The Structure of the Upper Class: A Social Space Approach

Abstract

This article seeks to identify the internal divisions within the ‘upper class’ of Norway, defined as comprising different types of property owners, top executives and business managers. Bourdieu’s concepts of social space and forms of capital are applied to construct a social space of the Norwegian economic upper class by subjecting 12 indicators of capital to Multiple Correspondence Analysis. Central issues in the sociology of elites and upper classes are addressed, including the role of educational credentials in upper class reproduction, and the salience of divisions by social origin. The article reveals a maintained division between owners and employees (managers, executives, business professionals) in an age of ‘financialisation’. Furthermore, the divisions established are related to the segmentation of the upper class by occupation and industry.

Keywords

Bourdieu capitalists class analysis elites forms of capital geometric data analysis multiple correspondence analysis social differentiation social space upper class

Class analysis typically deals with the relations between social classes. Relations within classes are less frequently examined. Analyses that consider only inter-class relations may risk treating classes as monolithic categories, neglecting their internal heterogeneity. Needless to say, any class category will be diverse in some respects. What is crucial is whether classes are divided in sociologically meaningful ways. In this article, I study the internal differentiation of the Norwegian upper class, conceived as people living off property, as well as the top strata of executives, managers and business professionals.

Even though class analyses more often deal with inter-class relations, there are some notable studies of intra-class relations in the literature. Weber emphasized the divergent sources of social power among the propertied classes, ‘according to the kind of property that is usable for returns’. This led him to differentiate the propertied class into rentiers and entrepreneurs (2009[1946]: 182). Dahrendorf held that the emergence of ‘post-capitalist society’ pitted managers against owners, with a ‘decomposition of capital’ fragmenting the upper class (1959: 41–8). Parkin argued that the bourgeoisie in capitalist society relied on two distinct exclusionary devices: education and the institutions of private property (1979: 48). Similarly, Bourdieu underlined how different forms of capital stood in opposition to each other, as competing bases of social power, and that this ‘chiasmatic structure’ helped explain the various struggles between dominant groups (1996: 264–5). Furthermore, Bourdieu developed a rigorous conceptual and methodical approach to address these questions, through the coupling of his theoretical concept of social space with the technique of Multiple Correspondence Analysis (MCA). This serves as the chief methodical inspiration for the present study.

Drawing inspiration from these authors, I examine potential internal divisions within the Norwegian economic upper class as of 2005. The analytic strategy is explorative, seeking to uncover the particular structure of internal divisions by going ‘bottom-up’, using MCA to determine which important differences can be uncovered. Based on existing literature, three lines of division are to be expected: division between owners and managers; division by the possession of educational credentials; and division by social origin. Having established the potential existence and structure of these divisions, their relation to the occupations and industries of the upper class members is examined.

The article answers Savage and Williams’ recent appeal to pick up Bourdieu’s analytical tools to investigate the composition of elites (Savage and Williams, 2008: 15). Social space, and the closely related conceptualization of forms of capital, offers powerful means to deal with intra-class relations. Social space is best known from its application in Distinction, where it signified something like a class structure (Bourdieu, 1984: 114–29). Using the social space approach to examine internal class cleavages draws inspiration from several of Bourdieu’s own applications (1988, 1996). Several other researchers have applied social spaces to examine elite groups, notably Lebaron (2001, 2008) on economists and central bankers, and Hjellbrekke, Korsnes et al. (2003, 2007) on the field of power. Like all these researchers, I use MCA (Le Roux and Rouanet, 2010) to construct a social space of positions. The data are drawn from Norwegian tax, employment and education registers.

Upper Classes and Internal Differentiation

‘Upper class’, like class itself, is a contested concept. Opinions differ as to whom or what should be considered the upper class in contemporary societies. In the present context, it must suffice to say that the present study is based on a fairly ‘orthodox’ conception, in terms of class theory: the upper class is understood as consisting of various owner capitalists, as well as the top strata of employed executives, managers and business professionals. This involves a partial concession to the ‘managerialist’ position, that the rise of impersonal ownership turned managers into the new upper class. While these claims were overstated, it should be conceded that the developments identified by the managerialists indeed empowered the employed executives and managers, warranting them being considered part of the upper class. In these respects, I follow John Scott’s (1997) interpretation. Through the operationalization (specified below), this includes the top strata of various groups engaged in finance.

Three potential types of division can be expected.

Opposition between managers and owners: Many authors expected that the rise of impersonal ownership and the ensuing bureaucratization of firms would entail a ‘managerial revolution’, with salaried managers and executives becoming the new dominant class (see discussions in Scott, 1997; Zeitlin, 1989). Dahrendorf held that the opposition between these groups would decompose the entire capitalist class (1959: 41–8). However exaggerated these claims may have been, Scott argues that executives and top managers should indeed be considered parts of the capitalist classes in contemporary societies (1997: 274–80).

Division by educational capital: The education system is underlined as important for class reproduction in closure theory (Murphy, 1988: 161–91; Parkin, 1979: 54–60), and credentials constitute one of the core components of cultural capital (Bourdieu, 1986: 247–8). Scott follows Bourdieu in arguing that, with the rise of impersonal property, the possession of credentials and cultural resources is becoming increasingly important for the reproduction of the capitalist class (1997: 292–310). Nevertheless, contemporary upper classes consist of both highly educated and less educated groups, and this might constitute an internal opposition.

Division by inherited capital: The social origin of elites is a salient theme in the literature (e.g. Giddens and Stanworth, 1974; Hartmann, 2000; Useem and Karabel, 1986). Earlier research found the Norwegian business elite to be disproportionally recruited from the upper classes (Gulbrandsen et al., 2002; Mastekaasa, 2004). However, it seems reasonable to expect a range of different origins within this class. Higher social origin might constitute an important asset or resource, most obviously in the case of inherited firms, stocks or wealth. This ‘volume of inherited capital’ is identified as an important line of division by Bourdieu (1996).

My main objective is thus to establish whether these lines of division can be detected in the Norwegian upper class, and to explore how these types of division relate to each other. Modelling a social space from the upper class members’ capital profiles will establish what divisions exist, and whether some of them overlap. Occupations and industries are projected onto this space, revealing how they relate to divisions in capital profiles. If they overlap in meaningful ways, this would reinforce the supposition that these are actual social divisions, since occupation and industry are typical indicators of differing fractions of the upper class (such as finance and manufacturing).

Social capital or networks arguably constitute the most central theme in the study of business elites (e.g. Carroll, 2004; Domhoff, 2010; Harvey and Maclean, 2008; Maclean et al., 2006; Useem, 1984). No detailed analysis of networks is undertaken here, but access to resources through the family is considered a part of a person’s capital stock. The data used offer two key indicators of family-based social capital: siblings’ economic and educational capital. Hansen (2009) has shown that siblings’ resources importantly affect the economic success of people with law degrees, even when the effects of social class origin are held constant.

I use Bourdieu’s concept of social space to model social differentiation of the upper class. Social space amounts to a relational ‘social topology, […] an analysis of relative positions and of the objective relations between these positions’ (Bourdieu, 1989: 16). The structures of social spaces are shaped by the distribution of, and relation between, various forms of capital; that is, scarce resources that may be ‘invested’ in particular fields to gain advantages. Capital may come in many variants, but the main forms are economic, cultural, social and symbolic (Bourdieu, 1986). When agents are situated near each other in social space, this means that they have similar capital profiles – their social positions are similar.

In Homo Academicus and The State Nobility, Bourdieu and his colleagues studied divisions between university professors and corporate heads by correspondence analyses of a set of capital indicators (Bourdieu, 1988: 271–6, 1996: 300–8, 340–9). In these studies, one basic assumption is that the volume and composition of one’s capital is a key component of one’s power and social position. Social space is the means to capture and display the system of such positions. A host of researchers have applied these notions to different questions (for a brief review see Lebaron, 2009: 23–5). Such an approach focuses on how the accumulation and conversion of capitals constitute a key ‘mechanism’ in class and stratification. Agents’ strategies are conditioned by their locations in social space (and fields), and their stock of capital is a crucial resource for gaining advantages (Savage et al., 2005).

Data and Method

While enjoying a certain increase in popularity (e.g. Bennett et al., 2009), MCA is still not widely applied in sociology. Nevertheless, it fits the bill for the questions addressed here. First, the common relational properties between social space and MCA make it particularly suited for modelling social spaces (Bourdieu, 1996: 263–4). The spaces produced by MCA are relational: the position of any individual or category within the space is only meaningful in relation to those of all others, just as any position in social space is intelligible only in terms of its relationship to other positions. Second, as a technique of pattern detection and data reduction it allows the inductive determination of what divisions are important. In this, MCA differs from mainstream techniques which test a hypothetical model on the data. Third, MCA can deal with many variables and is devised for categorical data, while making no assumption about distributions and relationships in the data (Le Roux and Rouanet, 2010).

This study uses 2005 data, compiled from official registers administered by Statistics Norway and containing information on the entire Norwegian population. These registers provide reliable data on occupation, income, assets and education, as well as the means to link individuals to parents and siblings. The registers encompass all persons living in Norway born between 1955 and 1970. This amounts to some 1,231,640 individuals, 48.6 per cent women and 51.4 per cent men.

A subpopulation of these data is drawn consisting of people considered to belong to the economic upper class. By drawing on and simplifying Scott’s identification of four capitalist class situations – entrepreneur, rentier, executive and finance capitalists (1997: 278–9) – I define the upper class as consisting of large property owners and top executives, managers and business professionals. Property owners are operationalized in two steps: identifying those 1) who draw most of their income from property (capital income, income from self-employment),¹ and 2) whose ‘property income’ exceeds NOK (Norwegian Kroner) 1 million annually (€125,786 in December 2005). ‘Executive capitalists’ are operationalized by selecting a range of managers, executives and business professionals by occupation, and then selecting those who earn more than NOK 1 million annually.² The dataset does not constitute a random sample, but a specific population³ of 1.2 per cent of the Norwegian population – 14,732 individuals, 86.3 per cent men and 13.7 per cent women. Apart from its strong male dominance, the category is characterized by very large assets, moderately high education levels, and a clear over-representation of higher social origins (see Table 1).

Table 1.

Marginal distributions of all variables. Data provided by Statistics Norway. Variables measured in 2004-2005, parents’ income measured in 1965–1973 (assets in NOK – Norwegian Krone). N = 14,732

Inherited capital	Counts	%	Economic capital	Counts	%	Cultural capital	Counts	%	Family-based social capital	Counts	%	Supplementary variables	Counts	%
Social origin			Wage			Education level			Number of siblings with high education			Occupational groups
Managers/executives	3567	24.21	0 to 499,999	6373	43.26	1 (Primary, none or lower)	508	3.45	0	4429	30.06	Directors and chief executives	3038	20.62
Professionals, academics	1279	8.68	500,000 to 999,999	3094	21.00	2 (Uncompleted secondary)	1441	9.78	1	5180	35.16	Production and operations dept. managers	1410	9.57
Engineers/administrators	1671	11.34	1,000,000 to 1,499,999	3880	26.34	3 (Vocational)	1396	9.48	2	2670	18.12	Other dept. managers	1493	10.13
Medium level, private sector	1061	7.20	1,500,000 and higher	1383	9.39	4 (Upper secondary)	2280	15.48	3–6	890	6.04	General managers of small enterprises	1048	7.11
Medium level, public sector	529	3.59	Capital income			5 (Lower university/college)	5604	38.04	No siblings	904	6.14	Business professionals	676	4.59
Small firm managers	983	6.67	Minus to −1	671	4.55	6 (Higher university/college)	3381	22.95	missing category	659	4.47	Executive officers in administration, business services, social work and entertainment	1405	9.54
Lower level employees, service	846	5.74	0 to 9999	2783	18.89	missing category	122	0.83	Number of siblings with top incomes			Other occupations	2194	14.89
Farmers, etc.	861	5.84	10,000 to 999,999	5192	35.24	Educational field			0	6850	46.50	missing category	3468	23.54
Skilled workers	1212	8.23	1,000,000 to 1,999,999	3232	21.94	General	1963	13.32	1	4211	28.58	Industry classification
Unskilled workers	2128	14.44	2,000,000 to 3 999,999	1599	10.85	Humanist	567	3.85	2	1416	9.61	Manufacturing	1869	12.69
missing category	595	4.04	4,000,000 and higher	1255	8.52	Teacher, pedagogical	190	1.29	3–5	417	2.83	Wholesale and retail trade; repair of motor vehicles and motorcycles Transportation and storage	770 2419	5.23 16.42
Parents’ relative income level			Self-employed income			Social studies, law	995	6.75	No siblings	904	6.14	Information and communication	581	3.94
0	1098	7.45	Minus to −1	990	6.72	Economic/admin	5331	36.19	missing category	934	6.34	Financial and insurance activities	1089	7.39
1	1292	8.77	0	9359	63.53	Science, technical	3786	25.70				Professional, scientific and technical activities	2646	17.96
2	1832	12.44	1 to 999,999	1654	11.23	Health	1014	6.88				Other industry	1889	12.82
3	5107	34.67	1,000,000 to 1499	1559	10.58	Primary sector	263	1.79				Unclassifiable (incl. missing)	3469	23.55
4	4910	33.33	1,500,000 and higher	1170	7.94	Transport, security	424	2.88
missing category	493	3.35	Assets			Unclassifiable	76	0.52
Parents’ highest education level			0 to 999,999	4582	31.10	missing category	123	0.83
1 (Primary, none or lower)	1444	9.80	1,000,000 to 3999	5703	38.71
2 (Uncompleted secondary)	4369	29.66	4,000,000 to 8,999,999	2571	17.45
3 (Vocational)	1382	9.38	9,000,000 to 14,999,999	861	5.84
4 (Upper secondary)	1524	10.34	15,000,000 and higher	980	6.65
5 (Lower university/college)	3104	21.07	missing category	35	0.24
6 (Higher university/college)	1996	13.55
7 (Unknown)	151	1.02
missing category	762	5.17
Parents’ education field
General	3246	22.03
Humanist	301	2.04
Teacher, pedagogical	1334	9.06
Social studies, law	594	4.03
Economic/admin	2262	15.35
Science, technical	4024	27.31
Health	572	3.88
Primary sector	616	4.18
Transport, security	877	5.95
Unclassifiable	144	0.98
missing category	762	5.17

For a social space analysis, one needs to choose variables that serve as indicators of capital. Emphasis will be placed on economic and cultural capital, both personal and inherited, and on proxies for family-based social capital. Economic capital is measured using 20 active categories: wages (4), capital income (6), self-employed income (5) and assets (5). Cultural capital is measured using 24 active categories: educational level (6), educational field (5), parents’ highest educational level (6) and educational field of parent (7). Family-based social capital is measured using eight active categories: number of siblings with higher education (4), and number of siblings in the tenth income decile (4). Inherited capital is measured using 15 active categories: father’s occupational class (10) and parents’ relative income (5). Parents’ education also provides a measure of inherited capital. The supplementary variables concerning occupation and industry (not capital indicators) are coded, respectively, from the ISCO-88 based STYRK code (SSB, 1998) and on the NACE rev. 2 based Standard Industrial Classification (SSB, 2008). All variables are drawn from data from 2005, except for the measures of inherited capital: social origin is measured as father’s occupational class in 1980; parents’ relative income is measured as mean income when the individual was aged 10–18; parents’ education is measured as their highest education obtained as of 2004, using the parent with the highest level.⁴ Marginal distributions of all variables with active and passive categories are given in Table 1.

A space of social positions in the upper class is constructed by analysing 12 indicators of capital using MCA. Distance and proximity between points in this space – measured relatively in standard deviations – represent difference and similarity of capital profiles. Individuals that appear close to each other have similar capital profiles, and categories that often go together appear close to each other. The goal is to sum up as much as possible of the variance in the Individuals x Categories table with as few dimensions as possible, while at the same time obtaining the best possible representation of the main structures in the data. Dimensions appear in descending order by the amount of variance (eigenvalues) accounted for (Le Roux and Rouanet, 2010). For example, in Bourdieu’s analysis of a range of variables on the lifestyles of dominant classes, the first dimension represented the contrast between the tastes of the fractions rich in cultural and economic capital, respectively, and the second represented differences of taste between the inheritors and the newcomers in the class. Once the space is constructed, supplementary/illustrative variables on occupation and industry are projected into it, so as to visualize how these characteristics are related to the structure of the space. This is equivalent to the procedure in Distinction, where indicators of social position were projected onto the space of life-styles (Bourdieu, 1984: 262).

The Space of the Upper Class

Three axes are retained for interpretation and further analysis. Axis 1 accounts for 38.6 per cent; axis 2 for 23.9 per cent; and axis 3 for 12.7 per cent of the total variance.⁵ This article focuses primarily on the two first axes. For interpretation, I inspect ‘explaining points’, categories that contribute above average to shaping the axes. Figure 1 shows explaining points for axis 1 and Figure 2 for axis 2. Axis 3 is not displayed, but the appendix table shows all explanatory points with contributions per axis. Distances in the space are relative: each coordinate has no inherent meaning, but denotes relative distance and proximity to other points.

Figure 1.

Explaining points for axis 1, plane 1-2. Percentages refer to rates of modified eigenvalues (variance accounted for). Size of points relative to frequency of categories; units used are NOK (Norwegian Krone)

Figure 2.

Explaining points for axis 2, plane 1-2. Percentages refer to rates of modified eigenvalues (variance accounted for). Size of points relative to frequency of categories; units used are NOK (Norwegian Krone)

The first axis contrasts volume of inherited and educational capital. On the left of this axis we find the upper class members of high social origin and high educational level, and on the right we find their opposites. The axis is primarily shaped by origin variables – parents’ income, education and occupational class – but also by respondents’ educational levels as well as number of highly educated siblings. Educational level and number of educated siblings follow the contrast between high and low origins. This suggests three things: first, a contrast between high and low social origins is the key line of division in the economic upper class; second, educational levels follow the same lines; lastly, having educated siblings also overlaps with the division by origin. Typically, then, upper class people of high origins have high education and also highly educated siblings, whereas the reverse holds true for those of lower origin.

The second axis separates owners (bottom) from employees (top) by their source of economic capital, which also overlaps with a contrast between social origins in the business world (executives, managers, small firm owners) and origins outside it. Indicators of economic capital – wages, capital and self-employed income, assets – are primary in shaping the axis. At the bottom we find those receiving large capital incomes and holding large assets, and those of business social origin – that is, parents who are executives and managers or running small firms. At the top of the diagram we find those depending on salaries rather than capital income, and having social origins outside the business world. Being from managerial or small-firm owner origins corresponds to large capital income and assets, whereas working class origins are clustered around low or no such income. Capital income and assets differ from wages and formal positions in that the former may be subject to direct inheritance. Axis two then suggests that there is more social closure around property than positions.

The number of siblings with top incomes maps out on axis 2 as well, showing that having wealthy siblings is more common among the owners of higher origins. People of lower origins typically have lower volumes of economic capital, and more often have wages or salaries than income from property. Moreover, those from lower origins lack the economic and cultural resources available through the family networks of the higher strata. Class reproduction seems linked specifically to capital incomes and assets, testifying to the importance of the direct inheritance in these processes. Thus, in factorial plane 1–2 we find the core inheritors in the lower left section of the space.

The explaining points for axis 3 can be found in the supplementary table (Appendix 1). It modifies the main divisions by distinguishing groups based on self-employment from the rest. It is mostly shaped by self-employed income, educational level, social origin and wages. Considered in plane 1–3 (not shown), we can distinguish three groups: those of farming origins with relatively low education and reliant on self-employed income; those with somewhat higher education within economics and administration, social origins in the private sector at various levels and wages between 1 and 1.5 million NOK; and, lastly, those with higher education and with highly educated parents and siblings, in social science, health and law – i.e. a self-employed cultural capital and/or professional fraction.

To summarize, three principal potential lines of division are established. The primary division is by volume of inherited capital, measured by occupation, income and education. This overlaps with level of education, as well as siblings’ education. The second division is between owners and managers. This overlaps with an opposition between managerial and working class origins. The third axis modifies these divisions by separating groups relying on self-employed income from the others, serving to identify two particular fractions: a self-employed cultural capital fraction and a self-employed fraction of farming origin.

Occupation and Industry

Supplementary variables are used to examine whether the divisions in capital profiles overlap with segmentation of the upper class by industry and occupation. To judge how well these categories are separated on an axis, one considers the deviation between the coordinates of the categories’ mean points: Less than 0.5 is regarded as small, above 0.5 as notable, and above 1 as large. In the cloud of individuals, concentration ellipses give a geometric summary of the distribution of categories. Along each axis the ellipse delineates two standard deviations from the category’s mean point. This helps determine the dispersal of the different occupations and industries (Le Roux and Rouanet, 2010: 70–1).

Figure 3 shows the mean points of the supplementary variables ‘occupation’ and ‘industry’ in factorial plane 1–2. Figure 4 displays their concentration ellipses in the cloud of individuals. Figure 3 thus reveals average differences, whereas Figure 4 summarizes the dispersion of the individuals with these properties. Managers of small firms are on the right, meaning that these positions are associated with lower volumes of inherited and educational capital. On the left we find most managerial, executive and business professional positions – save for directors and chief executives – indicating that these occupations are relatively more distinct for the fractions rich in inherited and educational capital. Note that the concentration ellipses demonstrate that we are not dealing with clear-cut divisions.

Figure 3.

Supplementary points for occupation and industrial classification in plane 1-2 Percentages refer to rates of modified eigenvalues (variance accounted for)

Figure 4.

Concentration ellipses for occupation and industrial classification in the cloud of individuals, plane 1-2. Percentages refer to rates of modified eigenvalues (variance accounted for)

As regards industries, ‘wholesale and retail trade of motor vehicles’ is most distinctive for those with low volumes of both inherited and educational capital (right side), whereas ‘finance/insurance’ is most distinctive for those with high volumes of the same. It would seem these industries mark out distinct fractions of the upper class, as deviations here are >1. From the concentration ellipses it is clear that the categories are reasonably well separated. Many industries are placed close to the centre on the first axis, indicating that they are not distinct in their capital profiles, i.e. more neutral in terms of origin or education. Interestingly, it appears that ‘finance and insurance’ is also clearly separated from those involved in ‘manufacturing’.

There is not much distinction between the industries along the second and third axes, except from ‘wholesale and retail trade of motor vehicles’ which appears on the negative side of the third axis, close to those of self-employed and of farming origin. Upper class members of working class origin typically cluster around smaller businesses, particularly in the wholesale and retail trade of motor vehicles. This contrasts with those of higher origins, who more often work as executive officers, managers or business professionals in financing and insurance or in professional, scientific and technical services. Being involved in manufacturing, information/communication and transport and storage appears more neutral in these respects.

The ellipses in Figure 4 show that the segmentation of the upper class by occupation and industry is only partial: executive officers and business professionals seem to be the most socially exclusive, while most of the occupations are more evenly distributed. Similarly, finance seems to be the most exclusive industry and motor vehicles the least so.

To sum up, plane 1–2 can be read as follows. In the upper left quadrant, we find the educated upper class members of privileged origins, although from outside the business world, employed as executives, managers and business professionals, with finance/insurance and professional/scientific services as their most characteristic industry. In the lower left quadrant we find the inheritors proper, with privileged origins from the business world, with large capital incomes and assets. Top right we find the ‘self-made’ employees in the business world, of worker and farmer origins. In the bottom right we find the ascendant sons and daughters of small firm owners, who themselves live off property income and manage small firms, most characteristically dealing in motor vehicles.

Concluding Discussion

These analyses turn up lines of division echoing salient themes in the literature on elites and upper strata: the contrast between high and low origins, divisions by educational credentials, and between owners and managers. It is indeed difficult to assess whether the divisions uncovered here correspond to actual social divisions. The fact they are meaningful in light of theory and research in the field, and in certain respects overlap with occupation and industry, merits serious consideration

By way of conclusion I would like to focus on four points, gradually moving from issues most immediately related to my own findings on to points of more general significance. First, this study highlights the importance of the educational system for class reproduction. In Bourdieu’s and Scott’s work, educational credentials are seen as being of rising importance in the reproduction of the upper class. Accordingly, educational credentials function more as a means of staying in the upper class for persons originating in it, than a route into the upper class for those of lower origins. This sits well with the present study’s finding that, in general, the volume of educational capital follows the volume of inherited capital. But since this refers to origins rich in both cultural and economic capital, this demonstrates the importance of educational credentials for both capitalist class reproduction and the inflow into the economic upper class from other elite origins. Furthermore, this corresponds with a Norwegian business newspaper’s report that several prominent, lower class celebrity capitalists in Norway boast that they have no higher education, and downplay the value of education as a means to success.⁶ However, in what might appear as recognition of the importance of education for class reproduction, their own children take up educations in business.

The correspondence between high volumes of both inherited and educational capital points to the importance of conversion strategies (Bourdieu, 1984: 125–32, 1996: 277–99): families may facilitate class reproduction by converting between forms of capital. Two types of conversion strategies seem important: either converting (inherited) economic capital into education capital, in turn facilitating the reverse conversion (as when children of wealthy origin take up exclusive business educations, and subsequently land a position as executive or board member); or, converting one’s inherited cultural capital into acquired educational capital and then into economic capital (as when children of origins rich in cultural capital get a business degree and find work in finance). In the case of the economic upper class, there seems to be a close connection between the two, potentially competing, exclusionary devices of property and education, to use Parkin’s terminology. The value of inherited capital, whether economic or cultural, seems to be secured through education, providing the ‘inheritors’ with the legitimacy offered by the supposedly merit-reflecting credentials. Parkin identified ‘a permanent tension within this class resulting from the need to legitimate itself by preserving openness of access, and the desire to reproduce itself socially by resort to closure on the basis of descent’ (1979: 47). The (re)conversions of capital seem to let them have their cake and eat it too.

Second, this study demonstrates the salience of divisions by social origin. This connects with the traditional interests of both the sociology of elites and of stratification: the recruitment into elites or upper strata. Most such analyses demonstrate an over-representation of higher origins in the higher classes. While such findings undoubtedly are important, one needs to broaden the scope to recognize that social origin is not simply a factor affecting the chance of entry into upper strata, but is also a source of internal differentiation. The present analysis indicates that different origins correspond to different fractions of the upper class. Whereas upper class members of higher origin cluster around business elite positions – particularly business professionals and finance – those of working class origins more typically cluster around what appear to be less central positions within the upper class, such as a manager of small firms and dealing in motor vehicles. This resonates with Bourdieu’s findings, and supports his emphasis on understanding social origin as a resource and an advantage – that is, as inherited capital – rather than merely an ‘ascribed status’.

Third, this study provides occasion to reflect on the changes ushered in by ‘financialisation’, a process held to be transforming and challenging established elites (e.g. Savage and Williams, 2008). However this may be, the present study shows that the old division between executives and managers, and various forms of owner capitalists, is still with us. These two fractions are clearly separated by their source of economic capital – salaries or capital income. This would seem to recall the debates on ‘managerialism’: Dahrendorf believed that the development of the joint stock ownership form, and the ensuing bureaucratization of the modern corporation, would ‘decompose capital’. The ‘managerialists’ conceived of this as a conflict of interest: managers would feather their own nests, while the owners attended to the long-term interests of the business. But it might be reasoned that financialization could be transforming the old manager-owner divide: as various funds take on a more important role in the ownership of contemporary business, the executives might indeed be facing owners with only a short-term interest in returns. If this is so, the managerialist divide might be persisting in transfigured form. While Bottomore may have been right that the ‘[t]he top executives and the owners of property are so intimately connected as to form a single social group’ (1993: 63), the same might not hold for the relation between top executives in business, various fund brokers, and the beneficiaries of these funds – this last would be characterized by capital incomes.

Moreover, even though this analysis by no means exhausts the question, we can gather something about the ‘class character’ of the financial industries. The present study indicates that those engaged in finance and insurance possess high volumes of inherited capital. Even though financialization surely affects the operations of capitalist economies and firms, perhaps spawning new lines of conflict, the fact that these groups seem to originate in the ‘old elites’ should warrant careful attention to their role in contemporary power relations. There seems to be good reason to expect tensions between various fractions of the business elite (found on the left side of Figure 3), as Savage and Williams argue, but we should be attentive to how such classical status markers as origin and education might lend them some cohesion.

Fourth, economic capital is not ‘one thing’, neatly measured by adding up whatever one has of monetary value. The simple notion that you can have much or little economic capital, glosses over the fact that where it comes from – property, stocks, salaries, wages etc. – is of paramount importance for understanding class relations. This surfaces in the present study in the contrast between owners (capital income) and executives/managers (wages/salaries). In contemporary capitalism, differentiating the sources of economic capital is even more important: significant groups of the population of western societies depend on welfare transferences, others on incomes from financial activities. A simple, ‘additive’ notion of economic capital leads away from a proper grasp of class relations in an age of financialized capitalism and mass unemployment. This is particularly so for the sociology of upper classes: it is central to distinguish between the forms of economic capital and power held by owner capitalists, executives and financial capitalists or rentiers, something which, admittedly, could only be partially accomplished with the data used here.

Lastly, I wish to make the theoretical-methodical point that despite the wide influence of Bourdieu’s work, his emphasis on quantification is still somewhat underappreciated (Lebaron, 2009). The work of Bourdieu offers not only stimulating concepts and theories, but also an approach to quantification consistent with an understanding of the social as relational. Coupled with the recent innovations in Geometric Data Analysis (Le Roux and Rouanet, 2010) the ‘social space approach’ constitutes a powerful and novel approach to class and stratification (e.g. Prieur et al., 2008), and to social differentiation more generally.

Footnotes

Appendix

Appendix Table 1.

Explaining point (categories with contributions above average) axes 1 to 3. Sorted in declining order by contribution. Units used are NOK (Norwegian Krone).

	Axis 1	Contr.	Axis 2	Contr.	Axis 3	Contr.
+	Parents’ edu level 1	7.73	Assets: 0 to 999,999	8.03	Edu level 6	8.26
	Parents’ edu field: General	6.06	Capital: 0 to 9999	6.73	Self-emp. income: 1,000,000 to 1,499,999	7.73
	Social origin: Unskilled workers	5.29	Wage: 1,000,000 to 1,499,999	6.70	Parents’ edu level 6	6.92
	Siblings high edu 0	4.76	Siblings with top income 0	2.71	Wage: 0 to 499,999	6.70
	Edu level 2	3.54	Social origin: Unskilled workers	1.88	Social origin: Professionals, academics	6.29
	Parents’ edu level 2	3.16	Capital: 10,000 to 999,999	1.67	Self-emp. income: 1,500,000 and higher	5.04
	Edu level 3	2.87	Self-employed inc: 1 to 999,999	1.50	Social origin: Farmers, etc.	2.70
	Parents’ income 0	2.60			Parents’ edu field: Primary sector	1.84
	Parents’ income 1	2.47			Assets: 1,000,000 to 3,999,999	1.68
	Capital: 1,000,000 to 1,999,999	2.30
	Social origin: farmers, etc.	2.18
	Edu level 1	2.16
	Wage 0 to 499,999	1.83
	Parents’ income 2	1.62
	Edu field: General	1.52
	Assets: 4,000,000 to 8,999,999	1.51
−	Parents’ edu level 6	6.81	Capital: 4,000,000 and higher	9.01	Self-emp. income 0	4.33
	Parents’ income 4	4.98	Assets: 15,000,000 and higher	8.42	Edu level 4	4.09
	Social origin: Professionals, academics	4.35	Social origin: Managers/executives	5.37	Wage 1,000,000 to 1,499,999	4.05
	Parents’ edu level 5	3.79	Parents’ income 4	4.45	Parents’ edu field: Economic/admin	3.55
	Edu level 6	2.78	Siblings with top income 2	3.83	Parents’ edu level 4	3.24
	Social origin: Engineers/administrators	2.66	Capital: 2,000,000 to 3,999,999	3.47	Edu level 5	3.04
	Wage: 1,000,000 to 1,499,999	2.11	Wage: 0 to 499,999	3.28	Edu field: General	2.34
	Siblings high edu 2	1.83	Assets: 4,000,000 to 8,999,999	3.14	Assets 0 to 999,999	2.18
	Assets 0 to 999,999	1.77	Assets: 9,000,000 to 14,999,999	2.94	Social origin: Managers/executives	1.98
	Edu level 5	1.75	Siblings with top income 3–5	1.96
	Parents’ edu field: Science, technical	1.51	Parents’ edu field: Economic/admin	1.81
	Social origin: Managers/executives	0.72	Social origin: Small firm managers	1.75
			Capital: 1,000,000 to 1,999,999	1.74

Acknowledgements

I am grateful to Marianne Nordli Hansen and Johs Hjellbrekke for their invaluable help and guidance. I have benefitted greatly from feedback from Vegard Jarness and Marte Mangset, as well as Sociology’s referees. An earlier version of this article was presented at the Elites in an Egalitarian Society seminar in Paris, October 2010 and the Thirty Years after Distinction conference in Paris, November 2010. I am grateful to the audience at these presentations, not least Mairi Maclean and Craig Calhoun, for their comments. Thanks also to Juliet Munden for generously proofreading the manuscript. The data used are kindly provided by Statistics Norway (SSB) for the project Educational Careers at the Department of Sociology and Human Geography, University of Oslo.

Notes

Magne Flemmen is Research Fellow at the Department of Sociology and Human Geography, University of Oslo, Norway. His primary research interests are in class analysis, social inequality and theories of class. His ongoing PhD project is on the application of the theoretical and methodical approaches of Pierre Bourdieu to empirical and theoretical issues in class analysis.

References

Bennett

Savage

Silva

Warde

Gayo-Cal

Wright

(2009) Culture, Class, Distinction. London: Routledge.

Bottomore

(1993) Elites and Society. London: Routledge.

Bourdieu

(1984) Distinction: A Social Critique of the Judgement of Taste. London: Routledge.

Bourdieu

(1986) The forms of capital. In: Richardson

(ed.) Handbook for Theory and Research for the Sociology of Education. New York: Greenwood.

Bourdieu

(1988) Homo Academicus. Cambridge: Polity Press.

Bourdieu

(1989) Social space and symbolic power. Sociological Theory 17: 14–25.

Bourdieu

(1996) The State Nobility. Cambridge: Polity Press.

Carroll

(2004) Corporate Power in a Globalizing World: A Study in Elite Social Organization. Ontario: Oxford University Press.

Dahrendorf

(1959) Class and Class Conflict in Industrial Society. London: Routledge & Kegan Paul.

10.

Domhoff

(2010) Who Rules America? New York: McGraw Hill.

11.

Giddens

Stanworth

(eds) (1974) Elites and Power in British Society. Cambridge: Cambridge University Press.

12.

Gulbrandsen

Engelstad

Klausen

Skjeie

Teigen

Østerud

(2002) Norske makteliter [Norwegian power elites]. Oslo: Gyldendal.

13.

Hansen

(1995) Class and Inequality in Norway. Oslo: Institutt for Samfunnsforskning.

14.

Hansen

(2009) Sosiale nettverk og karriere – tilfellet jurister [Social networks and careers – the case of jurists]. Søkelys på arbeidsmarkedet 26: 45–58.

15.

Hartmann

(2000) Class-specific habitus and the social reproduction of the business elite in Germany and France. Sociological Review 48: 262–82.

16.

Harvey

Maclean

(2008) Capital theory and the dynamics of elite business networks in Britain and France. In: Savage

Williams

(eds) Remembering Elites. Malden, MA: Blackwell, 103–20.

17.

Hjellbrekke

Korsnes

(2003) Det norske maktfeltet: interne strukturar, homogamitendensar, mobilitetsbaner og -barrierar i norske elitar [The Norwegian field of power: Homogamy tendences, mobility trajectories and barriers in Norwegian elites]. Oslo: Makt- og demokratiutgreiinga 1998–2003.

18.

Hjellbrekke

Le Roux

Korsnes

Lebaron

Rosenlund

Rouanet

(2007) The Norwegian field of power anno 2000. European Societies 9(2): 245–73.

19.

Le Roux

Rouanet

(2010) Multiple Correspondence Analysis. Thousand Oaks, CA: Sage.

20.

Lebaron

(2001) Economists and the economic order. The field of economists and the field of power in France. European Societies 3: 91–110.

21.

Lebaron

(2008) Central bankers in the contemporary global field of power: A ‘social space’ approach. In: Savage

Williams

(eds) Remembering Elites. Malden, MA: Blackwell, 121–44.

22.

Lebaron

(2009) How Bourdieu ‘quantified’ Bourdieu: The geometric modelling of data. In: Robson

Sanders

(eds) Quantifying Theory: Pierre Bourdieu. Berlin: Springer.

23.

Maclean

Harvey

Press

Cadbury

(2006) Business Elites and Corporate Governance in France and the UK. Basingstoke: Palgrave Macmillan.

24.

Mastekaasa

(2004) Social origins and recruitment to Norwegian business and public sector elites. European Sociological Review 20: 221–35.

25.

Murphy

(1988) Social Closure: The Theory of Monopolization and Exclusion. Oxford: Clarendon Press.

26.

Parkin

(1979) Marxism and Class Theory: A Bourgeois Critique. London: Tavistock.

27.

Prieur

Rosenlund

Skjøtt-Larsen

(2008) Cultural capital today: A case study from Denmark. Poetics 36(1): 45–71.

28.

Savage

Williams

(eds) (2008) Remembering Elites. Malden, MA: Blackwell.

29.

Savage

Warde

Devine

(2005) Capitals, assets and resources: Some critical issues. British Journal of Sociology 56(1): 31–47.

30.

Scott

(1997) Corporate Business and Capitalist Classes. Oxford: Oxford University Press.

31.

SSB (1998) Standard for yrkesklassifisering. Norges offisielle statistikk [Norwegian official statistics]. Oslo-Kongsvinger: SSB.

32.

SSB (2001) Norsk standard for utdanningsgruppering [Norwegian standards for education grouping]. Oslo: Statistisk sentralbyrå.

33.

SSB (2008) Standard for næringsgruppering [Standard for industrial classification]. Oslo-Kongsvinger: SSB.

34.

Useem

(1984) The Inner Circle: Large Corporations and the Rise of Business Political Activity in the US and UK. New York: Oxford University Press.

35.

Useem

Karabel

(1986) Pathways to top corporate management. American Sociological Review 51: 184–200.

36.

Weber

(2009[1946]) Class, status, party. In: Gerth

Mills

(eds) From Max Weber. Essays in Sociology. New York: Routledge.

37.

Zeitlin

(1989) The Large Corporation and Contemporary Classes. Cambridge: Polity Press.