The Palme Assassination: A Prime Minister, a Pistol, and No Answer

Just after eleven o’clock on the night of 28 February 1986, Olof Palme and his wife Lisbeth were walking home from a cinema in central Stockholm. They had been to a late showing of a Swedish comedy, The Mozart Brothers, on Sveavägen, and they were doing something that tells you almost everything about the country Palme governed: the Prime Minister of Sweden was walking home through the middle of his capital, at night, with no bodyguard, because he had waved his security detail away earlier that evening. They were an ordinary couple on a cold street. At the corner of Sveavägen and Tunnelgatan a man stepped up behind them and fired a heavy revolver twice. One bullet went through Palme’s back and severed his spine and aorta; he was dead within minutes on the pavement. The second grazed Lisbeth. The gunman ran up a flight of steps into the dark of the old town and was gone, and in a sense he has never stopped running, because thirty-four years of investigation never laid a hand on him.

What makes the Palme case so fertile for conspiracy is not that the theories are wild. It is that the ground they grow in is real. This is a killing where the official investigation genuinely failed, where the man at its head genuinely chased the wrong quarry for a year, where evidence genuinely went missing and a murder weapon was genuinely never recovered. Before we ever reach the exotic solutions — apartheid hit squads, rogue policemen, arms-dealing vendettas — we have to sit with the documented, sourced, embarrassing history of an investigation that broke down in public. That failure is the kernel. Everything else grew around it.

The most botched murder inquiry in Swedish history

Start with the scene itself, because the errors began there. When the shots were fired, Stockholm’s police response was slow and disorganised. No cordon was thrown around the district in time. The all-points bulletin went out with confused descriptions. Officers and volunteers trampled the site. Most notoriously, the two bullets — the physical heart of any ballistics case — were not recovered by the police at all. They were found days later by members of the public: one by a passer-by near the scene, another weeks afterwards. A prime minister had been assassinated in the centre of the capital and the police could not secure two spent rounds.

The investigation was placed under Hans Holmér, the Stockholm county police chief, a confident and media-friendly figure who became the public face of the hunt. Holmér fixed early and hard on a single theory: that the killing was the work of the PKK, the Kurdistan Workers’ Party, which Sweden had at one point classified as a terrorist organisation. This became known as the “Kurd track” (Kurdspåret), and Holmér pursued it with a tunnel vision that starved every other lead of oxygen. In January 1987 he mounted a large coordinated operation against Kurdish suspects. It collapsed. There was no case. The prosecutors were furious that the raids had been launched on such thin grounds, and the whole edifice fell in on Holmér, who resigned from the investigation that spring. A year had been lost chasing a theory that produced nothing, and the trail — such as it was — had gone cold while he chased it.

This is worth holding onto, because it is the part of the story that is not in dispute and never was. The kernel of the Palme mythology is a documented institutional failure: a slow response, lost evidence, a fixated lead investigator, a year burned on a dead end. You do not need to invent a conspiracy to explain why the case was never solved. Incompetence, on this scale and this public, is explanation enough for the vacuum. But a vacuum around a murdered head of state does not stay empty.

The one man they charged

Out of the wreckage of the Holmér period, the investigation eventually produced a suspect: Christer Pettersson, a Stockholm drifter with a history of alcoholism, drug use and violence, including a conviction for a manslaughter committed with a bayonet years before. Pettersson had no clear political motive and no established connection to Palme. What he had was proximity to the milieu around Sveavägen and, above all, an identification from the one witness who mattered most — Lisbeth Palme.

In 1989 Pettersson was convicted of the murder by the Stockholm district court, largely on the strength of Lisbeth’s identification. Then the Court of Appeal overturned the conviction and acquitted him the same year. The reasons cut to the bone of the whole case. Lisbeth’s identification had been made in circumstances that a defence lawyer could drive a lorry through: she reportedly knew before the line-up that the police suspected an alcoholic, and Pettersson was the obvious such figure in the row of men presented to her. No murder weapon tied him to the crime. No convincing motive was ever established. The appellate court could not stand the conviction up, and it fell. Pettersson walked, gave rambling and contradictory interviews for the rest of his life, and died in 2004 after a fall. Sweden had tried one man for the killing of its prime minister, and the courts had let him go.

So now there are two things the record establishes beyond argument. The investigation failed. And the one prosecution it produced could not survive appeal. Both are true, both are sourced, and both are exactly the conditions under which grand theories flourish. When the ordinary machinery of justice visibly cannot close a case, people reach for extraordinary explanations, and here the extraordinary explanations had real material to work with.

The theories with teeth

Some conspiracy theories are spun from nothing. The serious Palme theories are spun from genuine facts about who Olof Palme was and whom he had angered, which is what makes them so difficult to wave away.

Palme was not a quiet caretaker prime minister. He was one of the most internationally combative social-democratic leaders of the Cold War. He denounced the American bombing of Hanoi in terms that froze Swedish-US relations for years. He was a loud, sustained, effective enemy of apartheid South Africa, channelling Swedish support to the anti-apartheid cause and speaking against the regime on the world stage. He was entangled, as head of government, in the Bofors affair — the Swedish arms manufacturer’s dealings, including a vast and corruption-tainted artillery contract with India and questions about weapons reaching Iran during the Iran-Iraq war. Any one of these made him enemies with the means and the ruthlessness to kill.

From this came the durable theories. The South African track is the most persistent: a former South African policeman, the notorious Eugene de Kock, testified after apartheid that his country’s security services were behind the killing, and the name of the operative Craig Williamson and a project sometimes called “Operation Longreach” circulate in every account. The police track (polisspåret) holds that right-wing officers within the Swedish police, who despised Palme, were involved or protective of whoever was — a theory made more than paranoid by the demonstrable hostility some officers bore him and by the very failures of the investigation they controlled. The Bofors and arms-dealing theories tie the murder to the weapons trade Palme was positioned to expose or obstruct. Each of these draws on something real: a real enmity, a real intelligence capability, a real institutional rot.

This is the fork — the precise place where the sourced history stops and the leap begins. It is one thing to say, truthfully, that apartheid South Africa ran assassination programmes, that Palme was its enemy, and that de Kock made an accusation. It is another to say that this proves South Africa killed Palme. The de Kock testimony has never been corroborated to the standard a court needs; the “Longreach” strand has never produced the operative, the order, the weapon. The same is true of the police track and the arms track. Each begins in documented reality — a real capability, a real motive — and then crosses, on faith, into a specific solved conspiracy that the evidence does not actually reach. The theories are not stupid. They are premature. They take the genuine kernel of Palme’s dangerous enemies and treat plausibility as proof.

The same shape recurs across this desk. A real, documented state capacity for murder — the kind proven at Katyn, where a government really did massacre thousands and lie for fifty years — teaches people, correctly, that governments are capable of monstrous things. The error is to carry that true lesson into a case where the specific evidence has not been done. Palme’s enemies were real in the way Beria’s crime was real. What is missing is the signed memorandum.

The answer that arrived without proof

In June 2020, thirty-four years after the shots, the Swedish chief prosecutor Krister Petersson gave a press conference that was meant to close the case. His conclusion was Stig Engström, a graphic designer at the Skandia insurance company whose offices stood near the murder site. Engström had been at the scene that night — he was known for years as “the Skandia Man” — and had given accounts of his own movements that shifted and inflated his role over time. He had a background that gave him access to firearms circles, held anti-Palme political views, and, in the prosecutor’s reconstruction, fit the movements and description of the gunman better than any other candidate. Engström had died in 2000, by his own hand.

Because the suspect was dead, the case could not go to trial, and the prosecutor formally closed the investigation. There was no confession, no recovered weapon, no forensic tie — the murder revolver was never found, and the bullets that the public had recovered decades earlier could not be matched to a gun that did not exist in evidence. What the prosecutor offered was a coherent narrative and a name, presented as the most probable answer that could now be given. To many Swedes it landed as an anticlimax that could not truly satisfy anyone: a solution you cannot test, attached to a man who cannot answer, closing a wound that had been open a third of a century.

And that is the honest ending, the one the case actually earns. The Palme assassination endures as a conspiracy engine because its origin is a real failure, its suspects have real motives, and its official conclusion arrives without the proof that would let anyone lay it down. A country that could imagine its prime minister walking home unguarded had also to imagine that anyone could kill him and vanish — and when the state that failed to catch the killer finally names one, without evidence to convict, it does not dispel the theories. It leaves them exactly where they have always lived, in the gap between what is plausible and what can be shown. The grief here is not for a lie, as it is with the Zinoviev Letter or another manufactured document. It is for a truth that was there to be found, on a cold street in 1986, and was lost by the people whose one job was to find it.