Resumé

The purpose of this study is to analyze the impact of political socialization and
life course on the ideological involvement of two cohorts of Norwegian men. Our
main thesis is that socialization in a period dominated by political and social
conflicts will produce a higher level of ideological involvement than socialization
in a period with a low level of political and social conflicts. But the higher level
of ideological involvement can be modified by the life course experiences of the
cohorts. We are only able to discuss some aspects of this rather general thesis in
this article, comparing two cohorts born in 1912 and 1932.

1. Introduction

There is a relatively strong tradition in sociology and political science which identifies age cohorts as generations according to the specific historic forces operating during the 'formation years'. The classic contribution is of course Karl Manheim's essay on The Problem of Generations (Mannheim 1952). He sees the changing intellectual, social and political circumstances in society as capable of producing different political and intellectual attitudes among the youth socialized in a specific period. Mannheim's analysis focuses especially on intellectual groups and organized political groups, or what he calls 'generational units'.

One of the central questions in a theory of generations that seeks to build upon Mannheim's work is to define the 'formation years' more precisely: Mannheim speaks about 'youth' without qualifying the term. Others have proposed more definite intervals of agt;. In an authoritative article, Marvin Rintala sees the age period between 17 and 25 years as crucial for the formation of political attitudes and ideology (Rintala 1968). T. Allen Lambert is very close to Rintala, arguing that the most important interval for development of political-cultural consciousness is from 18 to 26 years

(Lambert 1972, pp. 24-25). And it seems that this interval, with small modifications, can be taken as valid for studies in our culture. The bulk of empirical studies utilizing the concept of generation or age-cohort has stressed the significance of the first vote and the specific political circurnstances of the years surrounding this vote. An important exception to this generalization is Peter Loewenberg in his study of'the nazi youth cohort'. He sees experiences during childhood in the First World War as crucial, forming the susceptibility to the nazi ideology in this cohort (Loewenberg 1971).

More recently there has been an important development in empirical studies that look at the general pattern of values of cohorts socialized under radically different economic and social conditions.¹ Cohorts socialized in the depression years of the 1920s and 1930s prefer material values to a higher extent than cohorts socialized in the post war period of better economic conditions. Ronald Inglehart characterizes the latter cohorts as Post-Materialist in their value preferences. He further sees the process of cohort replacement as a continuous 'value-revolution': Older, materialist cohorts are being replaced by younger cohorts oriented more to non-material values (concern with self-expression and the quality of life). We see the cohort approach as fruitful for the study of a wide range of questions concerning political attitudes and behaviour. In this article we limit the analysis to the question of the effect of cohort-specific socialization and cohort-specific life courses on the level of ideological involvement. Before we present our data, we shall comment briefly on the cohort model, a definition of what we call ideological involvement, and then specify our hypotheses.

2. The Cohort Model

The cohort model is commonly seen as a three-factor model.² There are three separate factors which can produce, and thus explain, the variance in the dependent variable. Both age, the cohort factor, and period can have separate effects on the dependent variable. There is a fundamental problem of separating these general factors. If we study two or more cohorts at the same period, the members of the cohorts will have different ages, and we cannot be sure that differences between the cohorts are caused by age or by the cohort factor. And if we study cohorts at the same age this will be at different time-periods, and we cannot be sure that this can explain the variance. The discussion of how to handle the problem of identification

has been rather extensive, and we shall not go into the controversy here.3

Since our data consist of interviews collected at the same time, the period factor is unproblematic. But the two cohorts born in 1912 and 1932 were aged 59 and 32 when interviewed. While the objective is to see if we can explain differences between the two cohorts from the cohort factor itself, in the discussion of our findings we must return to the question of a possible influence of age on the level of ideological involvement in the cohorts.

It is evident from most of the studies that it is the cohort factor itself which raises the most serious difficulties. We have already mentioned the problem of defining the crucial years of impact in the socialization process, but this is only one part of the problem. To assume that the formation of cohort-specific properties is determined in a limited time-period during youth is of course a simplification. Socialization will normally take place throughout the whole lifespan. But, as Gøsta Carlsson and Katarina Karlsson point out, it can for a number of reasons be of less importance (Carlsson and Karlsson 1970). The individual tends to remain on the first chosen track of values and behaviour. In other words, a model ofrigidity is proposed. And it is this assumption of rigidity which underlies most of the research on political attitudes, ideology, and behaviour. While appreciating much of the work done under the assumption of a preserved cohort socialization, we will try to add a new dimension to the cohort factor. Just as cohorts may have specific sociali2;ation periods, they may also have specific life courses. By this we mean that each cohort lives through a period which is particularly its own. Only members of the same cohort have their youth, middle age, and old age at the same time-periods. Therefore the life courses of cohorts may be seen as unique. We should not overemphasize this uniqueness, especially if we think of age cohorts as being born in years following each other. In this case there are continuous transitions between cohorts.

We can trace the life course of a cohort in many ways. We can analyze cohorts by emphasizing the changes in the historical context through which the cohorts live. This can be changes in the economy, social welfare, political conflicts in society, etc. It is seldom satisfying to analyze the life courses of cohorts only in such a contextual manner. To give a more complete picture of the life course of a cohort, we can go further and collect data on the members of the cohort itself. This can be done for example by collecting life histories of the members of the cohort.⁴ Our thesis is that the specific life course of a cohort will influence the attitudes and behaviour of the members of the cohort. In other words we cannot say

that the cohort factor is determined by socialization alone we must take
the complete life course of the cohort into account.

3. Ideological Involvement

As the central element in the concept of ideological involvement, we here take the holding of attitudes which characterize society as sociopolitically divided or politicized. Another use of this concept would define ideological involvement as involvement with systematic ideologies (of which communism and fascism are extremes). It is likely that a high degree of involvement with systematic ideologies also will reflect a society which is sociopolitically divided and politicized.

Our definition of the concept of ideological involvement is partly related to the much debated thesis of'deideologization', so popular in the debates of the early 1960s.⁵ This thesis (which the political development in Western Europe and in the United States during the later 1960s has so dramatically falsified, at least in its most universal formulation), postulated a levelling off of class differences in post-industrial society, and the replacement of politics as a way of solving conflicts by professional or technical solutions.

The operationalizations are restricted due to the post facto character of the data. The questions we use deal with the perception of class-cleavages in society, people's identification with a particular class, the perception of the distance between the main political blocks, the development of the Labor party from the 1930s until today, and the politicization of the unions.

We consider people who hold the opinion that there are classes in society today as more involved than those who think that classes do not exist. And we consider individuals who can place themselves as a member of a class and are sure of this position as more involved than those who are not able to do so. Class is one of the strongest forces which can divide society. Consequently we assume that attitudes which stress class divisions and class positions as relevant for describing society or own position in society as a measure of the ideological involvement of the individual.

A person who believes that the distance between the main political blocks (the Labor party and the bourgeois parties) is great is more involved than one who believes that the distance is small or unimportant. An individual holding the opinion that the Labor party is much the same now as it was in the 1930s is regarded as more involved than one who thinks that the Labor party has moved to the centre (or in a conservative direction).

We will assume that to hold the opinion that the Labor party has not changed fundamentally since the 192>0s also implies that the party is regarded as a clearer socialist or working-class party. To retain such a picture of the Labor party we here take as an involved stance.

Finally, we will analyze two questions on the relation between the unions and politics. On the first question we consider a person as involved if he believes that unions should be active in politics. On the second question we consider a person as involved if he believes it is right that unions can be collectively affiliated with the Labor party.

4. Hypotheses

Working from the assumption of cohort-specific socialization, we expect cohorts socialized in periods with a high level of political conflict and clearcut alternatives in politics to show a higher level of ideological involvement than cohorts socialized in periods with both a low conflict level and a tendency of cooperation across the main dividing lines in party politics. For Norway we see the period between the two World Wars as an example of a period with a high level of political conflict and manifest ideologization of political debate. The period following the Second World War (until around 1960) we see as a {period with a low level of political conflict and also with a dying-off of manifest ideology. Consequently we expect the 1912-cohort to be more ideologically involved than the 1932cohort because of differences in the socialization periods.

If we follow the other line of argumentation from the discussion of the cohort factor, the result will be rather different. The attitudes and opinions of a cohort can also be influenced by its life course. In our case we can especially note the contrast factors of the life course of the older cohort.⁶

This cohort has experienced both the interwar- and the post-war period. In politics and ideology the members of this cohort have experienced both a period with a high level of conflict and one with a low level of conflict. We assume that the life course and the life-course experiences of this cohort will also have significant influence upon their current attitudes and opinions. What we especially focus on in connection with ideological involvement is the modifying influence of the contrast factors of the life course. Differences today in class structure, between the political parties etc., will be interpreted in the light of the past. The 1912-cohort will understand political differences at present, and the development of differences and contradictions, from a frame of reference determined by the

contrast between past and present. Therefore it is reasonable to expect this cohort to show a lower level of ideological involvement than the 1932-cohort, since the latter lacks the concrete experience of the contrast between the pre- and post-war periods. But this gives the opposite result of that assumed from the socialization argument. Our answer to this dilemma is that we see, as already mentioned, the effect of the life course as only a modifying factor. This means that the basic effect of cohort socialization will persist over time, but not with an equal degree of consistency for all dimensions of ideological involvement. We expect the modifying effect of the contrast factors of the life course of the 1912-cohort to be stronger in dimensions where there is a more or less explicit reference to change, that is where there are trends in differences and contradictions.

The relevance of the Second World War is important. Both cohorts have experienced the war, but at a different age. The 1932-cohort had its childhood during the war years, for the 1912-cohort the war came in adult life. The impact of war on ideological involvement is most evident in its ability to create national unity. This means that traditional sociopolitical conflicts will be reduced. If we assume that war reduced the ideological involvement of the citizens, this effect will be strongest for the 1912-cohort. The younger cohort is too young to be directly influenced, at least when we focus on political ideology.

It is perhaps appropriate at this point in the discussion to emphasize that the life-course factor is general: it operates on all cohorts. The 1932-cohort has a life course of its own and this may have relevance for a number of variables. In relation to ideological involvement, however, we expect the contrast factors in the life course of the 1912-cohort to be stronger, and thus to have a greater impact as a modifying force upon the level of ideological involvement today. In general we can expect a number of different relations between socialization and life course. Our hypotheses in this article are restricted to the analysis of differences in ideological involvement between two cohorts.

To sum up, we expect the 1912-cohort to be the more ideologically involved cohort. On dimensions with more or less explicit reference i:o comparisons between earlier periods in the life of the cohort and the present situation, we expect the difference between the cohorts on the question of ideological involvement to be less. And we can also think of the relation between cohorts being reversed if the modifying influence of the life course as a reference factor is strong. In Table 1 we give an overview of the discussion up to this point.

5. The Data

The data were collected in 1971 as part of a study of the living conditions in Norwegian society over the past 50 years.⁷ The measurement of ideology (including ideological involvement) was not a focal point of the research, hence our data are limited. The sample was drawn from all men of two cohorts bom in 1912 and 1932, living at the time of the interviewing in the city of Trondheim. The data consist of complete interviews with 285 men born in 1912 and 203 born in 1932.

Even if we omit possible influence of age on ideological involvement, there remain a number of factors which must be taken into account when comparing the cohorts. The cohort factor and the age factor account for differences in background variables of the cohorts. And since it is possible that background variables will influence our dependent variable, we

must control for occupation and education. In addition we shall also
control for political preference.

We shall first present data where the life course is assumed to have a
strong modifying influence, and then data where the influence of cohortspecific
socialization is assumed to be persistent.

On the question of whether there are classes in the Norwegian society today, the highest positive response came from the 1932-cohort. As can be expected from a theory of interest (based on place in the class structure), workers show the highest positive response. And we also see that voters preferring the left socialists or Labor give a more positive response than the supporters of the bourgeois parties. This can also be expected on the basis of differences in ideology between the parties: the socialist and social-democratic parties being more eager to analyze fundamental properties of society in class terms. Though the 1912-cohort was socialized in a period of socio-economic crises and manifest class conflicts, it shows a lower level of ideological involvement as measured by this variable. In accordance with our expectations, this is not surprising, since we also expect the life course to play a role. Though socialization during a situation of crisis and class polarization is likely to produce a picture of society divided by class, this will be modified by a life course which has 'followed' the emergence of the 'welfare society'. The life course reference then

produces a situation where it will become more and more difficult to retain the picture of a class-divided society. The modifying effect of the life course is hence important in order to understand the degree to which the 1912-cohort sees contemporary Norwegian society as a class society or

since we have formulated it more weakly, a society in which classes exist. This life-course reference is much weaker for the 1932-cohort because it lacks the immediate experience of the contrast between the pre- and post-war periods.

On the question of party distance, we take the involved persons to be those who think that the distance between the bourgeois parties and the Labor party is 'very great' or 'great'. Only a minority in each cohort holds this opinion. One cannot say from our data that there are any significant differences. As for the question on classes in contemporary society we assume that the socialization effect has been modified by life-course experiences. We assume that the socialization of the 1912-cohort in the turbulent late 1920s and 1930s has produced an image of a great distance between the main party blocks. However, this image of party distance will appear in the light of their experience of the development of a higher degree of consensus between the blocks. The result of this process is that the cohorts seem to be about equally involved. The youngest cohort is more ideologically involved over the question of the development of the Labor party. This means that this cohort sees the party as not having moved in a conservative direction. This relation holds for all occupational groups except for the self-employed and the professionals, and for the highest educational group. A possible explanation of this pattern can be that the group with the highest education evaluates the development of the Labor party from a more 'intellectual' point of view. A more scientific evaluation of the Labor party could hardly escape the conclusion that the party has moved to the right over the decades. It seems that the members of the 1932-cohort with lowest levels of education are more prone to think that the Labor party has not changed fundamentally. There is no correlation with education in the 1912-cohort. A trivariate breakdown by political preference and education shows that the pattern is different in the cohorts ⁸ While there is no relation between opinion on the development of the Labor party and education for the 1912-cohort, the breakdown shows education to be the important variable for the 1932-cohort. The more education the higher the percentage in the cohort which thinks that the Labor party has moved in a conservative direction. If we explain the high percentage in the 1912-cohort which believes that the Labor party has moved in a conservative direction by the life-course contrast of this cohort, we can understand the role of education for the 1932-cohort in an analogous manner. Immediate life-course experience of the actual historical period causes members of the older cohort to evaluate the development of the Labor party as a movement to the right. Education, as a

mediated form of experience of the same historical period, causes the
members of the younger cohort to hold the same view on the development
of the Labor party after the 19305.

All the three aspects of ideological involvement in Table 2 show a pattern of no difference, or a slight tendency towards more involvement by the 1932-cohort. This might seem unexpected on the basis of arguments derived from a theory of cohorts, in which socialization under specific historic circumstances is seen as crucial for the forming of ideology. However, when we extend the cohort model to take life-course experiences into account, this becomes understandable. Although the older cohort is socialized in a period when differences between the classes were clear, when antagonism between the Labor party and the bourgeois parties was evident, and when the Labor party took a revolutionary stance, the postulated ideological involvement resulting from socialization in this context may have been modified through the life course of the cohort. The modification can give results which show the youngest cohort to be most involved.

We can now turn to those questions which lack explicit time reference. Here comparison with earlier periods in the life course is not of immediate relevance. We have postulated the 1912-cohort to be the more involved since it will retain its basic socialization. We present the data in Table 3.

The question of being sure or unsure of own class identification is asked only of persons who could place themselves in one of two given classes: working class and middle class. Our reasoning is that class identification would be more common where an individual is socialized in a society with prevailing class antagonisms. As to our basic hypotheses, this means that the older cohort is formed by the class struggle in the period between the wars, and therefore it is likely that its members will tend to be sure of their class identification. The younger cohort, we see as formed in a period with

a lower level of manifest class-struggle. It is likely that the members of this cohort will have difficulties in clearly identifying themselves with a class. The question does not allow for a time reference, and we consequently expect the socialization effect to be persistent. But we may remind the reader that we see a modifying effect of the life course working also here. The life-course reference of the 1912-cohort will influence their class identication, and it seems reasonable to think that it will weaken a particular identification as compared to the identification in earlier periods of life. But we see the modifying effect as weaker compared to variables where the comparison with earlier periods is more evident. The two questions on the politicization of the labor unions show roughly the same pattern. There is a clear tendency in the direction of the 1912-cohort to be more in favor of a close connection between activities in the unions and the politics of the parties. For the bourgeois-voters only there is no difference between the cohorts on the active-union issue. Very few of this group think that the unions should be active in politics. The question of collective membership provides one further exception to the overall pattern; this time for the political group at the opposite end of the spectrum, the Left socialists. The most interesting poini: to note is perhaps the difference between the Labor voters of the two cohorts. While 51 percent of the Labor voters in the 1912-cohort say that the unions should be active in politics, the corresponding figure for the 1932-cohort is only 35 percent. As for the question of collective affiliation, the percentages in the two cohorts are 59 and 32 respectively. If we assume this tendency to be general, it can only mean that replacement of the older Labor cohorts by younger cohorts in the unions also leads to a weakening of the ties between the unions and the party. Further analysis shows that it is among the members of unions that opinions on the politicization of the unions are most divided.

The second part of our data analysis gives a picture somewhat different from that of the first. While the latter showed that there existed no differences between the cohorts, or, if so, one in favor of the 1932-cohort being most involved, the result was reversed in the eecond part of the analysis. Now it is the 1912-cohort which seems to be most involved. But the apparent contradiction of these findings disappears in the light of our theory. The assumed higher ideological involvement of the 1912-cohort due to the socialization of this cohort in the inter-war period will be modified by the life-course experiences of this cohort. These experiences give the members of the cohort a frame of reference for cognitions and evaluations of the development of society.

6. Discussion

On the basis of our limited set of data (few variables, small N's) we have i;o be careful about presenting any definitive conclusions. It seems, however, that the extension of the cohort model may have been fruitful. The traditional approach, where only socialization effects are taken into account, gives only one part of the cohort argument. Cohorts do not only have specific socialization periods, they also have specific life courses. In this article we have tried to use these two aspects of the cohort model to explain the differences of ideological involvement in two cohorts. As the analysis shows, the socialization factor and the life-course factor do not necessarily produce the same results. In our case they work against each other. But this cannot be generalized: for other variables the result may be otherwise.

One of the factors not controlled for in our analysis is age. At the time of the interviewing, the cohorts were aged 59 and 39. The most common assumption about age and ideology is that age is positively correlated to conservatism (Foner 1972). Though this assumption can be questioned, it is perhaps relevant to the problem of ideological involvement. At least some of the aspects of ideological involvement which we have discussed here, can be related to the radical-conservative dimension. The two questions that seem most reasonable for this dimension are those on classes in Norwegian society today and the question about political activities in the unions. If we expect these questions to measure a radical-conservative dimension, we will also expect them to be related both to occupation (with workers holding the most radical attitudes), and to party preference (with socialists holding the most radical attitudes). Moreover on both these questions we see ideological involvement as the radical stance. Tables 2 and 3 seem to confirm these expectations. But when we compare the cohorts, the younger cohort is more radical on the first variable and the older cohort is more radical on the second. Allthough this can seem a superficial way of argument, an explanation from age differences appears to be less rewarding than our extended cohort approach.

To understand the ideology of people we have to analyze them as members of political generations. The past is living in the present generations and has a significant influence on their attitudes and behavior. The political ideology of a generation reflects the historical experiences of its members. If we accept that there was a general trend toward deideologization and political consensus in Norway in the first two decades following the Second World War, this can in our model be accounted for by two different processes: socialization and life course. The lower level of ideological

involvement of the generation maturing in post-war society is due to the fact that this generation was politically socialized in a period with political cooperation and diminishing class cleavages. The deideologization of the pre-war generation was caused mainly by this generation's experience of the welfare society, which forms a contrast to the social and economic conditions of the 19305. For the still younger cohorts which had their youth in the last half of the 19605, these two forces of deideologization had lost their influence: a more ideologically involved generation was born.

---

NOTES

1. The most important contributions have been made by Ronald Inglehart. A summing up of his research is found in his latest book (Inglehart 1977).

2. One of the more precise statements on the nature of the cohort model is found in Karen Oppenheim Mason et al. (1973).

3. See, for example, Glenn (1976), Mason et al. (1976).

4. Life histories were collected at the Department of Sociology and Social Studies, University of Trondheim, under the direction of Tore Lindbekk, and at the Institute of Applied Social Science, Oslo, under the direction of Natalie Rogoff Ramsøy.

5. The clearest statment of this thesis dorthe Norwegian political system is found in Torgersen (1962).

6. A discussion of the possibility of using life course as a reference factor basedon the well known reference group theory is found in Listhaug (1975).

7. A more extensive discussion of the data is given in Listhaug (1977).

8. See Listhaug (1977), table 24.

REFERENCES

Carlsson, Gösta and Karlsson, Katarina 1970. 'Age, Cohorts and the Generation of Generations',
American Sociological Review 35, 710-718.

Foner, Anne 1972. 'The Polity'. Mathilda White Riley et al.: Aging and Society, vol. 3.
Russel Sage Foundation, New York.

Glenn, Norval D. 1976. 'Comment'. American Sociological Review 41, 900-908.

Inglehart, Ronald 1977. The Silent Revolution: Changing Values and Political Styles among
Western Publics. Princeton Univ. Press, Princeton N.J.

Lambert, T. Allen 1972. 'Generation and Change. Toward a Theory of Generations as a
Force in Historical Process'. Youth and Society 4, 21-46.

Listhaug, Ola 1975. Livsløpet som referansefaktor (Life Course as a Reference Factor) Univ.
of Trondheim, Dept. of Sociology and Social Studies. Mimeo.

Listhaug, Ola 1977. Sosialisering livsløp og ideologisk involvering (Socialization, Life
Course and Ideological Involvement) ISS rapport nr. 4. Univ. of Trondheim, Dept of
Sociology and Social Studies.

Loewenberg, Peter 1971. 'The Psychohistorical Origins of the Nazi Youth Cohort'. American
Historical Review 76, 1457-1502.

Mannheim, Karl 1952. 'The Problem of Generations', in Essays on the Sociology of Knowledge.
Routledge & Kegan Paul, London.

Mason, Karen Oppenheim et al. 1973. 'Some Methodological Issues in Cohort Analysis of
Archival Data'. American Sociological Review 38, 248-258.

Mason, William M. et al. 1976: 'Replies'. American Sociological Review 41, 900-908.

Rintala, Marvin 1968. 'Political Generations', in Encyclopedia ofthe Social Sciences, vol. 6,
pp. 92-%. Macmillan/Free Press, New York.

Torgersen, Ulf 1962. 'The Trend towards Political Concensus: the Case of Norway'. Acta
Sociology a 6, 159-172.