Eighteen people in this archive were released from custody and then killed. Each release was a decision, a parole board, a hospital discharge, a sentence served out, made by somebody who had the file in front of them.
Edmund Kemper killed his grandparents at fifteen. He was sent to a state hospital, and released at twenty-one into the care of his mother, over the objection of the doctors who had treated him. He then killed six young women, his mother, and her friend in California in 1972 and 1973.
Willem van Eijk killed two women in the Netherlands in the early 1970s and served eighteen years under a hospital order. He was released in 1990 as a treatment success, the phrase belongs to the system that let him out, not to us, and killed three more women in Groningen.
Jack Unterweger is the case that should be uncomfortable for anyone who writes. He was serving life for murder in Austria and began publishing from prison: poems, an autobiography, a play. Austrian writers and intellectuals took him up as proof that a man could be remade, and campaigned for his release. He was paroled in 1990, worked as a journalist covering crime, and killed at least nine more women in Austria, Czechoslovakia and Los Angeles.
The rest divide by how long the state waited before letting go.
Richard Biegenwald killed a shopkeeper in New Jersey in 1958 at eighteen and was released in 1975 after seventeen years, then killed at least five young people. Terry Blair served twenty-one years for beating and strangling his pregnant ex-girlfriend, was paroled in 2003, and killed six women in the same city within about a year. Louise Peete was convicted of murder in California in 1920, paroled after eighteen years, and convicted again in 1944 of killing the woman who had taken her in. Thomas Neill Cream was convicted in Illinois in 1881 and released after ten years; he killed four women in Lambeth and was hanged in 1892.
Then there are the short intervals, which are the ones that read worst.
Jerome Dennis was paroled in November 1991 after ten years for rape and killed five women in East Orange and Newark within the following five months. Nicholas Wiley was released as a Level 3 sex offender, New York's highest risk category, and killed three women within months. Mauricio Rodriguez Silva was paroled in May 1984 and killed three people within weeks. Daniel Siebert served six years for manslaughter, was paroled, and murdered women and two small children in Alabama within a year. Vincent Groves, Joseph Miller and Silvo Plut each killed again within months of release. Leslie Allen Williams was a paroled rapist when he abducted and killed four teenage girls in southeastern Michigan.
Two do not fit and are worth separating out. Daniel Camargo Barbosa was not released in any ordinary sense, he escaped a Colombian prison colony and the state recorded him as gone. Ricardo Caputo killed four women between 1971 and 1977, spent seventeen years at liberty without being caught at all, and then walked into a lawyer's office and surrendered.
What the fifteen genuine releases have in common is not that anyone was careless. Kemper's doctors objected and were overruled. Wiley was released under the highest risk classification his state had, which means the system had assessed him accurately and let him go anyway. Van Eijk was discharged because the treatment was judged to have worked. Unterweger had a public campaign behind him.
Brandon Tholmer is the eighteenth and the clearest. In 1975 he raped and robbed a woman of about eighty in Los Angeles and tried to strangle her. He was convicted, formally classified a mentally disordered sex offender, committed to Patton State Hospital for treatment, and released on parole in 1979. The killings began in January 1981, and the four elderly women he was convicted of murdering all lived alone within about a mile of his flat. What separates his case from the others here is the paperwork: the state had not merely assessed him and taken a view, it had given him a legal classification for what he had done to an elderly woman, and then discharged him. The jury that convicted him in 1986 declined to sentence him to death, several jurors saying afterwards that they had been moved by the conditions of his own life, and the defence had put before them a diagnosis of borderline intellectual disability and years in juvenile institutions from the age of eleven. The same facts that had shaped his treatment were the ones that spared him.
In none of these cases did the state fail to know who it was dealing with. It knew, wrote it down, and released them on the strength of a judgement about the future that turned out to be wrong. The interval between release and the next death, five months, a year, within weeks, is the only part that varies.
In one case the machinery was rebuilt afterwards. Until 1992 the Michigan parole board was made up of professional civil servants who could release a prisoner serving life after ten years, and it was that board which had let Leslie Allen Williams out in 1990, seven years into a sentence for sexual assault. He killed nine months later. After his confession the governor abolished the board and replaced it with ten political appointees. The family of Melissa and Michelle Urbin, aged fourteen and sixteen, said publicly that the board had killed their daughters. It is the only release in this list that changed how the decision would be made for everybody who came after.