Smith v R.

The prosecution has a discretion whether to call or tender a witness, but it must be exercised to secure a fair trial. It should normally call witnesses with direct evidence of primary facts unless their evidence is reasonably regarded as incredible. The prosecution is not required to call a witness merely to provide material for attacking the credibility of its own witnesses, and an appellate court will generally intervene only where the prosecutor has erred in principle.

Fresh evidence will not render a conviction unsafe unless it is credible and materially undermines the evidence supporting guilt. The court must assess its likely effect in the context of the whole case.

The appellant appealed against his 1997 conviction for murder at Norwich Crown Court. The Crown case relied substantially on evidence that the appellant and his co-defendant had confessed to Paul Courtman and Steven Sparkes.

Adam Crossland had made statements alleging a further confession by the appellant, but later alleged that Courtman had pressured him to give false evidence. The Crown did not call Crossland at trial, although he was available to the defence. The appellant relied on Crossland’s later evidence, further psychiatric material concerning the co-defendant, alleged witness-management failures, a corrected misdirection, and uncalled good-character evidence.

The central issue was whether those matters, individually or cumulatively, rendered the conviction unsafe.

1. Appeal dismissed. The court was satisfied that the conviction was safe.

2. The trial judge correctly permitted the Crown not to call Crossland. Applying the principles stated in R v Russell-Jones [1995] 1Cr App R 538, the prosecutor’s discretion was constrained by the need for a fair trial, but a witness need not be called merely to supply material for an attack on the Crown’s own case. The Crown had made Crossland available to the defence. Although the court expressed surprise that the prosecution had not obtained a further statement after his letter, that did not make the decision wrong in principle.

3. The fresh evidence confirmed that Crossland was not credible. Had he given evidence, he would have been cross-examined on his detailed earlier statements. That would have damaged, rather than assisted, the appellant’s case. The prosecutor’s judgment and the judge’s ruling were therefore correct.

4. The later psychiatric material concerning Groves did not materially undermine the evidence identifying the appellant as the other participant. It could reinforce the conclusion that Groves had participated, and the jury had already received some evidence of his mental health. Later assertions by Groves that he alone committed the murder did not make the appellant’s conviction unsafe.

5. The complaints concerning Sergeant Stansbury, the corrected reference to material from Crossland’s statement, and uncalled good-character witnesses did not affect the safety of the conviction. Considered individually and cumulatively, none justified intervention.

• Court of Appeal (Criminal Division): Appeal against conviction dismissed; the conviction was held safe.
• Crown Court at Norwich: The appellant was convicted of murder before Judge Hyam and a jury on 13 October 1997.