Social and Cultural Distinctions in Aalborg, Denmark Annick Prieur, University of Aalborg, [email protected] Lennart Rosenlund, University of Stavanger, [email protected]

Project title and website: Contemporary Patterns of Social Differentiation – The Case of Aalborg (COMPAS). www.socsci.aau.dk/compas

PAPER TO “RETHINKING INEQUALITIES – 7TH ESA CONFERENCE, TORUN, POLAND SEPTEMBER 2005”. WORKSHOP ON SOCIOLOGY OF CONSUMPTION

This paper will present some findings from an ongoing study of the city of Aalborg in Denmark. The overall aim of the study is to enhance the knowledge of social differentiation as a multidimensional phenomenon and analyze its consequences in contemporary society through a thorough empirical study of one city based on a multifaceted methodological approach: a survey, an analysis of register data and qualitative interviews. The project has a point of departure within Pierre Bourdieu’s sociology, and will seek to explore the potentials and limitations of his model of social differentiation. We have found a considerable confirmation of his model, in the sense that our survey, subjected to correspondence analysis, reveals a series of differences that follow the distribution of economic and cultural capital, and, further, that there is a strong correspondence between structural positions and cultural, ethical and political dispositions. An axis going from strong local “roots” to having emigrated from other regions in Denmark seems, however, to structure many lifestyle choices, too. It is quite striking that the cultural elite seems to disengage from local cultural activities, a finding that contrasts with Rosenlund’s earlier analysis of the Norwegian city Stavanger. Furthermore, the traditional high culture seems to play a quite marginal part in Aalborg, so that elements traditionally associated with popular culture play a complicated role within the social distinctions. These questions have, however, not yet been thoroughly analyzed. This paper will give a general outline of the social space in Aalborg. Background The background for this project is on one hand a society that has undergone radical changes over the last generations: continued urbanization, de-industrialization, decrease in blue-collar sector and a simultaneous increase in white-collar sector, growth in public sector, increased level of education,

1

entry of many women into the labor market, new immigrant population etc. On the other hand there is an evolution in theoretical sociology, where some persist to claim that social class still is an important differentiating factor, while a number of other analysts claim this factor has been replaced by new patterns of differentiation, and others again claim a new individualism has rendered all ideas about such patterns obsolete. The overall aim of this project is to enhance the knowledge of social differentiation as a multidimensional phenomenon and analyze its consequences in contemporary society through a thorough empirical study of one city - Aalborg in Denmark – based on a multifaceted methodological approach: a survey, an analysis of register data and qualitative interviews. Further, the aim is to craft improved tools for such analyses. Based on data gathered through different sources, current patterns of social differentiation will be uncovered, as well as patterns of sociogeographical segregation. At a theoretical level, the findings from the project will be used to contribute to ongoing debates on social differentiation. The project has, however, a point of departure within Pierre Bourdieu’s sociology, and will seek to explore the potentials and limitations of his model of social differentiation. The temporal and spatial focus of the project is the current development of a city that has undergone huge transformations in the post-war period. In the first half of the 20th century, the economy was to a very strong degree based on traditional industry and on commerce. The decline of the relative importance of industry and craftsmanship started in the middle of the century, all while administrative, service and liberal occupations increased their share (Christensen & Topholm 1990: 338). The composition of the industry in Aalborg was rather narrow, with a dominance of agroalimentary industry (a tobacco factory being the biggest workplace in the city in the beginning of this period) and production of construction materials (particularly cement). Due to structural rationalization on a national level, the agro-alimentary part decreased (and the tobacco factory closed down), while the cement industry kept on growing, for a while together with the shipbuilding yard, which in 1972 was the second workplace in the city. The biggest was now, and ever since, the municipality of Aalborg. The categories of the official statistics have changed, as well as the borders of the municipality. This makes comparison over time difficult. We have, however, a few indicators of the profound changes in the city’s class composition. Out of the work force in Aalborg in 1960 57 % were classified as workers (Statistisk Årbog for Aalborg 1965, p. 40). In 1989, 30 % of the work force were classified as workers (Statistisk Årbog for Aalborg 1991 p. 17). In 1989, the sectors of financing, education and research, health and social affairs, public administration and other service occupations employed exactly 50 % of the working population (op.cit.: 18). This evolution from blue to white collar has continued into the 21st century. The industrial sector has diminished not only in relative figures, but also in absolute figures. The fastest increasing sectors in these tables are “social work” and the “business service etc.”. The decreasing sectors are agriculture, industry (particularly “iron and metal industry”), and public administration. The ongoing changes of the structure of differentiation may, however, not be read out of these statistics in any detail, as the categorizations they use are outdated; they express the logic of industrial society and not the educational differentiation of contemporary society. A special analysis of the sector of information and communication technology given us by the bureau for statistics and analyses in the municipality reveals the fast growth, from 5 714 persons working within this sector in 1993 to 9 095 in 1999, a 59 % increase, compared to a 16,5 % increase at a national level, and a 40 % increase at a regional level. The University has been a core actor in the development of the city. The University, which received its first students in 1974, has today 11 000 students and 1 500 employees. Many

2

joint ventures involving the university and private companies have taken place in recent time, and especially led to growth within communication technology and currently also medical technology (with a coming establishment of a university hospital). The description above traces out some of the profound changes this society has witnessed over a generation: urbanization, de-industrialization, decrease in blue-collar sector, increase in white-collar sector, growth in public sector, expansion of the educational system with an increased level of schooling among its population as a consequence, entry of many women into the labor market, a new immigrant population (although still small) etc. This makes Aalborg a perfect case to explore transformations in patterns of social differentiation: If class as stratification looses importance (at an objective and/or a subjective level) due to de-industrialization and the so-called new knowledgebased society – we will see it in Aalborg. If new patterns of social differentiation other than those related to class emerge – we will find them in Aalborg. This transformation process is not, however, the focus point of this paper, as we have not yet analyzed our register data (covering the period from 1980 to 2002). Here we will give a snapshot of contemporary social and cultural distinctions in Aalborg, based on survey data. Bourdieu’s model of social differentiation Bourdieu’s theoretical and research production has yielded a massive body of knowledge about the forms and functions of class systems, symbolic violence, education, language, gender, media and a host of other issues central to social life. Most of the social scientists that use Bourdieu, however, focus on the theoretical and philosophical implications of Bourdieu´s ideas, and how his approach may apply to a diversity of areas (sport, fashion, masculinity, television, education, etc.), and tend to overlook the potentialities in his work for analyzing structures and dynamics of change in contemporary society, and have little to say about the overall (methodological and theoretical) approach of Bourdieu. Bourdieu’s book Distinction (1979, English ed. 1984) contains his contribution to the study of the dynamics of social divisions in contemporary society and its interrelationship with formation and divisions of lifestyles. The book explores the relationship between the objective structure of social positions on the one hand (economic and educational assets), and on the other hand lifestyles, i.e. patterns of consumption and preferences in cultural, moral and political matters. There is, on the one hand, an objective system of social positions in which each and all of us have positions, which have bearings on the way we form our lives. On the other hand, there is social practice, or human symbolic activities. These two separate universes of social reality have been conceptualized into two interconnected space constructs: the space of social positions (for short the social space), and the space of lifestyles. The construction of them has been undertaken by the help of correspondence analysis, which is a statistical aid that enables the analyst to unveil and display visually, underlying basic structures in a large and amorphous dataset. The “raw material” for the construction is based on an analysis of principally two forms of capital: economic and cultural. Economic capital consists of access to money or assets that may be transposed into money. Cultural capital (or “informational capital”) exists in an embodied state (i.e. cultural competence, habitus), an objectified state (i.e. books, art etc.) and an institutionalized state (i.e. through educational certificates etc.), and is mainly acquired in the family and in school (Bourdieu 1986). Bourdieu charts in Distinction the variation of positions and of lifestyles, and interprets the axes: The first dimension, the one that accounts for the largest part of the variation, is interpreted as the total volume of capital – how “rich” the social agents are in both economic and cultural capital. The second axis is interpreted as a dimension of capital composition, which refers

3

to the relative amount of the two main forms of capital (and accounts for the differences in lifestyles between for instance teachers and business men). The third dimension, by order of importance (but not presented in a two-dimensional diagram), is a time-dimension referring to the trajectory: the social agents’ history of stability or mobility in a social position (which for instance may explain that newcomers in the economic bourgeoisie may have other consumption patterns than those who have inherited their positions). Bourdieu goes on to depict the different main classes, and within these, the different fractions of the classes: groups with about the same social position and common lifestyles. The analysis is sustained by various other data sources: interviews and observations, photographs, newspaper readings, a series of other surveys and opinion polls etc. Bourdieu underlined that his model was based on a study of one case (France) among all socially possible cases. He was careful not to generalize findings from this analysis to other societies, but regarded his model as a generative model, and explicitly encouraged researchers in other societies to see if the model could serve for constructing other social spaces than the French (Bourdieu 1998). Few have taken up the challenge. Lennart Rosenlund’s (2000) study of Stavanger is internationally the closest one may come to an application of the “analytic tool-box” from Distinction in systematic empirical research. Other applications of this analytic model have either a much more limited scope - Blasius & Winkler (1989), Koch (1996), Hjellbrekke (2000) - or are very superficial – Dahl (1997). There are, however, a couple of very rich studies that also challenge Bourdieu’s model, but without submitting their data to the same kind of analysis: Tony Bennett, John Froww and Michael Emmisons study from Australia (Bennett et. al. 1999), and the ongoing study on cultural capital and social exclusion in Great Britain, a cooperation between Tony Bennett, Modesto Gayo, Mike Savage, Elisabeth da Silva, Alan Warde, and David Wright. What emerges from Rosenlund’s study, is basically a confirmation of Bourdieu’s model. It appears that the very same social forces Bourdieu identified in France in the 1960s and 1970s have their counterparts in the community of Stavanger in the middle of the 1990s. An ongoing analysis by Rosenlund of nationwide data indicates that this also holds for social divisions and the formation of lifestyles at the national level. This does not mean, however, that Norway resembles France, neither that Stavanger is a “typical” Norwegian city. The point is that, albeit all the differences, the structuring dimensions of these societies are basically the same. But, for instance, cultural capital did not have the same role in the Norwegian social space as in the French (or more precisely; it has a more recent role), the relative size of the different class fractions are not the same, and their taste, consumption patterns opinions and preferences are not the same at all; they are just structured according to a similar logic. The idea of this project is to confront Bourdieu’s theoretical framework and methodological approach with a Danish social reality, and thereby provide an arena for testing the applicability of his model. Are the processes of social differentiation in this community meaningfully described with Bourdieu’s concepts of overall volume of capital and composition of capital (the relative amount of economic and cultural capital)? Are these two principles responsible for the production and maintenance of inequalities and social distances in an advanced and differentiated society? And do the principles found to apply in France 25 years ago for explaining the differentiation in formation and articulations of perceptions, classifications and other expressions of lifestyle apply also in this society? The overall aim of our project is to obtain knowledge about current patterns of social differentiation in a Danish context, and an understanding of the structuring forces behind these patterns. This means that we do not wish to carry out a simple replica of Distinction, but an application of its model in order to challenge it and elaborate it according to our empirical findings, and according to a series of theoretically based critical points.

4

Data and data analysis Our working hypothesis is that Bourdieu’s model, with certain modifications, will prove to be applicable also in Aalborg. But it is of course of utmost importance to keep an open eye for the possibility that the structuring principles of Aalborg in 2004 are not the same as they were in France 30 years earlier or in Stavanger 10 years earlier. In order to ensure the open eye, the methods chosen are basically inductive. They will give a possibility for discoveries and surprises, and thereby also for findings that will support competing theories. There are three main sources of data: survey data, register data, and interviews. In this paper we only exploit the survey data. These survey data are drawn from the population in Aalborg. From a sample of 1600 residents between 18 and 75, 1174 were interviewed (73,38 %). Of the remaining 326, 204 refused to participate (12,75 %), 181 were not reached (they had no phone, were not at home in the interview period, had moved or died), and the remaining 41 could not be interviewed due to illness, handicap or language difficulties. We chose to conduct the interviews by phone, as this method today is judged to be the one with the best outreach, and as it is far cheaper than face-to-face interviews. This choice, however, put limitations both on the number of questions and the kind of questions to be asked. The following analysis has the status of “work in progress”. The data collected have been subjected to a number of analyses in order to reveal the hidden structures. Different variables have been examined thoroughly by including them in the analyses and then kept or eliminated, depending on their contribution. Further, different principles of coding have been tested. The solution that we present here is in some respects better than other solutions, in other respects not. Altogether ten variables were chosen based Bourdieu’s choices when constructing the social space in Distinction (Bourdieu 1984:126 – 131); and they are all different indicators of forms of capital. They fall into three groups. The first group of variables comprehends positional variables: the vocations of the father when the respondent was fifteen years of age, and the vocation of the respondents themselves and the type of employment (public/private). The variable describing the father’s work gives a rough indication of the social positions and the social environment under which the respondents were brought up, and combined with the positions of the respondents, they may constitute different types of intergenerational social trajectories. A variable that describes geographical trajectories has been also included. This group of variables comprises of altogether 21 categories. The second group of variables consists of four indicators of economic capital. Thus, one variable gives an indication of whether the respondent possesses shares, bonds, art collection etc. Further, the household income is included in the analysis, as well as ownership and worth of holiday cottage and, finally, ownership and worth of car(s). The indicators of economic capital comprise of nineteen categories altogether. Two indicators of cultural capital constitute the third group of variables. We here have the educational level of the father, which is viewed as indicator of inherited cultural capital. (We also have information about the mother, but in our analyses so far, including this information has not altered the distribution.) Finally, the educational backgrounds of the respondents themselves were grouped into seven different categories. The following table gives an overview of all ten analyzed variables and their frequencies.

Table 1: Analyzed variables and categories. (next page)

5

VARIABLES AND CATEGORIES

Type of community brought up in

Label of categories

Count

Active/passive

City Town Countryside missing category

560 281 283 49

560 281 283 Passive

% 48,4 24,3 24,5 4,2

Count 315 321 218 249 5 10 2 2 3 34 15

Active/passive 315 321 218 249 Passive Passive Passive Passive Passive Passive Passive

27,2 27,7 18,8 21,5 0,4 0,9 0,2 0,2 0,3 2,9 1,3

Label of categories

Count

Active/passive

Basic level Vocational training I Vocational training II Short university training Intermediate university training Long university training Other missing category

399 26 468 27 120 59 31 44

399 26 468 27 120 59 Passive Passive

Label of categories

Count

Active/passive

Technician Physician Biological work Health work University teacher Teacher Journalist, artist Accountant, HR manager Public adm. Top level Manager Insurance work Office work Farming, fishery Transport Metal work Construction work Precision handicraft

35 4 14 29 32 31 12 39 1 37 5 98 5 25 21 39 6

35 Passive Passive 29 32 31 Passive 39 Passive 37 Passive 98 Passive 25 Passive 39 Passive

Father’s vocation Label of categories Father employer Father manager Father skilled worker Father unskilled worker Without work Early retirement Pension Work at home Disability pension Other missing category Father’s education

34,5 2,2 40,4 2,3 10,4 5,1 2,7 3,8

Respondent’s vocation (ISCO)

3,0 0,3 1,2 2,5 2,8 2,7 1,0 3,4 0,1 3,2 0,4 8,5 0,4 2,2 1,8 3,4 0,5

6

A multiple correspondence analysis (MCA) was carried out on these 53 different categories, or modalities, derived from the ten variables. 1 The correspondence analysis constructs a space with few dimensions (one dimension = a line, two dimensions = a plane, three dimensions = a physical space) which optimally “fits” the dispersion among the analyzed properties. The first dimension “explains” the greatest part of the total variance (eigenvalues); the second dimension explains a somewhat smaller amount, the third still an additional amount etc. These amounts of explained variance by the dimensions might be calculated as percentages of the total. The analysis gave the following result:2 Table 2: Eigenvalues, modified rates and cumulated modified rates for the two first principal axes. Axis 1 Eigenvalues (variance of axes) 0.2357 Modified rates % 31,96 Cumulated modified rates % 31,96

Axis 2 0.2175 23,95 55,91

The results of the analysis show that the two first extracted principal axes together explain a bit more that half of the total variance. We will not here explain the statistical “machinery” of the MCA, only describe its functioning by the help of a simple metaphor. Figure 1 contains the two first principal axes produced by a MCA of the data. In this “map” all the respondents have been projected according to their co-ordinates on the two first dimensions. All respondents, represented here as points in the plane, are characterized by a profile of attributes corresponding to their values on each of the ten chosen variables. On the basis of these values – the profiles of attributes - the co-ordinates of the respondents have been calculated. The plane has been constructed in such a way that the maximum “difference” or “contrast” has been created, with regard to attribute profiles, between those respondents situated on the right-hand side of the plane and those on the left-hand side. This is the first dimension, the horizontal one. Similarly, the second largest “difference” with regard to attribute profile, which is possible to create, is represented by the “difference” or “contrast” between the respondents situated in the top part of the plane, and those at the bottom. This second dimension, the vertical one, represents this difference. If these points where real people standing on a football pitch, and we could walk around among them, we would be able to observe that wherever we go, the surrounding respondents will resemble each other; in the sense that they will have many similar attributes. All regions and segments of the plane are characterized by homogeneity of attributes among the respondents. The distribution of these attributes could then be summarized and established simply by asking, for instance, those who have basic level education to raise their hands. The point in the middle of them would then be the center of gravity for this attribute. Then those respondents who have vocational training could be asked to raise their hands, in order to establish the center of gravity for this new 1

The method is presented, among others, by Lebart et al (1984), Greenacre (1984) and (1995), Greenacre and Blasius (1995), Jambu (1983) and (1991), Benzécri (1969, 1992) and in Norwegian: Clausen (1989), Rosenlund (1992) and (1995) and Hjellbrekke (1999), LeRoux & Rouanet (2004). 2 Multiple correspondence analysis underestimates grossly the amounts of explained variance by the principal dimensions. In order to make its results commensurable with those of multiple regression analysis, for instance, Benzécri - the inventor of the technique – has proposed a formula to re-calculate them, to make them more realistic. See Greenacre 1995:145 ff. The given amount of explained variance has been obtained by applying this formula.

7

group of respondents with this attribute in common. In this way we may go through the whole list of attributes - all 53 categories - and establish their positions in the plane. Then the patterns of the points describing the statistical relationships between the attributes can be contemplated3. This is what we are going to do on the following pages, analyzing the characteristics of the social space. Figure 1. Projections of the respondents in the Aalborgian social space, plane 1 –2.4

Interpretation of the principal axes Axis one We shall now examine the “content” of the extracted two principal axes. The presentation is organized around one graph and one table that contain the basic result for each of the axes. We shall examine a plane consisting of axis one and two.

3

This metaphorical description is not quite correct. When projecting the points representing the categories onto the same plan as the individuals, there occurs a difference in scale. 4 The percentages given in the graph are the non-modified rates of explained variance. Their sizes are much lower than the ones presented in table 2. The modified rates are preferred since they better in gauging the true rates of explained variance. (See Le Roux & Rouanet 2004, pp xxx)

8

Figure 2: Plane 1- 2. Categories (21) contributing the most to axis 1. The sizes of the markers are proportionate to the frequencies of the categories they represent.

Table 3: Interpretation of axis 1, 21 modalities contributing the most to axis. Variables are ranked according to decreasing contribution (ctr.). Contribution of variables and modalities are in percent. Variables Respondent’s education

Ctr of variables 19,3

Household income

18,7

Father’s education

14,2

Father’s vocation when respondent 15 years Respondent’s vocation DISCO

13,3

Value of car(s)

11,3

Value of shares, bonds, art-collection etc. Value of holiday cottage Type pf community brought up in Type of employment

3,6

13,1

3,2 2,5

Modalitties Left Univ. intermed. Hum, soc. 580-750DKK + 750DKK Univ. interm. + Univ long Manager Manager + Accountant +techn. + Univ. teacher Value 6 + value 5 Yes

Right Basic level

Ctr of modalities % Left Right 4,0 11,1

150-250 DKK

8,8 + 2,8

3,8

Basic level

¤,6 + 2,3

6,0

Unskilled worker Police, Fire brigade

5,8

&,9

3,8+2,1+1,3+1,3

1,6

Value 1

4,3 + 2,3

2,6

2,7

> DKK 800.000

2,7 Countryside

1,8

0,5

9

The table and the figure give a very clear impression of a social hierarchical axis. It has the bottom (“low class”) to the right and the top (“high class”) to the left (it would have been more pedagogic to turn the graph so that those who in our imagery are at the top also would be so in the graph, but the program – SPAD – has its own logic, which we cannot influence …). Thus, with regard to the education of the respondents – which is the variable that has contributed the most to the first axis – we find that the strongest opposition is between the categories university, intermediate level, Humanistic/social science on the one hand and education basic level on the other. Further, with regard to household income - the second most contributing variable – the strongest opposition comes from the categories 580-750.000 DKK and above 750.000 DKK on the one hand, and 150250.000 DKK, the second lowest income category, on the other. (The lowest category was eliminated from the analysis because of its detrimental effect on the solution; it is an “outlier”.) The characteristics linked to the fathers also fit with this hierarchical logic. On the left side are respondents whose fathers had university education, intermediate or long, while those whose fathers had education on basic level are found on the right side. These results clearly indicate a reproductive pattern; the descendents from an upper social echelon in the parent generation are over represented in the present dominant class, while those who had their social origin in more modest social environment are over represented in the contemporary lower social classes. The categories describing work positions of the respondents do also stick with the hierarchical nature of this first axis. On the left side we find typically “high status” professions: managers, accountants, technicians and university teachers, which all contribute significantly to the axis, while on the right we find polices and firemen. Finally, a second indicator of economic capital does also follow the hierarchical nature of the axis. The respondents who have expensive cars (car value 6 and car value 5) are found on the left side of the map, while those who do not have a car at all (car value 1), are positioned on the right side. Further, when examining the complete print out table from the analysis, one may note that all variables of ordinal nature are nicely ordered monotonously with increasing values when moving from the right to the left. The first extracted dimension is hierarchical in nature and corresponds nicely with the capital volume dimension that Bourdieu presents in Distinction as the first and most important principle of social differentiation. Axis two Let us turn to the interpretation of the second principal axis extracted in the multiple correspondence analysis of the ten mentioned variables. The following graph (fig. 3) and table (4) give the main results. The structure of the second most important axis is of a very different nature than that of the first. What it has in common with the first, is that the respondents’ education is the most contributing variable, this time even more than in the interpretation of the first principal axis. Almost one fourth of the variance of the dimension may be ascribed to this variable. Again the strongest opposition goes between university, intermediate level, and humanistic/social science on the one hand and education basic level on the other. Next in strength is the variable describing the respondent’s vocation, our application of the DISCO category system. This system has many categories, many more than are retained in our analysis (those excluded were smaller than 2 %). On this variable it is the categories technicians and construction workers that are the significant categories as opposed to university teacher, teacher and health workers along the second dimension of the solution. Type of employment, the third variable in order of importance, adds to this (10,8 %) and underlines a significant aspect of this second axis: the importance of working life as a structuring factor. Together the two variables “explain” more than one fourth of the variance of the dimension.

10

Figure 3: Plane 1- 2. Categories (18) contributing the most to axis 2. The sizes of the markers are proportionate to the frequencies of the categories they represent.

Table 4: Interpretation of axis 2, 19 modalities contributing the most to axis. Variables are ranked according to decreasing contribution. Contribution of variables and modalities are in percent. Variables Respondent’s education

Respondent’s DISCO

vocation

Ctr variables 23,0

of

Modalitties Bottom Vocational

17,7

Technician Builder

Type of employment Value of car(s) Father’s education

10,8 10,7 10,4

Private Value 6 Basic level

Father’s vocation when respondent 15 years Value of summer/winterhouse Value of shares, bonds, art-collection etc. Household income Type pf community brought up in

7,8 6,7 6,6

+

> 800.000 DKK Yes

Top Univ. Interm Hum/Soc + Econ/Adm + Univ long Hum/Soc Health + Univ. teacher + teacher Public Value 1 Univers. Intermediate + Univers. long Manager

Ctr of modalities Bottom Top 2,2 12,4 + 2,6 + 4,2

2,0 + 1,9

3,3 +2,0 +4,0

6,0 3,2 3,8

4,7 6,4 3,3 + 2,8 4,5

4,8 5,7

4,1 1,9

11

Next variable in order of importance is one of the indicators of economic capital: value of car(s). The variable has been constructed on the basis of two questions in the questionnaire, one on the number of cars that one posses, the other on the expected sales price of the most expensive car. The two extreme categories belonging to this variable, car value 1 (no car) and car value 6, are positioned opposite each other along the second dimension. The category indicating the highest value is positioned at the bottom (left) and the lowest at the top (right). In between these extreme categories the intermediate categories are ordered in a monotonous fashion along the axis increasing in value as one move along the axis downward. The ordering of the indicator of inherited cultural capital – father’s education – which also has contributed significantly to the second axis, is structured so that the categories that indicate longer and more prestigious educational certificates are positioned at the top of the map (to the left), while those that indicate more modest educational careers – vocational training - are placed at the bottom (to the right). In between the extreme categories (the highest – lowest) the intermediate categories are nicely ordered in a monotonous fashion (categories not shown in the figure). The other indicators of economic capital are ordered in an “opposite” manner; they are ordered inversely, compared to father’s education, but they follow the pattern of the value of car(s). Thus, the highest value of holiday cottage (> DKK 800.000) is found in the bottom region (left), so is the indicator of economic property shares and bonds, yes. Although the indicator household income does not meet the threshold value for contributing to this axis, it is still ordered monotonously along the second axis. All in all the lowest values of the indicators of economic capital are found in the top region (right), the highest values are positioned at the bottom left. In between the extreme values the intermediate ones are ordered monotonously, increasing in value when moving along the second axis downwards (not shown in the figure). When keeping an eye on the indicators of cultural capital (the education of the father and of the respondents) simultaneously in the same movement, we note that these indicators are inversely ordered compared to those indicating economic capital. The second axis is clearly a capital composition axis, very similar of nature to the one described by Bourdieu in Distinction.

A two-dimensional space The first and second axes form a two-dimensional space that we will name the social space of Aalborg, as the dimensions are the same as the two independently operating principles of social differentiation that Bourdieu described as the French social space in Distinction. The analysis accounted for above is one of several that have been carried out and examined. In all analyses undertaken the same or very similar four dimensions have appeared, although their eigenvalues have differed. The first dimension has always been one that is hierarchically ordered, one that could be described as a volume dimension, composed of indicators of different forms of capital. which are monotonously ordered, increasing in value when moving along the dimension, from the right to the left. The second dimension has been in previous analyses just as it is in the present, one that distributes indicators of economic capital and of inherited cultural capital (education of the father) in an inversely fashion along it; at the pole where indicators of cultural capital are high, indicators of economic capital are low, and vice versa. The second dimension has a capital structure character; it describes a “two-component” attribute, where one consists of inherited cultural capital, the other of economic capital. These two dimensions are of the same nature as those analyzed by Bourdieu in his various works, and which he identifies as two independently working principles of social differentiation in the “advanced and differentiated” society. They manifest themselves in society at

12

large, as he shows in Distinction, where differences and oppositions related to the “social space” follow the same logic as differences and oppositions related to lifestyles in the wide sense of the word, as they do in our own study of the Aalborgian social space. Both universes are structured according to the same two principles of social differentiation. (These principles are also present when Bourdieu analyses smaller social entities, such as “the field of power” - Bourdieu 1989. Here the two principles have been reversed in degree of importance (pp 375); it is the capital structure principle that has the strongest impact.) The variables used to construct the social space have, of course, a complex relationship with other so-called “independent” variables. Before we start on the analysis of the space of lifestyles, we want to situate gender, age and geographical origins in the social space constructed above, so that one can keep in mind that an attribute we in the following will link, for instance, to high cultural capital, will also be closer to women than to men, to younger people than to older, and to people with their origins in other places than Aalborg. There are very few immigrants from other countries in Aalborg, but there seems to be an interesting opposition between people having their origins in Aalborg and people coming from other parts of Denmark – with the educated elite largely belonging to the latter group. The complex intersections of these variables will not be analyses in this paper. Figure 4. Gender, age and origin as illustrative points, added to the social space.

13

The space of lifestyles in Aalborg When the space of social positions has been constructed, this can be used as a “predictive map” (Lebart et al 1984:108 ff.) to show characteristic traits with regard to other attributes than those which have been used to construct it. Such attributes - variables and categories - are projected onto the map as so-called supplementary or illustrative points. This technique has been used in constructing the space of lifestyles. This means that the life style variables do not contribute to the distribution, they are just plotted in so that they may be understood in relation to the capital distribution in the scheme over the local social space. First we have a look at the regular reading of newspapers, where the distribution is exactly as one would have expected from Bourdieu’s theories (Figure 5). Information, Politiken and Weekendavisen are all newspapers with an intellectual image, the two former are also politically rather left oriented. No surprise they are most popular among those with high cultural capital. Berlingske Tidende and Jyllandsposten are kind of general “pro-establishment” newspapers, while Børsen is the Danish equivalent to Financial Times – and is mostly read by people with high economic capital. Ekstrabladet and BT are colored newspapers with big headlines – equivalents to Sun. They find their highest popularity among people with low capital. Figure 4. Regular reading of eight newspapers, as illustrative points plotted into the social space.

Next we have a look at which TV-programs the respondents tell us they watch on a regular basis (figure 6). Here the sharp distinctions go between high and low cultural capital – those with high capital enjoy the absurd, ironic entertainment of The Boys from Angora, while those with low cultural capital enjoy the competition Who wants to be a millionaire, watch a health program (The Doctor’s Desk) and have interest for local news as well as for news about the royal family, which they find more interesting than the presidential election in the US. (The survey was carried out in

14

September 2004 – at a time when some of us were very interested in the expected outcome of the presidential election in USA, all while one of the Danish princes announced he was going to divorce from a very popular princess.) The reality shows Robinson and Strandvejsvillaen were also most popular in this region of the social space. The Money Magazine (economic news) is a popular program within the economic elite, just as we expected, and actually it is the only program where the interest for it clearly “follows the money”. Figure 6. Self reported frequent watching of selected TV-programs, as illustrative points plotted into the social space

We included a thoroughgoing news program with background information and debate (Deadline) in our survey, expecting to catch the cultural elite’s preferences, but it did not give exactly what we expected. Much more people said they watched it than what we know, from the TV-companies audience overviews, is true! From the mentioned overviews we for instance know that the reality show Robinson that week was seen by about four times more people than Deadline was, but in our survey Robinson comes out with 18 % self reported frequent viewers compared to 44 % for Deadline. We don’t say that the over reporting of Deadline compared to Robinson is due to people lying, but probably they do not need to watch Deadline so often in order to perceive themselves as

15

people who watch this program, all while they may watch Robinson without identifying themselves as people who watch Robinson. And when asked about TV-preferences by an interviewer, they certainly want to give an impression of themselves as cultivated people – which proves that symbolic dominance still works. The cultural elite’s taste is a dominant taste in the sense that it sets a standard for other people, and influences their perception of their own practices. Music taste reveals to a certain degree the same pattern, although age comes in as a particularly strong determinant for the distribution of likes and dislikes. At first sight, the distribution of music “likes” seemed to confirm the “cultural omnivore”- hypotheses: that people with high cultural capital may have a broad specter of “likes”, including “popular” cultural choices. The cultural choices one would, on the background of the analysis in Distinction, have considered “popular” meaning “low capital”, seem to have become “popular” in the sense of being “general”, shared by a majority. We have, for instance, to a certain surprise, found that the two highest scores for “likes” go to Thomas Helmig, a Danish rock artist with 63 % and DAD (formerly Disneyland After Dark), a Danish hard rock band, with 48 % - while Bizet’s opera Carmen, which we included in order to catch the middle class convenient choice (e.g. “one knows one should enjoy opera, but would perhaps not be able to distinguish the advanced from the déclassé”), only got 34 % likes – and 41 % dislikes!”. When including the “dislikes”, however, a clearer structuring appeared. Figure 7. Likes and dislikes for musical artists, as illustrative points plotted into the social space

16

Tamra Rosanes is a Danish-American country & western singer, while Kandis is a Danish light pop band. The cultural elite does perhaps not “know what they should like” anymore, but they certainly know what they should dislike. (It is an error that Carmen and DAD are not included in the graph, but we could not solve the problem in June when this is written.) When it comes to the uses of the city’s cultural institutions or events and sports and leisure facilities, the sharp opposition goes between those who use these institutions and facilities and those who don’t – rather than between the kinds of facilities. Figure 8. Uses of the city’s cultural institutions or events and sports and leisure facilities, as illustrative points plotted into the social space

Visits to the art museum, to the theatre, to the symphony orchestra, to the annual opera or jazz festival are all clearly located within the zone of cultural capital. Attending evening classes seems to

17

be most popular among people with a balanced capital above the average level, while – the most surprising finding perhaps - attending sport events is clearly located within the zone of economic capital. Gender and age certainly have some influence (see the localization of gender and age in fig.4), as participation in the annual carnival is most popular among the youngest, and the sports among men, but even the ladies’ handball, which has a considerable female audience, comes out on “the economic side”. The only choice that is more frequent at the low capital side is going to the trotting races, but it should be noted that this is one of the low frequent choices (87 % never attend the trotting races, so only the opera festival with 93 % never attending is less “popular” in that sense of the word). What we know from another analysis not included here, however, is that the sports are the only of these activities where the elite with local origins are more present than the elite that has immigrated to Aalborg from other regions. Actually, the expensive cultural institutions like the museum, theatre and orchestra, exist in Aalborg for the people who do not originate from the city. The same structuring goes for choices considered more personal or private – like the arrangement and decoration of one’s home. The respondents were asked whether a series of 11 characteristics applied to one’s own home (9 are included in the diagram). Figure 9. Characteristics that apply to one’s home, as illustrative points plotted into the social space.

18

We can see that those who report that they fill their homes with books, do not fill them with ornaments. Those who hang original art on their walls, tend to consider their homes as modern, and do definitely not consider them as ordinary. “Ordinary” can be a positive characteristic among “ordinary” people, but not in the high capital fractions. The choices that demand economic capital – room, exclusivity and architect-designed furniture – are, as expected, located within the economic corner. All but one choice seem to be more structured by total capital volume (traditional hierarchy) than capital composition. The exception is the preference for order – a clean and tidy home –, which follows the capital composition axis, and seems not at all to be influenced by the volume of capital. With a bit of self-awareness, we are many intellectuals who would admit that we are not orderly people – and some of us would even consider the lack of sense of order as a virtue, and would proudly oppose ourselves to “law & order”-people, so the finding is not a surprise. Still, to live with lack of space and quite a lot of disorder is also a generational phenomenon. Many of those who have a particularly high score on cultural capital and a particularly low on economic capital are students. Figure 10. Opinions and values (5 questions) plotted into the social space as illustrative points.

19

It gives meaning also to analyze opinions and values as structured by the same dimensions. Here it clearly seems like level of cultural capital is the most important for most of the selected opinions and values – like whether the Danish aid to developing countries is too small or not, whether employers should choose Danes when there is shortage of jobs or not, whether there are too many who do not want to work or not, and whether one would prefer having a male boss over a female. The only of these value questions that “follows the money” is the question that goes directly to economic distribution: Whether increased wage differences would improve Danish economy or not. There is a general trend within political science to consider voting as less related to class structure than earlier – that voting is now issue voting instead of class voting, or that it is related to e more profound value orientation that is considered independent of class structure. But as the graph above shows, values and opinions on diverse issues seem to be related to class structure, albeit in a complex way. Even if we in our study find it difficult to draw a simple line between left and right voting, and relate this to class, there should be no doubt that voting could be understood in the light of a more complex, contemporary class structure. Figure 11. Voting for parliamentary elections as illustrative points plotted into the social space

A traditional vision of classes does not apply well to the left-right divide in Danish politics, if we see this traditional vision as corresponding to the axis of total volume of capital (the horizontal axis). But two of the three parties considered to be left in politics, namely Socialdemokratiet (Social

20

Democrats, have had the government in several and long periods) and Socialistisk Folkeparti, which is to the left of the social democrats, are situated on the side with total capital volume less than average. Four of the six parties considered to be in the center or at the right side of the spectrum are situated in the area with total capital above the average. Among these are the two parties governing today: Venstre, originally the farmers’ party, today a right wing, liberal party, and Det konsevative Folkeparti, a party with a more traditional conservative ideology. They are situated close to the small Centrumdemokraterne, a party in the center of the traditional spectrum. Radikale Venstre is also ideologically a non-socialist and social liberal party, but cooperated with the Social democrats in the last government, just as they have cooperated with parties to the right of them in earlier governements They are more profiled on values than on economic questions, and they have an image as a party for intellectuals - not without reason, as one can see. Their position in the landscape, far from the other non-socialist parties, gives us a hint about how the other dimension of the social space works with respect to political choices: They are close to Enhedslisten, a party that was formed of what was left at the radical left after the breakdown of the wall in Berlin. Apparently their voters vote against their own economic interests. We already knew this party’s voters had the highest level of education of all parties, but are still surprised the difference in total capital is so big between this party and the other party left of the Social Democrats – Socialistisk Folkeparti. The small Kristendemokraterne – The Christian Democrats – is one of the two non-socialist parties that are situated within the area for low capital. They are at the cultural pole within this area, while the other, Dansk Folkeparti – The Danish People’s Party, a right wing nationalist party – is over to the economic side. It should be no surprise that the Social Democrats have lost a lot of voters to the Danish People’s Party, as their voters resemble each other in many aspects. They also resemble those who did not vote or do not want to tell what they voted. We believe some of the latter actually voted for the People’s Party, as their support seems to be underreported (only 5,6 % self reported voters for this party, which is about half of the percentage they got in both the preceding and the following parliamentary election). Conclusion? Well, not yet, but … We will have to do a more thorough analysis of our findings, and confront them with findings from other studies, before we draw any conclusion. But let’s say that we take these findings as indications that economic and cultural capital structure many life style choices in Denmark today. The complex interplay with other structuring factors, particularly gender and age, needs, however; a further empirical analysis, and the meaning of the uncovered differences needs to be documented through qualitative studies. And after all that, we may start to discuss whether class is a meaningful term for designing the differences uncovered.

Literature Bennett, Tony; Frow, John & Emmison, Michael (1999): Accounting for Tastes: Australian Everyday Culture. Cambridge, Cambridge University Press. Benzécri J. P. (1969): “Statistical Analysis as a Tool to Make Pattern Emerge from Data”. In: Natanabe, S. (ed.): Methodologies of Pattern Recognition.. New York: Academic Press. Benzécri J. P. (1992): Correspondence Analysis Handbook. New York. Blasius, Jürg & J. Winkler (1989) „Gibt es die „Feinen Unterschiede“? Eine empirisches Überprüfung der Bourdieuschen Theorie.“ Kölner Zeitschrift für Soziologie und Sozialpsychologie no. 41: 72-94.

21

Bourdieu, Pierre (1984) Distinction. A Social Critique of the Judgment of Taste. London: Routledge and Kegan Paul. Bourdieu, Pierre (1986) “The Forms of Capital.” In: J. G. Richardson (ed.) Handbook of Theory and Research for the Sociology of Education. NY: Greeenwood. Bourdieu, Pierre (1998) Practical Reason: On the Theory of Action. Cambridge: Polity. Bourdieu, Pierre (1996) State Nobility. Cambridge: Polity. Christensen, Per Bo & Jens Topholm (1990) Aalborg under stilstand og fremgang fra 1814 til 1970. Aalborg: Aalborg Kommune. Clausen, Sten-Erik (1989): Korrespondanseanalyse: Grafisk representasjon av frekvenstabeller. NIBR-rapport 1989:14. Dahl, Henrik (1997) Hvis din nabo var en bil. København: Akademisk forlag. Greenacre Michael J. (1984): Theory and Applications of Correspondence Analysis. London: Academic Press, INC. Greenacre Michael J. (1995): Correspondence Analysis in Practice. London: Academic Press, INC. Greenacre Michael J., Blasius, Jörg (1995): Correspondence Analysis in the Social Sciences. London: Academic Press, INC Hjellbrekke, J. (1999): Innføring i korrespondanseanalyse. Hjellbrekke, J. (2000) Studying mobility trajectories in the Norwegian Social Space. University of Bergen: Department of Sociology. Jambu, M., Lebeaux, M-O (1983): Cluster Analysis and Data Analysis. Oxford: North Holland Publishing Company. Jambu, M.(1991): Exploratory and Multivariate Data Analysis. San Diego: Academic Press. Koch, M. (1996) “Class and Taste. Bourdieu’s contribution to the Analysis of Social Structure and Social Space.” International Journal of Contemporary Sociology. Vol. 33 no. 2. Lebart, L., Morineau, A. Warwick (1984): Multivariate Descriptive Statistical Analysis. Correspondence Analysis and Related Techniques for Large Matrices. New York: John Wiley & Sons LeRoux, B. & Rouanet H. (2004): Geometric Data Analysis. Dordrecht, Kluwer. Rosenlund, L. (1992): Korrespondanseanalyse. Metode for kvantitativ behandling av kvalitative data. Arbeidspapirer fra Høgskolesentret i Rogaland, nr 156. Rosenlund, Lennart (1995): Korrespondanseanalyse. Dataanalysens magiske øye. Sosiologisk tidskrift nr 1 (1995), 55-78. Rosenlund, Lennart (2000) Social Structures and Change : Applying Pierre Bourdieu’s Approach and Analytic Framework. Working papers 85/2000. Stavanger: Stavanger University College. Statistisk Årbog for Aalborg 1965, 1991, 2001. Aalborg: Aalborg statistiske kontor.

22