Palme

Who killed Swedish Prime Minister, Olof Palme?

Olof Palme, the former Prime minister of Sweden, was murdered on February 28, 1986, shot in the back without bodyguards following a cinema visit. The first English language account of the circumstances around the murder, “Blood on the Snow”, was published by the medical historian Jan Bondeson in 2005. Because of its consequences for European history, Bondeson regards the Palme murder as being along with the Kennedy assassination as one of the “crimes of the century”. The crime is still not solved and it is possible that, during the twenty-four years that have passed, the murderer himself has died and that no conviction is possible.  Yet, given the appropriate resources, the question concerning who the murderer was, can still be answered. One ground for optimism is that the case has never been subjected to the form of analysis that uses the now well-established principles of forensic psychology that form the basis for the present article.

A major hinder to taking any initiative of this sort has been that in Sweden a sizable proportion – perhaps the majority – of the population have come to accept that a violent alcoholic, Christer Pettersson, who died some years ago, was the perpetrator, although he had been acquitted by the High Court of Appeals (Svea hovrätt). In spite of this acquittal, even the police investigators appear to have continued to regard Pettersson as the main suspect. The Chief Public Prosecutor even tried to reopen the case against Pettersson but the Supreme Court decided that no new important evidence had been submitted, and declared that there were still alternative hypotheses that had not been fully explored by the Chief Prosecutor in the case. 

As a consequence of this state of affairs, the case has come to be seen in Sweden as an open wound that was better left alone in order to heal. This sense of resignation may however be giving way to the need for a re-evaluation of contemporary methods of investigation. One reason is that another high profiled case (that of the alleged Serial murderer Thomas Quick) is currently revealing the enormous shortcomings in the investigatory methods used by the police and prosecutors in Sweden from the 1980s onwards. These shortcomings were apparent in the Palme investigations and the most damning was the absence of any systematic use of forensic psychology. The current article applies some of the principles of forensic and witness psychology to the case in order to take us beyond these shortcomings and show that the extant findings do in fact give clear pointers to a resolution.

Despite the above shortcomings, there was one witness psychologist, Astrid Holgerson, involved in the case, although she was used by the defence counsel and acted only in a limited capacity. Nevertheless, it was essentially Holgerson’s testimony that led to the acquittal of Christer Pettersson. Holgerson succeeded in demolishing the main evidence against Pettersson, which had come from the Prime Minister’s widow, Lisbet Palme.  Lisbet Palme had been present at the murder scene and subsequently at the trial identified Pettersson with “great certainty” as being the murderer. It was her certainty that succeeded in convincing both the general public and the police investigators that the identification was correct. Yet, Holgerson was able to factually show that that the memory of the event had most probably been influenced by what she had read in the press about Pettersson and had created a cognitive illusion of certainty.

In the USA there are several cases of false convictions made with equal or greater certainty that are now on record, where the convicted men have been later acquitted when DNA analyses have proven their innocence. In most of these cases the false convictions were due to false identifications. In 1956, long before formal studies revealed the fallibility of human judgments, Alfred Hitchcock’s had produced the film The Wrong Man showing that justice could be completely mislead in this way.  In England the problem of using witness identifications as a sole basis for securing convictions, was subject to a judicial investigation that resulted in the Devlin Report, a report that influenced the Swedish Court of Appeal in its decision to free Pettersson.

The legal discounting of Lisbet Palme´s testimony is in full accordance with the scientific analyses that have been made, concerning the witness psychological analysis of how her statements came to centre on Pettersson. It is also in accordance with a later experimental investigation using the video recording made of the actual identification parade in which Lisbet Palme picked out Pettersson. The video was shown to American students, who did not know the identity of the suspect. When they had no further information about the suspect, their success at picking out Pettersson as the suspect was no greater than expected by chance, but when they were told, that the suspect was an alcoholic – a fact that had been released to the Swedish public and was known to Lisbet Palme prior to the line-up – no less than 75 percent of the students chose Pettersson.  Lisbet Palme´s spontaneous comment when she saw the line-up was: ”One can immediately see who is an alcoholic.”

In spite of both the legal and the scientific reservations concerning the value of Lisbet Palme´s testimony, the police investigators, the media and the general public seem to keep to their gut feeling that her identification of Pettersson was after all correct. An explanation for this may be that when a crime victim is confronted with a possible perpetrator, the victim can become easily overwhelmed by the feelings that he or she experienced at the time of the crime. These feelings then become transformed into a certainty that the suspect must be the perpetrator, and the feelings are passed on to those who hear the witness if they are unaware of these psychological mechanisms.

In her expert testimony on the case, Holgerson made the crucial point that it is the memory recall given in the first interrogation and not subsequent ones, that should carry weight and guide the investigation. This interrogation should be  carried out as soon as possible. Using this principle, the three more serious books on the case[1] agreed that a strong argument can be made that the initial description of the murderer which Lisbet Palme gave, did not at all fit Pettersson but fitted rather that of another person, Anders Björkman. Björkman had been innocently walking behind Lisbet and Olof Palme and happened to be in her line of vision immediately after the murder.

If Christer Pettersson was not the murderer, then who was? The former police inspector Börje Wingren led some of the very first interrogations that were carried out during the weeks immediately following the murder in 1986. It is the results of this investigation that provide the basis for his claim to know who the murderer was. Wingren argues in his book He Shot Olof Palme (from 1993, which is still only available in Swedish) that the true murderer was actually the very first suspect, the so-called ”33 year old man”, who was arrested twelve days after the murder, but who became later eliminated from the investigation. Wingren´s account of the facts in this case seems not just plausible but convincing because he uses original witnesses statements (which he copied on resigning from his appointment) rather than relying on memory or on secondary sources. This makes it all the more curious that his book, even in Sweden, is virtually unknown. But then there is probably a simple all too human reason for this:  Wingren´s conclusion is seen to be at variance with the popular and perhaps now all too convenient opinion that Pettersson was the murderer.

Let us briefly summarize Börje Wingren´s account of facts. Olof Palme was murdered at 23.21, on February 28, 1986, on the street Sveavägen in Stockholm. Earlier in the same evening the ”33 year old man”, Victor Gunnarsson, who we will call by his initials, VG, was observed in a café a few hundred meters away. It is established that VG talked at length to three women in a café, close to the place of the murder, and they described his appearance, his clothes and his shoulder bag. The witnesses told the investigators that VG had been rather belligerent in expressing his hatred for Olof Palme (which other witnesses also testified to). They had remarked that he was not drunk but he ”smelled of garlic”. The three women left the café at 22.35 with VG still there, which was about 45 minutes before the murder. About 15-20 minutes after the murder, two other witnesses observed a person who closely matched VG´s description and clothing when he sneaked into a cinema. Wingren calls this a ”classical hiding place” for a person like VG who loved Hollywood movies. Wingren then gives the reader a coherent chain of evidence about what happened.

VG was dressed in a green knee-length pilot jacket with a hood, tight jeans and high boots. The original first statements – and once again it is these first statements which are of course of prime value – from various witnesses who saw the murderer running away, correspond well to the description of VG and his clothes (but not with those of Pettersson and his clothes). A student nurse, Anna Hage, who tried to give first aid to Olof Palme after he was shot, described how the murderer ran away with fast, vigorous steps and that he was dressed in a knee-length jacket, probably with a hood. Another witness, Lars Jeppsson, who was in an adjoining street, saw the murderer run past this street toward a long flight of stairs (in total 89 steps) and then observed how he ran quickly up the stairs with 2-3 steps at a time.  This would imply that the murderer was vigorous and in good physical condition which is inconsistent with that of Pettersson who was an overweight alcoholic.  What is further if not finally damning for the whole case for the prosecution of Pettersson, is a curious coincidence (not mentioned in Wingren’s book). Lars Jeppsson happened to live in the same small suburban area and told he knew that Pettersson was an easily recognisable down and out street corner alcoholic. Crucially, Jeppsson did not recognise Pettersson and did not believe the murderer he saw and eventually unsuccessfully ran after up the flight of stairs, was Christer Pettersson.

The murderer, after running up the stairs, was observed by a Finnish witness, Yvonne Nieminen, who later described that the running man’s knee-long jacket was at that time unbuttoned and that he was trying to close or open a little bag. Winberg believes that the murderer was then hiding the bag with the weapon in it under his jacket. Shortly after the witness Nieminen´s observation, yet another witness,  called “S”, observed a running man from the back with a half long jacket. S later identified the suspect VG from a witness parade, although with some hesitation in the choice between VG and another person in the line-up. Some minutes after the murder, yet another witness, this time an African called “ID”, who was working as an unlicensed cab driver gave some further crucial information. In a side street from where the murderer had last been seen, he had been stopped by a man who was insistent on wanting the witness to drive him from there. The witness ID described that this man had a jacket hood over his head, and something hidden under his jacket, and he had only one glove.  Remarkably when the police took in VG for interrogation, he had only one glove and apparently no explanation for this.

What may be the most crucial evidence of all, is that this witness ID spontaneously remarked that the man had smelled of garlic. When ID refused to drive him, the man ran into a cemetery from where it is only a short distance to the cinema where VG was observed, according to Wingren. The witness even appeared to have picked out VG from a witness line-up. However an internal dispute then arose between the police master and the chief prosecutor over a breach of protocol in the procedure with the outcome that VG was released. VG later procured an alibi for the crucial minutes during the murder, but Wingren claims that this was probably a false alibi. VG was never brought to trial. 

Since we know that initial memory processes use areas on the brain that are adjacent to those responsible for smell (our most primitive sensory modality) the garlic smell should have been regarded as a reliable memory experience. From a psychological perspective, corresponding as it does with the observed smell of garlic from VG’s breath in the café prior to the murder, it becomes a rather valuable piece of evidence.

Winberg used VG´s jacket as a source of more incriminating evidence. One witness, called “JA”, from the vantage point of his car in its stopping position at a red light, had seen the murderer run away. When JA was shown VG´s jacket along with six other jackets he pointed out VG´s jacket from among others as belonging to the murderer. According to JA, the murderer had tight jeans, and ran with long, vigorous steps ”like an elk”. Another witness, LL (who first called the alarm number) remembered that the murderer had jeans that did not reach the feet, and he, too, selected VG´s jacket as that of the murderer. The witness Anna Hage (the nursing student mentioned above who tried to give first aid to Olof Palme) was shown the pictures of the jackets and also pointed out the same jacket as the other witnesses. She had not seen the face of the murderer, and she was shown photos of VG and the others in the identification parade with the upper parts covered so she only saw the men from the knees and down. Under these conditions, she immediately pointed out VG. As it happened VG had actually kept this pilot jacket stored during the winter until the day before the murder. There were some small traces of detonating composition on the jacket, but the quantities were considered to be too small to be used as certain incriminating evidence.

An interesting detail is that VG, although he otherwise normally had a moustache, shaved this off shortly before the day of the murder. Winberg calls this ”the moustache trick”, which means that the perpetrator changes his appearance at the time of a crime, so that it will later be difficult to recognize him, when the moustache has grown again. The witnesses who saw him on the days before the murder all agreed that he did not have a moustache, but VG flatly denied having shaved it off. Again, the moustache trick may have been inspired by VG’s interest in American action movies.

If there was so much in the way of circumstantial evidence, why was VG released? Two things seem to have been decisive: First, the key witness ID had been shown pictures of VG and some others before the proper identification parade and did not identify VG from the photo. This is of course a serious procedural mistake, but it does not diminish the value of the testimonies from the other witnesses. Second, two and a half months after the murder, VG was able to provide a possible alibi for the critical time 22.35-23.35 of the murder. He claimed to have met two boys at the café, and these boys indeed confirmed that they had met him. Nevertheless, there appears to be some uncertainty about his alibi. Two and half months had passed before they were interviewed so Winberg believes that it could easily have been on another occasion, especially given that they said that VG had showed them his camera, and VG himself had said that he did not bring his camera with him on the night of the Palme murder.

Another important psychological principle that the police investigation failed to use is that of the devil’s advocate, meaning that evidence against the prevailing hypothesis should be reviewed and given serious consideration. If we apply this here, then we need to examine the credibility of the testimony of ID, the African cab driver:  ID had been forced by the police to confront VG and was emotionally affected by it. Some days later he reported seeing a murder suspect known as  “the phantom man” and this gave the police reason to dismiss his whole testimony as a lively fantasy. The phantom man was in reality a sketch by an artist of someone she saw in the neighbourhood of the murder place and who might have been the murderer. The sketch was press-released and its vagueness led to an epidemic of sightings throughout Sweden. Given these circumstances, it seems unreasonable to use this as the pretext for dismissing all of ID’s testimony.

Are there other explanations for the garlic identification that does not implicate VG as the murderer? Police interviews often give unintentional cues to witnesses to give the wished for answers especially if the same officers are carrying the investigations.  However in this case there were different officers responsible for interviewing the girls from the café and for ID.  Moreover, if the record of the interview is correct, then responses about the garlic statement came quite spontaneously without cuing.

Some scepticism is however justified towards the use of witness parades. In Sweden witness parades are used as standard procedures in major crime investigations. Yet they are often deficient as to the elimination of potential cues that might give away who is the suspect. This is not to say that witness recall is valueless but in order to increase their value a rather special approach known as the cognitive interview is needed.  This is a supportive open-ended technique of interviewing as opposed to the customary interrogative methods. It has been shown that witnesses do indeed recall wrong memories of minor details such as clothes or hair colour but they are nevertheless often accurate as concerns the main events or any unusual aspects of it that occur. Chance or coincidental events can of course occur and be grossly misleading. For instance, although the eating of garlic is not a dominant habit in Sweden, it can always be reasoned that this is just such a curious coincidence. Yet taken together with the glove, the camera bag, and the coat, coincidence in this case does not seem to be a plausible explanation.

There are three further arguments that the “devils advocate” could still use to discount VG and to implicate Pettersson in the murder. The first is that the murder was described as having a special gait by running with a limp and Pettersson did have such odd way of jogging.  Yet if initial interviews are used as source material, then this appears to be a myth created by newspaper reports.  Borgnäs in reviewing this (page 56) writes: “Not one of the 22 murder place witnesses say in their first interviews that ” the murderer should limp, or have a special gait”.

A similar myth concerns the so-called “Grand man”. This was a man who had been standing and staring in an unpleasant manner outside the cinema  “Grand” which the Palme couple visited before their fateful 300 metres walk. Pettersson is said to have been this “Grand man” because of his staring eyes and threatening stance and he would fit this, rather than VG who generally looked more pleasant. Even the High Court, in freeing Pettersson because it was not proven the Grand man was the murderer, concluded that it was proven beyond reasonable doubt that Pettersson was identical with the Grand man.

Before writing his book on the Palme murder (“En iskall vind blåste genom Sverige”), the journalist Lars Borgnäs consulted one of the world’s leading forensic psychologists, Gisli Gudjonsson. Using the advice that priority should be given to the information contained in the original interviews, Borgnäs arrived at a clearer scenario as to what various witnessed had observed. From this, he was able to conclude that the person observed at the cinema did not closely resemble Pettersson. The star witness in this context, the air traffic controller, Lars-Erik Eriksson, who was waiting to pick up his parents from the cinema, had become so threatened by this Grand man that he locked his car door: He picked out Pettersson from an identity parade as being the Grand man but this was two years later after a lot of information had been fed to the press. Neither did the descriptions of the Grand man correspond much with those of a man seen following the Palme couple that were given by a Yugoslav immigrant Ljubisa Najic serving at a hot dog stand on route to the murder place.

It is said with certainty that the visit to the cinema was decided at the last moment and the decision to walk home was more or less fortuitous. The question then naturally arises how could the murderer have known that the Palmes were at the cinema? Winberg mentions that VG was a regular guest at a restaurant, which was situated only 60 meters from Olof Palme´s entrance door to his home in the Old Town of Stockholm. VG would often stay there for hours. He also had delivered newspapers on the same street (but on the other side of the street from Palme´s home). This gives a possible scenario in which VG, having observed the cinema visit, spends time waiting at the café a few hundred metres away from the murder place. It was here he was observed in a belligerent state, expressing his hatred and apparently even making threats towards Palme. The indications are that he left the café after 22.35 and if this is so, it was well in time to have arrived in the vicinity of the cinema and then to have followed the Palmes on route to the chosen murder place.

Another argument that has been put forward (for example by social historian Jan Bondeson) against VG’s involvement is that “there is no evidence he ever possessed a firearm”. Yet as Wingren notes referring to a photograph taken in California, VG is seen standing on a shooting-range holding a revolver of a similar type to the murder weapon. VG even applied for permission to carry a weapon for a security company (an application which was declined).

If then there are good reasons for believing that VG may have been the actual murderer, it was surely a crucial mistake to eliminate him from the investigation. Winberg does not conceal his disappointment with the prosecutor for this decision, and unfortunately it is this, his emotional attacks on the prosecutor, which makes the book somewhat difficult to read. Two objective aspects do stand out and give Winberg’s account credibility. He had access to and makes use of the original case records that were denied to other writers and in agreement with the earlier mentioned witness psychological principle he emphasises that the very first statements from the witnesses are the most important and that the first few days of an investigation often are decisive for the outcome.

Previous authors who have discussed the Palme murder have often by default arrived at conspiratorial theories; theories which have not been difficult to find, given Palme’s international involvement and his controversial status.  Wingren makes no such allusion and neither do we, although one of Wingren’s supporters (Anders Leopold) has provided his web site readers with the evidence for VG’s own claims that he had CIA connections.

Winberg´s book was published in August 1993. A few months later, on December 3, VG vanished without a trace from his new home in the U.S.A., where he had moved. He was found murdered a few weeks later, on January 7, 1994. He had been shot in the head, left without clothes and identification papers, and was identified merely from his watch and his ring. Since he had been eliminated from the Palme investigation, there would seem to be uncertainty as to whether a final confirmation of his identity was ever made. A man was later convicted of this murder, but an American citizen group has taken up the case in supporting the innocence of the convicted man. Apparently during the course of the trial, a witness testified that VG had confessed to, and even bragged to him about being the murderer of Palme.

The murder of a Prime Minister is not an ordinary crime; it is also an important historical event. As such any attempt at providing as accurate and true an account as possible of this event is clearly justified.  The suggestion proposed here is that future research into the identity of the murderer could be authorised to be carried out by scientists possessing the appropriate competence. Researchers with competence in law, historical methods, statement validity analyses and other disciplines could then collaborate in a joint attempt to clarify what actually happened. A scientific commission with broad scientific competence ought to be created with its mission to look at the forensic evidence that can help us answer the question, who killed the Prime Minister. The suggestion may come at an appropriate time. By all accounts the police investigations have come to a standstill, and a scientific re-analysis may well not only provide the concrete leads but also provide new guidelines for the future investigations of major crimes.

[1] Inuti Labyrinten published 1994 by K. Poutiainen and P. Poutiainen and Blood in the Snow published in 2005 by Jan Bondeson, En iskall vind drog genom Sverige, published in 2003 by Lars Borgnäs.