6.5 years;
  • positive steps to address anger-related issues (including seeking medical help and anger management);
  • a long-term stable relationship with strong partner support;
  • positive engagement with probation and employment (post-sentence information before the Court of Appeal).

This is not framed as creating a new doctrinal category of “delay mitigation”, but it is a practical clarification: even where delay is not shown to have harmed the offender in the Barrett sense, the passage of time may still weigh heavily where it demonstrates genuine, sustained rehabilitation and renders immediate custody years later particularly disruptive and arguably of reduced penal utility.

(d) Quantification of mitigation and the “2-year” suspension boundary

The court assessed that the maximum proper allowance for borderline personality disorder, delay, and rehabilitation was about 9 months ([34]), implicitly indicating that the Crown Court’s reduction (to 18 months) was excessive. However, because the appellate court could not be confident that an uplift above the starting point was required (given the factual uncertainty), it could not be confident that the correct sentence must exceed 2 years.

This mattered because suspension is generally available only up to 2 years’ custody. Once the possibility remained that a lawful sentence could be 2 years or less, it became harder—on the unduly lenient standard—to say the judge’s choice to suspend necessarily fell outside the range of permissible outcomes.

(e) Suspension under the Imposition Guideline

The court criticised the Crown Court’s failure to apply the guideline openly, then applied it itself. Starting from the premise that custody was unavoidable and would likely be between 21 and 24 months ([36]), it weighed the guideline factors:

  • Realistic prospect of rehabilitation: strongly supported by the PSR, long-term behavioural change, and probation engagement.
  • Strong personal mitigation: not overwhelming, but significant given the scale and consequences of the delay.
  • Impact on others: immediate custody would significantly harm the offender’s partner and two children.
  • Risk considerations: although the PSR suggested high residual partner-risk, the court considered the risk profile had changed materially.
  • Compliance: no history suggesting likely breach; recent compliance supported continued community management.

Applying R v Hussain [2019] EWCA Crim 1542, the court held that even if immediate custody might ordinarily be needed for punishment, the countervailing factors were sufficient to justify suspension—especially given the undesirability of immediate custody so long after the events, where the offender appears to have changed.

3) Impact

(a) Sentencing practice: reasons, findings, and guideline engagement

The decision reinforces (in emphatic terms) that where a jury verdict does not resolve key factual disputes relevant to sentencing, judges must:

  • state what facts are found and what is not relied upon;
  • explain how guideline category, starting point, adjustments, and mitigation were applied;
  • where suspension is imposed, explicitly apply the Imposition Guideline.

The practical consequence is procedural: inadequate remarks can prevent effective appellate review, including under s.36 references.

(b) Delay: a widened lens (without claiming a new rule)

While the court disclaimed precedential intent, its reasoning is likely to be cited as a reminder that delay may matter not only by “detriment to the offender” (Barrett), but also by producing reliable evidence of rehabilitation that changes the proportionality and utility of late custody—particularly when the delay is “inexcusable” and the offender has demonstrably reformed.

(c) Case management and fair labelling of the prosecution case

The judgment is also a cautionary tale about basic case preparation: the absence of a prosecution case summary and the lack of clear particulars contributed to “confusion” at trial and sentencing, and ultimately to uncertainty on appeal. Although the defence did not seek further particulars, the court’s narrative suggests that the criminal process works poorly—and review becomes difficult—when the alleged controlling/coercive course of conduct is not clearly articulated and then translated into sentencing findings.

Complex Concepts Simplified

Attorney General’s Reference (s.36 Criminal Justice Act 1988)
A procedure allowing the Attorney General (or Solicitor General) to ask the Court of Appeal to increase certain Crown Court sentences if they are “unduly lenient”. The threshold is high: the sentence must fall outside the range of sentences a judge could reasonably impose.
Controlling or coercive behaviour guideline categories (A1, etc.)
The Sentencing Council guideline uses culpability and harm categories. Category A indicates higher culpability (e.g., prolonged/persistent conduct, multiple methods, intent to maximise fear). Category 1 harm indicates the most serious harm impact. Each combination yields a starting point and range.
Why factual findings matter after a jury verdict
A jury often decides only “guilty” or “not guilty” on a broad charge and may not identify which allegations it accepted. The judge must then decide what conduct is proved for sentencing purposes (applying the criminal standard), because sentence depends on what happened.
Suspended sentence and the Imposition Guideline
A suspended sentence is custody that is not served immediately if the offender complies with conditions. The guideline requires the court to consider factors pointing towards suspension (rehabilitation prospects, impact on others, strong mitigation, etc.) and factors against (seriousness, risk, poor compliance history). Even where punishment suggests immediate custody, courts retain a discretion to suspend if the balance of factors justifies it.

Conclusion

[2026] EWCA Crim 427 is a sharply critical judgment about sentencing methodology: where the factual basis is unclear and suspension is in issue, findings and guideline reasoning are essential. Paradoxically, the very inadequacy criticised by the court also limited its ability to intervene on an unduly lenient reference.

Substantively, the court treated prolonged, unjustified delay as highly relevant not because it harmed the offender in the conventional sense described in R v Barrett [2025] EWCA Crim 840, but because it provided a long window in which rehabilitation was evidenced and immediate custody years later risked being both disruptive and of diminished sentencing value. Suspension was upheld after an explicit application of the Imposition Guideline, supported by R v Hussain [2019] EWCA Crim 1542.