'It's better that 10 guilty men go free than one innocent man be wrongly convicted'

The country's top appeal judges are failing to correct miscarriages of justice where they suspect the jury has come to a wrong verdict, the head of the body charged with investigating wrongful convictions has warned.

Professor Graham Zellick, the outgoing chairman of the Criminal Cases Review Commission (CCRC), said the Court of Appeal should order retrials in cases that have a "lurking doubt" about the safety of the conviction.

In an interview with The Independent, Professor Zellick also called on judges to prevent "very dubious" expert evidence, including lip-reading and ear-prints, being presented to the jury. He argued: "It is far better that 10 guilty men go free than one innocent man is wrongfully convicted ... We know from bitter experience that juries get things wrong. The Court of Appeal ought to be more active in quashing convictions even though there has not been any irregularity in the trial process."

He added that when he had raised this argument with members of thejudiciary he had been "admonished" for asking judges to second-guess the jury. "They tell me that in this country we have trial by jury, so who are they to go behind the verdict of the jury which has seen all the evidence? Well, I say we have trial by judge and jury, not just jury."

Professor Zellick, pictured below, who has been in charge of the commission's referrals to the Court of Appeal for the past five years, said a more interventionist approach would allow the court to order a retrial when judges were unhappy about the safety of a conviction. "The Court of Appeal is even more reluctant in 2008 than in the 1990s to quash convictions because they think they are unsafe. We are more deferential to a jury now than in the 1990s when things were going wrong," he said.

Professor Zellick, a professor of law at University College London, cited the wrongful conviction of the solicitor Sally Clark, 42, as an example of the consequences of judicial reluctance to free immediately an innocent victim of a miscarriage. Mrs Clark was jailed in 1999 for killing her 11-week-old son Christopher in December 1996, and eight-week-old Harry in January 1998. An appeal in 2000 failed, but she was freed in 2003 after a fresh appeal following a referral from the CCRC. She was found dead last year. "Sally Clark should never have been convicted," said Professor Zellick. "She should have succeeded at her first appeal. It should never have taken two years' work by us [CCRC] and a referral before she was released, by which time she was broken in mind and body." The jury at her trial was told by an expert witness, Professor Sir Roy Meadow, that the probability of two natural unexplained cot deaths in a family was 73 million to one. Other experts said the odds were about 200 to one. Although not criticising the standing of the expert witnesses in the case, Professor Zellick said juries were not always capable of deciding between diverse expert opinions. "There have been miscarriages of justice caused by experts whose expertise is somewhat suspect. We are too casual about expert evidence in the criminal justice system." He said he was particularly concerned about lip-reading and ear-print evidence, which he described as "very dubious" for a jury.

The tragedy of the case of Sally Clark, who died last year after finally being cleared of killing her two baby sons, was exacerbated by the failings of the criminal appeals process. The 42-year-old solicitor spent nearly four years behind bars. By the time the Criminal Cases Review Commission had intervened and sent the case back to the Court of Appeal, Mrs Clark was already a broken woman. Her suffering, believes Professor Graham Zellick, the retiring head of the commission, could have been relieved earlier if the Court of Appeal had taken a more interventionist approach the first time around. Instead of waiting to see compelling new evidence, Professor Zellick argues that the Court of Appeal should quash convictions where they simply believe the prosecution's case has not been proved.

Robert Verkaik

*Lip reading

In a case currently at the CCRC, prosecution evidence turned on the lip-reading testimony of an expert witness who interpreted what was being said in a silent video. During a review of the evidence, it was discovered the video had sound. An ensuing transcription of the conversation bore no resemblance to what the expert witness had claimed.

*Earprints

Mark Dallagher, 30, spent nearly seven years in jail protesting his innocence after an Old Bailey jury was told by an expert, a Dutch police inspector, that earprints found at a murder crime scene matched his exactly. After the Court of Appeal ordered a retrial, a DNA sample from the earprint proved it was not Mr Dallagher's.