Ralph Dahrendorf’s conflict theory of social differentiation and elite theory 1


Ralph Dahrendorf arguably ranks amongst the most influential social theorists of the former century. Given this significance, it is worth revisiting his conflict theory from the angle of what kind of social differentiation conception it adopts. Social differentiation is here meant as the most general term, of which social stratification and class theory are principal subtypes. This clarification is needed owing to the persistent conflation of social differentiation with social stratification in which usage it is the latter concept that is located at the top, social class, for instance, being its variety or, worse still, part. Meanwhile, there is a world of difference between a stratum and a class. First and formost, social stratification is by definition a hierarchy ranking individuals or categories according to some criteria, such as income, prestige, etc. Social, or should we say, socio-economic classes are interrelated in a more complex fashion, irreducible to any ladder-like structure.

Furthermore, strata are universal both in space and in time, that is to say, they can be located throughout society, while classes are social groups rooted in the economy, and, secondly, the latter, as opposed to stratification, are born in a certain historical period and are not supposed to persist eternally. With those notions in mind, we can commence our analysis thanks to which it will be possible, as already demonstrated above, to raise a number of important theoretical and substantive issues.

So let us assume as a tentative working hypothesis, to be verified, refuted, or modified in the course of our investigation, that Dahrendorf’s conflict theory(1959) represents a mixed system having some properties of both alternative approaches.Darendorf’s starting point is that neither structural functionalism nor Marxism alone provides an acceptable perspective on advanced society. He claims that structural functionalists neglect realities of social conflict and that Marx defined class too narrowly and in a historically-specific context. Furthermore, he believes that traditional Marxism ignores consensus and integration in modern social structures.

There are a number of serious problems with those statements. Firstly, the said division of social theories into integration and conflict ones, or, in another wording, statics and dynamics, equilibrium and change etc. fails to take account of the fundamental fact that each structural theory contains at least implicitly a theory of change and the reverse is also true – there can be no theory of change which would not refer at least tacitly to a definite conception of structure. If one defines a structure and identifies within it a set of core components, one by the same token points to the most likely source of a qualitative change of that structural whole. And, similarly, you cannot speak about change without assuming what is subject to that change. Dahrendorf’s characterisation is thus unfair, Parsons’ structural-functional theory of social system, even ignoring, at the moment, his writings on evolution of societies, includes a better or worse theory of social change. Likewise, Marx’s theory of conflict and development is inextricably interwoven with his detailed conception of the structure and functioning of a society undergoing given transformations. Michio Morishima argues that “Marx should […] be ranked as high as Walras in the history of mathematical economics.

It has rarely been pointed out that the general equilibrium theory was formulated independently and simultaneously by Walras and Marx” (1978).

Secondly, Dahrendorf commits an error which is, to be sure, frequent, but this fact does by noe means justify him – A bipolar conception of class is indeed over-simplified, but it is doubtful whether it could be attributed to Marx.

The source of this misconception is obviously The Communist Manifesto, and its famous statement: “Our epoch, the epoch of the bourgeoisie[…] has simplified class antagonisms. Society as a whole is more and more splitting up into two great hostile camps, into two great classes directly facing each other—bourgeoisie and proletariat”.

This claim, however, can be interpreted in different ways. Amongst others, it can be regarded as a prediction of the future development rather than as a statement of fact.Even in the same Marx and Engels’ work one can find sentences pointing to a more complex picture of societal differentiation. This view is also even more clearly present in Marx’s historical works. In „18th Brumaire of Louis Bonaparte” he writes, for example, about „aristocracy of finance, the industrial bourgeoisie, the middle class, the petty bourgeoisie”. Thus, according to this account the bourgeois class should be further subdivided. And in his unfinished chapter on class in the 3rd volume of “Capital” Marx mentions at least one large class, that of landowners, and suggests that the class structure of capitalist society is in fact more complicated .

The question of scientific fairness may be important, but what is crucial from the perspective of Dahrendorf’s own analytical framework is that he adops the very approach he criticises and declaratively rejects.

Be that as it may, Dahrendorf combines elements from both of these perspectives to develop his own theory concerning class conflict in postcapitalist society, as it is dubbed in his work.

Dahrendorf claims that capitalism has undergone major changes since Marx initially developed his theory on class conflict. This new system of capitalism, which he identifies as postcapitalism, is characterised by diverse class structure and a fluid system of power relations. Thus, it involves a much more complex system of inequality. Dahrendorf contends that postcapitalist society that has institutionalized class conflict into state and economic spheres. For example, class conflict has been habituated through unions, collective bargaining, the court system, and legislative debate. In effect, the severe class strife typical of Marx’s time is not longer relevant.

Dahrendorf believed, however, that Marx’s theory could be updated to reflect modern society. He rejects Marx’s two class system as too simplistic and overly focused on property ownership. Due to the rise of the joint stock company, ownership does not necessarily reflect control of economic production in modern society. Instead of describing the fundamental differences of class in terms of property, Dahrendorf argues that we must “replace the possession, or non-possession, of effective private property by the exercise of, or exclusion from, authority as the criterion of class formation: „By social class shall be understood such organised or unorganised collectivities of individuals as share manifest or latent interests arising from and related to the authority structure of inperatively coordinated associations. It follows from the definitions of latent and manifest interests that social classes are always conflict groups” (Dahrendorf 1959:238).

According to Dahrendorf, Marx’s notion of class is justifiable because in his time capitalism was dominated by owner-managed firms where ownership and authority were concentrated in the same hands. In contemporary economy, however, the most representative form of business organisation is a joint-stock company with dispersed share ownership. In this situation control over the means of production is wielded by professional managers, and not by legal owners.This shows, in Dahrendorf’s opinion,that the priority order of ownership and power should be reversed, it is no longer, as in Marx’s time, that ownership entails authority, but, contrariwise, property is subordinated to authority, is its special case.

First and formemost, the thesis that is the corporate managers who are in charge in companies is by no means unanimously held; many property rights theorists maintain that the fact that it is them that are entrusted with the day-to-day management does not matter, since at the end of the day the interest of the shareholders who are true owners, after all, does take precedence owing to a variety of mechanisms that ensure such an outcome.

Looking at the matter from a more theoretical angle, it may be noted that Dahrendorf is merely the most known proponent of this view which is shared, inter alia, by such neo-Weberians as Parkin (1979:46) and Giddens(1981:60) who treat exploitation as but a subspecies of the more general phenomenon of domination. As will be argued, this view of the relationship between ownership and control is false.

That the conception being discussed can make strange bedfellows is shown by the two following examples of Marxist or near-Marxist writers: Poulantzas (1978: 18-9) and Scott (1979: 32), who define real or „effective ” ownership as control.

Similarly, in one of his polemics with market socialism Hayek writes: “If the community is the owner of all material resources of production, somebody will have to exercise this right for it, at least in so far as the distribution and the control of the use of these resources is concerned”(1989: 135). But, a few sentences later, it turns out that this “central authority would simply have rights of ownership of all real resources”(Hayek, 1989: 135). It is one thing, of course, to attribute ownership of the means of production to the community, and quite another to vest it in the state, the centre or some other institution. Moreover, although Hayek is not completely clear on this point, he seems to equate ownership with decision-making or power over the means of production. In the present writer’s opinion, this notion is erroneous, as it refers to consequences or prerequisites of what Berle and Means call beneficial ownership (1969: 8), and not to this ownership as such. To argue that ownership consists in either “the power to assign the means of production to given uses or in the power to dispose of the products obtained” (Poulantzas, 1979: 18) is like staging Hamlet without Hamlet. Making decisions determining the use of the object is not the same thing as actually benefiting from that use. It is not the same thing also in the sense that those who exercise control over given assets need not be those who enjoy the fruits of these assets. The above argument applies even more explicitly to Mises, who expressly states that “ownership is the right of disposal” (1936:517).

That this treatment is mistaken, can be most easily shown on a number of specific exaamples. The circumstance that an executive of a public library makes a decision where concretely , say, Russian literature should be stored, and which room should accumulate English fiction and poetry and so on and so forth does not, to be sure, transform her into an owner of these resources and the building itself. Similarly, while city authorities may take a decision that a definite street must be closed to traffic, it does not render them private owners of the street involved.

Besides, even if one accepts that property equates power or control, it does not alter the fact that this property yields income which calls for an adequate term. In other words, Dahrendorf’s solution would amount in this case just to a semantic issue. In addition, his treatment of hired managers as allegedly having anything common with ownership of capital is empirically and theoretically misplaced. It does not take consideration of a variety of equity-based forms of executive compensation, including stock options, and even more critically, of the size of their pay which, for this very reason, cannot be viewed as simply pay for work, salary reflecting their productive contribution or whatever. A simple comparison of executive earnings with shareholders’ incomes in the form of dividens and/or capital gains on the one hand and , say, army commanders whose job is, if anything, more stressful and responsible than one of a corporate manager will show that the former comprise a large surplus over what can be counted as their earned compensation. Where do such surpluses come from? The only answer possible is that from the same source as owners’ profits, that is, from the surplus value.

On the basis of his assumptions, Dahrendorf argues that society can be split up into the „command class” and the „obey class” and class conflict should refer to situations of struggle between those with authority and those without. However, there are several serious problems with that notion. Firstly, Dahrendorf claims that all conflicts only involve two contending parties. This view of conflict appears too simplistic to apply to advanced society, the very same inaccuracy Dahrendorf accuses Marx of.

Furthermore, on the basis of Dahrendorf’s theoretical premises, one can in fact distinguish innumerable classes. For for Dahrendorf classes are present in each so-called imperatively-co-ordinated group, be it an non-professional theatrical troupe, football club or a business corporation. In each such case one can discern two opposed groups : superiors and the subordinated. Needless to say, this leads to rather odd conclusions. From his definition of social class , if we see all authority relations as class relations, it follows that a conflict between parent and child, for instance, is a class conflict. Besides, he fails to establish the difference between authority resulting from truly legitimate power and authority stemming from a situation where a subordinate is regularly obedient to a superior for other reasons. This lack of distinction even more multiplies based on those authority relations social classes, thus allowing for an infinite number of classes, which in turn removes significant meaning from the concept of class. There is more to this than that, though. Now if all the members of the former groups wer to be included in the broad societal ruling class, it does not end the matter. For an awkward question arises: should all without exception members of the so-called command classes in each of a variety of imperatively co-ordinated associations be included in the overall societal command class? This raises doubts as almost each such individual may belong to another group, this time in the character of a subordinate. This must mean that Dahrendorf’s supposedly cohesive societal superior class extinguishes. But let us suppose that, for the sake of argument, all such persons of a binary and inconsistent – from the standpoint of the conflict theory of class – membership are excluded from the picture. This does not salvage Dahrendorf’s theory either. For one must ask what social non-trivial characteristics and interests have in common, say, company presidents, prime ministers, heads of tasks groups, university chairpersons, orchestra conductors, heads of a multitude of non-governmental organisations, mafia godfathers and so forth. An answer is again straightforward: nothing. Thus, we must conclude that an attempt to construct an over-arching class of authority holders on Dahrendorf’s analytical grounds has failed.It is not only because the German theorist views his units of social differentiations in hierarchical terms that we do hesitate to refer to them as classes. For Dahrendorf adopts another assumption incompatible with ones underlying class theories. He claims, namely, in full accord, to be sure, with his own framework of reference that his conflict classes are historically universal. And finally, his conflict classes are not exclusively economic groupings, as in classic Marxian or Weberian theory. And the circumstance that they can be found throughout society likens his approach to that of stratification. And this is true despite his(orgiven the situation, his supporters and followers’) likely protestations to the contrary, as in Dahrendorf’s opinion the concept of stratum is static, whereas the concept of class is dynamic. This is a typical half-truth. It is the case, to be sure, that social strata, being analytical categories and not real entities cannot create social groups in the same sense that social classes(and social estates, for that matter, as we term social groups grounded in non-economic property relations) due to their common socio-economic position in the structure of property relations and common interests stemming from it can. But this statement hardly refers to his own superficial units misleadingly termed social classes.

The above critique does not mean downplaying on our part social relevance of power or authority relations. On the contrary, in our own account of social differentiation, as representing non-economic property relations and thereby a foundation of a social estate. A more detailed presentation of the theory concerned, however, goes beyond the confines of the present paper.

To return to the question of the theoretical character of Dahrendorf’s conception, it is arguable that it bears close resemblance to elite theory. It is within the framework of that theory that one finds a dichotomous division opopulation into the ruling elite, sometimes termed also ruling class(e.g. by Mosca), and the rest or masses subject to that overarching authority. Whilst the concept of elites is also used in other meanings and contexts, it can be argued, following, that “the word 'elite’ should be used only in relation to those groups that have a degree of power. Some but not all of the groups indiscriminately described as elites are holders of power; and my argument is that the concept should be limited to such groups” (2008).

From a political point of view, one should be grateful to the elite theorists who ruthlessly revealed the shortcomings of democratic regimes obtaining in the advanced capitalist countries. Their perspective is surely superior to the pluralistic ideology. But elite theory is not the only alternative to the latter, as one of the most influential elite theorists concedes: “the Marxian doctrine of class struggle surely was then, and certainly is now, closer to reality than any assumed harmony of interests”(Mills 1956), and his brother in arms expresses his view on the salience of class even in the very title of his book(Domhoff 1996). This may suggest that the debunking zeal of elite writers can lead them astray. It is one thing to state what is a trivial truism that in various organisations, such as the state, corporations and so on there is always one group in charge and another subject to their authority, and quite another to posit that all those groups somehow forma coherent whole. As this supposed ruling elite is by definition invisible(Savage and Williams write about the “glaring invisibility of elites”(2008)), all crucial decisions are taken behind the scenes, the proof of its real existence is difficult to provide. This empirical weakness is related to the theoretical premises of elite theory in another respect as well.

The notion of elite revolves around power, yet this crucial concept is ill-defined or not defined at all, which makes it possible to include in the ruling elite wielders of very different sorts of power and also those who wield no power at all. It is for that reason, too, that the power elite, the ruling elite etc. is so differently conceived by particular theorists, and in the end may become quite elusive. Compare this with class theory where, despite all theoretical differences, there is consensus as to at least certain key concepts such as the capitalist class.

Class is indeed one of the problems elite theory, as some elements of the power elite are clearly socio-economic classes, e.g. corporate managers. Other groups included in that concept are in turn social estates. The following excerpt from Mills fits in our theory of estates very well: “They are inside an apparatus of prerogative and graded privilege in which they have been economically secure and unworried. […] such striving for status as they have known has been within an unambiguous and well-organized hierarchy of status, in which each knows his place and remains within it” (1956).

A few words should be devoted to another theorist whose approach is similar to that of Dahrendorf in two crucial respects: both in the choice of the specific foundation of class and in a heteregeneous quality of a classificatory scheme built on that basis, i.e. related to stratification theory. “Lenski adopts a similar position to Dahrendorf, although he tends to pursue a more eclectic usage of „class,” including a variety of other dimensions besides authority. Lenski (1966) first defines class broadly as „an aggregation of persons in a society who stand in a similar position with respect to some form of power, privilege, or prestige” (1966:75). He then goes on to say that „if our goal is to answer the question 'who gets what and why?’. power classes must be our chief concern,” where power class is defined as „an aggregation of persons in a society who stand in a similar position” (Wright 1979).

Wright rightly notes that “authority definitions of class tend to see authority itself as a unidimensional relation of domination/subordination within a given organisation. No systematic theoretical distinctions are made concerning the object of authority. What matters is having authority or power; little is said about how it is used. Conceptions of class in terms of authority relations thus tend to emphasize the form of class relations over the content of those relations.

Finally, because of this formal character of the conception of class, authority definitions generally do not provide a sustained account of why social conflict should be structured around authority relations. Implicitly, one of two arguments is usually made. Either it is assumed that human beings somehow have an intrinsic drive for power for its own sake, and thus the division between the powerful and the powerless intrinsically constitutes the basis for social cleavage; or it is argued that power and authority enable the powerful to appropriate various kinds of resources, and that as a result the powerless will attempt to gain power for instrumental reasons. The evidence for the first of these assumptions is particularly weak. People may have an intrinsic drive to control their own lives, but there is little evidence that most people have a basic need or drive to control other people’s lives. In any event, empirically most struggles over power are struggles over the use of power, not simply the fact of power. The second assumption is thus more plausible. But in order for it to provide a sound basis for an explanation of the relationship of authority to social conflict, it is necessary to develop a systematic theory of the relationship between authority and the appropriation of resources. Most discussions of authority lack such an account” (1979).

Given the above criticism, it might appear somewhat surprising that the American sociologist introduces the concept critisised by him to his own definition of class. The explanation of that puzzle lies inWright’s hope that his approach avoids the drawbacks identified above. He states: “Classes, in these terms, are most pivotally defined by the relations of appropriation of the surplus product and secondarily defined by the relations of control over the technical division of labor and relations of authority” (1979:18).

However, the above definition is in effect akin to Dahrendorf’s conflict approach. Adding authority to exploitation as an additional dimension of class relations does not go beyond the bipolar view of the class structure as composed of antagonistic classes which, for reasons discussed also below, is untenable; what is more, this new conceptual addition even strengthents this view as relations of authority are inherently hierarchic. In addition, originality of Wright’s position is problematic not only owing to its affinity to that adopted by Dahrendorf, but also because Wright (1979, 1985) obtains class categories by cross-classifying property with authority in the manner earlier proposed by Ossowski (1963).

Lest our contentions be understood in purely formal or semantic terms, we must make an important reservation. One may or may not concur with our view of socio-economic classes as grounded in economic property relations, arguing instead, like Dahrendorf, for instance, that the foundation of class realtionships lies in authority rather than ownership. Still, one need somewhat to frame those social groups based on ownership, even if it is not them that are termed social classes. Thus, supposing, for the sake of argument, that the theoretical foundation of class is to be found in economic, as distinct from legal, property, it is blatantly clear that no society, and in particular modern capitalist society cannot be reduced to just two classes. First and formost,there are quite numerous owners and at the same time operators of their own means of economic activity, i.e. production, commerce, transport, finance and services, which in our own theory of class are termed, boorrowing this specific term from Weber, the autocephalous class in order to avoid non-substantive ideological and political associations linked to the concept of the petty bourgeoisie, traditionally used in the Marxist literature. While the latter term is in our opinion simply inconvenient, and not unsound in principle, this criticism fully applies to a widely used term of the middle class/es. First and formost, it is not any class in the proper sense of the word at all. Rather, it refers to a, better or worse, identified social stratum or a cluster of such strata, whether the criterion of its alleged existence be income, life style, or whatever. Some writers use instead a pair of concepts, distinguishing between the old and new middle class. In such a situation the former refers to what traditional Marxists term the petty bougeoisie, whilst the latter suffers from all the limitations identified above. From our perspective the grouping concerned may include certain social classes, but also social estates, i.e. non-economic groupings, and its usefulness in sociological research is extremely problematic, to say the least, as it blurs rather than highlights key social divisions.

The golden age of elite research associated with such names as C. W. Mills or domhoff did not last forever.

Until the early 1970s, the study of social inequality placed great emphasis on the significance of elites. In the United States, the most eminent critical sociologist of his day, C Wright Mills, forged his reputation on his inspiring account of The Power Elite (1950), and these ideas were taken up by other leading scholars, notably Maurice Zeitlin (1989) and Richard Useem (1984). In the United Kingdom, the research programme on social mobility and stratification loosely associated with David Glass at the LSE placed great emphasis on understanding the role of elite professions, leading to studies of senior civil servants and related groups. Studies of political elites were undertaken most notably by Guttsman (1963). “The classic tradition of elite studies had a common methodological belief in the value of the focused case study, using a mixture of documentary sources, in-depth interviews, and ethnographic analysis. It was this pluralism which was to be dispelled by the rise of survey analysis in the 1960s, and which by the 1970s had become the hegemonic means of studying social inequality.

In the United States, Blau and

The sample survey abstracts individuals from their context, and allows them to be arrayed through classification, none of which however, are amenable to researching 'small’ groups – such as elites. Those within the class-structural tradition differentiated classes which had sufficient numbers in them to allow statistically significant findings to be drawn using methods such as log linear modelling. Goldthorpe’s most 'elite’ class, the 'higher service class’ thus contains as many as 14.3 per cent of the male population in the UK, even in 1972; anything smaller, as he readily acknowledges, is not amenable to survey analysis (see the debate between Penn 1981 and Goldthorpe, 1981).

Within the status-attainment tradition, it was possible to give a high-status score to smaller occupational groups; but this had the effect of defining elites not as distinctive social entities but as the 'apex’ of a status hierarchy. Interest in elites as specific social formations was subordinated to a concern with unravelling the determinants of 'who gets ahead’. Furthermore, this approach tended to define elite groups in terms of their social exclusiveness, rather than their wealth or political power.[…] One of the unanticipated consequences of the rise of the sample survey, as the most powerful and legitimate social science research method of the late 20th century, is that elites become opaque from within its purview. It necessarily focuses on large social aggregates and/or decontextualized individuals.” (Savage, Williams 2008).

Thus, in this view the demise of traditional elite studies is largely accounted for the rise of orthodox, positivist or neo-positivist social science. “A central feature of this shift was the insistence by quantitative social scientists that the sample survey was the central research tool for analysing social inequality. Given their small size and invisibility within national sample surveys, elites thereby slipped from view. The other side of this [movement] comes from structuralist and post-structuralist social theory. The anti-humanism which was central to the structuralist movement of the 1960s led to a rejection of the focus on visible, human, elites, signalled most famously in Poulantzas’s (1973) critique of Miliband’s (1968) account of the capitalist state. Foucault famously built on, and reworked, this reasoning through his critique of 'sovereign power’ and his insistence that capillary power was central to contemporary liberal and neo-liberal governmentality. Actor network theory and currents within science and technology studies (STS) have further insisted on the distributed, local, and mobile character of socio-technical relations, thereby rejecting any obvious appeal to an 'elite’ acting as a 'deus ex machina’ which orchestrates society. Acting together, these two different [processes] have theoretically and methodologically 'whipped the carpet’ away from elite studies which became deeply unfashionable right across the social sciences from the mid 1970s onwards. (Savage, Williams 2008).

This does not mean that the interest in elites has completely WANED. John Scott has undertaken an ambitious analytical study whose utility for our purposes lies, amongst others, in its attempt to systematise the issues involved in the theory of elites that have been mentioned above.

At the very outset Scott addresses the issue of the over-inclusiveness of the central concept of elite theory: “The weakness of elite research over the last three decades can, perhaps, be seen as a reaction to the overstated claims that had been made for the idea of the elite and as an implicit acceptance of the many critical attacks levelled at it. The claim that elite researchers tended to overstate the power and cohesion of elites was unintentionally reinforced by the tendency of sociologists to use the word indiscriminately. At the height of its popularity almost any powerful, advantaged, qualified, privileged, or superior group or category might be described as an elite. The term became one of the most general – and, therefore, one of the most meaningless – terms used in descriptive studies. It was applied to such diverse groups as politicians, bishops, intelligent people, aristocrats, lawyers, and successful criminals. Not surprisingly, elite research attracted sustained criticism. The most vociferous critics were those pluralists who challenged what they called 'positional’ studies in the name of a more dynamic approach to power and decision-making (Dahl, 1958, 1961, 1966; Polsby, 1962). It was such challenges that led many of those involved in power structure research to eschew the word 'elite’ in their substantive studies. The idea simply seemed to carry too much unwanted intellectual baggage to warrant its continued use” (2008).

Together with Savage and Williams’ account emphasising the methodological issues, we have thus a more or less complete picture of the recent fortunes and misfortunes of elite theory. Scott, however, does not confine himself to the above acknowledgment of the present unsatisfactory state of elite theory, but sets out to repair it.

“It is now time to reassess that conclusion. A number of powerful and sophisticated studies (Carroll, 2004; MacLean et al., 2006) have once again made it central to their concerns. The term can play an important part in sociological research; and its meaning must be narrowed down and refined so that it can be retained as a powerful analytical concept and not inflated beyond its legitimate use. Elites must be distinguished from all those other social groups with which they are often confused; and their relations with other groups with which they may often be associated in real-world situations must be clarified. […] My concern will be to try to distinguish precisely what forms of power give rise to the formation of elites and, therefore, to set some limits on the ways in which the term should be used” (Scott 2008).

Scott explains that “this emphasis on power means that people with high IQ, for example, do not constitute an elite in any sociologically meaningful sense. They may be very significant in many walks of life and it may be very important to study them, but they are not a category to which the word 'elite’ should be applied.

Similarly, highly-paid occupational groups should not be described as elites simply because of their high pay, however privileged or advantaged they may be. Such groups become elites only if their intelligence or high pay becomes a basis for significant power. To label superior or advantaged groups indiscriminately as 'elites’ is to make it more difficult to study them, as it implies spurious similarities among them and with other groups. Such an approach masks their specific features and destroys all distinctiveness that the elite concept can have. Clarifying the concept of an elite, therefore, can help us to study both elites and those other groups with which they tend to be confused” (2008).

One cannot agree more, but there remains one important problem: how is the foundation of elites, i.e. power to be defined? Scott is well aware of the centrality of this issue:

„Elites are most usefully seen, then, in relation to the holding and exercising of power. This implicitly raises a further problem, as power itself is a much-contested idea and has been defined in numerous different ways. […] Power, in its most general sense, can be seen as the production of causal effects, and social power is an agent’s intentional use of causal powers to affect the conduct of other agents. At its simplest, then, social power is a bipartite relation between two agents, one of whom is the 'principal’ or paramount agent, and the other the 'subaltern’ or subordinate agent. The principal has or exercises power, while the subaltern is affected by the power of a principal” (2008).

This view of power is essentially sound, as it points out that there must be two sides to the power equation. Scott, however, expands on this basic idea:

“In the mainstream of power research, investigators have largely been concerned with the actual exercise of power by a principal over a subaltern: power consists in actually making someone do something. In a second stream of research, on the other hand, attention has been given to a principal’s capacity or potential to do something. From this latter point of view, the central significance of power is to be found in the ability that certain actors have to facilitate things.

The paradigm for power relations in the mainstream view is the exercise of decision-making powers in a state through the use of electoral and administrative mechanisms. This view of power is extended to other kinds of sovereign organization, such as business enterprises, universities, and churches. According to this point of view, principals are those who make others do what they would not otherwise do. Conversely, agents may resist the attempts of others to place them in subaltern positions by making them act against their own wishes and preferences. In sovereign organizations, power relations are asymmetrical and are organized around the conflicting interests and goals of the participants. [Within…] the second stream of power research, […] researchers have focussed on the cultural construction of institutional structures far more than they have on the relational structures. […] Power is not concentrated in sovereign organizations but is diffused throughout a society, and so must be seen as a collective property of systems of co-operating actors. Instead of the repressive aspects of power, which tend to figure in the mainstream, the second stream stresses its facilitative or 'productive’ aspects. […] It is the combination of mainstream and second stream approaches to power that provides the basis for developing a nuanced understanding of the various forms that power can take. Each stream has highlighted different, but complementary, sets of mechanisms, and it is important to develop an understanding of these mechanisms, working from the elementary forms to the more complex patterns of domination that are found in states, economic structures, and associations”(2008).

So far, so good; one cannot sensibly object to an attempt to work out a more nuanced concept of power and on that basis – elites.

Scott goes on to say that “there are two elementary forms of social power. Corrective influence, analysed principally within the mainstream, depends on the rational calculations made by agents and operates through punishments and rewards. Persuasive influence, a principal theme of the second stream, depends on the offering of arguments, appeals, and reasons that lead subalterns to believe that it is appropriate to act in one way rather than another. The two main forms of corrective influence are force and manipulation. Where force involves the use of negative physical sanctions to prevent the actions of subalterns, manipulation involves both positive and negative sanctions (eg, money, credit, and access to employment) as ways of influencing subaltern decisions. The two main forms of persuasive influence are signfification and legitimation, operating respectively through shared cognitive meanings and shared value commitments. These discursive meanings make a particular course of action seem necessary or emotionally appropriate.

These elementary forms of power are the building blocks from which more fully developed power relations are built. (2008)

The above statement does not respect Scott’s own premise that one should make one’s concepts as clear as possible. His notion of power, as laid out above, does not meet that requirement, since it blurs the distinction between influence and power proper. Power can be understood as an ability to initiate, modify or stop other people’s action by means of coercion. This coercion may be in fact applied or tacit, but nevertheless its presence is pivotal to the implementation of a given power relation. Influence, on the other hand, refers to the same effects as above, but effected within a voluntary social relation. However controversial his theory of communication media may be in other respects, Parsons, to whom Scott is referring anyway, does not confuse the two. Meanwhile, Scott in another passage speaks of power as manifested at the level of interpersonal relations and resulting from personal traits of the individuals involved only. In terms of our understanding of power, such an account does not make sense; it is only influence that can be at issue. Indeed, that approach seems more consistent with Scott’s general methodological programme rejecting any obliteration of differences between power and non-power relations.



The analysis of Dahrendorf’s conflict theory has been useful in that it has permitted us not only to determine the theoretical camp or stream to which his conception of social differentiation actually belongs, but to consider and, hopefully, cast some light on a number of saliennt theoretical questions pertaining to class theory in particular and social differentiation in general.

Let us not prejudge the matter, though, and look at the continuation of Scott’s argument:

These developed power relations occur in a number of modalities as structures of domination, forms of counteraction to domination, and the more amorphous patterns of interpersonal power that are rooted in face-to-face relations. […]Domination is power that is structured into stable and enduring relations of control, and four forms of domination [can be identified]. Coercion and inducement are structures of constraint through corrective influence. Expertise and command, on the other hand, are discursively based structures of authority built through persuasive influence. Constraint is the form of domination that Weber (1914) referred to as 'domination by virtue of a constellation of interests’ and that Giddens (1979:100–101) called 'allocative domination’. Constraint exists where principals are able to influence subalterns by determining the action alternatives open to them, either by direct force and repression or by offering inducements that influence a subaltern’s calculations. Within the overall distribution of resources, the resources controlled by the principal determine the constellation of interests faced by principal and subaltern and within which both must act. Domination through discursive formation is what Giddens called 'authoritative domination’. Authority exists where principals influence subalterns through persuasion rooted in the institutionalized commitments, loyalties, and trust that organize command and expertise.

In relations of coercion, action alternatives are restricted through direct force or repression. Subalterns are coerced by power that exists independently of their preferences or wishes. They must take account of it in their subjective assessment of their situation, but it does not depend upon their giving it any discursive justification. Inducement, also, operates through the subjectivity of participants and without discursive justification. In this case, however, it relies on the preferences and desires of the subalterns by influencing the calculations that they make about how to act in particular situations.

The leaders of an invading army of conquest, for example, may coerce a population into compliance through threatening or actually using violence against them. Bank managers, on the other hand, may induce clients to invest by altering rates of interest and other conditions attached to loans. (2008)

The above passage has not corrected what in our view is an error, and what, by the way, appears to be, at least implicitly, recognised by Scott in that he speaks of “influencing”, still in the context of power, however. Scott adds that Agents who have internalized prevailing cultural values will tend to identify with those who occupy positions of domination defined in terms of these values.

This internalization and identification defines the powers of command and expertise available to principals in relation to subalterns. Relations of command are those where internalized values structure both the rights of principals to give orders and the corresponding obligations for subalterns to obey. Subalterns willingly comply because they are committed to a belief in the legitimacy of a specific command and of those who issue commands. Legitimacy exists when there is a belief that a pattern of domination is right, correct, justified, or valid (Held, 1989:102; Beetham, 1991:10–12). Relations of expertise are those where knowledge that is monopolized by one group is accepted by others as a legitimate basis on which they can offer authoritative expert advice. Subalterns are not obliged to treat this advice as an instruction, but there is a presumption that the experts can be trusted to offer valid and reliable guidance that ought to be followed. An executive manager in a bureaucratic hierarchy may hold a position of command over junior employees, while a lawyer or accountant may exercise expertise through the technical advice that he or she can offer to clients. (2008)

Again, a terminological problem arises; initially, contrary to the established tradition, Scott did not used the criterion of legitimacy in his definition of domination, but later he refers to this very criterion. It remains to justify, therefore, whether the stability of a domination relation is , and if yes, to what extent, tied to its legitimation. On the basis on the above considerations which, as we have pointed out, are not without their problems, Scott addresses his central issue:

This delineation and clarification of the forms of power and domination has had a purpose: to enable me to set out a defensible and useable concept of 'elite’ as a specific kind of group involved in holding and exercising power. Specifically, elites are to be defined in relation to the structures of domination that constitute them. Elites are those groups that hold or exercise domination within a society or within a particular area of social life. Corresponding to the four forms of domination are four types of elite . […] Coercive elites and inducing elites are based in allocative control over resources. Coercers and inducers derive their power from the constraints that flow from the distribution of the resources involved in force and manipulation. They are the elites that Pareto (1916) referred to as the 'lions’ and the 'foxes’, using the language of fables. Expert and commanding elites are based in relations of authority. Experts and commanders derive their power from the discursive formation of signifying and legitimating principals and subalterns. Emulating Pareto’s language, it can be suggested that the experts be referred to as 'owls’ and the commanders as 'bears’. Coercive and inducing elites can be identified in purely formal terms by the resources under their control. Those who control access to the use of the means of violence have the ability to coerce others into conformity and to act against their wishes, desires, and interests. Those with financial and industrial assets organized as economic capital are able to induce others to conform by influencing their rational, self-interested calculations of personal or group advantage.

Expert and commanding elites can be identified by the particular symbols and social meanings that they monopolize. Expert elites are those whose specialized bodies of technical knowledge are organized into 'professional’ structures and practices. Lawyers, accountants, doctors, and investment advisers, for example, may all be involved in persuasive power on the basis of a claimed and accepted expertise. Commanding elites are those who legitimately occupy the top administrative positions in institutional hierarchies of management and control. In contemporary societies this characteristically takes the form of what Weber described as bureaucracy. Such 'top’ bureaucratic positions are institutionally defined as those that carry strategic significance for a particular organization or form of association.

These four ideal types of elite overlap with each other in concrete situations and may only rarely be distinguishable in their particular forms. The analytical distinctions are, however, important to make as it is only through analysis that the complex interdependence of factors may be investigated. Thus, a commanding elite may also possess coercive powers that provide an ultimate, last-resort back-up for their authority. They may also be able to gain personal control over those resources that give them enhanced life chances and the ability to manipulate the behaviour of others. Holders of commanding positions in business enterprises – the executives and top managers – are especially likely to have further powers of financial inducement available to them. Under some conditions, however, it is possible for the mechanisms that structure the powers of particular elites to achieve a degree of differentiation from each other. In these circumstances, members of some or all of the types of elite may be able to act as autonomous and specialized agents. A commanding elite in the state, for example, may face a challenge from the coercive power held by the military forces of another state or from the financial inducements that can be offered by criminal syndicates. The concrete configuration of power in any particular situation is always a matter for empirical investigation; but such investigations must rest on a clear delineation of the various types of domination and their bases. (2008)

Scott’s statements are consistent with his earlier claims, which otherwise should count as their merit.This does mean, however, that they do not pose problems. In our view, for instance, commands should be viewed not as a distinct mode of power, or domination, for that matter, but rather as a means of it. A soldier obeys a command of his commander which is for the former a form of subordination to power, and for the latter – of exercising power. Lastly, Scott addresses another important issue considered above:

In its most general sense, then, the term 'elite’ is most meaningfully and usefully applied to those who occupy the most powerful positions in structures of domination. Elites can be identified in any society by identifying these structural positions. As occupants of a purely formal category, the members of an elite need have few bonds of interaction or association and may not exist as a cohesive and solidaristic social group. Such solidarity occurs only if social mobility, leisure time socializing, education, intermarriage, and other social relations are such that the members of an elite are tied together in regular and recurrent patterns of association. Only then are they likely to show any unity or to develop common forms of outlook and social consciousness.

A key area of elite research is to examine whether these links of background and recruitment exist and to chart their consequences for elite consciousness and commonality of action.

This elite structuring is especially likely to occur where recruitment to elite positions reflects larger processes of class and status formation. Elites are analytically distinguishable from social classes and status groups, no matter how entwined they may be in real situations. One of the recurrent problems in elite research, however, has been the tendency to confuse these concepts and to use them interchangeably. “Economic elite” and “capitalist class”, for example, may be used interchangeably to describe various privileged, advantaged, or powerful economic groups. This tendency must be resisted if the analytical power of the elite concept is to be retained, as this is the only basis on which the dynamics of power can be clearly understood.

The resources involved in holding and exercising power are also relevant to the formation of class and status situations. It was for this reason that Weber held that class and status were to be seen as aspects of the distribution of power. When access to material resources is structured through property and market relations, the resource distribution forms the 'class situations’ that determine the life chances of their occupants and become the bases of social class formation (Scott, 1996). Similarly, the symbolic resources of social prestige that comprise cultural capital may be formed into 'status situations’ that determine styles of life and become the bases for the formation of social estates and status groups. Class structures are differentiated by divisions of property and employment that are the bases of the inequalities of wealth and life chances measured in class schema. Status situations are constituted by cultural definitions of factors such as gender, religion, and ethnicity that become the bases for judgements of social superiority and inferiority. (2008).

The preceding corroborates, unfortunately, our earlier charge regarding the over-stretched concept of power. Scott fails to substantiate his claim according to which in Weber both classes and estates or (status groups, as he mistakenly terms them) are structured by power relations.

The term “Soziale Stünde”, which figures prominently in Weber’s theory of social differentiation is, as a rule, translated into English erroneuously, as ‘status groups’, or even ‘status classes’. Meanwhile, the term has altogether different connotation, and its association with feudalism or the old epoch of the Middle Ages is, in the eyes both great otherwise rivals any drawback, but, rather, the merit of the concept, as it allows to pinpoint such contemporary social groups that bear resemblance to their conceptual forebears in important respects. Another faux pas of Scott is that he associates this Weberian concept with the concept of cultural capital which Weber would certainly abhorred.

More to the point, Weberian “powers of disposal” have little in common with power as based on coercion. And power is at best one of the dimensions of estate positions. Our own conception of social classes and estates does not commit such errors; it takes account of power as one of the extra-economic relations determining estate positions, and, on the other hand, avoids the conflation of power or control with economic property which forms the basis of social classes.

Within the framework of our own theory of social differentiation the concept of an elite is essentially an alien body which is redundant, since its functions can be fulfilled by other concepts. The notion of an elite would make sense only if could be established that within a society there exists a single coherent group in the sense of classic elite theories. This, however, is precisely where the rub lies. Even Scott in his otherwise fine paper has failed to provide any proof of the existence of such a group.As one should never say “never”, this does not mean, of course, that some empirical research will not achieve that goal, although, honestly, given a range of theoretical and empirical difficulties, this remains a distant prospect. This much is, in effect, conceded by the author being discussed himself:

The various elites of a society may overlap and combine to form a single, overarching elite. Those who occupy the leading positions of command and expertise within a state, an established church, and in capitalist enterprises, for example, may be forged into a single concrete elite, though it may still be fruitful to distinguish the varying mechanisms of domination in each of these areas.

Mosca introduced the term classe dirigente or 'ruling class’ to describe this kind of ruling minority. In view of the confusions surrounding the language of 'class’, however, it is preferable to follow Pareto’s (1916) terminological innovation, and use the word 'elite’. […] It must not be assumed, however, that elite unity is the norm. Even specialized political and economic elites may be internally divided along ideological, religious, ethnic, or other lines, and these divisions may preclude them from achieving any overall solidarity or from forming part of a larger ruling elite. Factions may divide a political elite to such an extent that it is better to consider it as comprising two or more rival sub-elites whose conflict and tension may be an important source of change. Michael Moran […] demonstrates the fragmentation within the business elite in Britain that developed during the 1980s and 1990s and some of the difficulties this poses for class-wide representation of business interests.[…] Elites can exist at various levels of a society and so are distinguishable by their degree of power”(Scott 2008).


As suggested in the introduction to the paper, it has been possible to pursue a range of interesting ends. Firstly, upon closer scrutiny, Dahrendorf’s approach to social differentiation, albeit bearing some resemblance to stratification and class theories, is most closely related to elite theory. The disjunction between apparent not only in Dahrendorf’s case, between the kind of terminological apparatus used and the real character of the analytical framework concerned has allowed us to clear up some misunderstanding one comes across in the field of social differentiation. Furthermore,an in-depth analysis of the conceptual underpinnings of elite theories also helped clarify some relationships between the key concept concerned and some related terms such as power, social class or estate. So it would seem that the output is rather satisfactory.

Jacek Tittenbrun


Arendt, H., (1969), On Violence, New York: Harcourt Brace and World.

Bachrach, P. and Baratz, M. S., (1962), 'The two faces of power’ in Scott, John (ed.), Power, Volume 2, London: Routledge.

Beetham, D., (1991), The Legitimation of Power, Houndmills: Macmillan.

Castells, M., (1996), The Rise of the Network Society, Volume 1 of the Information Age: Economy, Society and Culture, Oxford: Blackwell Publishers.

Dahl, R. A., (1958), 'A critique of the ruling elite model’ in Scott, John (ed.), (1990), The Sociology of Elites, Volume 1, Aldershot: Edward Elgar Publishing.

Dahl, R. A., (1961), Who Governs? New Haven: Yale University Press.

Dahl, R. A., (1966), 'Further reflections on „the elitist theory of democracy” ’, American Political Science Review 60: 296–305.

Dahl, R. A., (1971), Polyarchy: Participation and Opposition, New Haven: Yale University Press.

Dahrendorf, Ralf. (1959) Class and Class Conflict in Industrial Society. Stanford: Stanford University Press.

Domhoff, G. W., (1967), Who Rules America? Englewood Cliffs: Prentice Hall.

Domhoff, G. William. 1996. State Autonomy or Class Dominance? Case Studies on Policy Making in America. New York: Aldine de Gruyter.

Dooley, P. C., (1969), 'The interlocking directorate’ in Scott, John (ed.), (1990), The Sociology of Elites, Volume 3, Aldershot: Edward Elgar Publishing.

Fennema, M., (1982), International Networks of Banks and Industry, Hague: Martinus Nijhof.

Foucault, M., (1975), Discipline and Punish, (1977), London: Allen Lane.

Foucault, M., (1975–6), Society Must Be Defended, (2003), Harmondsworth:Penguin.

Foucault, M., (1976), The History of Sexuality, Volume 1: An Introduction, (1980), New York: Vintage Books.

Foucault, M., (1982), 'The subject and power’ in Scott, John (ed.), (1994), Power, Volume 1, London: Routledge.

Giddens, A., (1972), 'Elites in the British class structure’ in Scott, John (ed.), (1990), The Sociology of Elites, Volume 1, Aldershot: Edward Elgar Publishing.

Giddens, A., (1979), Central Problems in Social Theory, London: Macmillan.

Goldthorpe, J. H., (1987), Social Mobility and Class Structure (originally 1980), Oxford: Clarendon Press.

Gouldner, A. W., (1954), Patterns of Industrial Bureaucracy, New York: Free Press.

Held, D., (1989), Political Theory and the Modern State, Cambridge: Polity Press.

Hunter, F., (1953), Community Power Structure, Chapel Hill: University of North Carolina Press.

Michels, R., (1927), First Lectures in Political Sociology, New York: Harper and Row.

Miliband, R., (1969), The State in Capitalist Society, London: Weidenfeld and Nicolson.

Mills, C. W., (1956), The Power Elite, New York: Oxford University Press.

Mintz, B., (1975), 'The President’s cabinet, 1897–1972: A contribution to the power structure debate’, in Scott, John (ed.), (1990), The Sociology of Elites, Volume 2, Aldershot: Edward Elgar Publishing.

Mintz, B. and Schwartz, M., (1985), The Power Structure of American Business, Chicago: Chicago University Press.

Morishima, M. (1978), The Introduction to Max’s Economics: A Dual Theory Of Value and Growth. Cambridge: Cambridge University Press.

Mosca, G., (1896), 'Elementi di scienza politica, volume one’ in Mosca, G. (ed.), (1939), The Ruling Class, Chapters 1–11, New York: McGraw Hill.

Mosca, G., (1923), 'Elementi di scienza politica, volume two’ in Mosca, G. (ed.), (1939), The Ruling Class, Chapters 12–17, New York: McGraw Hill.

Pareto, V., (1901), The Rise and Fall of Elites, (1968), New York: Bedminster Press.

Pareto, V., (1916), A Treatise on General Sociology, (1963), New York: Dover.

Parsons, T., (1963), 'On the Concept of Political Power’, Proceedings of the American Philosophical Society 107: 232–262.

Polsby, N. W., (1962), Community Power and Political Theory, second edition 1980, New Haven: Yale University Press.

Poulantzas, Nicos. 1975. Classes in Contemporary Capitalism. London: New Left Books.

Savage, m., K. Williams (2008), Elites: remembered in capitalism and forgotten by social sciences, Sociological Review Volume 56, Issue s1, Pages 1-24.

Scott, J., (ed.), (1990), The Sociology of Elites, Three Volumes, Cheltenham: Edward Elgar Publishing.

Scott, J., (ed.), (1994) Power, Three volumes, London: Routledge.

Scott, J., (1996), Stratification and Power: Structures of Class, Status and Command, Cambridge: Polity Press.

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

Scott, J., (2001), Power, Cambridge: Polity Press.

Scott, J. (2008), Modes of power and the re-conceptualization of elites, Sociological Review Volume 56, Issue s1, pages 25-43.

Sklair, L., (2001), The Transnational Capitalist Class, Oxford: Blackwell Publishing.

Useem, M., (1984), The Inner Circle, New York: Oxford University Press.

Useem, M., (1993), Executive Defence: Shareholder Power and Corporate Reorganization, Cambridge, Mass.: Harvard University Press.

Wright, Erik Olin. 1979: Class Structure and Income Determination. New York: New York: Academic Press.

Źródło pierwodruku: „Innovative Issues and Approaches in the Social Sciences”, September 2013.

Pozostaw odpowiedź

Witryna wykorzystuje Akismet, aby ograniczyć spam. Dowiedz się więcej jak przetwarzane są dane komentarzy.

Komentarz do “Ralph Dahrendorf’s conflict theory of social differentiation and elite theory