concept of class. If class and class struggle are to be made compatible with a strategy that displaces the opposition between capital and labour from its pivotal role, it is necessary to redefine class itself in such a way that the relations of exploitation cease to be ‘dominant’ in the determination of class. Poulantzas achieves this reformulation, and in the process succeeds by definition in reducing the working class to such minute proportions that any strategy not based on ‘popular alliances’ appears recklessly irresponsible.
The most important element in Poulantzas’s theory of class is his discussion of the ‘new petty bourgeoisie’. The question of the petty bourgeoisie, as Poulantzas points out, stands ‘at the centre of current debates’ on class structure and is of critical strategic importance.9 Considerable debate has surrounded not only the class situation of ‘traditional’ petty-bourgeois traders, shopkeepers, craftsmen, but most particularly the ‘new middle classes’ or ‘intermediate strata’, wage-earning commercial and bank employees, office and service workers, certain professional groups – that is, ‘white-collar’ or ‘tertiary-sector’ workers. These two ‘petty bourgeoisies’ are the main constituents of the popular alliance with the working class, those which together with the working class constitute the ‘people’ or ‘popular masses’. To locate them correctly in the class structure of contemporary capitalism has been a major preoccupation of Eurocommunist strategists and theoreticians. Poulantzas stresses the strategic importance of the theoretical debate, the necessity of accurately identifying the class position of these groups ‘in order to establish a correct basis for the popular alliance …’.10
Poulantzas begins by attacking two general approaches to the question of these ‘new wage-earning groups’, lumping together some very disparate arguments in each of the two categories. The first approach is that which dissolves these groups into either the proletariat or the bourgeoisie, or both. The second general ‘tendency’ is what Poulantzas calls ‘the theory of the middle class’, a politically motivated theory according to which both bourgeoisie and proletariat are being mixed together in the ‘stew’ of an increasingly dominant middle group, ‘the region where the class struggle is dissolved’.11 Most of these theories are intended to dilute the concepts of class and class struggle altogether. From the point of view of Marxist theory and socialist strategy, there is only one theory, among the several included in these two categories, which represents a serious challenge to Poulantzas’s own: the theory which assimilates the new wage-earning groups to the working class, arguing that white-collar workers have been increasingly ‘proletarianized’. We shall return in a moment to Poulantzas’s reasons for dismissing this approach.
Poulantzas then turns to the solution proposed by the PCF in its political strategy of the ‘anti-monopoly’ alliance. Like Poulantzas himself, the PCF line rejects the ‘dissolution of the wage-earning groupings into the working class’,12 but it denies their class-specificity altogether and allows them to remain in a classless grey area as ‘intermediate strata’. Poulantzas attacks this refusal to identify the class situation of the new wage-earning ‘strata’. It is, he suggests, an abdication to bourgeois stratification theory and is inconsistent with the fundamental Marxist proposition that ‘the division into class forms [is] the frame of reference for every social stratification’. The principle that ‘classes are the basic groups in the “historic process”’is incompatible with ‘the possibility that other groups exist parallel and external to classes …’.13
It should be noted immediately that Poulantzas’s criticism of the PCF line on the ‘new wage-earning groups’ does not strike at its roots either theoretically or practically. In fact, his argument proceeds not as a rejection of PCF principles but, again, as an attempt to supply them with a sounder theoretical foundation, albeit somewhat to the left of the main party line. A truly Marxist theorization of popular alliances must, he argues, be based on a definition of class which grants these ‘strata’ their own class position instead of allowing them to stand outside class. The significant point, however, is that this class position is not to be found within the working class. In other words, Poulantzas is seeking a more clearly Marxist theoretical support for the Eurocommunist conception of an alliance between a narrowly defined working class and non-working-class popular forces.
Why, then, does Poulantzas, in common with the PCF, refuse to accept the theory which ‘dissolves’ these ‘strata’ into the working class? This theory, which he attributes primarily to C. Wright Mills, has been developed more recently in unambiguously Marxist ways by Harry Braverman and others. Poulantzas, however, apparently regards it as a departure from Marxism – for example, on the grounds that it makes the wage the relevant criterion of the working class, thereby making the mode of distribution the central determinant of class.14 (It is perhaps significant that Poulantzas focuses on the wage as a mode of distribution and not as a mode of exploitation – as we shall see in a moment.) He argues further that by assimilating these groups to the working class, this view promotes reformist and social-democratic tendencies. To identify the interests of ‘intermediate strata’ with those of the working class is to distort working-class interests, accommodating them to more backward, less revolutionary elements.15 A political strategy based on the hegemony of the working class and its revolutionary interests, he maintains, demands the exclusion of these backward elements from the ranks of the working class.
On the face of it, then, Poulantzas’s refusal to accept the proletarianization of white-collar workers appears to be directed in favour of a revolutionary stance and the hegemony of the working class which alone is ‘revolutionary to the end’.16 He even criticizes the PCF analysis on the grounds that, despite its refusal to accept this dissolution, it courts the same danger by neglecting to identify the specific class interests of the new wage-earning strata and hence their divergences from working-class interests. It is true that he fails to explain how these dangers will be averted by a ‘working-class’ party whose object is precisely to dilute its working-class character by directly representing other class interests, but let us leave aside this question for the moment. Let us pursue the implications of his own theory of the ‘new petty bourgeoisie’ to see whether it does, in fact, represent an attempt to keep exploitative class relations, class struggle, and the interests of the working class at the centre of Marxist class analysis and socialist practice.
For Poulantzas, the primary structural criterion for distinguishing between the working class and the new petty bourgeoisie at first seems to be the distinction between productive and unproductive labour. The ‘unproductive’ character of white-collar work separates these groups from the ‘productive’ working class. Poulantzas proceeds on the assumption that Marx himself applied this criterion and marked off the ‘essential boundaries’ of the working class by confining it to productive labour. Now it can be shown convincingly that Marx never intended the distinction to be used in this way.17 In any case, Marx never said that he did so intend it, and Poulantzas never demonstrates that this is what he meant. He bases his argument on a misreading of Marx. He quotes Marx as saying ‘Every productive worker is a wage-earner, but it does not follow that every wage-earner is a productive worker.’18 Poulantzas takes this to mean something rather different: ‘as Marx puts it,’ he says, putting words into Marx’s mouth, ‘if every agent belonging to the working class is a wage-earner, this does not necessarily mean that every wage-earner belongs to the working class.’ The two propositions are, of course, not at all the same, nor does Poulantzas argue that the one entails the other. He simply assumes it – i.e. he assumes precisely what needs to be proved, that ‘agent belonging to the working class’ is synonymous with ‘productive worker’. He can then go on to demonstrate that various groups do not belong to the working class simply by demonstrating that they are not, according to Marx’s definition (at least as he interprets that definition), productive workers.
Why this distinction – as important as it may be for other reasons – should be regarded as the basis of a class division is never made clear. It is not clear why this distinction should override the fact that, like the ‘blue-collar’
