Resumé

The question of political realignment and dealignment in advanced industrial democracies has been the subject of considerable comparative analysis. The present study takes the literature regarding the 'new politics' thesis as the point of departure and examines the relative importance of different political cleavages. The traditional socio-structural cleavages in the Rokkan/Lipset model for political polarization in industrial society are contrasted with an extended ideological cleavage model which includes five different dimensions. Using data from a nationwide Norwegian survey, the findings indicate that although there is a clear tendency towards a generation-based structural dealignment in accordance with 'new politics' literature, the thesis that the materialist/postmaterialist has taken over as the dominant polarization pattern in Norway is not generally confirmed. The 'old left-right' ideological cleavages - which we have called Leftist/Rightist Materialism appears to be of greater importance, even in the post-war generations.

Political Cleavages and Political Realignment: The 'New Politics' Perspective

During the last 10-15 years there has been considerable discussion on the potential long term transformation of the political cleavage structure accompanying changes in socio-economic structure from the industrial to the post-industrial One of the most interesting and intellectually stimulating perspectives in this area is the so-called lnew politics' thesis regarding a basic transformation of the cleavage structure in advanced industrial societies. The: concept of the 'new politics supporters' - first of all Ronald Inglehart, but also Russel Dalton and Scott Flanagan - is based on a thesis of a long term change in the relative importance of different political cleavage patterns (political realignment) the industrial/post-industrial society transformation.

The 'new politics' is identified as the post-materialist value syndrome. In a series of articles and the book The Silent Revolution (1977), Ronald Inglehart argues convincingly that this value syndrome is a basic value orientation opposed to more traditional materialist political values. The Silent Revolution' is the gradual value change along the materialist/post-materialist dimension (hereafter

MPM-dimension) which is explained by generational replacement, the growth of the new middle class and the spread of higher education. In summary, the support of the new value syndrome is strongest in the postwar generations, the new middle class and the higher educated strata according to the silent revolution thesis.

This perspective on the social base of the post-materialist versus materialist value priorities is convincingly supported by an impressive empirical documentation most West-European countries (Inglehart 1977). According to a Norwegian survey from 1981 which constitutes the empirical base of the present article (see below), these social strata are the most post-materialist groups in Norway. There is, however, one important aspect of these general findings which is very pronounced in this material: The middle class post-materialist orientation is mainly limited to the dominant part of this stratum engaged in the public sector (Knutsen 1983, 1985 ch. 8).

The post-materialist value syndrome is further connected to a. psychological base: The main explanation for the spread of post-materialist values is the satisfaction of lower level needs in Abraham Maslow's need hierarchy. Increased economic growth and social security have led to fulfilment of psychological and physical needs for larger sections of the population in their formative years, and these fulfilments have resulted in a search for higher level needs (aesthetic, belonging and esteem needs) which predispose these sections to post-materialist political value priorities. This perspective is generally confirmed in the Norwegian survey. There is a relatively strong congruity between personal values which are assumed to tap the Maslowian need hierarchy more directly, and political MPMvalues & Knutsen 1985).

The new politics perspective is, however, not only a thesis of a new value dimension and its social and psychological base, it is also an integrated perspective which tries to connect the MPM-dimension to changes in the political cleavage structures of advanced industrial societies. Inglehart, who advances this perspective we can call a thesis on the consequences of the Silent Revolution, takes a developmental point of departure: Different kinds of societies are characterized different types of variables which define the cleavage structure of that society. These cleavages tend to persist (or 'freeze' in Lipset's and Rokkan's terminology), although the given structural conditions that gave birth to them have changed or ceased to exist.

Inglehart uses the terms pre-industrial, industrial and post-industrial variables
distinguish between the three different cleavage types (Inglehart 1977,
181-182):

- Pre-industrial cleavages are more or less ascriptive variables such as religion,
language, race, and territorial identity.

- Industrial cleavages are related to the polarization pattern underlying industrial

societies: status and hierarchical variables such as income, occupation,
education and union membership.

- The post-industrial cleavages reflect - according to Inglehart - 'individual level
values, particularly those basedon post-economic needs', and it is of course
this cleavage which is reflected by the MPM-dimension.

The MPM-dimension will increasingly take over as the dominant polarization pattern (political cleavage) in post-industrial society; i.e. it will be the most important variable for explaining party preference, political attitudes and behaviour.

More generally, the developmental perspective may be viewed as a transformation cleavage structure from ascriptive (pre-industrial) variables, and achieved or hierarchical (social status) variables to post-industrial (political value) variables. An important implication of this perspective thus involves a reduction in the importance of the traditional sociostructural cleavages in the well-known Lipset-Rokkan model for 'freezing' political alignments. New value cleavages are gradually replacing the traditional social cleavage model or - in Ronald Inglehart's words - there is a change from a 'Class-based to a value based pattern of Political Polarization' (Inglehart 1984, 26-33).

The pre-industrial cleavages are, however, explicitly given a more persistent character. They are not exclusively assumed to weaken; some of them are expected to revive as a consequence of the spread of post-materialism (the language cleavage; territorial cleavages) (Inglehart 1977, ch. 8).

The cleavage model in Inglehart's theoretical discussion remains relatively simple. The structural model relates to the ascriptive and hierarchical variables mentioned above. In the empirical test this model is represented by religious affiliation activity) and social class. The new post-industrial variables are represented by the MPM-dimension which is considered identical to the ideological or value cleavage dominant in post-industrial society.

Treating the two structural variables as the dominant cleavages in industrial society is of course in accordance with the dominant cleavage model in political sociology, but when introducing the MPM-values as the new cleavage dimension which already has become the dominant cleavage dimension in western democracies, cautious warning is needed. If political value priorities are introduced as relatively permanent political cleavages in addition to the structural model, one should take into consideration the dominant ideological or value pattern in the typical industrial society. The MPM-dimension can scarcely be considered as the only ideological cleavage in an extended cleavage model which includes ideological cleavages in addition to the structural model.

Inglehart seems to be well aware of this. In a recent article he reasserts the view
that the political polarization in industrial society 'was a direct reflection of social

class conflict', but adds that 'the key issue underlying the Left - Right polarization was conflict over ownership of the means of production and the distribution of income' (Inglehart 1984, 25). In this article there is a more explicit discussion of the 'old' issue or ideological cleavages in addition to the more traditional structural cleavages underlying industrial society, although they are not introducedin multivariate analysis where the relative importance of different cleavages is tested empirically.

By introducing value based or ideological political cleavages on equal terms with
the more commonly used structural cleavages, the new politics supporters have
introduced a new cleavage type which needs to be treated more systematically.

The purpose of this article is to test this perspective ofthe politicalconsequences the Silent Revolution in a Norwegian context. We will try to identify several ideological political cleavages which we consider central to Norwegian political culture, and examine whether the conclusions of the new politics literature are seriously challenged by our extended ideological cleavage approach. Although comparative European data might have been preferable, the relevant indicators are not uniformly obtainable. Our analysis may therefore be considered as a test of these important aspects of the new politics perspective in a Norwegian context, as Norway is not included in the empirical analysis in new politics literature.

Types of Political Cleavages

Political cleavages may be defined as relatively stable patterns of polarization which can explain variance in our dependent variables, party preferences, political attitudes and identifications (Inglehart 1984, Zuckerman 1975). The cleavage variables can therefore be looked upon as the independent variables, and the hypothesis of a change in cleavage structure can be formulated as a thesis incorporating those cleavage variables having the largest causal effects with regard to the political phenomena in question.

We have above discussed the value or ideological cleavages as an alternative to the dominant socio-structural cleavage model. In brief, we use Milton Rokeach's definitions of values as our point of departure: 'A value is an enduring belief that a specific mode of conduct or end-state of existence is personally or socially preferable to an opposed mode of conduct or end-state of existence' (Rokeach 1973, 5). Values are therefore prescriptive beliefs which set behaviouralnorms determine attitudes we should adopt towards specific objects and situations. In our approach we are concerned with terminal values (endstatesof in contrast to instrumental (modes of conduct) values. Our concentration on political values emphasizes the kind of end-states people prefer in society or would like to see implemented through the political system. This

contrasts with more personal values which are end-states in the individual's own
life (Rokeach 1973, 5-8).

When we relate political values to ideologies, we are primarily interested in the evaluative or prescriptive aspects of ideologies, while the cognitive aspect - i.e. whether individuals are characterized by 'ideological thinking' or not (Converse 1964) - is left open. We can say that we are focusing attention on ideologies as organizations of perscriptive political beliefs (political values) which may control behavior patterns (Minar 1961, Hamilton 1983, Sainsbury 1983). More precisely we take Philip Converse's (1964) approach, where ideologies are studied as constraint among presumably coherent ideational (value) elements, as our point of departure. A set of constrained political values therefore constitutes an ideology, and sets of opposed value priorities which are empirically related constitute sets of opposed ideological orientations (Lafferty & Knutsen 1984).

In our approach there may exist more than one set of opposed ideologies in a given political culture. Therefore, in a political cleavage perspective there may be multidimensional sets of ideologies or value-based patterns of political polarization. ideological polarization patterns may be viewed as cross-cutting and sometimes overlapping cleavages in the same way as the more common structural cleavages.¹

We defined political cleavages as 'relatively stable patterns of polarization' above. Since ideologies consist of basic preferences for different political endstates political values), we can consider them as central in individuals' political belief systems, and we can expect them to transcend attitudes towards different political objects and situations which may be changed much more easily. The value or ideological approach to political cleavages presumes that these cleavage dimensions are relatively stable and enduring sets of beliefs in individuals' world view.

Cleavage Models: Different; Structural and Ideological Cleavages

Having argued that political cleavages may be viewed as conflicting value sets, we will take a closer look at the new politics thesis and extend the simple cleavage model in this research tradition by introducing different ideological cleavages. The new politics thesis may be portrayed as having both a general and a more specific aspect.

The general aspect concerns the relative importance of the structural and ideological cleavages in post-industrial societies. It is assumed that ideological cleavages will increase in importance in most western democracies, simultaneouslyto decline in relevance of the socio-structural model. The new interest in the former model may also be seen as a revival of interest in ideologies after

"the end of ideology' debate and as a search for a new cleavage model as the structural model explains less and less variance in political phenomena. In a causal perspective it may be argued that political values and ideologies arise out of structural conditions and that the strength of the ideological model may be strongly reduced when controlling for the structural variables (i.e. the relationship is spurious). To understand the relationship between the different cleavage models we have constructed a complete cleavage model in causal terms. (Fig. 1).

The model suggests that the new politics thesis rests on two fundamental
empirical assumptions:

1. The direct path from social structure to political attitudes (path 1) should not
be strong, and the same should hold for the indirect effect via the ideological
model.

2. On the other hand, the ideological model is expected to have a large explanatory power (path 3) which is stronger than that of the structural model, even when the latter model is controlled for. This implies that the ideological model should not be strongly correlated with the structural variables (path 2).²

The specific aspect concerns the new politics literature's emphasis on the MPMdimension; is this dimension among the ideological variables which is expected to have the most significant causal effect. It is also the increased importance of MPM-vaJues which is the explanation of the structural dealignments and realignments post-industrial societies. According to the new politics thesis, it is the materialist political values of the lower status groups that explain their status quo or conservative attitudes, while the higher strata's post-materialism explains their radicalism and change orientations.

Our extended cleavage model includes two main types of variables - structural
and ideological:

The structural type3 consists of the ascriptive variables, age and sex; and the
most important hierarchical status variables, education, occupation and income.4

The latter are referred to as the socio-structural or status variables below. They
may be said to be a direct reflection of the essential cleavages of classical
industrial society.

In our ideological model we have included four dimensions in addition to the MPM-dimension. As mentioned above, Inglehart in his recent analysis of 'changing cleavages' in advanced societies has discussed opposing economic orientations regarding such questions as ownership and regulation of production, and the distribution of income. In Inglehart's work these opposing orientations are referred to as economic issues, while the MPM-dimension is called values. As we see it, however, underlying the basic materialist struggle in industrial society is a pattern of ideological or value-based polarization concerning desired political end-states which obviously deserve to be thought of as political values. We have called these value systems 'socialist' and 'bourgeois'¹ or - to make use of the logic of the new politics supporters - Leftist and Rightist Materialism in consideration of the fact that the motivational core here is opposing materialist ideologies. These value systems concern the question of control and distribution in the economy and the production sphere and have been measured by four pairs of forced-choice value opposites. These we have named general governmental regulation, economic regulation, worker control and income equality. The value alternatives and the index construction of Leftist versus Rightist Materialism are shown in Appendix 1.

Our original intention was to include a value-indicator tapping support for the welfare state in this dimension. Support for versus opposition to a high level of taxation in order to maintain and develop the welfare state did not, however, correlate with the other indicators. The conflict over the welfare state is only loosely related to the other aspects of leftist versus rightist materialist values in the Norwegian context (Lafferty & Knutsen 1984, 355-356, Knutsen 1985,93-97). have, therefore, kept Welfare Support as a separate value dimension.

Another materialist value dimension is reflected in conflicting views regarding subsidies and prices for agricultural and other primary economic products (Rokkan 389-390). This value dimension is measured by a single indicator which contrasts the need for an economic return for farmers with the welfare of consumers and taxpayers (Farmers' Subsidies).

The moral-religious value dimension is measured by a general question
tapping the respondents' emphasis on the role of Christian values in essential
political decisions (Christian Values).

All of these ideological cleavages may be said to have their structural counterpart social class and affiliation with religious institutions. Whether the correlations between the ideological cleavages and different political attitudes can be traced back to the structural cleavages is a question that in the following we shall make the subject of empirical examination.

The MPM-dimension is measured by the most recent version of the Inglehart

value batteries. Our procedure in administering these value batteries differed slightly from the procedure followed by Inglehart and associates; we have simply asked our respondent to rank each of the three four-item sets but without probing further for an overall ranking. Nevertheless, our data must be considered comparableto used by Inglehart in documenting the Silent Revolution.⁵

Details regarding operationalizations and construction of indices are presented
in Appendix 1.

The Dependent Attitude Variables and the New Politics Thesis

According to the new politics theory, it is not voting and party preference which are in the forefront of the cleavage transformation process. The decline of class voting - which is seen as being strongly related to the increasing importance of the MPM-dimension - is only one aspect of a broader transformation. Furthermore:

... in some ways, class voting patterns give an underestimated impression of what has been going on. For voting behavior is shaped, to a considerable extent, by an internalized sense of political party identification in given individuals, and by institutional ties between given parties and given social networks, such as labour unions and church. In so far as this is true, voting behavior has a good deal of inertia; it does not necessarily respond to current conditions, but may continue to reflect old alignments long after the circumstances that gave rise to them have changed. Other indicators of political polarization may reflect the dynamics of contemporary politics more directly (Inglehart 1984,31).

Inglehart here indicates that the freezing of party alignments, which Lipset & Rokkan (1967) paid so much attention to in their classic essay, continues to make the old structural cleavage factors of major importance. At the same time, however, the MPM-dimension will play a central role in connection with other political orientations which are not so strongly bound to institutionalized organizations and partisan loyalties (Inglehart 1984, 33, Dalton & Flanagan 1982,9-12).

Four basic attitudes and identities are discussed by the new politics adherents
in various articles, and the role of the MPM-dimension has been empirically
documented for three of these attitudes and identities.

One of these is the attitude towards social change. According to the new politics thesis, support for radical political and social change no longer comes from the working class, but from people with post-materialist political values (Inglehart 1977,470-471). While the working class is increasingly integrated in the established materialist value structure in Western societies, the post-materialistswant change in these materialist structures and the outputs from the political system. Since support for social change seems to be only loosely constrained by party ties, Inglehart expects this basic attitude primarily to reflect individual value priorities (Inglehart 1984, 63). In a multivariate analysis based

on data from the European Community surveys, he has confirmed that the
MPM-dimension has a stronger effect than both religion and social class on
support for social change (Inglehart 1984, 66-67).

In our Norwegian national survey,⁶ this attitude was measured by asking the respondents to indicate their preference among the following four alternatives regarding social changes: radical change, gradual reform, status quo, and reform repeal (see Appendix II for the exact wording).

Another variable that the new politics literature pays a great deal of attention
to is left-right identity. In Inglehart's words:

... the political meaning of'Left' and 'Right' ... has been changing. With the rise of new issues, identification with the 'Left' increasingly would come to connote support for new courses such as environmentalism, with a diminishing tendency to evoke the classical issues such as nationalization industry. Similarly, self-placement on the left would have a declining linkage with working-class (Inglehart 1984, 31).

According to Inglehart's own empirical work, church attendance has the strongest effect on 'Left-Right identity', followed by MPM-values and social class (Inglehart 1984, 67). The classical economic values which are assumed to be of reduced importance are, however, not included in the model. (See also Inglehart Klingemann 1976,257, where the old ideological cleavage also is assumed to be of minor importance.)

We have measured left-right identity by the well-known 10 point left-right selfplacement
also employed in the other European surveys (Eurobarometers).

According to the new politics literature, the rise of the MPM-dimension as an important political cleavage is, furthermore, connected to the declining political support for political authorities and institutions in Western democracies (the socalled of authority'). This relationship has, however, not been convincingly demonstrated empirically, so it remains an open question if the decline in political trust and confidence is attributable to the MPM-dimension (Inglehart 1977a, 15 and 303-307, Inglehart 1977b).

We have measured Political support by six indicators intended to tap support versus alienation regarding the general political system and the politicians, and degrees of confidence in the power of basic political institutions (see Appendix II).

Finally, the MPM-dimension is related to the 'participatory revolution' (Kaase 1980, Barnes & Kaase 1979). The increase in protest activites - or unconventional' participation - predominant among the young, well-to-do, more educated strata, is explained by the MPM-dimension. Inglehart confirms empirically - by the Political Action data - that along with age and education the MPM-dimension does indeed have a relatively strong causal effect on 'protest potential' in Western societies (Inglehart 1979, see also Inglehart 1981, table 5, 891).

In the Norwegian survey design, attitudes towards direct actions and participation such activities are kept separate; in this analysis we include only the attitudinal aspect of the 'participatory' revolution. Attitudes towards unconventional (AUNPA) were measured by asking our respondents if they were 'very positive', 'somewhat positive', 'somewhat negative' or 'very negative' to eight different types of political action. The types of action involved are listed in Appendix 11.

Empirical Analysis: Political Cleavages and Political Attitudes

Let us start the empirical analysis by comparing the total explanatory power of
the two cleavage models (the structural and the ideological) with consideration to
the four dependent attitude variables.

Table 1 indicates that the structural model, which includes age and sex in addition to the three status variables, has a modest explanatory power in relationship at least three of the dependent attitudes. This model has its greatest explanatory power with regard to AUNPA (explained variance 18.6%). However, is the highest educated strata and the younger generations that are most permissive in accepting direct actions, and age and education appears to be almost alone in contributing to the relatively high explanatory power of the sociostructural (not shown in the table).

Standing alone, the explanatory power of the ideological cleavage model is stronger than the socio-structural model with regard to each of the four attitudes. This picture is only slightly modified when the structural model is controlled for as prior variables; only with respect to AUNPA does this change the relative importance the two models.

To obtain a more precise impression of the various effects, we decided to probe further into the causal mechanisms by focusing on a more simple model (with fewer variables). We therefore included only one structural variable (education) and three ideological variables (Leftist Materialism, Welfare Support and MPM). The selection of these variables is based on both empirical and theoretical criteria.

Education is an important structural variable in post-industrial society - and also turns out to be the status variable with the strongest causal effects. Likewise, the three ideological variables have the strongest causal effects on every attitude. In fact, when the other value dimensions are controlled for, Christian Values and Farmers' Subsidies have only modest or insignificant effects.

The four causal diagrams are presented in Fig. 2, see page 246.

Let us first comment on the correlations among the ideological variables. These correlations are moderate, the largest being between Leftist Materialism and MPM (r=.29), indicating that post-materialists are more leftist on values related to economic equality and control over production. The lowest correlation is between Welfare Support and Leftist Materialism (r=.lB), indicating that Welfare Support is more strongly correlated with MPM (r=.25) than with Leftist Materialism.

One implication of these findings is that the three ideological cleavages are partly overlapping ideological cleavages, i.e. some of their bivariate correlations with the dependent attitude variables may be of a non-causal character, not only because of prior structural variables, but because of their interrelatedness with other ideological variables.

The models also indicate interesting variations in the correlations between education and the ideological variables. The higher educated strata are more post-materialist and favour the welfare state to a larger extent than do those with lower education, but they are also more rightist on the left-right Materialist Scale - although this is only a weak tendency. This indicates that the assumed dominant ideological cleavage in industrial society is only loosely related to social status variables in post-industrial society.1

A closer examination of the path diagrams reveals that the new politics thesis with regard to the importance of the MPM-dimension is generally not confirmed. Although we have found support for the general thesis that ideological cleavages are stronger than structural cleavages in terms of explanatory power, the MPMdimension a significant stronger causal effect than the other ideological dimensions on only one of the attitudes in question.

It is - surprisingly, given the above discussion of the new politics literature - the old Left-Right Materialist ideologies which have the overwhelmingly strongest causal effect on Social Change and Left-Right Scale. Although these; attitudes are only weakly related to 'industrial' structural and hierarchical variables, Leftist Materialism has a dominant causal effect in clear contradiction to what is assumed in the new politics literature.

Examining the path diagram for Political Support, we see that Leftist Materialism has no significant causal effect, but it is here that Welfare Support and the MPM-dimension have the strongest causal effects and are of almost equal importance (betas .16 and -.18). Post-materialists and those who reject the high level of social welfare, are the most political alienated of all ideological groups, although the effect is not evident. Political distrust seems to be related to discontent with the relatively high level of social welfare in Norway. This finding corresponds to that expressed in other studies where discontent with the welfare state is considered as a populist rebellion against bureaucracy and the perceived waste and inefficiency in bureaucratic organizations (Nielsen 1979, ch. 6, Sänkiaho ⁸ This value dimension is - according to the causal model - of almost equal importance compared to the post-materialist value syndrome.9

The MPM-index ranks dominantly first only with regard to AUNPA, where its correlation and causal effect are considerable. Here it appears that we have found a very important and pronounced consequence of the post-materialist value syndrome in accordance with the new politics theory. From the model it appears that the MPM-dimension partly explains why the higher educated strata are more positive to direct actions: The indirect effect via MPM is .09 (i.e. one third of education's total effect). Most of the effect of education is, however, direct - i.e. not transmitted via the value model.¹⁰

The pattern for AUNPA is, however, not the dominant pattern so far. The perspective of the MPM-dimension as the most important political cleavage in post-industrial society, is generally not confirmed. We have seen that the omission of the Leftist Materialist-dimension from the ideological cleavage model, appears to give a rather misleading impression of what is going on: While it is true that social class - or (more broadly) the hierarchical social cleavage variables - have declined and are of marginal importance, the old ideological cleavage between bourgeois and socialist world views is still of utmost importance when we consider Left-Right Scale and Social Change.

Party Preference, Political Cleavages and the New Politics Thesis

The new politics thesis with regard to party preference can be summed up in three
arguments:

1. Although the clear decline in class voting is well documented in most Western
democracies, the old structural alignments are still working due to political traditions,
socialization, party identification etc.

2. The MPM-dimension will therefore be of only moderate importance for explaining preference, although the tension between the party systems and value orientation (MPM-values) in Western democracies seems to be associated with dealignments and realignments in different countries (Inglehart 1984, 68).

3. The MPM-dimension is, however, an important explanation of the decline in class voting since it tends to reverse the 'industrial' relationship between social structure and party choise (Inglehart 1984,29). The 'post-materialist' left parties and the 'new left' tendencies within 'old left' parties are increasingly replacing the old left (parties and tendencies).

Again, only minor attention is given in this discussion to the old ideological cleavages. This is especially problematic in relation to party preference because recent economic uncertainty has undoubtedly brought the traditional economic values related to economic equality and regulations versus classical liberal right values related to market economy and economic differentiation back on the political agenda. This seems to have occurred without an accompanying return of the old structural class and social status polarization with regard to party preference. attitudes to the welfare state have also recently been very pronounced on the political agenda in most Western democracies.

Let us first look at the traditional division between socialist and bourgeois

parties in Norway. If we dichotomize party preference along this division, the structural model explains 19.1% variance,¹¹ while the complete ideological model explains 31.2% variance. The unique contribution of the latter model is 22.2% when the structural model is controlled for. Although almost one third of the uncontrolled explanatory power of the ideological model is spurious, its total causal effect is greater than that of the structural model. This indicates that for party preference the importance of the value model is larger than the new politics literature contends.

Which dimension(s) in the ideological model contributes to this pattern? The
answer is Leftist Materialism, which has an overwhelmingly strong effect on
socialist versus bourgeois party preference. (See Table 2.)

Leftist Materialism alone contributes 24% of the explained variance (r=.49) and this effect is only slightly reduced when the other structural and ideological variables are controlled for. The effect of MPM and Welfare Support is very minor compared to Leftist Materialism. Support for Christian Values and Farmers' Subsidies leads to bourgeois party preference and has stronger causal effects than the former variables (MPM and Welfare Support).

There is then, no indication of the post-materialist value syndrome having a strong effect on preference for socialist parties. The explanation is, of course, that the 'old industrial left-ideology' is dominant among socialist voters and especially in the old Left party, per se - The Labour Party.

While Leftist Materialism divides the parties along the socialist/bourgeois boundary in Norwegian politics, the MPM and Welfare Support dimensions cut across this division in a similar manner (Knutsen 1985, ch. 13). It is the Socialist Left and the Liberal Party which have the most post-materialist and welfare support electorate, while the Labour and Conservative Parties have much of the same location along these dimensions; both parties have a relatively materialist electorate and the degree of welfare state support is relatively low. If we dichotomize party preference according to the preference of 'green' parties

(Liberal and Socialist Left) versus all other parties, the explanatory power of the total structural model is 12.4%, while the corresponding percentage is 19.8% for the ideological model (this is reduced to 12.8% when the structural model is controlled for). Of the structural variables, education and age have the greatest causal effects, while occupation and income have minor, although still significant effects. The 'structural' location for the green parties seems to come from the predicted groups according to new politics theory; it is the more educated and middle class, especially in the postwar generations, that give their support for these parties.12

Among the ideological cleavage dimension, the three variables in our previous
path analysis are again most important (see Fig. 3).

The MPM-dimension is overwhelmingly the most important ideological cleavage dimension, while the causal effects of the two other dimensions are minor when this variable is controlled for. About one third of the effect of education is indirect via the ideological model (.08). Almost all of this effect is transmitted via MPM; we can therefore say that it is partly the MPM-dimension that explains why the higher social strata support the green parties to a larger extent than the lower strata.

Structural and Ideological Cleavages in Different Generations

The long term transformation of cleavage structure in advanced societies is partly

explained as a generation effect. According to the new politics literature, it is within the post-war generations that the structural variables have lost their dominant importance. They maintain, however, a strong effect in the pre-war generations, where class antagonism dominated the political socialization and experiences in their formative years (especially for the so-called 'depression generations' which grew up in the 1930'5). In the post-war context, these experiencesand patterns have lost some of their meaning, and in these generations people more or less choose their political ideologies, life styles and political identifications irrespective of their class background. Political values and ideologies are here expected to play a more independent role for social and political preferences. This independent role is first of all identified with the MPM-dimension according to the new politics thesis (Dalton & Flanagan 1982, 8-9).

If there is a basic generation based cleavage transformation in accordance with
the thesis of new politics, we therefore should expect that:

1. The socio-structural model will be of modest significance in the post-war generations, while it will be considerably more significant in the prewar generations. the relationship between the 'industrial' cleavage variables and various political phenomena will be turned 'upside down' in the postwar generation in the sense that it is the higher social strata which are most radical and change oriented.

2. The MPM-dimension will have its strongest effects in the post-war generation
where it has taken over as the dominant pattern of polarization.

3. The MPM-dimension will play an important role as an intervening variable
and as an explanation for the generation-based structural dealignments and realignments.

We shall here test these theses on the basis of our four political attitudes and party preference dichotomized in the two directions discussed above. The post-war and pre-war generations are delimited to 18-34 years (born 1946-61) and 45-65 years (born 1915-35), respectively. Other age groups (35-44 years) are omitted in this context. In the upper part of Tables 3 and 4 (see below) we have illustrated the explanatory power for the socio-structural model.¹³

The socio-structural model has greater explanatory power in the pre-war generations for Left-Right Scale, Political Support and, first of all, socialist versus bourgeois party preference. For all these phenomena there is a clear tendency towards a generation based structural dealignment: In the pre-war generations the lower status groups are 'left'-oriented, politically distrustful and prefer socialist political parties to a much larger extent than the higher status

groups. These status differences are minor in the post-war generations. ConcerningSocial the results are more in accordance with the 'new politics' thesis than the table implies, because it is the lower status groups that are more changed oriented in the pre-war generations, whereas it is the opposite in the post-war generations. The effects in both generations are, however, minor.

On AUNPA and support for the two green parties the effect of the structural model is larger in the post-war generations. In these cases the higher strata favour direct actions and green parties in both generations, but the status differences are strongest in the post-war generations. For these orientations there is, then, a generation structural alignment taking place in Norwegian political culture.

When studying more closely the total impact of the ideological model, it is worth noting that this model explains more variance in all basic political attitudes and party preferences than the socio-structural model even in the pre-war generations. It may be that class (or social status) cleavages were strong in the typical industrial society, but that society was not unaware of ideological or value cleavages, if we consider the results from the pre-war generations as an indication of this cleavage structure.¹⁴ When we compare the relative importance of the ideological model in the two generations, the main tendency is that this model explains more variance in the post-war generation. In the uncontrolled models the exceptions are Political Support and socialist versus bourgeois party preference; in the complete causal model only the former is the exception. In this model the effects are approximately the same for Social Change and socialist/ bourgeois party preference. For the three other phenomena - Left-Right Scale, AUNPA and green party preference - the tendency for the ideological model to explain most variance in the post-war generations is very clear.

The question is then: Is it the MPM-dimension that contributes most to this
generation based ideological alignment and can this dimension explain the
status dealignments and alignments which we have observed above?

Since there seem to be two main tendencies for respectively three and two of
the discussed dependent variables, we have chosen to present and comment on
each of these tendencies separately.ls

The first tendency involves the three dependent variables where we have seen that
Leftist Materialism is the most important political cleavage for the whole sample:
Left-Right Scale, Social Change and Socialist versus bourgeois party preference.

From Table 3, sections A, B and C, we observe that the importance of Leftist Materialism is not dramatically reduced when comparing the pre-war and postwargenerations. dimension is clearly the most important political cleavage even in the latter generations. The MPM-dimension is more higly correlated with the three phenomena and has its largest causal effects for Left- Right Scale and Socialist-bourgeois voting in the post-war generations compared to those of the pre-war period. The MPM-dimension contributes to the higher

explanatory power for the ideological model in the post-war generations, but
even here its relative importance is at best second to Leftist Materialism.16

This overall impression is dramatically changed for AUNPA and green party preference (Table 4, sections A and B). We have seen that there is a generation based value and status alignment for these variables. The table clearly indicates that this is mainly due to the increased importance of the MPM-dimension in the post-war generations although the other ideological variables also have their largest causal effects in these generations. The MPM-cleavage together with education comprises the two dominant cleavages in these age groups.17

Discussion

The findings from the Norwegian survey have led to mixed conclusions with regard to the new politics thesis. The main findings contradict this thesis: In the case of Left-Right Scale, Social Change and Socialist versus Bourgeois party preference, there is no evidence that the MPM-dimension has taken over - or in fact will take over - as the dominant polarization pattern in Norway. The important consequences of the Silent Revolution are limited to priorities of 'green' parties versus others and acceptance of direct political actions, which is part of the participatory revolution. For these phenomena the new politics thesis is verified on almost every point: MPM-values are the most important political cleavage and its increased importance in post-war generations contributes to a value alignment, also (as an intervening variable) to a generational based structural alignment. The MPM-dimension has also major importance (together with Welfare Support) with regard to political Support (trust) versus alienation.

We can also make a more precise conclusion as regards what aspects of the post-materialist value syndrome contribute most in this respect. We have shown elsewhere that the MPM-dimension consists of two quite distinct subdimensions; one dimension which contrasts democratic and participating versus authoritarian or 'tough-minded' values, while the other contrasts preoccupation with economic growth and security versus environmentalist and non-materialist concerns in a narrow sense (Knutsen 1985, 141-157).¹⁸ Of these subdimensions, which we have called Populist-Humanism and Economic Post-Materialism respectively, it is clearly the former that is most important for all political phenomena examined here (Knutsen 1985, ch. 12 and 13). The green post-materialist value syndrome is much less consequential, so it is mainly the participatory, antiauthoritarian of the post-materialist value syndrome which make the MPM-dimension an important political cleavage in Norway - in contrast to the emphasis on the growth/environmental subdimension implicit in the new politics literature.

For Left-Right Scale and Socialist-Bourgeois party preference we find a clear
tendency towards a generation based class or more broadly social status dealignmentin

alignmentinNorwegian political culture. The MPM-dimension is, however, not an important explanation in this context. The main point here is that it is not correct to take it for granted that the decline of class cleavage implies that the traditional conflict between socialist and bourgeois (Leftist versus Rightist Materialism) ideologies is diminishing, something new politics theory appears to contend. What has changed is the social location of these opposed ideologies, although they still remain the dominant ideological cjeavage in Norwegian political culture. We see no immediate reason for this not also applying to other West European political cultures, had the relevant value indicators been available and included in multivariate analysis.

We have shown that social status variables and Leftist Materialism are not strongly related in the post-war generations. Their cross-cutting character in these generations may imply that they will not travel together in post-industrial societies: New social groups (the more educated strata, the new middle class in the public sector) seem to have acquired leftist materialist values, while part of the younger working class seems to be purveyers of bourgeois or rightist materialist values (Knutsen 1985, ch. 6, Knutsen 1985b). The problem with new politics theory in this context is that social status cleavages and Leftist Materialism are not analytically discussed as different cleavages and treated independently as such in multivariate analysis.

As observers of Scandinavian (possibly European) politics are aware, the prime ideological battle is still closely linked to degrees of state regulation and control, wage differentiation versus economic equality, workers' participation in the firm etc. It is this obvious political reality which is reflected in our analysis of the political mass culture in Norway.

We have made a further analysis of the importance of the different aspects in our Leftist Materialist Index and the different dependent variables discussed above. This analysis indicates that the highest correlations and the highest causal effects are found between worker control and income equality with regard to the various dependent variables (although not shown here). These, then, are the most important value elements within the larger opposing materialist value syndrome.

Generally, ideological cleavages have a surprisingly high explanatory power compared to the traditional social cleavage model, even in the pre-war generations. This unexpected finding is important and indicates a need for new directions in political cleavage research. We can conclude that the importance of the social cleavage model has been overestimated and has received excessive attention at the expense of other cleavage models in this research tradition.

Ideological cleavages seem to be even more important in post-industrial society if we can take the results from the post-war generations as a prediction of the cleavage structure in this society. Instead of postulating a thesis with regard to which particular ideological cleavage will dominate post-industrial society, our findings support a general perspective where increasing weight is attached to all

political ideologies and value sets. This perspective is one of increasing/rø/to*/ rationality or value consciousness in the sense that the post-war generations appear to hold political attitudes in accordance or consistence with their basic political values. The dominant social determinist cleavage model in industrial society - based on social status and ascriptive variables - seems to be declining, accompanying the process of generational replacement.

---

NOTES

1. Cross-cutting ideological cleavages may be defined as different opposed value sets or ideological dimensions which are uncorrelated or lowly correlated, while overlapping ideological dimensions are relatively highly correlated.

2. If the ideological model is strongly correlated with structural variables, the result may be a strong indirect effect from social structure to political attitudes via political ideologies. Another empirical result which contradicts the 'new politics' thesis is - of course - that the effect on the ideological model is strongly reduced when the structural variables are controlled for (i.e. the effects are spurious).

3. Originally the structural model included the pre-industrial center-periphery variables, which were strongly emphasized in the Rokkari/Valen model for structural cleavages in Norwegian politics (region, the rural-urban axis and types of communes) (Rokkan & Valen 1964, Rokkan 1967). We also included membership in religious organizations as a measure of institutionalized religious 'belongingness'. These variables surprisingly did not have any significant causal effects on the political attitudes and identifications discussed below.

4. E^ducation is a5 value variable ranking from no education beyond the compulsory level via different kinds of vocational training (3 and 4) to university and college education (5). The income variable is reported family income. The occupation variable has 7 categories and is used as a nominel level variable in the analysis: 1 - Currently nonemployed (housewives, students, pensioners, etc.) 2 - Workers (skilied and unskilled) 3 - Lower-level functionaries 4 - Middle-level functionaries 5 - Higher-level functionaries 6 - Self-employed in business and professions 7 - Self-employed in primary sector A discussion of the classification and a detailed list of the occupations included in each category is given in Lafferty & Knutsen 1982.

5. In 'The Changing Structure of Political Cleavages' Inglehart introduces a set of items designed to tap the 'New Politics' dimension (such as stronger measures against terrorism, development of nuclear energy, stronger defence efforts). These new politics issues are not included in our MPM-index. We see, however, no reason for departing from using the normal value batteries which have dominated this field.

6. The project 'Democracy in Norway: Participation and Basic Values' has been led by Professor William Lafferty and jointly financed by the Norwegian Research Council for Science and the Humanities, the Ministry of Local Government and Labour and the Institute of Political Science at the University of Oslo. The survey was conducted by the Central Bureau of Statistics during January-February 1981 on the basis of a sample drawn from the national population register. For reasons of design the age limits of the sample have been set at 18 and 65. The sample size is 1170.

---

7. The correlation with social class (working versus new middle class) is somewhat larger (-.17), but this does not change the overall picture. Social class is operationalized by contrasting category 2 versus categories 4 and 5, on the occupation variable (see footnote 4).

8. From the model we can see that it is the higher educated strata which have the highest degree of political trust. Some of this effect is transmitted via Welfare Support (.04). The indirect effect is, however, counterweighed by a negative indirect effect from education via MPM to Political Support (-05). This means the Welfare Support partly explains the lower status groups lower degree of political support, while this is counterbalanced by the MPM-dimension so that the 'netto' indirect effects are minor.

9. The two value dimensions have their greatest impact on different aspects of political support: parties, while MPM-values are strongest related to 'cynicism' with regard to the power of political institutions like Parliament, government and elections (see Appendix II for the operationalization of the different indicators).

10. There are also other interesting indirect effects in the models which are not commented upon.

11. In addition to the socio-structural variables, the 'pre-industriaF variables membership in religious organizations and region have significant causal effects in this model.

12. The causal effect of income is negative, which means that low income goes together with strong support for the 'green' parties. This is related to the fact that 70% of the supporters of these parties are located in the public sector (Knutsen, 1985,552-555). The horizontal cleavage between private and the public sector is not given any attention in this article, but is also extremely important for materialist versus post-materialist value preference and other valueand (Knutsen 1985b).

13. Because the new politics thesis primarily is related to the status variables in this context, we have only included the status variables education, occupation and income in the structural models here.

14. This interpretation is of course not unproblematic. It may be that strong 'period effects' are involved, which have led these generations to be less polarized along class lines.

15. Political Support does not belong to any of these tendencies and is not commented upon in the text, although the results from the analysis are presented in Table 4.

16. Above we have criticized the new politics approach for not having paid sufficient attention to the ideological cleavages in industrial society (especially leftist versus rightist materialist ideologies). We can perhaps understand this lack of attention as an expression of the assumption these ideologies are strongly constrained by social class and consequently will have minor independent causal importance. This may be the reason for omitting these ideologies from multivariate analysis. The importance of holding these dimensions separate is clearly seen in a generation perspective. In the pre-war generations, social status and Leftist Materialism are relatively strongly correlated, but there is no significant correlation between these assumed interrelated cleavages in the post-war generations. The three status variables explain 8.7% variance in the Leftist Materialism Scale in the pre-war generations, and only 1.7% in the post-war generations. This means that while they are to some extent overlapping cleavages in the pre-war generations, they are crosscutting cleavages in the post-war generations. This generation difference explains why it is possible to have a generation based social status dealignment, while the importance of Leftist Materialism is largely unchanged in the respective generations: The Left-Right Materialist ideologies are disconnected from social status cleavages, but are still the dominant cleavage for these three phenomena in the post-war generations.

17. The generational structural alignment seems to a certain extent to be explained by the MPMdimension these generations: MPM is an intervening variable to a larger extent because it is higher correlated with education and has a stronger direct causal effect in these generations, and about 30-40% of the education variable's causal effects are transmitted via MPM-values in the post-war generations (not shown in the table).

18. These dimensions correspond respectively to A) the physical security needs versus the belongingness aesthetical needs, and B) the sustenance needs versus the self-actualization needs in Maslow's need hierarchy. In a rotated factor analysis two distinct factors load on the values intended to tap these respective needs.

18. These dimensions correspond respectively to A) the physical security needs versus the belongingness aesthetical needs, and B) the sustenance needs versus the self-actualization needs in Maslow's need hierarchy. In a rotated factor analysis two distinct factors load on the values intended to tap these respective needs.

APPENDIX 1: Operationalization of Political Value Priorities and Index-Constructions

Most of the value priorities are measured by 'forced choice'-value opposites where the respondents
were requested to choose 'the statement being closest to their own view'.

Leftist Materialist Scale

General Governmental Regulations

(L) Laws, guidelines and restrictions on the part of the authorities are necessary for the promotion
of equality and justice.

(R) Detailed regulation by the authorities in a number of areas is a serious threat to personal freedom
should be limited as much as possible.

Economic Regulations

(L) Regulation of business by the state is completely necessary to make the economy function for
the best of everyone.

(R) In order for the economy to function for best of everyone, it should -as far as possible -be
allowed to develop without interference by the state.

Worker Control

(R) The owners and shareholders in companies should have the final word when important
decisions are to be made. It is they who have invested money in the firm and who have the right
of private property.

(L) It is the workers and functionaries who have created value at the workplace through their
labour. They should, therefore, have the final word in important decisions.

Income Equality

(R) In order to fill important positions with qualified people, and to strengthen the will to work on
the job, wage differentials should be increased.

(L) In order to even out social and economic differences in society, we have to have a greater
degree of income equality.

The Leftist Materialist-index was constructed as a simple (unweighted) additive index based on these four indicators. It has values from 1 to 9. High Score means leftist priorities. The index has 9 values because each indicator was recoded in the following way before the index construction: 1 = Rightist priorities, 2 = 'don't know', 'difficult to decide', 3 = Leftist priorities.

Welfare Support was measured by the following item:

- In order to retain and develop the welfare state, we must be willing to give up a relatively large share
of our incomes to the government in the form of taxes and fees.

- The government should reduce its expenses for social welfare so that the individual can retain a
larger share of one's own income.

Farmers' Subsidies was measured by a single forced choice item:

- In order to prevent migration from rural districts and to preserve agricultural industries, fanners
must be guaranteed satisfactory returns for their products.

- Fanners economic returns cannot be more important than the welfare of other groups in society.
Consumers and taxpayers must pay in the form of higher prices on food products and through state
subsidies.

The first alternatives were given the highest score on the two indices and don't know (8% on each
item) were given a neutral score.

Christian Values were measured by the following question:

What is your opinion of basic Christian values in politics? When important political decisions have to
be made, do you consider that these should play:

4. An important role
3. A certain role

2. A small role
1. No role at all

High score on Christian Values means a positive orientation towards Christian principles in politics.

The MPM-dimension was measured by three four-item sets. In each set the respondents were asked
to rank 4 values, one for each level in Maslow's need hierarchy.

The question was formulated as follows:

There has been a good deal of talk lately about the goals this country should aim at during the next; 10 years. Here is a list of different goals which different people might prefer. Which of these goals do you believe to be most important? And which do you think is next most important? And the third mos; important?

(A) 1) Maintaining law and order in the country

2) Giving people more to say in important national political decisions
3) Fighting rising prices

4) Protecting the freedom of speech

If you had to choose in the same way between the goals on this card, which would you prefer first'.'
And as number two? And number three?

(B) 1) Maintaining a high rate of economic growth

2) Making sure this country has a strong defence

3) Giving people more to say in decisions on the job and in the local community

4) Protecting nature from pollution

Here is one more list. Which goal do you prefer as number one here? And as number two? And as
number three?

(C) 1) The fight against crime

2) Making for a less impersonal, more human society
3) A more stable economy

4) Wishing for a society where ideas count more than money

Scores were assigned each value choice on the basis of the following scale:

Most important (4)

Next most important (3)

Third most important (2)

Value not chosen (1)

The MPM-index was then constructed as follows: First, we constructed four separate indices for
each of the presumed four levels of the Maslows hierarchy (as described in Inglehart 1977, p. 42).
Each of these indices contains three values:

Economic Security needs: A3, 81, C3

Physical Security: Al, 82, Cl

Belongingness and esteem needs: A2, 83, C2

Self-actualization needs: A4, 84, C4

Next, we constructed two indices for the subdimensions of economic security versus self-actualization Post-Materialism), and for physical security versus belongingness (Populist-Humanism) in the conclusion. This was done additively after reversing the index score for economic security and physical security. Finally, we added the two subdivision-indices to attain the final MPM-index. The scores were then collapsed to a 7-value index.

APPENDIX II: The Political Attitude Measures

Attitudes towards Social Change

On this card we have four opinions as to how society should be improved. Please choose the opinion
you agree with most.

- We must radically change our entire form of social organization
- We must continue to improve our society gradually with the help of reforms
- We should take a 'reform pause'

- Several of the reforms which have been enacted recently should be reversed because they create
more problems than they solve.

The first (radical) alternative is given the highest score (4) and the last the lowest score (1) in the
correlation and multivariate analyses

Political Support

The six indicators are statements where the respondent should choose between the alternatives 'agree
strongly', 'agree', 'disagree', and 'disagree strongly'.

General Political System Support

(A) Personally I have great respect and trust in the Norwegian political system.

(B) There are great differences between how I think our political system should function and how it
really functions.

Trust in Politicians

(C) Those who make decisions in the Parliament (Storting) do not care about ordinary people's
opinions and beliefs.

(D) Most of the political parties act in accordance with their promises made at election time.

Support for Basic National Institutions

(E) Elections play an important role in the making of political decisions in this country.

(F) The real power in this country is situated outside our Parliament and government.

On each item those who 'agree strongly' were given a score of 5, those who 'agree' a score of 4, 'don t knows' a score of 3, those who 'disagree' a score of 2 and those who 'disagree strongly' a score of 1. On items B, C and F these scores were reversed so that 5 == 1,4 = 2 etc. The index for Political Support was constructed by adding these scores and collapsing two and two categories.

The unweighted additive index has 13 values and high value on the index indicates high degree of
political support.

Attitudes towards Unconventional Participation (AUNPA)

The eight political action types were:

- signing a petition

- participation in unofficial strikes
- taking part in a demonstration

- participation in a local protest action (for examples against traffic, pollution etc.)
- participation in an action campaign

- participation in a boycott (for example purchases, payment of rates)
- use of civil disobedience to impede implementation of official decisions
- make financial contributions to action carripaigns

The respondents were asked if they were 'very positive', 'somewhat positive', 'somewhat negative' or 'very negative' to each of the eight political action types. The scores were added by given those who were 'very positive' a score of 5, those who were 'somewhat positive' a score of 4, the 'don't knows' a score of 3, those who were 'somewhat negative' a score of 2 and those who were 'very negative' a score of 1. The index had originally 33 values. Three and three categories were collapsed and a neutral middle category held separate.

Correlation and factor analysis of the Political Support items and attitudes towards unconventional
participation items confirm that each of these sets taps an underlying empirical dimension.

REFERENCES

Barnes, S. & Kaase M. eds. 1979. Political Action - Mass Participation in Five Western Democracies.
Hills: Sage.

Converse, P. 1964. 'The Nature of Belief Systems in Mass Publics', pp. 206-261 in Apter, D., ed.,
Ideology and Discontent. New York: The Free Press.

Dalton, R. & Flanagan, S. 1982. 'The Changing Content of Ideological Beliefs in Western Europe,
the United States and Japan'. Paper delivered at the Annual General Meeting of the American
Political Science Association, Denver, Sept. 2-5.

Hamilton, M. B. 1983. 'The Elements of the Concept of Ideology'. Paper prepared for the ECPR
Workshop on 'Ideology after the End of Ideology Debate', Freiburg, West Germany, March 20-25.

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

Inglehart, R. 1977b. 'Political Dissatisfaction and Mass Support for Social Change in Advanced
Industrial Societies', Comparative Political Studies 10, 455-472.

Inglehart, R. 1979. 'Political Action: The Impact of Values, Cognitive Level, and Social Background'.
12. in Barnes, S. and Kaase, M. eds., Political Action. Beverly Hills: Sage.

Inglehart, R. 1984. 'The Changing Structure of political Cleavages in Western Societies'. Ch. 2 in
Dalton et al., eds., Electoral Change in Advanced Industrial Societies. Princeton, New Jersey:
Princeton University Press.

Inglehart, R. & Klingemann, H. D. 1976. 'Party Identification, Ideological Preference and the
Left-Right Dimension among Western Mass Publics', ch. 13 in I. Budge et al., eds., Party Identification
Beyond. London: John Wiley & Sons.

Inglehart, R. 1981. 'Post-Materialism in an Environment of Insecurity', American Political Science
Review 75, 880-900.

Knutsen, O. 1983. 'Post-materialisme, middelklassen og elitegrupper i Norge', Tidsskrift for samfunnsforskning
123-154.

Knutsen, O. 1985. Politiske verdier, konfliktlinjer og ideologi - den norske politiske kulturen i
komparativt perspektiv. Oslo: Institute of Political Science.

Knutsen, O. 1985b. Sosiale klasser og politiske verdier i Norge - 'Middelklassen' i den offentlige
sektor som 'den nye klasse\ Report no. 19 from the project on 'Democracy in Norway: Participation
Basic Values'. Institute of Political Science, University of Oslo.

Lafferty,W. & Knutsen, 0.1982. Yrkesom variabel: Noen erfaringermed åbruke'Nordisk yrkesklassifisering' grunnlagfor oppbyggning av yrkesvariabel og etforslag til løsning. Report no. 6 from the project 'Democracy in Norway: Participation and Basic Values'. Institute of Political Science, University of Oslo.

Lafferty, W. & Knutsen, O. 1984. 'Leftist and Rightist Ideology in a Social Democratic State: An
Analysis of Norway in the Midst of the Conservative Resurgence', British Journal of Political
Science 14,345-367.

Lafferty, W. & Knutsen, O. 1985. 'Post-Materialism in a Social Democratic State. An Analysis of
the Distinctiveness and Congruity of the Inglehart Value Syndrome in Norway', Comparative
Political Studies 17,411-430.

Minar, D. W. 1961. 'Ideology and Political Behavior', Midwest Journal of Political Science 5,317-331.

Nielsen, H. J. 1979. Politiske holdninger og Fremskridsstemme. København: Forlaget Politiske
Studier.

Rokeach, M. 1973. The Nature of Human Values. New York: The Free Press

Rokkan, S. 1967. 'Geography, religion and Social Class. Cross-cutting Cleavages in Norwegian
Polities'. Ch. 8 in Lipset, S. M. & Rokkan, S., eds., Cleavage Structure and Voter Alignments.
New York: The Free Press.

Rokkan, S. & Valen, H. 1964. 'Regional Contrasts in Norwegian Polities'. In Allardt, E. and
Littunen, Y., eds., Cleavages, Ideologies and Party Systems. Helsinki: Westermarch Society.

Sainsbury, D. 1983. 'The End of Ideology Debate Revisited: A Preliminary Look at Definitional and
Theoretical Issues'. Paper presented for the ECPR Workshop on 'Ideology after the End of
Ideology Debate', Freiburg, West-Germany, March 20-25.

Sänkiaho, R. 1984. 'Political Remobilization in Welfare States', Ch. 3 in Dalton et al., eds.,
Electoral Change in Advanced Industrial Democracies - Realignment or Dealignment?
Princeton, New Jersey: Princeton University Press.

Valen, H. 1981. Valg og Politikk. Oslo: NKS-forlaget.

Valen, H. & Aardal B. O. 1983. Et valg i perspektiv - en studie av Stortingsvalget 1981. Oslo:
Statistisk Sentralbyrå.

Zuckerman. A. 1975. 'Political Cleavages: A Conceptual and Theoretical Analysis', British Journal
of Political Science 5, 231-248.