Notes on Poulantzas and social classes posted by Richard Seymour

I have written previously on Poulantzas’ theory on social classes. Here I will develop the argument in a different direction.  While inevitably covering some of the same material, I will largely set aside the problematic analysis of the ‘new petty bourgeoisie’ and the centrist conclusions Poulantzas drew from it, with which I was somewhat preoccupied in the previous post.

I. Poulantzas can best be understood as, in Peter Thomas’s words, “a theorist of political and institutional praxis who ‘traversed’ the Althusserian problematic in order to concretize research questions founded upon a different [Gramscian] terrain”. Poulantzas’ work was first and foremost an attempt to theorise the political practices, strategies and tactics of communist parties, using at first the categories of Althusserian marxism to define his object but then moving beyond those categories into the Gramscian terrain of conjuncture and praxis.

II. Social classes are determined principally, but not exclusively, by their place in production (also, with a slightly different emphasis, ‘the economic place’, ‘place in the economic sphere’). But political and ideological relations also have a determining role: "everything happens as if social classes were the result of an ensemble of structures and of their relations, firstly at the economic level, secondly at the political level, and thirdly at the ideological level". Thus, social classes are defined by their place in the matrix of social practices; this corresponds to ‘the structural determination of classes’.

III. Classes only exist in the class struggle, which determines the whole field of social practices. Class determination is thus certainly ‘structural’, but above all relational. Class determination must be distinguished from ‘class position in the conjuncture’: members of the proletariat can take up bourgeois or petty bourgeois class positions, just as members of the middle class can take up proletarian or bourgeois class positions. But their class determination is not altered by this.



IV. The type of determining role taken by economic, political and ideological relations in Poulantzas's theory of classes is related to two concepts taken over from Althusser. The first is the concept of structural causality, distinguished from mechanical causality (in which factor A causes event B causes factor C etc. in an infinite chain) and expressive or teleological causality (in which the features of a social formation are caused by an essence, be it material or ideal). Structural causality refers to the effects of the social whole on its parts. It is the complex organisation of the whole, rather than an essence, which exerts these effects: a present-absent cause, analogous to the way in which the 'causes' of a theatre scene are absent as far as the audience is concerned, but actually constantly present, and always constituting and organising the scene. This is much the way in which class formation must be understood, in terms of "the overall effects of the structure on the field of social relations and on the social division of labour." The second concept is the notion of a social totality as a structure in dominance, which means that its elements each have their specific effectivity, ie none are mere epiphenomena, but they are related to one another asymmetrically so that one element is dominant. Which element is dominant in a given mode of production is fixed by the economic level - so, for example, while the type of surplus extraction in feudalism requires the dominance of politics, the form of extraction in capitalism requires the dominance of the economy. In this sense, the social totality consists of a complex articulation of levels or instances (economic, political and ideological), each of which will have an effect in the constitution of classes, with the economic level having the primary determining role.

V. The economic sphere, as the primary determining level, consists of the production/consumption/division of the social product. In this sphere, relations in the production process, such as the labour-capital relation in the capitalist mode of production, are decisive, and from these flow visible class attributes such as income differentials, as well as less immediately tangible class capacities and strategic advantages or inhibitions. The production process is never purely technical, but consists of a unity between the labour process and the social relations between different productive agents.

Poulantzas distinguished between two types of relation in the production process. The first hinged on the owners and the means of production, and the second on the workers and the means of labour.  Depending on the mode of production, the respective classes would have different degrees of economic ownership (the real power of a class, distinct from legal ownership, to dispose of economic resources for various uses) and economic possession (the power of a class to organise and determine labour processes).  For example, under feudalism, the ruling class enjoyed full economic (and juridical) ownership, but it did not have full economic possession because the peasants retained significant control over their means of labour (land) and how it was operationalised in the labour process.  By contrast, capitalist control extends both to the ownership of the means of production and possession of the labour process.

At any rate, these are the most important production relations which determine the characteristics of the exploiting and exploited classes in any given mode of production.  This is where Poulantzas made one of his most important, but controversial, delimitations: if the production process was one of exploitation and struggle, productive labour was that which gave rise to surplus-value and was thus exploited.  Those wage workers who were not directly involved in the production of surplus value - service industry workers - were not exploited in this way and were thus not part of the working class.  Poulantzas acknowledged that those wage earners he excluded from the working class were nonetheless still exploited in the sense of contributing surplus labour; their wages were still equivalent to the cost of reproducing their labour rather than the total value of their labour.  The arbitrary nature of the distinction introduced here was evident where Poulantzas referred to transport workers as proletarians because of their necessary role in ensuring the realization of surplus value by bringing commodities to the market.  This logic can be extended insofar as capitalism has an extended reproduction.  Take the public sector.  Workers there by and large do not contribute directly to the production of surplus value.  Rather, they facilitate the realization of surplus value in the present, while also ensuring the future extraction of surplus value - by, for example, sustaining the circulation of fictitious commodities.  These are just as vital to the circuit of capital as the labour performed by transport workers.  In my opinion, it makes much more sense to speak in terms of a core, and an extended working class; the former largely consisting of 'productive' labourers (those generating surplus value), the latter of 'non-productive' labourers, but also of pensioners, the unemployed, and so on.

VI. But, as we have seen, the social division of labour is not structured by the economic level alone.  It could only be so if the economic level (production for profit, mediated through exchange) was a self-sufficient, self-reproducing sphere.  This cannot be the case since, for example, it relies on 'fictitious commodities' (land and raw materials, labour, money and forms of knowledge), which are bought and sold as if they were commodities but are not produced in the first instance in a profit-driven process.  This means that the economy is always-already mutually articulated with other systems - ecological, juridical, political, etc. - which are essential for its expanded reproduction.  Political and ideological relations are thus important not only in the determination of classes, but also in the determination of strata and fractions within classes.

In this connection, Poulantzas made a valid point about the determination of the 'new petty bourgeosie'.  In discussion wage-earning technicians, he argued that they occupied a seemingly 'contradictory' class location, insofar as they both contributed to the production of surplus value, and at the same time partook of authority in the overseeing of the labour process "and its despotic organization".  Their primary role was not to produce surplus value, but to assist in extracting surplus value from the labour force.  These workers were thus separate from the working class, and constituted a middle class.  However, the despotic organization of production, in which capital strives to monopolize knowledge of the labour process and subordinate labour to its authority, is political, juridical and ideological, not merely a technical arrangement.  It is the form of presence, within the production process, of capital's political dominance over labour.

VII. Ideology's determining role in the formation of social classes cannot be reduced to 'consciousness', false or otherwise.  The concept of 'class consciousness' derives from a marxist gloss on bourgeois social theory, according to which social action is governed by normative elements called 'values'.  Values corresponding to social classes derive from their function within the social totality, and ensure their continued reproduction within it. A marxist gloss on this 'imputes' values to classes irrespective of the ideas actually held by their members, based not on their functional contribution to the social totality but rather their role as genetic principles in an historical process.  (Cf Lukacs: "class consciousness consists in fact of the appropriate and rational reactions ‘imputed’ ... to a particular typical position in the process of production. This consciousness is ... neither the sum nor the average of what is thought or felt by the single individuals who make up the class. And yet the historically significant actions of the class as a whole are determined in the last resort by this consciousness and not by the thought of the individual – and these actions can be understood only by reference to this consciousness."). From this arises the dubious concept of 'false consciousness', which Poulantzas rejected.

If anything, per Althusser, he treated ideology as an unconscious structure of domination imposed on agents from without. And so far as the determination of classes is concerned, the major form of ideological dominance exercised by capital was centred on the division between mental and manual labour.  The hoarding of expertise by capital ensured its ideological dominance over subaltern classes, naturalising its domination and 'proving' the inability of workers to organise production by themselves.  Poulantzas argued that the role of mental labourers in securing that dominance meant that they were petty bourgeois elements who aligned ideologically with the bourgeoisie.  Since he included low-level clerks in this, it seems fair to say that this would also include groups like teachers, whose role is to secure capitalist ideological dominance on a nation-wide rather than company- or industry-wide scale.  But here, there is a problem.  Teachers in most advanced capitalist societies do not appear to align ideologically with the bourgeoisie.  They are increasingly among the most militant, left-wing workers.  This can only be explained by introducing a distinction with Poulantzas did not, which is between mental labour as an executive process and mental labour as a menial, subordinated process.  The 'proletarianisation' of intellectual labourers precisely takes the form of subordinating their intellectual labour, depriving it of autonomy, and depriving its agents of real authority.  The social division of labour centres really, not on the mental/manual division, but on the executive/menial division.

VIII. Likewise, one cannot interpret class strata, fractions and social groups without referring to political and ideological dimensions.  Beginning with strata, these can't be understood on the basis of economic differentiations within the labour force - 'skilled' vs 'unskilled' labour, for example, or high wage vs low wage labour.  This is partly because 'skill' is profoundly politically and ideologically determined - think of how 'skill' has been racialised and gendered in various contexts.  This table from E O Wright's Classes (1985), which nonetheless takes issue with Poulantzas on several key problems, illustrates the point:


Just as important, however, is that differentiations within a class do not necessarily coincide directly with positions in the organization of labour.  Wage differentiations have a role, but looking at these closely it's clear that this issue is also structured by politics and ideology.  In the abstract, wage rates are determined by the cost of the reproduction of labour.  But the specific wage rate is never set purely by 'technical' necessities.  Marx includes in his factors in this cost the 'level of the development of civilization', meaning that there are norms and expectations produced by the development of the forces of production, as well as traditions of struggle, around which struggle takes place.  At the same time, employers often raise wages for groups of workers whose labour is not inherently more expensive to reproduce, thus suggesting that the purpose is political control of the labour force.

In the same way, the fractionalisation of the bourgeoisie has an economic logic, as fractions such as manufacturing and finance have a different relationship to surplus value and different positions in the market, but political and ideological differences can also have an impact.  Poulantzas's example here was the distinction between 'national' and 'comprador' bourgeoisies in the age of imperialism - the former's interests linked more closely to national economic development, the latter bound up with foreign imperialist capital.  The distinction did not strictly correspond with economic interest, since the internationalization of capital and the interpenetration of capitals tended to erode such economic differences.  This meant that the division in the peripheral societies (if you'll permit the term) between 'comprador' and 'national' bourgeoisies was to a significant extent political and ideological.  Certainly, in the advanced capitalist societies Poulantzas thought it was impossible to find a 'national' bourgeoisie that was in practice opposed to US imperialism.  I do not think that there is an important division between 'national' and 'comprador' bourgeoisies in this era, even in the peripheral societies.   In most cases, the division is between fractions with different global and regional orientations, different imperialist and sub-imperialist alliances, etc.  One could argue that in an imperialist society like the UK, for example, there has been a division, by no means strict, between Atlanticist and Europeanist capital, which is a division that takes place not simply or neatly between medium and large capital, or between manufacturing and finance, but also along long-term strategic lines expressed at a political and ideological level.

Finally, in this section, a 'social category' is a group of workers that does not form a class or a class fraction but which, due to certain political or ideological factors, might have a certain unity of their own.  One can think of the police, whose members are distributed across classes (a bourgeois leadership, and a middle class rank and file), yet are unified in their relationship to the state and their role in the moral-ideological regulation of society.  There is no inevitability about such unity.  It could be argued that the civil service was once a much more cohesive 'social category' unified by its duty of public service, but neoliberal reforms have exacerbated class divisions within it, and provoked a hitherto unknown degree of militancy among the rank and file.  It is important to recognise the role of 'social categories' in the course of class formation, and class struggles since, strategically, they can be obstacles to the unity of one or other class in its political action; and just as suddenly, a fissure in the 'social category' can result in its proletarian members forming a militant vanguard.

IX. One other factor which is of huge importance is the direct role of the state in class division and stratification.  Clearly, politics and ideology has had a pronounced role in the determination of classes in the public sector.  An example will suffice to show this can work.  The UK is very much closer to the US than its continental counterparts, inasmuch as only a fifth of total employment is accounted for by the state.  Clearly a much wider periphery of employees has a direct interest in public expenditure since a very large number of firms are dependent on government contracts.  But those working directly for the state, rather than simply in the state's field of operation, are more directly exposed to the political organisation of classes.  Take, for example, the reforms to the civil service rolled out by New Labour.  The last government largely succeeded in its neoliberal 'streamlining' of civil servants, reducing total employees to close to its target of 480,000.  Simultaneously, it increased the number of Senior Civil Servants by 35%.  This is actually a trend wherever public sector employment is cut: the ranks of managerial and supervisory staff are significantly increased: partly, one must assume, to manage any potential militancy as a consequence of cuts, but also to secure greater control of the labour process and thus increase the pace of work. 

But the state's role, as indicated, goes beyond the public sector.  For an example, here is a very current and pressing question raised by the coalition government's efforts, and by the effects of the recessions.  Employment is not rising, except among a single sector - the self-employed.  Now, the question is, do developments such as this signal a strange resurgence of the traditional petty bourgeoisie?  Does it attest to an upsurge of entrepreneurial activity, getting-on-bikes and so on?  It seems unlikely to say the least.  I would suggest, just as a working hypothesis, that most of this self-employment is fictitious, a bit of legal legerdemain to get around legislation and pay well below the minimum wage, or deny certain rights or conditions.  The people who deliver goods, or work in pyramid schemes, are often legally self-employed.  But the real relationships of economic ownership and possession are such that the means of production are owned and control by a firm, the labour process itself reduced to a series of elements easily controlled by the firm.  I'm aware of how this works, because I have worked as a 'self-employed' market researcher in the past.  Here, the state's action has taken the form of expanding the pool of precarious labour, while providing the juridical basis for its intensified exploitation.

X.  The real site of action of social classes is the social formation.  A social formation is not the same thing as a mode of production.  In a simple capitalist mode of production, there would tend to be only the two 'fundamental' classes - the bourgeoisie and the proletariat - with the pre-capitalist remnants experiencing a sad diminuendo over time. But despite the growing simplification of classes in the advanced capitalist countries, social formations are usually not simple; they consist of elements and residues of pre-capitalist modes of production under the international dominance of the capitalist mode of production. As a result, in real social formations you might find the remnants of old classes, petty commodity producers, peasants, and so on. Moreover, in the development of capitalism, it has had a tendency to develop middle classes that replace the old, dwindling petty bourgeoisie. Poulantzas mis-diagnosed the 'new petty bourgeoisie', but there is no doubt that such a class did come into existence as capital developed new managerial and technical apparatuses to control the work process more effectively and suppress militancy on the shop floor.  Finally, of course, the processes of class decomposition and recomposition ensure that there will be groups in the process of moving between classes over a loong period.  This is why in concrete struggles, there will tend to be in practice various forms of class alliance.

In addition, of course, the social formation is not simply composed of its productive processes, or its social classes.  Poulantzas never abandoned the central focus on class, and particularly on the nature of class alliances.  But if, as suggested above, the productive sphere is mutually articulated with other spheres (family reproduction, religion, ecology, etc), and has a tendency to dominate those other spheres and subordinate them to its logic, then the extended reproduction of capitalism must necessarily continually multiply and reconstitute social antagonisms.  The fundamental antagonism will always be class, even if the principal antagonism in any given situation may not be - but it will always itself be structured by other antagonisms, and one must always look toward other sources of political action if one is engaged in an anticapitalist project.  Finding these sources, and unifying them under the hegemony of the working class is where the analysis would shift, for Poulantzas, from an Althusserian problematic to a Gramscian one.