Resumé

Sweden's Prime Minister Olof Palme was shot dead, on a street in central Stockholm, on his way home from the cinema late in the evening of 28 February 1986. The Swedish public reacted with grief and horror. The emotional reactions to Prime Minister Palme's assassination were greater than expected. However, placed in an international context they are, nevertheless, weak. In the analysis, the situation six and four years after the assassination will be compared with the situation three weeks after the event. The focus is on: what role, if any, the assassination of Prime Minister Olof Palme has played in the way in which Swedes and immigrants express their views on a number of important issues related to the murder and what effect, if any, it might have on the Swedish political culture and on the trust of the Swedes in the political and judicial system. The final argument that can be presented from this study of the connection between exposure to a dramatic event, such as the murder of a prime minister, and children's and adults' political values, is that the emotional effect of the assassination fades away fairly quickly and is replaced by a much more vague and unclear structural effect related to the total impact of the assassination seen as a dramatic event of national importance. This kind of structural effect on the political culture in a country can never be clearly described and analysed for the simple reason that an effect of this magnitude is almost impossible to control and isolate from other experiences.

Introduction

Sweden's Prime Minister Olof Palme was shot dead, in a street in central Stockholm, on his way home from the cinema late in the evening of 28 February 1986. The Swedish public reacted with grief and horror. The emotional reactions to Prime Minister Palme's assassination were greater than expected. However, placed in an international context they are, nevertheless, relatively weak.

People who had moved to Sweden reacted more emotionally than nonimmigrants. immigrants expressed feelings similar to those observed in the USA in connection with President Kennedy's assassination (Sigel 1970). For example, 44 percent of first generation immigrants said they cried compared with only 24 percent of native-born Swedes.

Strikingly, many immigrants felt as if they had lost a close and dear
friend and were anxious about how Sweden would carry on without Olof
Palme's leadership. Compared with Swedes, they were more concerned

about what would happen to Palme's work for world peace, about Sweden's relations with other countries, as well as future help to developing countries. The reactions voiced by second-, as opposed to first-generation immigrants were, on the other hand, less person-orientated. Thus, what is usually called a resocialization effect, i.e. an adaption to the host country's political culture of collective rather than person-orientated leadership, was clearly expressed.

The results presented above are from a survey conducted less than three weeks after the assassination of Prime Minister Palme (Gustafsson, Jacobsson & Johansson 1987). A second survey was carried out in January and February 1990, four years after the dramatic event. On both occasions the empirical part of the study was carried out in the cities of Luleå, Umeå and Södertälje. A school was selected in each city, and all the children attending it aged between 10 and 15 years were interviewed. In Luleå and Södertälje a questionnaire was also distributed to the children's parents. In Umeå, however, a random sample of adults living in the vicinity of the school in question was used as a substitute (see Appendix 1 for a presentation the empirical data).

What can be learnt from the second study of the effects on the Swedish political culture of the assassination of Prime Minister Olof Palme? In 1990, we found more covariation between different background factors such as, social group, place of birth, gender and our measures of political culture than in 1986. One reason is probably that, on the first occasion, both young and old people's answers to our questions were more strongly influenced by the assassination than they were four years later. By 1990, the "trauma" of the assassination had begun to fade, which is reflected in the considerable increase in the number of people answering that they did not believe "that a murderer should be convicted against his/her denial if the evidence is weak but still clear". On the whole, we found clear support for the traditional liberal position on law and justice. The questions of the value of proof or guilt in court and of personal integrity were linked together in a way that did not exist immediately after the murder.

The final argument which can be presented from this second study on the connection between exposure to a dramatic event, like the murder of a prime minister, and children's and adult's political values is that the emotional effect of the assassination fades away rather quickly and is replaced by much more vague and unclear structural effects. In this case the total impact of the dramatic event comprises a suspected murderer who could not be convicted, a Minister of Justice who was removed from office because she had authorized an investigation of the murder by a private investigator and many more questions concerning law enforcement. Structural of this kind on the political culture in a country can never be described and analysed as such because they will always mix with other experiences in relation to law and justice (Johansson 1991).

In 1992, a third study containing questions related to the assassination of the Prime Minister was undertaken. This time the study was a nationwide representative sample, n = 1307. In this essay, data from all three studies will be analysed. The aim is to elucidate to what extent dramatic events, such as murders of political leaders, give rise to new ways of thinking about politics. In the analysis, the situation six and four years after the assassination will be compared with the situation three weeks after the event. The focus is on: what role, if any, the assassination of Prime Minister Olof Palme has played in the way in which Swedes and immigrants express their views on a number of important issues related to the murder and what effect, if any, it might have on the Swedish political culture and on the trust of the Swedes in the political and judicial system.

Political Culture - Some Theoretical Considerations

The classical literature on political culture deals with how knowledge, values and behaviour of importance for a society are passed on from one generation to the next. Various significant persons, i.e. agents, are specified who are assumed to exert influence on children's and young people's knowledge, values and behaviour in political issues - political culture. Normally, interest is focused on parents, friends and teachers (see, for example, Rutter et al. 1969; Beck 1977). The mass media's function as an agent of socialization is mentioned in several works (see, for example, Johansson 1985) and effects of the surrounding society on youngsters' political culture are discussed by Gustafsson (1972).

What is meant by political culture? One of the first definitions of the concept was made by Gabriel Almond in 1956. His initial formulation defined political culture as the "particular pattern of orientations to political action" (Almond 1956, 396). This early definition highlights political action. Very few later definitions have such a focus. Even Almond revised his early definition when planning the Civic Culture study. The definition introduced in the Civic Culture study is frequently used when the concept is discussed in the literature:

The term political culture thus refers to the specifically political orientations - attitudes toward the political system and its various parts, and attitudes toward the role of the self in the system. We speak of a political culture just as we can speak of an economic culture or a religious culture. It is a set of orientations toward a special set of social objects and processes (Almond & Verba 1963, 13).

No empirical guidelines are provided in this definition. It only emphasizes
that the objects of study are the orientations towards the political system
and the role of the self in that system.

There are variations in the definitions of political culture given by

different researchers and no "single" definition of this concept has been accepted by the scientific community. However, the following elements are common in most definitions. First, knowledge and information on the political systeirfs structures and functions are transmitted together with knowledge of those values and norms upon which that knowledge is based - "political cognitions". Next, the process is assumed to result in the development of assessments of the political system and its functions - "political orientations". Finally, political activity can also arise within the group or nation to which one belongs - "political participation" (Johansson 1985).

The effects we will measure here relate to how people cope with the uncertainty around the assassination and how their trust in the judicial and political systems of Sweden is affected. In relation to the latter aspect a comparison will be made with Great Britain.

Independent Variables

In this study one important independent variable is generation. We work with children aged 12 to 15 and their parents as our two generations. In our first report (Gustafsson, Jacobsson & Johansson 1987), we discussed two different ways of interpreting differences between generations. According earlier theories of political socialization, the adult generation's thoughts, attitudes, values and behaviour were the norm that the growing generation should acquire in order to be accepted as mature (Sigel 1965). This theory is based on the assumption, and even on empirical evidence from that time (see, for example, Langton 1969 or Dawson & Prewitt 1969), that the parents were the dominant political socialization agent for the child. Differences between parents and children were, therefore, often interpreted as a lack of political maturity among the younger generation (Easton & Dennis 1970).

However, studies from the same time period also showed that it was plausible that the youngsters' own parents were perhaps not always the main source of political values and information. Many other circumstances affecting the youngster could be of equal importance, for example, various "surrounding" factors in the local community (Gustafsson 1972). Some studies suggested that the total number of socialization agents who were in contact with the youngster was of importance (see Beck 1977) for his/her understanding of the political culture. It has later been shown that youngsters be more influenced by the lifestyle of the parents' generation than by their own parents (Jacobsson 1994). The political tradition is also often transferred through something vague that we can call "structural mechanisms", i.e. time-bounded experiences or dramatic events that affect the thinking of people in political matters (Johansson 1991).

In this article we are confronted with the difficulty of measuring the effects of a dramatic event - the assassination of the Swedish Prime Minister Olof Palme - on the political culture among ordinary people. In a report on President J. F. Kennedy's assassination Professor Roberta S. Sigel (1970) argued that:

When ... an event is of crisis or pseudo-crisis dimensions, it is a particularly propitious setting in which to study political socialization. A political crisis is by definition a situation in which existing norms or values are under attack. It thus affords an opportunity to test how firmly children share in adult political consensus and how much they have internalized the norms of the existing political order, i.e., how politically socialized they have become.

The murder of a prime minister can, of course, give rise to many questions which have effects on basic societal norms, e.g. political culture (see also Edelman & Simon 1971; Marvick & Marvick 1971). It may not be possible to pinpoint one reason for the change, but the whole trauma can have a structural effect. The assassination of Prime Minister Palme raised a number of questions and created many doubts, beliefs and attitudes related to the murder, the legal process and the strength of the democratic system - i.e. trust in the political system. It is possible to summarize the debate under four headings.

First, we have the murder as such. Why did it happened? Who did it? How could it happen in friendly Sweden? and so on. The second set of questions concerns the police investigation of the murder. Will they find the murderer? Will there be sufficient evidence to convict the killer? Did the police act in the right way on the night of the murder? etc. A third set of questions, which has presumably opened many people's eyes with regard to the judicial system, concerns the way in which the courts have been involved in the murder trial. Is the right man being brought to trial? Have all the court rules been followed? Did the Prime Minister's wife, Lisbeth Palme, sabotage the court procedures by not taking part in line-ups at the police station? When the suspected man was released, did the court demand greater proof of guilt because it was a prime minister that had been killed? and so on. The last, but probably the most frustrating set of questions, relates to the internal, highly technical, legal dispute between the police, the courts and the politicians in the standing parliamentary committee for dealing with different aspects of the constitution - konstitutionsutskottet. Questions have concerned: the rules for tapping telephones; whether a police investigation such as this can be above the law and can be called a state of emergency and who makes that decision; the level of government involvement, e.g. information and acceptance; and of course the so-called "Ebbe Carlsson affair", which is about the private police work carried out by the publisher Ebbe Carlsson - a former top civil servant in the Ministry of Justice who obtained a letter of recommendation from the Minister of

Justice, Anna-Greta Lejon, and also claimed that he worked with the
approval of many senior police commissioners.

The mass media coverage of all these side-effects of the assassination has been very extensive. However, despite the media attention, most of the above questions have been left without any concrete answers. Those answers that have been given have often been theoretical and presented in a way that leaves them open to discussion both among the public and in legal quarters. One hypothesis is that all this uncertainty together with the assassination can be seen as having a structural effect on the political culture in Sweden.

An attempt is now made to test whether people's beliefs and attitudes have been changed by or after the assassination. Have people's evaluations of why the Prime Minster was assassinated changed over the years? Can shifts in the answering pattern be explained by the fact that no murderer has been arrested, brought to court and convicted? Has the trust in the system of justice changed in the years following the dramatic event? Have Swedes greater or lesser trust in the system of justice compared with British people? Did the assassination cause any shift in political interest? These are all questions that will be addressed below.

The Threat or Why was He Assassinated?

In the study that was conducted three weeks after the murder, we found that the absolute majority of the Swedish people believed that the murderer intended to kill the Prime Minister, i.e. in their view the threat was real and the assassination was not a mistake. Almost 40 percent were of the opinion that a madman had killed Prime Minister Palme. This can be contrasted with President Kennedy's assassination when only about 15 percent of Americans interviewed considered that the accused person - Lee Harvey Oswald - was a madman (Sigel 1970). One explanation as to why Swedes rank higher in this respect is probably linked to the fact that no murderer had been arrested so there was no person to judge three weeks after the murder. In the United States, the respondents answered a question about Lee Harvey Oswald and not about an unknown killer.

The people had no clear picture in 1986 of the reason why Prime Minister
Palme was killed. About one-third believed that it was because of his
domestic policies, but his international peace work ranked almost as high.

In 1990 there was still frustration over why he was assassinated. There was no substantial change in the answering pattern compared with 1986. The same two reasons were given most frequently - his domestic policies and his international work for peace - but this time there was a shift so

that his international work was perceived as the most likely reason why he
was killed. This is also valid for 1992.

In our first study, we argued that one reason why Swedes were so calm after the assassination was their belief that it was the Social Democratic Party, and not Olof Palme as a person, that was in power and that Swedes recognize the whole party structure and its potential for peaceful and undramatic change. Our assumption turned out to be correct. The Social Democratic Party did fairly well in the elections in 1988 and stayed in power. It has continued on the same main path with Ingvar Carlsson as chairman of the party. There has been little visible change in its politics but, in spite of this, the party has been subject to a great deal of criticism and has been attacked for moving to the right in the domestic domain by both party members and supporters. Social Democratic domestic policies, however, did not change as dramatically as did Sweden's role in the international arena during the four years of Social Democratic governments following the assassination. This change of political focus in, what can be

called, perceptible politics may be the reason why Olof Palme's internationalpeace now ranks highest as a reason why he was killed. When asked about a concrete policy area such as immigration policy, which was also discussed a great deal during his time in office, very few people mention his policy in this area as a reason for his assassination. The Social Democrats lost the election in 1991 and Sweden got a right-wing government which remained in power in 1992 when our last study was conducted.

In the absence of a murderer, the number who think that the murderer was a madman has increased among young people, whereas the number answering "no" has increased among the parents and other adults from 23 percent in 1986 to 37 percent in 1992. The shift is explained by a decline in the number of people who answered "yes", from 37 percent in 1986 after the assassination to only 18 percent in 1992, six years later.

In 1986 we found no difference in relation to the question; "Why was Prime Minister Palme assassinated" between the Swedish respondents and immigrant respondents, between men and women, or between persons with different social backgrounds. The only differences we discovered were some irregular variations in the number of "perhaps" answers. The analyses for 1990 and 1992 provide basically the same results but there are some interesting differences. Persons born in Sweden believe to a lesser degree than immigrants that the Prime Minister was shot because of his domestic policies. It should be noted that we find this difference in the most frequently given reason for his assassination in both 1990 and 1992. Male immigrants, to a greater extent than Swedes, also believe that it was his policy towards immigrants that might have been the assassin's motive. This trend is very dominant among young people and is also valid for those who do not believe that his immigration policy was of any importance for what happened; Swedes are overrepresented in the group who do not think that this policy sphere had anything to do with his assassination. The hypothesis that someone disliked him personally and therefore assassinated him meets with very little support and the denial of this motive is strongest among women who have immigrated to Sweden. When the respondents are divided into different social groups, it becomes clear in both 1990 and 1992 that male blue-collar workers are more highly represented than expected among people who believe that he was assassinated because of his domestic policies or that someone disliked him personally. There is a clear trend for women of low social status to believe that he was killed because of his policies in Sweden. This may possibly be related to the fact that women from lower social groups view his programme on social policy as very threatening to the traditional patriarchal society (see Hernes 1987 or Pateman 1991).

The differences in the answering pattern between 1986, 1990 and 1992
indicate that people are, on the later occasions, more uncertain about the

reason why he was killed. The picture will probably become more and more diffuse as time goes by and no murderer is arrested and convicted for the assassination. The absence of a murderer will probably contribute to a great deal of speculation about, and presumably even effect the attitudes towards, the police and the judicial system. The failure to solve the assassination of such a prominent member of society as the Prime Minister is likely to raise many questions which will keep the trauma alive (see also Rothstein 1971). For instance, five years after the assassination there were programmes on radio and TV and articles in newspapers in which all aspects of the assassination and the murder hunt were discussed. Books have even been published on this theme. All this publicity will probably contribute towards prolonging the state of trauma. Examples of this kind of trauma exploitation can be found, from time to time, in American newspapers and books in relation to the assassination of PresidentJohn Kennedy. The speculations in the US arise from the assassinationof suspected murderer and, because of that, the absence of a convicted murderer and a solved case. The unsolved murder case in Sweden will probably, in the long run, give rise to the same kind of conjectures as those in the US and the more time that elapses between the murder and the articles written the more difficult will it be to have an objective discussion.

Trust in the Judicial System

Sweden can be characterized as an open society, but also as very controlled and closed society. There are a great many regulations that protect the individual from intervention by the state and other individuals. At the same time, the Swedish state is very collective in its nature. Society "looks after" its citizens in all stages of life. Nevertheless, even this collective part of life is regulated in such a way as to allow the individual to control what is going on. In a country where so much is regulated in order to keep society open, the kind of attitude the public has towards the controlling judicial system is important.

The police investigation following the assassination of Prime Minister Palme has shed a great deal of light on the way the judicial system works. Much of the discussion has focused on a comparison between the way in which the police and the courts actually operate and the way in which they are supposed to operate according to Swedish law.

One of the difficult questions that has been discussed in relation to the judicial system is the right for the police to tap people's telephones in order to get information. More than 75 percent (see Table 2 below) think that the police should have the right to tap the telephones of suspected

murderers. A much more divided attitude for both children and adults is found when the same statement concerns ordinary people suspected of shielding murderers. The proportion of people not wanting to allow the police to tap telephones increases and in 1986 the number of children who want to forbid this kind of activity is almost as large as the number who think it is right to bug ordinary people's phones, although it is less in the case of adults. By 1990, the view that the police should not have the right to tap ordinary people's telephones has gained support among both children and adults. Protecting the integrity of ordinary people seems to be importantto and parents alike.

Two of the statements focus on murder trials. The number of people who state that murderers should be entitled to a fair trial has increased among both the young and the elderly in 1990 compared with 1986. Even the number of people who argue against the statement that "murderers should be sentenced despite their denial if the evidence is weak" has increased.

In 1986 we found very few differences related to gender and citizenship. Among well educated persons and people who in some respect could be considered to belong to the upper strata of society, we found a greater belief in the principle of liberal justice as measured in the statements above. When asked in 1990, those very questions result in much greater diversity.

Boys are more, and girls less, frequently represented than expected among those who express the view that murderers should be convicted against their denial even if the evidence is weak. The same pattern is found for the statement that the police should have the right to tap the phones of people who are suspected of assisting murderers. Male immigrants are also found in greater numbers than expected among those who believe that murderers should be convicted against their denial, if the evidence is weak, and males born in Sweden are clearly underrepresented in this group. The same pattern is valid for women.

When this statement is controlled for social group, we find that very few in the two highest social groups believe that a murderer should be convicted against his denial, if the evidence is weak. In lower social strata this is a more common opinion. The same pattern is found for fathers whereas it is only valid for mothers from working-class homes. It is very clear from the analysis that the statement about the value of proof in murder trials has divided the respondents into subgroups more strongly in 1990 than directly after the assassination in 1986. One possible explanation is perhaps that the discussions concerning the judicial process following the assassination have made people more aware of all the problems and conflicts in the judicial system.

In 1992 we replicated questions about the system of justice that had been
used in Britain. In Table 3 below, we compare our results with empirical
data for Britain (excluding Northern Ireland) from 1987 and 1992.

We can see clearly from the table that both Swedish and British citizens were in favour of stiffer sentences for people who break the law. This is interesting because it shows a strong belief in a traditional and conservative way of looking at crime prevention. We know from research in this field that there is only a weak relationship between crime prevention and the length of the punishment and that longer and stiffer sentences, in many cases, have instead the opposite effect (see, for example, Brody 1980). Swedes have a stronger belief in the law-making system. They believe that it is possible to determine when a particular law is unjust and, when this is the case, almost two-thirds of the Swedes think that the law should not be obeyed. British people have a much more formal attitude to the law and most think that even if a law is unfair it should always be obeyed.

The greatest difference between the two countries concerns attitudes relating to the death penalty. In this case two-thirds of the Swedes are against capital punishment, whereas the same proportion of British people approve of the death penalty as the best form of punishment for some crimes. One reason for the considerable difference between the two countries this respect might be found in the structure of crimes in the two countries. In Sweden there are no terrorist actions - despite the murder of Prime Minister Palme - whilst in Britain bomb alerts, disturbing at least part of the population, had become something of a daily occurrence. The absolute numbers of murders are also higher in Britain than in Sweden. All this, together with the fact that Britain has had a Conservative government over 20 years expressing concern for law and order and advocating stiffer sentences, may explain the difference between the two countries. It is also worth keeping in mind that Britain has a tradition of using the death penalty as an ultimate resort.

Finally, in both Sweden and Britain about two-thirds of the population trust the police not to try to bend the rules in order to get a conviction. In Sweden the remaining 33 percent are more negative than in Britain. Almost 22 percent of the Swedes think that the police do bend the rules from time to time in order to get binding evidence. The corresponding figure for Britain is 7 percent. This can very well be an effect of the debate concerning the police investigation of the assassination of Prime Minister Palme.

The impact of the independent variables differs somewhat from one statement to the other. There is a clear relationship between social group and attitudes to stiffer sentences for people who break the law in Sweden and Britain alike. The higher the social status, the more critical the attitudes towards stiffer penalty levels. In Britain, older people seem to be more in favour of stiffer sentences. A corresponding relationship cannot be found for Sweden. Women in Britain are also more inclined to accept higher sentences for crimes: again such a relationship does not exist in Sweden.

Immigrants to Sweden are more in favour of stiffer sentences than native

Swedes. Immigrants also believe more often than Swedes that the law
should be obeyed even if it is wrong and unjust.

Men, in both Britain and Sweden, believe more often than women that the law should be obeyed even if it is wrong and unjust. In Britain, but not in Sweden, there is a relationship between age and attitudes on this item. The older the British citizens are, the more often they think that the law should be obeyed even if it is wrong. Once again we find clear trends for social class, but these are different in the two countries. In Sweden people with higher social status are less inclined to agree that the law should always be obeyed than those from lower social groups. In Britain we find the reverse pattern.

With regard to the death penalty, there is a slight tendency in Sweden for men to be more in favour than women but not in Britain. The explanation this relationship is that male immigrants to Sweden support this proposition much more often than Swedes, which affects the result for gender. The expected pattern that people from lower social groups are more in favour of the death penalty than those from higher social groups is valid for both Britain and Sweden. In Britain, this pattern also applies for age. The older people are, the more inclined they are to accept the death penalty as an appropriate punishment for some crimes. In Sweden, on the other hand, the reversed pattern emerges but when the data are controlled for ethnic background the same pattern emerges as that in Britain.

Our final comparative question addressed itself to the confidence in the police force. And again for Sweden, the immigrant group differs in response from the Swedish-born population. Swedes trust the police to a much higher degree, 72 percent as opposed to 54 percent for the immigrants. Twenty percent of the native-born Swedes and 27 percent of the immigrants openly declare that they distrust the police force. The result is not unexpected. Many immigrants have probably had bad experiences from contacts with the police both in their home countries and with the Swedish police in connection with their immigration to Sweden. The only strong covariation we find for this statement is with social background and it applies in both countries. The pattern is that the higher the social background the respondent has, the more confident he or she is that the police will not try to bend the rules in order to get a conviction. The result can probably be explained in the same way as that accounting for the behaviour of the immigrants. People from higher social backgrounds have, in general, been less involved with the police with regard to criminal offences than people from lower social backgrounds (see, for example, Ahlberg 1992) and have, therefore, a more positive view of the police force.

When condensing our comparison of the opinions of people in Britain
and Sweden on the legal framework in the two countries it is clear that the

basic political culture in the two countries has affected the results. In general, however, there are many similarities between the attitudes held by the British and the Swedes. The differences relating to culture concern a greater acceptance of the death penalty and a more formal way of looking at the legal system in Britain than in Sweden. The considerable difference between Swedes and immigrants to Sweden in these matters is also interesting.A proportion of the immigrants have attitudes that are more infiuenced by experiences from their home country than are influenced by the Swedish culture.

Expressed Political Interest and Future Private Goals

Children's and parents' political interests have been measured by asking them to state whether they intend to participate in the political process. The first aspect concerns their disposition to join a political party in the near future. The second question is formulated as follows: Do you plan to join any sort of voluntary organization, e.g. an environmental group? Both children and adults have a more positive attitude towards becoming a member of a political party in 1990 than in 1986 (see Table 4 below). The proportion, in both cohorts, who can conceive of themselves as party members in the future has increased substantially. The most interesting shift can be noted among adults. More than one third of them are ready to join a political party. This marked shift is very difficult to explain. The same strong increase can be observed when the respondents are asked about their willingness to join an environmental organization. More than every other person states that they can think of themselves as members of an organization of this sort. This positive attitude towards active participationin political activity can also be found in our last question. This was constructed to measure one aspect of the so-called post-materialist values, i.e. the importance of material values versus life fulfilment values such as one's own ideals (see Inglehart 1977). Our question is phrased: "What is most important for you in the future, to earn a lot of money or work for an ideal in which you believe?" Among both the children and the adults the proportion expressing an interest in working for an ideal is fairly constant over time. The largest shift over time can be found among the children. Most of the children who in 1986 stated that they "did not know" have, in 1990, switched to the position that it is most important to make a lot of money. Even the adults show the same pattern, but the tendency is weaker and for them the post-material values are still the most important. The strong shift to material values among the children has turned them into a very mixed group. The proportion is about the same in

both the materialist and post-materialist groups. In conclusion, the overall picture in 1990 still looks promising for a future society in which there is a high level of participation by interested citizens, if the expressed political interest is to be used as the predictor.

In 1986, the variations between different groups in relation to expressed political interest were very conventional. Men and boys were more engaged than women and girls, people with a high level of education were more active than those with a low level of education, etc., but there were no variations relating to ethnic background. Four years later, in 1990, men are still more positive than women to the idea of joining a political party in the near future. Women seem to be more attracted to new movements and can, in a greater proportion than men, think of themselves as members of an environmental organization in the future (see, for example, Eduards 1992 and SOU 1990:44). This attitude is also reflected in the fact that women are in the majority among those who want to work for an ideal, whereas men aim at making money. When it comes to the child cohort, boys make more negative statements than girls about membership of political parties. In the same way as in the adult cohort, girls are more in favour of being a member of an environmental organization than boys. Only 10 percent of the girls say that they cannot see themselves as members of an environmental organization. The corresponding figure for boys is 35 percent. When it comes to the question of what is most important in the future - making money or working for an ideal - boys more often think of

themselves as hard-working money-makers, whereas girls put more stress
on working for ideals they believe are important.

Another dimension concerns place of birth. Boys and girls born in Sweden are more negative to membership in political parties than youngsters have immigrated to Sweden. The opposite pattern is valid for environmental organizations. The tendency is very strong, every second child born in Sweden compared with every fourth child not born in Sweden. The same pattern can be found for this statement for the older generation when they are divided into those born in Sweden and immigrants. The pattern is clear but weaker. Finally, when it comes to the question of making money or working for an ideal, most Swedish women express the view that it is more important to work for an ideal they believe in than to make a lot of money. The difference between men and boys born in Sweden or abroad is not statistically significant in this respect.

When the results are controlled for social group, it is clear, both for youngsters and for men, that the higher the social group they belong to the more positive they are to membership both in political parties and in environmental organizations. It also seems that work satisfaction is of greater importance the higher one's social group. The women do not comply with this pattern. Instead, they are more positive to membership in environmental organizations, the higher their social status. Social group does not seem to have any importance when women express what is most important for them in the future, to earn a lot of money or to work for an ideal they believe in.

There is no doubt whatsoever that the picture in 1990 and 1992 is much more diffuse than that of 1986. The results have in some sense gone back to a normal pattern - e.g. differences between social groups, between Swedes and immigrants and between boys and girls. The variation is not totally new but is more distinct than in 1986.

Is there an Effect on the Political Culture?

What can we learn from these studies of the effects of the assassination of Prime Minister Olof Palme on Swedish political culture? In 1990 and 1992, we found more covariation between different background factors, such as social group and country of birth, and our measures of political culture than in 1986. One reason is probably that on the first occasion people, both young and old, were more emotionally influenced by the assassination when answering our questions than they were four and six years later. Four years after the assassination, the trauma of an unsolved murder case was also beginning to grow. Articles can be found in newspapers, TV programmes discuss the situation critically and books have been published

(see, for example, Holmér 1988 and Elmbrant 1989). Finally, news about the murder investigation still gets front page or headline coverage after all these years even when it is trivial and of no importance in relation to the chances of convicting a murderer. The considerable increase in people answering that they do not believe "that a murderer should be convicted against his/her denial, if the evidence is weak but still clear" can be viewed as an expression of the situation with no suspected or convicted murderer four years after the tragic event. On the whole, we find a clear growth in support for the traditional liberal standpoints on law and justice. The questions of the value of proof and personal integrity have been linked together by the respondents in a way that did not exist in the highly emotional situation directly after the murder. The hypothesis that all the uncertainty around the assassination could be viewed as having a structural effect on the beliefs and attitudes of Swedes in areas related to the assassinationand, a broader scale, the political culture is, at least to some degree, verified by the results in sections 3 to 5 above. We have found a greater uncertainty about the reason why he was killed; greater belief in traditional codes for law and justice; and an increased readiness to participatein and voluntary organizations as well as an expressed interest to work for a non-material future goal rather than purely to make a lot of money.

The unanswered question of why he was assassinated is built around two assumptions. According to the first, the threat was aimed at Olof Palme as a person in his political role and according to the second version, the threat was directed at the Swedish government of which Prime Minister Olof Palme was the highest ranking representative. If the threat was of the second type, then it was aimed at the whole Swedish society and the danger is that such an act of terrorism can be repeated. When asked about the two assumptions, most people linked their answer to the political person Olof Palme. So in that sense, the entire Swedish political system was not in danger; the threat was aimed at a very outspoken and dominant politician. Again people's emotions and beliefs about the person rather than about the political system can be traced in the answers.

The final argument that can be presented from this study of the connection exposure to a dramatic event, such as the murder of a prime minister, and children's and adults' political values, is that the emotional effect of the assassination fades away fairly quickly and is replaced by a much more vague and unclear structural effect related to the total impact of the assassination seen as a dramatic event of national importance. This kind of structural effect on the political culture in a country can never be clearly described and analysed. The reason is that an effect of this magnitude is almost impossible to control and isolate from other experiences.

REFERENCES

Ahlberg, J. ed., 1992. Brottsutvecklingen 1991. Brottsförebyggande Rådet, Brå-rapport
1992:2, Stockholm.

Almond, G. 1956. "Comparative political systems", Journal of Politics 18, 391-401.

Almond, G. A. & Verba, S. 1963. The Civic Culture. Political Attitudes and Democracy in
Five Nations. New Jersey: Princeton University Press.

Beck, P. A. 1977. "The Role of Agents in Political Socialization", in Renshon, S. A., ed.,
Handbook of Political Socialization. New York: The Free Press.

Brody, S. 1980. "Sentences as Punishment", in Påföljdsval, straffmätning och straffvärde.
Brå-rapport 1980:2, Stockholm.

Dawson, R. E. & Prewitt, K. 1969. Political Socialization. Boston: Little, Brown and
Company.

Easton, D. & Dennis, J. 1970. "The Child's Image of Government", in Adler, N. &
Harrington, C, eds., The Learning of Political Behavior. Glenview, Illinois: Scott, Foresman
Company.

Edelman, M. & Simon, R. J. 1971. "Presidential Assassinations: Their Meaning and Impact
on American Society", in Crotty, W. J., ed., Assassinations and the Political Order. New
York: Harper & Row, Publishers.

Eduards, M. L. 1992. "Against the Rules of the Game. On the Importance of Woman's
Collective Actions", in Rethinking Change. Current Swedish Feminist Research. Stockholm:
Humanistisk-Samhällsvetenskapliga forskningsrådet.

Elmbrant, B. 1989. Palme. Stockholm: Författarförlaget Fischer & Rye.

Gustafsson, G. 1972. "Strukturomvandling och politisk socialisation", Umeå: Statsvetenskapliga
Umeå universitet.

Gustafsson, G., Jacobsson, J.-E. & Johansson, O. 1987. "Reaktioner på Palmemordet",
Politik 1987:1, Statsvetenskapliga föreningen, Umeå.

Hernes, H. M. 1987. Welfare State and Women Power- Essays in State Feminism. Norwegian
University Press, Oslo.

Holmér, H. 1988. Olof Palme är skjutenl Stockholm: Wahlström & Widstrand.

Inglehart, R. 1977. The Silent Revolution- Changing Values and Political Style Among Western
Publics. Princeton, New Jersey: Princeton University Press.

Jacobsson, J.-E. 1994. Politik, föräldrar och barn - en studie av föräldrarnas betydelse för
barns politiska kultur (Draft from a forthcoming dissertation.)

Johansson, O. 1985. Ungdom och massmedia, Forskningsrapport 1985:13. Umeå: Statsvetenskapliga
Umeå universitet.

Johansson, O. 1991. "Swedes' Reactions - 4 years later - to the assassination of the Swedish
Prime Minister Olof Palme", Research Report 1991:1. Umeå: Statsvetenskapliga institutionen,
universitet.

Langton, K. P. 1969. Political Socialization. London: Oxford University Press Inc.

Marvick, D. & Marvick, E. W. 1971. "The Political Consequences of Assassination", in
Crotty, W. J., ed., Assassinations and the Political Order. New York: Harper & Row,
Publishers.

Pateman, C. 1991. "A New Political Theory? Political Science, The Public and The Private".
Paper presented at IPCA XV World Congress, Buenos Aries, 21-25 July 1991.

Rothstein, D. A. 1971. "Presidential Assassination Syndrome: A Psychiatric Study of the
Threat, the Deed, and the Message", in Crotty, W. J., ed., Assassinations and the Political
Order. New York: Harper & Row, Publishers.

Rutter, M. et al. 1969. Fifteen Thousand Hours - Secondary Schools and Their Effects on
Children. London: Open Books Publishing Limited.

Sigel, R. S. 1965. "Assumptions About the Learning of Political Values", The Annals, 361,
1.

Sigel, R. S. 1970. "An Exploration Into Some Aspects of Political Socialization: School
Children's Reactions to the Death of a President", in Sigel, R. S., ed., Learning About
Politics. New York: Random House.

SOU 1990:44 Demokrati och Makt i Sverige. Maktutredningens huvudrapport. Stockholm:
Allmänna Förlaget.