The Blood Under the Paint: Fifteen Years to Find Who Killed Lynette White
Lynette Deborah White was born in Cardiff on 5 July 1967 and was twenty years old in the winter of 1988. She was a small woman, well known around Butetown, the old docklands district south of the city centre, where she worked the streets. She had a boyfriend, Stephen Miller. People who knew her described someone who was easy to talk to and who looked after herself as best she could in a life that gave her very little protection.
On the evening of Sunday 14 February 1988, Valentine's Day, she was working as usual in the streets around James Street. It was cold. The area was a grid of terraces, small businesses and closed-up shopfronts, and the women who worked there used rooms they could get access to. Number 7 James Street was a flat above a betting shop. Lynette had the use of it.
She was seen alive that evening in the ordinary course of her night. Some time after that, in the small hours, she went up to the flat with a client.
The post-mortem examination later placed her death in the early hours of 15 February.
Her body was found that night in the flat. She had been attacked with a knife with a ferocity that experienced officers struggled to describe. She had been stabbed more than fifty times. Her throat had been cut. Her hands and arms carried the injuries of someone who had fought hard and for some time against a man holding a blade.
The scene
The detectives who came into that small room were looking at something that, in forensic terms, should have been an easy case.
A frenzied knife attack in a confined space is a messy crime for the person committing it. When a hand slips on a wet handle the blade takes the fingers. Somewhere in that flat there was almost certainly the killer's own blood mixed with hers, and in 1988 the police did what the science of the day allowed: they collected, photographed, swabbed and bagged, and they sent samples away for grouping.
DNA profiling existed in 1988, but only just. It had been used in a criminal case for the first time in England the previous year and it required a relatively large, clean sample. Traces mixed with the victim's blood, degraded and smeared across a room, were beyond it.
So the samples went into storage, tested as far as the period allowed and then kept. That decision, taken by people who could not have known what it would be worth, is the reason there is a solved case to write about.
What the investigation did have was witnesses. In the hours after the murder, people in the area reported seeing a white man in the streets nearby, distressed, with cuts to his hands and blood on his clothing. That description was taken seriously enough for an artist's impression to be produced and circulated.
Nothing came of it. Weeks went by, then months. The murder of a young woman on Valentine's night, in a city that considered itself safe, generated exactly the kind of pressure that police forces of that era responded to badly. A very large team was working the case. The public wanted an arrest.
Where the inquiry turned
By the late autumn the direction of the investigation had shifted away from an unknown white man and towards men who had known Lynette.
The central figure became her boyfriend, Stephen Miller. He was interviewed repeatedly over a period of days. The interviews were tape recorded, which is the only reason anybody can now say with certainty what happened in them.
He denied any involvement. He kept denying it. By the count later made for the appeal, he denied it more than three hundred times, across sessions running to many hours, while officers told him repeatedly that they did not believe him. Stephen Miller had significant learning difficulties. Eventually, he stopped denying it, and produced an account in which he had been present and in which he named others.
In November 1988 five men were charged with the murder: Stephen Miller, Yusef Abdullahi, Tony Paris, John Actie and Ronnie Actie.
The forensic evidence recovered from the flat did not implicate any of them. That is not a detail added with hindsight. It was true at the time and it was known at the time.
The trial ran into 1990 and was, then, the longest murder trial in British legal history. In November 1990 the jury acquitted John Actie and Ronnie Actie, and convicted Stephen Miller, Yusef Abdullahi and Tony Paris. They were sentenced to life imprisonment, and they became known as the Cardiff Three.
The tapes
Two years later the Court of Appeal listened to the interview tapes.
The court's reaction is a matter of public record. The convictions were quashed in December 1992 on the basis that the questioning of Stephen Miller had been oppressive, that what he eventually said could not safely be treated as a confession at all, and that the case against the other two rested substantially on it. Three men who had spent years in prison walked out.
What happened next is the part of this story that is least often told. An acquittal on appeal does not automatically restart a murder inquiry. For years afterwards the case sat in an uncomfortable silence, and in some quarters the belief persisted that guilty men had been released on a technicality about interview conduct. The men themselves lived under that. The families of the wrongly convicted campaigned. Lynette White's family had been told who killed her, then told it was not true, and were given nothing to replace it.
Nobody was looking for anybody.
It is worth being precise about what that meant in practice. Between December 1992 and the reopening of the file, there was no active hunt for the person who had killed Lynette White. A murder had been committed in a British city, five men had been publicly identified as responsible, and the courts had established that the case against three of them was worthless. The consequence was not a renewed search. It was a gap.
The material that had been kept
What saved the case was that the physical evidence from 1988 had not been destroyed.
In 2000 South Wales Police opened a cold case review. The reasoning was straightforward. The original scene had produced blood that was not Lynette White's. In 1988 nothing could be done with it. Twelve years later, the techniques had changed out of recognition, and the samples were still in storage.
The review team went back to the flat itself. This is where the case turns on something almost banal. In the years since the murder the room had been redecorated. Fresh paint had gone over the woodwork. The team examined the skirting boards and took them away, and a scientist scraped back through the layers of paint, looking for what had been sealed underneath.
Beneath the paint there was blood. It had been missed in 1988, then covered over by a decorator who had no idea what he was painting on.
In January 2002, using a technique known as Second Generation Multiplex Plus, the laboratory produced a full DNA profile of an unknown man from that material.
The profile was run against the national DNA database. There was no match. The man who left it had never had his DNA taken.
The fourteen-year-old boy
At that point the inquiry did something that was still new.
Rather than accept a blank result, the team looked at near matches, profiles that were close to the unknown man's without being his. A DNA profile shared with a close blood relative resembles that relative's more than it resembles a stranger's, and a search for partial matches can therefore point not at the offender but at his family.
The search produced a boy of fourteen. He was on the database because of a minor matter. He was also, plainly, not the killer, because he had not been born when Lynette White died.
This is a slow and delicate way to work. A partial match is not evidence against anybody. It is a direction. It also means knocking on the doors of people who have done nothing, explaining that a relative of theirs may be a murderer, and asking them to provide a sample voluntarily. The team began working outwards through the boy's family, taking samples from male relatives and eliminating them one by one.
The trail moved along the paternal line and arrived at the boy's uncle: Jeffrey Gafoor, a security guard living quietly in Llanharan, west of Cardiff. He had been eighteen years old in February 1988.
His DNA was an exact match to the profile from under the paint.
He was arrested in February 2003. He attempted to take his own life before he could be interviewed, and survived. He then admitted the killing.
The account he gave was almost unbearably small. He had gone to the flat as a client. There had been an argument about thirty pounds. He had a knife with him and he used it, and then he ran. He had cut his own hands, which is why his blood was in the room. He had gone home, and for the following fifteen years he had lived without coming to police attention, in a job, a few miles from where it happened, while five men were charged and three were imprisoned for what he had done.
On 4 July 2003 at Cardiff Crown Court, Jeffrey Gafoor pleaded guilty to the murder of Lynette White. He was sentenced to life imprisonment with a minimum term of twelve years and eight months.
What followed the conviction
The conviction settled who killed Lynette White. It did not settle anything else.
A criminal investigation began into how the original inquiry had produced confessions and convictions against men whose DNA was nowhere in that room. In 2011 a group of former South Wales Police officers stood trial accused of perverting the course of justice. It was one of the largest police corruption prosecutions ever mounted in Britain, and it collapsed, after an estimated thirty million pounds of public money had been spent, over disclosure: documents that the court was told had been destroyed were later found intact in a storage unit. A second prosecution involving further retired officers was abandoned and the charges dropped.
No officer has ever been convicted in connection with the case. Everyone accused is entitled to the presumption of innocence, and in law they stand acquitted.
The Home Office commissioned an independent review by Richard Horwell QC into why the prosecution had failed. His report, published in July 2017, concluded that the collapse was caused by human error on the part of police and the Crown Prosecution Service rather than by any deliberate attempt to sabotage the trial.
Conclusions and Open Questions
The core of the case is closed. Jeffrey Gafoor is a convicted murderer who pleaded guilty, and the forensic evidence against him was overwhelming and independent of anything he said. Stephen Miller, Yusef Abdullahi and Tony Paris were innocent men who lost years of their lives, and John Actie and Ronnie Actie were innocent men who sat through the longest murder trial in the country's history. Yusef Abdullahi did not live long after his exoneration.
What the investigation teaches divides into things that are agreed and things that are not.
Almost everyone agrees on the forensic lesson, and it is the reason this case is taught. The evidence that eventually identified the killer existed on the first day. It was in the room while the original inquiry was interviewing the wrong men. It survived only because exhibits were retained rather than destroyed, and it became readable only because someone in 2000 decided to go back and look again at a case that officially had no suspects. A twelve-year-old sample and a length of painted skirting board did what a very large investigation had failed to do.
The second lesson concerns confessions, and here there is still argument. One view, effectively that of the Court of Appeal, is that the interviews themselves manufactured the case: that sustained pressure on a vulnerable man produced an account that then dragged four other men in behind it. Its strength is that the tapes exist and can be listened to, and that the account produced was contradicted by the physical evidence in the room. A competing view, advanced over the years by defenders of the original inquiry, is that officers were working within the standards of the time, under enormous public pressure, and that they genuinely believed they had the right men. Its strength is that interview practice in 1988 was not what it is now, and that hindsight is a poor guide to the state of the art. Its weakness is severe: the witness reports of a white man with cut hands existed from the beginning, and the forensic material from the scene never matched any of the accused. Believing something sincerely does not explain proceeding against men the evidence excluded.
The collapse of the 2011 prosecution remains contested. Richard Horwell's finding of human error rather than deliberate wrongdoing is the official conclusion and was reached after a lengthy independent review with access to the papers. Against it stands the simple shape of the events: a prosecution of police officers over the handling of evidence failed because of the handling of evidence, and material said to have been destroyed turned out to be sitting in storage. Horwell's report addresses that directly and rejects a conspiracy reading. Others, including some of those wrongly convicted, have never accepted it. There is no mechanism now by which that argument gets settled.
Several practical questions were never answered. Nobody has established why the early sightings of an injured white man in the streets around James Street did not hold the inquiry's attention. Nobody has explained how a case was built and taken to trial against five men when the blood in the room belonged to none of them. And nobody has publicly accounted for the fifteen years in which Jeffrey Gafoor, having killed a woman over thirty pounds, went to work, kept his head down, and let three men serve his sentence.
The last open question is not an investigative one. Lynette White was twenty. The case that carries her name is remembered as a landmark in forensic science and as one of the worst miscarriages of justice in modern British history, which means that almost everything written about it is about the men who were wrongly convicted and the officers who were later accused. Her family waited from 1988 until 2003 to be told the truth, and were told two different wrong things in between.