3.1 Precedents Cited (and How They Shaped the Decision)
R. v Horncastle [2009] UKSC 14; [2010] 1 Cr. App. R. 17 (and [2009] EWCA Crim 964; [2009] 2 Cr. App. R. 15)
The Court treated Horncastle as the constitutional and practical foundation for modern hearsay admissibility under the CJA 2003.
It adopted the paired formulation highlighted in Riat: hearsay is acceptable where it is either
“demonstrably reliable” or where its reliability is capable of proper testing and assessment.
The present case fell into the second category: the statements were not “demonstrably reliable” on every issue (not least because of an inconsistency),
so the correct question was whether the jury would have sufficient material to test and assess reliability safely.
Riat provided both the governing approach and the practical checklist.
The Court of Appeal in the present case did not depart from Riat;
rather, it recast the six steps to avoid the kind of “tunnel vision” that can arise if step (ii) (testing material) is treated too narrowly as “s.124 material only”,
and to ensure disclosure is operationally foregrounded.
The Court relied particularly on Riat at [18] (the expectation of “very full enquiries” and disclosure where prosecution hearsay is important),
and treated the prosecution’s failure to deploy obvious testing material (notably body-worn video first accounts) as a serious deficiency.
R v Ali [2024] EWCA Crim 77
The Court addressed a practical point: judges often rule on admissibility and, in the alternative, on exclusionary discretion.
It confirmed this is legitimate and often desirable.
Any contrary impression drawn from Ali should be corrected; the true difficulty in Ali was said to relate mainly to jury directions, not to the permissibility of alternative rulings.
R v.TM [2020] EWCA Crim 1343 [23]
TM reinforced that the section 125 evaluation must be conducted having regard to all the evidence.
The Court used this to justify a more confident approach at the admissibility stage:
the judge need not assume the worst about future testing at trial, because if—after all evidence and tailored directions—the hearsay remains so unconvincing that conviction would be unsafe,
s.125 supplies a mandatory safety valve.
3.2 Legal Reasoning
(a) The appellate standard under CJA 2003 s.67
The Court reminded itself it could only reverse the hearsay ruling if satisfied it was:
(a) wrong in law, (b) involved an error of law or principle, or (c) was not a ruling it was reasonable to make.
It ultimately concluded the ruling failed both as a matter of legal approach and reasonableness.
(b) The “critical flaw”: underweighting independent support and the real issues
The trial judge focused on inconsistencies and credibility deficits in the deceased witnesses’ accounts—particularly the later “change of account” about why the defendants were at the flat.
The Court of Appeal accepted the inconsistency was significant but held the judge misapplied the correct question:
could reliability be properly tested and assessed by the jury, in light of all admissible material and the actual issues in dispute?
Central to the Court’s reasoning was the interaction between:
-
Defence statements (used legitimately to define the disputed issues): the defendants largely accepted they had been in the flat for days and had travelled together.
This dramatically reduced the live significance of “stranger identification” risk.
-
Independent dovetailing/supporting evidence (videos, forensic links, car evidence, and other case features) supporting key elements of the hearsay and enabling assessment of whether the inconsistent part was a lie, a correction, or contextual reframing.
-
The fact that the changed account went largely to motive/context rather than the core question in several defences (participation/duress/role),
so its existence did not automatically render the identification/participation evidence incapable of assessment.
(c) Proper use of “weaknesses” in hearsay
The Court made a subtle but important point: the judge identified flaws that were apparent on the face of the statements.
Such flaws are not necessarily a reason to exclude; they may be reasons why the hearsay is capable of being tested—because the jury can be directed on them and evaluate them.
Exclusion is not the default response to credibility issues; section 124 exists precisely so credibility attacks can be deployed despite the maker’s absence.
(d) Disclosure failures and their correct procedural significance
The Court was critical of the prosecution’s failure to assemble and disclose obvious reliability-testing material,
including first accounts captured on body-worn video and materials explaining the circumstances of the later statement.
While acknowledging that such failures might, in an appropriate case, justify refusal of admission, the Court stressed that the correct course is often to
consider whether the interests of justice require directions for a proper disclosure process rather than immediate exclusion.
(e) The error in the judge’s alternative exclusion reasoning (PACE s.78 / CJA s.126)
The judge said that if wrong about the core “testing and assessment” conclusion, he would nevertheless exclude for similar reasons under PACE s.78 / CJA s.126.
The Court found this internally inconsistent: if the court is wrong and the hearsay can be safely tested and assessed, the judge must identify some additional, coherent basis for discretionary exclusion.
Simply repeating the rejected reasoning is not a proper step (v) analysis.
(f) A court-driven s.125 safety culture
The Court emphasised that section 125 is a statutory duty and should be considered by the court even if not raised by the parties.
It suggested it will commonly be best assessed at the end of the evidence (often close to when tailored directions are being finalised), so that “safety” is assessed
in light of the full evidential picture and the actual directions the jury will receive.