Green v R: Misapplication of the 2003 Consent Standard to Pre-2003 Offences and the Boundaries of Cross-Admissibility Directions

1. Introduction

The Court of Appeal’s decision in R v Green ([2025] EWCA Crim 923) revisits two fundamental areas of criminal trial practice: (i) the correct test for consent in historic sexual offences tried under the Sexual Offences Act 1956 (“the 1956 Act”), and (ii) when a trial judge must give a cross-admissibility/propensity direction where multiple complainants give evidence.

Peter Green, a 93-year-old appellant, sought to overturn convictions for rapes allegedly committed against an adult partner (C1) during the 1990s, and for sexual offences against her daughter (C2) when the latter was a child. The appeal turned on three core complaints:

  • Misdirection on consent – the judge told the jury to apply the stricter “reasonable belief” test from the Sexual Offences Act 2003 (“the 2003 Act”) instead of the “genuine (even if unreasonable) belief” standard applicable under the pre-2003 law.
  • Failure to give a formal cross-admissibility/bad character direction.
  • Lack of an express instruction that the jury should ignore the judge’s own comments if they disagreed with them.

Only the first complaint succeeded, leading to the quashing of the convictions on counts 1 and 2 (rapes of C1). Counts 3–8 (offending against C2) were upheld. No retrial was ordered.

2. Summary of the Judgment

  • Misdirection on Consent: Because the trial judge used the 2003 Act test, the jury might have failed to consider whether Mr Green genuinely (albeit unreasonably) believed C1 was consenting, as required by DPP v Morgan and s.1(2) of the 1956 Act. The Court could not be sure the verdicts on counts 1-2 were safe, so those convictions were quashed.
  • Cross-Admissibility: On the specific facts—mother-and-daughter complainants who had discussed the allegations, very different types of offending, and a defence theory of joint fabrication—the standard “treat each count separately” direction was adequate. No error, and no impact on counts 3-8.
  • Judicial Comments: Although best practice is to tell jurors they may disregard any judge’s comments, failure to do so did not render the trial unfair in this case.
  • Sentence and Retrial: The aggregate 24-year sentence was substituted with a 12-year custodial term (plus one-year extended licence) on the surviving counts. Given the appellant’s age and the passage of time, the Crown did not seek—and the court refused—a retrial or a restraining order on acquittal.

3. Analysis

3.1 Precedents Cited and Their Influence

  • DPP v Morgan [1976] AC 182 – House of Lords authority establishing that a genuine belief in consent, however unreasonable, suffices to negate mens rea for rape under the 1956 Act. This was the lynchpin for finding the trial judge had used the wrong test.
  • Sexual Offences (Amendment) Act 1976 – Statutory codification of the Morgan principle via s.1(2) (knowledge or recklessness as to consent).
  • R v H [2011] EWCA Crim 2344; R v Adams [2019] EWCA Crim 1363; R v AHC [2022] EWCA Crim 925 – trio of authorities on when cross-admissibility directions are mandatory. Green reconciles apparent tension by reaffirming the “fact-sensitive” approach: sometimes the orthodox separate-counts direction suffices, sometimes more is needed.
  • Bernard v The State of Trinidad and Tobago [2007] UKPC 34 – reminder that not every lapse in best practice renders a conviction unsafe; the appellate question is always overall safety.

3.2 The Court’s Legal Reasoning

  1. Consent Misdirection
    The Court accepted a concession from the Crown that the wrong legal test had been given. The crux was whether that error mattered. Because C1 and Green were in a consensual sexual relationship, there was some evidence—such as C1 pretending to climax—that could support a genuine belief in consent. Therefore, applying Hanson-style reasoning on misdirections, the Court could not conclude the verdicts were bound to be the same had the jury been properly directed.
  2. Cross-Admissibility
    Adopting the approach in H and endorsed in AHC, the Court reasoned that (a) the mother-daughter relationship and acknowledged discussions negated the “multiple independent complainants” logic; (b) the offending was of different types; (c) the defence needed to contrast the two sets of allegations to bolster its fabrication theory. A propensity or “mutual corroboration” direction would have undermined that defence. Hence, no miscarriage.
  3. Bad Character Evidence
    Most of the complained-of material was either “integral to the facts” (s.98 CJA 2003) or admitted by agreement. Because the jury’s task was essentially binary—believe the complainants or believe the defence’s fabrication claim—further refinement was unnecessary.
  4. Judicial Intervention
    The Court distinguished between a failure to invite the jury to disregard judicial comments (undesirable but not fatal) and an appearance of bias in the comments themselves (absent here).

3.3 Potential Impact of the Decision

  • Historic Sexual Offences Trials: Counsel and judges must scrupulously apply the correct statutory framework governing the period of alleged offending. Green is a salutary reminder that the 2003 Act standards cannot be retro-fitted to pre-2003 conduct.
  • Appeal Strategy: Identification of a material misdirection—especially an admitted one—will often justify an out-of-time appeal, even decades later, as here.
  • Cross-Admissibility Directions: The judgment tempers the enthusiasm generated by Adams. It confirms that the necessity for propensity directions remains fact-specific, and standard separate-counts guidance may still be adequate in many multiple-complainant cases.
  • Restraint Orders on Acquittal: The Court’s refusal (age, absence of recent contact, existing sentence) illuminates the high “necessity” threshold in s.5A Protection from Harassment Act 1997.

4. Complex Concepts Simplified

  • Consent Tests Pre- and Post-2003:
    • Pre-2003 (1956 Act): A defendant is not guilty of rape if he genuinely believed the complainant consented, even if that belief was unreasonable (Morgan).
    • Post-2003 (2003 Act): The belief must be both genuine and reasonable. Unreasonable beliefs no longer excuse.
  • Cross-Admissibility / Propensity Evidence: When more than one complainant testifies, the jury may (if properly directed) treat similarities as showing the defendant’s tendency or as reducing coincidence. Such directions must be handled carefully to avoid unfair prejudice.
  • Bad Character Evidence (CJA 2003): “Bad character” means evidence of misconduct other than the offence charged. Section 101 sets out gateways for admissibility; s.98 carves out conduct forming part of the indictment facts.
  • Unsafe Conviction Test: The Court of Appeal must quash if it thinks a conviction is unsafe. A misdirection does not automatically = unsafe; it must be potentially material to the verdict.

5. Conclusion

Green v R underscores two practical imperatives. First, historic sexual offence trials demand doctrinal accuracy: using the modern “reasonable belief” yardstick for 1990s conduct was enough to unsettle rape convictions, despite otherwise cogent evidence. Second, directions on cross-admissibility remain a matter of judicial discretion tethered to the facts—not an obligatory formula triggered by every multi-complainant indictment. The Court’s measured approach balances fairness to defendants with the realities of complex sexual offence trials, while its restraint-order analysis evidences ongoing sensitivity to proportionality where elderly or infirm appellants are concerned.

Practitioners should take from Green that meticulous attention to the temporal reach of substantive criminal statutes and to the nuanced deployment of jury directions is not mere pedantry but can be outcome-determinative.