structural similarities among workers. What is true is that the organization of production in industrial capitalism establishes various divisions among workers within the labour process which are determined not by the technical demands of the labour-process itself but by its capitalist character. These divisions often constitute obstacles to the formation of a unified class – even in the case of workers who belong to the same class by virtue of their relation to capital and exploitation. But it is not clear why the divisions cited by Poulantzas should be more decisive than any others that divide workers in the labour-process or disunite them in the process of class organization. It is not clear why such divisions should be regarded not simply as obstacles to unity or roadblocks in the difficult process of class-organization – a process riddled with obstacles even for blue-collar workers – but rather as definitive class barriers dividing members from non-members of the working class.20 In fact, Poulantzas’s theory seems unable to accommodate any process in the development of classes at all. There seems to be only a string of static, sometimes overlapping, class situations (locations? boxes?). This is a view which in itself would seem to have significant political implications.
If the ideological division between mental and manual workers within the exploited wage-earning groups does not correspond to any objective barrier directly determined by the relations of production between capital and labour, neither does it correspond to a real and insurmountable division of interest between these workers. The class interests of both groups are determined by the fact that they are directly exploited through the sale of their labour-power; these interests have to do in the first instance with the terms and conditions of that sale, and in the last with the elimination of capitalist relations of production altogether, both the ‘formal’ and the ‘real’ subjection of labour to capital. The different functions of these workers in the labour-process may create divisions among them, based in some cases on differences in their responsibilities, education, income, and so on;21 but these differences cannot be regarded as class divisions by any standard having to do with relations of production and exploitation. The ideological divisions between them are constituted less from the point of view of their own class interests than from the point of view of capital, which has an interest in keeping them apart. The imposition of capitalist ideology can certainly operate to discourage unity within the working class and interfere with the processes of class organization, but it can hardly be accepted as an absolute class barrier between different kinds of workers.
Poulantzas has thus presented a class analysis in which relations of exploitation are no longer decisive. This is in keeping with the fundamental principles of his theory. The relations of production and exploitation, according to him, belong to the ‘economic’ sphere which, as we have seen, though it ‘determines in the last instance’ may not be dominant in any given mode of production or social formation. This notion is carried over into the analysis of class.22 It now becomes clear that there are cases in which political or ideological factors ‘reign supreme’ in determining class. Poulantzas is saying more than simply that the formation of classes is always a political, ideological, and cultural process as well as an economic one, or that relations between classes are not only economic but also political and ideological. Nor, again, is he simply pointing to the special role of the ‘political’ where relations of production are themselves ‘politically’ organized. He is suggesting that ideological and political relations may actually take precedence over the relations of exploitation in the ‘objective’ constitution of classes, and that political or ideological divisions may represent essential class barriers. Again the relations of exploitation have been displaced.23
What, then, are the practical consequences of Poulantzas’s views on class? Why is it a matter of such critical importance whether or not white-collar workers are theoretically included in the working class? Poulantzas himself, as we have seen, maintains that it is strategically important to separate out the ‘new petty bourgeoisie’ in order to protect the revolutionary integrity and hegemony of the working class. There is, however, another way of looking at it. We have seen that for Poulantzas the relations of production are not decisive in determining the class situation of white-collar workers. The ‘new petty bourgeoisie’ is distinguished as a class on the basis of ideological divisions defined from the point of view of capital. In other words, they constitute a class insofar as they are absorbed into the hegemonic ideology of capitalism; and that absorption seems to be definitive: the new petty bourgeoisie can be made to adopt certain working-class positions – that is, their political attitudes can ‘polarize toward’ the proletariat; but they cannot be made part of the working class. These propositions are very different from the observation that the inclinations of white-collar workers to accept capitalist ideology may be stronger than those of blue-collar workers; that these inclinations constitute a problem for class organization, for the development of class consciousness, and for the building of class unity; and that they must be taken into account by any socialist strategy. For Poulantzas, it would appear that these inclinations represent a decisive class boundary; and this has significant strategic implications.
Despite Poulantzas’s criticism of PCF theory and strategy, his theory of class belongs to the ‘attempt of the theoreticians of Eurocommunism to reduce the weight of the Western proletariat to that of a minority within society…’.24 At a stroke of the pen, the proletariat is reduced from a comfortable majority in advanced capitalist countries to a rump group which must inevitably place class alliances at the top of its agenda. Poulantzas’s very definition of class in general and the ‘new petty bourgeoisie’ in particular displaces the focus of socialist strategy from creating a united working class to constructing ‘popular alliances’ based on class differences, even based on divisions imposed by capital. Any appeal to the ‘new petty bourgeoisie’, for example, must be directed not to its working-class interests but to its specific interests as a petty bourgeoisie.
The strategic implications become even clearer when this view of alliances is embodied in a particular conception of ‘working-class’ parties as organizations which do not simply form alliances with other groups and parties but directly represent other class interests. Poulantzas insists that ‘the polarization of the petty bourgeoisie towards proletarian class positions depends on the petty bourgeoisie being represented by the class-struggle organizations of the working class themselves … This means, firstly, that popular unity under the hegemony of the working class can only be based on the class difference between the classes and fractions that form part of the alliance …’25 This notion turns out to be a two-edged sword. On the one hand, it suggests that the popular forces should themselves be transformed in the process of struggle. That is why, argues Poulantzas, the alliance should be established ‘not by way of concessions, in the strict sense, by the working class to its allies taken as they are, but rather by the establishment of objectives which can transform these allies in the course of the uninterrupted struggle and its stages, account being taken of their specific class determination and the specific polarization that affects them’.26 On the other hand, the very idea that alliances must not be based merely on ‘concessions’ to allies ‘taken as they are’ also entails that working-class organizations must cease to be organizations of the working class. It now appears that it is not just the integrity of working-class interests that these organizations must protect, but also that of the petty bourgeoisie. Poulantzas now seems to be criticizing the PCF for taking the ‘popular masses’ too much for granted, instead of acknowledging the specificity of their various class interests. A ‘working-class’ party cannot simply make ‘concessions’ to elements outside itself from a vantage point consistently determined by working-class interests; it must actually represent other class interests – and this means establishing objectives addressed to these other class interests. This inevitably raises the question of the degree to which the ultimate objectives of socialism itself must be tailored to the measurements of cross-class alliances.
IV
The groundwork for a theorization of Eurocommunism was, then, already firmly laid in Classes in Contemporary Capitalism; but its logic and strategic implications were not fully worked out until Poulantzas wrote his last two major works: The Crisis of the Dictatorships (1975–76) – which may mark the critical turning point to the right – and State, Power, Socialism, published