Ivery much welcome Nicos Poulantzas’s critique of The State in Capitalist Society in the last issue of NLR: this is exactly the kind of discussion which is most likely to contribute to the elucidation of concepts and issues that are generally agreed on the Left to be of crucial importance for the socialist project, yet which have for a very long time received altogether inadequate attention, or even no attention at all. While some of Poulantzas’s criticisms are, as I shall try to show, unwarranted, my purpose in the following comments is only incidentally to ‘defend’ the book; my main purpose is rather to take up some general points which arise from his review and which seem to me of particular interest in the investigation of the nature and role of the state in capitalist society. I hope that others may be similarly provoked into entering the discussion.

The first such point concerns the question of method. Poulantzas suggests that, notwithstanding the book’s merits (about which he is more than generous) the analysis which it attempts is vitiated by the absence of a ‘problematic’ which would adequately situate the concrete data it presents. In effect, Poulantzas taxes me with what C. Wright Mills called ‘abstracted empiricism’, and with which I myself, as it happens, tax pluralist writers.footnote1 Poulantzas quite rightly states that ‘a precondition of any scientific approach to the “concrete” is to make explicit the epistemological principles of its own treatment of it’; and he then goes on to say that ‘Miliband nowhere deals with the Marxist theory of the state as such, although it is constantly implicit in his work’ (p. 69). In fact, I do quite explicitly give an outline of the Marxist theory of the statefootnote2 but undoubtedly do so very briefly. One reason for this, quite apart from the fact that I have discussed Marx’s theory of the state elsewhere,footnote3 is that, having outlined the Marxist theory of the state, I was concerned to set it against the dominant, democratic-pluralist view and to show the latter’s deficiences in the only way in which this seems to me to be possible, namely in empirical terms. It is perfectly proper for Poulantzas to stress the importance of an appropriate ‘problematic’ in such an undertaking; and it is probably true that mine is insufficiently elucidated; but since he notes that such a ‘problematic’ is ‘constantly implicit in my work’, I doubt that my exposition is quite as vitiated by empiricist deformations as he suggests; i.e. that the required ‘problematic’ is not absent from the work, and that I am not therefore led ‘to attack bourgeois ideologies of the State whilst placing [myself] on their own terrain’ (p. 69).

Poulantzas gives as an example of this alleged failing the fact that, while I maintain against pluralist writers the view that a plurality of élites does not exclude the existence of a ruling class (and I do in fact entitle one chapter ‘Economic Elites and Dominant Class’) I fail to provide a critique of the ideological notion of élite and do therefore place myself inside the ‘problematic’ which I seek to oppose. Here too, however, I doubt whether the comment is justified. I am aware of the degree to which the usage of certain words and concepts is ideologically and politically loaded, and indeed I provide a number of examples of their far from ‘innocent’ usage;footnote4 and I did in fact, for this very reason, hesitate to speak of ‘élites’. But I finally decided to do so, firstly because I thought, perhaps mistakenly, that it had by now acquired a sufficiently neutral connotation (incidentally, it may still have a much more ideological ring in its French usage than in its English one); and secondly because it seemed, in its neutral sense, the most convenient word at hand to suggest the basic point that, while there do exist such separate ‘élites’ inside the dominant class, which Poulantzas describes by the admittedly more neutral but rather weak word ‘fractions’, they are perfectly compatible with the existence of a dominant class, and are in fact parts of that class. He suggests that the ‘concrete reality’ concealed by the notion of ‘plural élites’ can only be grasped ‘if the very notion of elite is rejected’ (p. 70). I would say myself that the concrete reality can only be grasped if the concept of élite is turned against those who use it for apologetic purposes and shown to require integration into the concept of a dominant or ruling class: i.e. there are concepts of bourgeois social science which can be used for critical as well as for apologetic purposes. The enterprise may often be risky, but is sometimes legitimate and necessary.

However, the general point which Poulantzas raises goes far beyond the use of this or that concept. In fact, it concerns nothing less than the status of empirical enquiry and its relationship to theory. In this regard, I would readily grant that The State in Capitalist Society may be insufficiently ‘theoretical’ in the sense in which Poulantzas means it; but I also tend to think that his own approach, as suggested in his review and in his otherwise important book, Pouvoir Politique et Classes Sociales, a translation of which into English is urgently needed, errs in the opposite direction. To put the point plainly, I think it is possible, in this field at least, to be so profoundly concerned with the elaboration of an appropriate ‘problematic’ and with the avoidance of any contamination with opposed ‘problematics’, as to lose sight of the absolute necessity of empirical enquiry, and of the empirical demonstration of the falsity of these opposed and apologetic ‘problematics’. Poulantzas declares himself not to be against the study of the ‘concrete’: I would go much farther and suggest that, of course on the basis of an appropriate ‘problematic’, such a study of the concrete, is a sine qua non of the kind of ‘demystifying’ enterpiise which, he kindly suggests, my book accomplishes. After all, it was none other than Marx who stressed the importance of empirical validation (or invalidation) and who spent many years of his life in precisely such an undertaking; and while I do not suggest for a moment that Poulantzas is unaware of this fact, I do think that he, and the point also goes for Louis Althusser and his collaborators, may tend to give it rather less attention than it deserves. This, I must stress, Is not a crude (and false) contraposition of empiricist versus nonor anti-empiricist approaches: it is a matter of emphasis— but the emphasis is important.

Poulantzas’s critique of my approach also underlies other points of difference between us. But before dealing with these, I should like to take up very briefly what he calls ‘the false problem of managerialism’. Managerialism is a false problem in one sense, not in another. It is a false problem in the sense that the ‘motivations’ of managers (of which more in a moment) are not such as to distinguish the latter in any fundamental way from other members of the capitalist class: i.e. he and I are agreed that the thesis of the ‘soulful corporation’ is a mystification. But he also suggests that I attribute to the managers ‘an importance they do not possess’ (p. 72). This seems to me to underestimate the significance of the ‘managerial’ phenomenon in the internal organization of capitalist production (which, incidentally, Marx writing a hundred years ago, did not do).footnote5 Poulantzas for his own part chooses to stress ‘the differences and relations between fractions of capital’. But while these are important and need to be comprehended in an economic and political analysis of contemporary capitalism I would argue myself that the emphasis which he gives to these differences and relations may well obscure the underlying cohesion of these various elements—and may well play into the hands of those who focus on these differences in order to deny the fundamental cohesion of the capitalist class in the conditions of advanced capitalism.

More important, however, Poulantzas also suggests that I attach undue importance, indeed that I am altogether mistaken in attaching any importance to the ‘motivations’ of the managers. Thus, ‘the characterization of the existing social system as capitalist in no way depends on the motivations of the conduct of the managers... to characterize the class position of managers, one need not refer to the motivations of their conduct, but only to their place in production and their relation to the ownership of the means of production’ (p. 71). I think myself that one must refer to both not because managerial ‘motivations’ are in themselves critical (and Poulantzas is mistaken in believing that I think they are)footnote6 but precisely in order to show why they are not. By ignoring them altogether, one leaves a dangerous gap in the argument which needs to be put forward against managerialist apologetics. This is why, I take it, Baran and Sweezy, for instance, devote a good deal of attention to ‘business behaviour’ in their Monopoly Capital.

This issue of ‘motivations’ also arises, in a much more significant and far-reaching way, in connection with what I have called the state élite and its relation to the ruling class. Poulantzas notes that, in order to rebut the ideologies which affirm the neutrality of the state, I bring forward evidence to show that members of that class are themselves involved in government, and also show the degree to which those who man the command posts of the various parts of the state system are, by social origin, status, milieu (and, he might have added, ideological dispositions) connected with the ruling class. But, he also adds, this procedure, while having a ‘capital demystifying importance’,footnote7 is ‘not the most significant one’ (p. 72). His reason for saying this is so basic that I must here quote him at some length: ‘The relation between the bourgeois class and the State is an objective relation. This means that if the function of the State in a determinate social formation and the interests of the dominant class in this formation coincide, it is by reason of the system itself (p. 73).footnote8 Similarly, the members of the State apparatus ‘function according to a specific internal unity. Their class origin— class situation—recedes into the background in relation to that which unifies them—their class position: that is to say, the fact that they belong precisely to the State apparatus and that they have as their objective function the actualization of the role of the State. The totality of this role coincides with the interests of the ruling class’ (pp. 73–4).footnote9