Colin Barker Archive   |   Trotskyist Writers’ Index  |   ETOL Main Page

Colin Barker


A note on the theory of capitalist states [1]


From Capital & Class 4, Spring 1978.
Copied with thanks from the REDS – Die Roten Website.
Marked up by Einde O’Callaghan for the Encyclopaedia of Trotskyism On-Line (ETOL).

In Capital and Class 2, John Holloway and Sol Picciotto (hereafter H&P) published an important contribution to the ongoing debate on the Marxist theory of the state. The Note that follows criticises their article on a couple of points, so I should remark that their argument, taken as a whole, has very significant merits. In particular, its insistence that the capitalist state-form cannot be considered separately from the capital relation is exceptionally valuable.

To summarise my argument: I suggest that there is a very significant weakness in H&P’s article, a weakness which is however anything but peculiar to them. Their treatment of the state remains at an inappropriate level of abstraction, in particular in that it treats the state as if it existed only in the singular. Capitalism, however, is a world system of states, and the form that the capitalist state takes is the nation-state form. Any discussion, therefore, of the capitalist state form must take account of the state both as an apparatus of class domination and as an apparatus of competition between segments of the bourgeoisie. The failure of H&P, together with most other Marxist writing on the state, to integrate this perception into their account of the state is connected with a second problem in their article: their insistence on a conceptual separation between “state” and “capital”, such that state economic intervention seems to be utterly problematic for the capitalist state. Indeed, their account rules out as theoretically impossible (or at least very difficult) what has actually been happening historically – namely, the tendency within 20th century capitalism for capital to be organised directly by nation-states, thus taking the form of state capital.

Throughout their article, H&P refer to “the state” and “the capitalist state” in the singular. One might get the impression, from H&P as from a mass of other Marxist writing on the state, that capitalism has but one state. Where it is acknowledged that the beast is numerous, the implications of that very concrete fact are not developed at All. In fact, however, the very multiplicity of capitalist states is of great importance to the theory of the capitalist state form.

H&P quote Marx’s Capital Vol. III:

The specific economic form, in which unpaid surplus-labour is pumped out of the direct producers, determines the relationship of rulers and ruled ... Upon this, however, is founded the entire formation of the economic community which grows up out of the production relations themselves, thereby simultaneously its specific political form.

Now the central thrust of their argument (with which I agree) is that an adequate theory of the state depends on comprehending the form of the fundamental social relations of production. But, if we take Marx seriously, what is the “economic community that grows up out of the production relations themselves”? Taken as a whole, it is the world market, international capitalism, the global system of social relations that has grown up – for the first time in human history – on the foundation of the capital relation. And if we ask too, what is the “specific political form” that this economic community takes, we must answer that it is the set of nation-states that make up the “international political community” of world capitalism. The nation-state system is a product of capitalist development, and is global in scope in just the way that capital as a social relation is.

In this light, we should conclude that some of Marx’s summary accounts of the relation between capital and the state are, at best, misleading. Thus the Communist Manifesto:

The modern state is merely the executive committee for managing the common affairs of the whole bourgeoisie.

Take the last phrase: “the whole bourgeoisie”. The whole bourgeoisie is an international class, like the proletariat: both are formed as elements of a system of production relations that, from their commencement, were international in scope. The “common affairs” of the whole bourgeoisie cannot be managed by any single nation-state, yet this is the form that the capitalist state takes. [2] (Were this not so, the whole Marxist discussion of imperialism and nationalism would rest on thin air.)

H&P suggest correctly that the problem in the Marxist theory of the state is to locate the state in terms of the basic social relations of capitalist production, in the capital relation: “The starting point for the analysis of the capitalist state is thus capitalist society, not the state in general.” Capitalist society, we’ve suggested, is not coterminous with the geographical space occupied by any one nation- state, but is rather the world. The problem is, how do we relate the form of the capitalist state to the society of which it is part and expression? To answer this, we need to understand the capital relation properly. There is a tendency within Marxism to treat the “social relations of production” as being only the relations formed at the moment and point of production itself: in other words, to identify a mode of production with the immediate relations of exploitation in the productive process itself. [3] This, even if we restrict our discussion to capitalism alone, will not suffice. Marx’s critique of bourgeois podlitical economy rested on an analysis of the forms of alienation developed within capitalism, one crucial aspect being the characteristic division of labour within commodity production, in which social production is out of the control of the associated producers. Social relations within commodity production are governed by mutual competition and antagonism among the producers. The materialist critique is aimed at capitalist relations, not simply as alienated exploitative relations, but also and simultaneously as alienated competitive relations. The capital relation is a summary expression for the whole nexus of social relations, founded in the production of surplus-value, but embodied in the whole circuit of capital. That is, the capital relation is more than the moment of capitalist production: it is also the other moments of the capital circuit (exchange, realisation, price-formation, etc.).

Capital, according to Marx, can only exist as many capitals; through the interaction between the many capitals the principles of capital-in-general are realised. A single universal capital is a contradiction in terms.[4] It is thus characteristic of capitalism that it develops through competition, which competition is the source and expression of the anarchy of capitalist production. Hence, Marx argued, the social relations of capital have a dual form: anarchy and despotism. Between the many capitals there is anarchy; within each capital, despotism. Each relation, anarchy and despotism, is the condition of the other. [5]

Now if the capital relation has this form, and if the state is an aspect of the capital relation, we might expect to find in the state form elements of this dual determination. As we do. The nation-state, capitalism’s state form, is itself both a structure of despotism vis-à-vis its “subjects” and a structure of competition vis-à-vis its rivals. Its very form expresses the fact that the capitalist state is not something above and separate from the relations of capitalist production, but is itself directly part of those relations. Being anything but a state of the “whole bourgeoisie”, each nation-state is never more than a state of some capital(s), of a segment of the whole bourgeoisie. Moreover, to insist on the partial, national character of the capitalist state-form is not merely a matter of adding on another “factor” to the discussion of the state. The dual determination of the state is a permanent presence in all aspects of state policy and activity. In a capitalist world, it becomes ever more the case that “forgetting” the international dimension of the capitalist state system puts the theorist in the position of a one-handed violinist.

This is not the place to consider how the various nation-states were formed. All we need to note is that each separate nation-state represents an achieved fusion, or mobilisation, of a particular segment of the whole bourgeoisie and other classes into a nation, and a nation formed moreover in opposition to other nations. It may well be that the nation-state is a committee for managing the common affairs of its bourgeoisie – however we define “its bourgeoisie” and “its common affairs”. But it is a structure that is shaped, inter alia, by that bourgeoisie’s competitive struggles with other bourgeoisies, other nations, other capitals. More than an instrument of inter-class domination, the modern state is also an instrument of intra-c lass competition. That competition takes many forms, including trade, trans-national investment, imperialist domination, war, diplomacy, etc. And each nation’s ability to compete depends on the relative size of the capital(s) that fall(s) within its orbit, however we define that orbit.

In H&P’s account, however, the state appears one-sidedly as a relation of force directed principally at the working class:

This abstraction of relations of force from the immediate process of production and their necessary location (since class domination must ultimately rest on force) in an instance separate from individual capitals constitutes (historically and logically) the economic and the political as distinct, particularised forms of capitalist domination. (p. 79)

The state must appear in this one-sided form, since their whole article is concerned with an abstraction called “the state” whose connection with the actual states of the capitalist system is not adequately developed. It is as if we were to try to account for capital and its laws of motion without reference to its existence as many capitals.

The above quotation illustrates a further problem in H&P’s article. Their argument centres on the idea of the “separation” of the economic and the political as instances and forms. This “separation” is a point of principle with them: the concept of “state”, though they argue it is founded in the capital relation, is so separated in their analysis from “capital” that “state” and “capital” are two opposed notions. In their treatment, the state always stands outside the immediate valorisation process, and must stand outside:

... the state must remain essentially external to the process of accumulation. While the purpose of state action must be to promote the accumulation of capital, it must, by reason of its form, remain external to that process. (p. 96, my emphasis – CB)

That permanent externality of the state seems to me exceptionally dubious. Given the national form of the capitalist state, there is no “reason of its form” why it must be thus external to capitalist production and accumulation. H&P actually come close, in the last part of their article, to belying their own assumption about the separation of the “political” and the “economic”, the “state” and “capital”. For in their own account, it seems that “the autonomy of state action from the immediate demands of the valorisation process” is itself threatened by the development of capital, even though that autonomy of the state is crucial to their account and is “implicit in the particularisation of the state as a distinct form of capitalist domination”. As they note, the British state “intervenes directly in the production process, taking over particular industries and reorganising the actual process of value creation and exploitation”. For H&P this kind of development (in no way peculiar to Britain) is a fundamental contradiction in the state’s situation, for it threatens the “generality implicit in its form”. It is exactly this idea, that there is “generality implicit” in the capitalist state form, that I suggest is inadequate. In the sense in which H&P use the term, capitalism has no organised institution with generality implicit in its form. No state can have this implicit generality, for each state is merely a national state. Within capitalism, the only source of “generality” consists in its blind laws of motion, produced by the movements and interactions of the world’s constituent capitals and constituent nation- states. Capitalism in this sense is ungoverned, anarchic. No central institution governs it, only the movement of its parts. Its “generalities” are only outcomes of anarchic relations between competing capitals.

H&P are not, of course, alone in the view that “the state is not capital”, which is repeated by numbers of writers, including Altvater, Offe and Habermas. In all cases, the strict demarcation line drawn between “state” and “capital” rests on an account of the state form in which the state is treated in the singular. [6] In all cases, for the purpose of analysing the capitalist state, the bounds of capitalism are treated as coterminous with the national frontiers. That is, rather than seeing capitalist society as a global “social formation , as a real totality, the world is seen as a set of capitalist societies, a mere agglomeration and not a unity.

If we hold firmly to the merely national character of the capitalist state, and reject the (residual Hegelian?) idea that the state has some principle of generality implicit in its very form, then we can abandon the highly restrictive assumption that “state” and “capital” are mutually exclusive terms. In which case, the discovery that some particular state is becoming more and more directly involved in the valorisation process is not in itself an implicit threat to its own logic or anything of the sort. And the kind of developments to which H&P refer can be accounted for in simpler and altogether more straightforward terms.

Consider for example the limitations on state action to which H&P very properly draw attention. These limitations they derive simply from “the state’s structural relation to, and separation from the immediate process of exploitation”. Yet these limitations also derive, surely, more simply from the inability of any single nation-state to manage the world economy. In this respect it is notable that David Yaffe’s account (cited by H&P) of the growth of the state apparatus in terms of “the state’s postwar commitment to full employment” is subject to just the same limitations as Keynesianism, in that it fails to explain how the British (or any other) state goes about encashing its “commitment to full employment”. Roosevelt in the 1930s seems to have been as subjectively committed to full employment as were Attlee and Churchill after the war, but he failed. What brought “full employment” to western capitalism was not simply the separate “commitments” of the various national governments, but also – and crucially – the global situations which, while it was the outcome of their various national decisions, was yet more than the sum of the parts. War, we should perhaps remember, restored full employment ...

In their last section, H&P do make reference to the phenomenon of the system of capitalist states, but without recognising the problem it poses for Marxist theory. Their only theorisation of the question refers to processes of “convergence” affecting capitalist states, through processes of combined and uneven development and inter-state imitation. As presented, their “convergence thesis” is rather dubious. There may well have been elements of “imitation” in, say, German or Japanese capitalist development in the 19th century, but that imitation occurred not in contemplative isolation but as a response to pressures exerted on them by the more highly developed capitals of Britain and the USA. The term “imitation” suggests mere repetition, sameness, and is inappropriate: the German and Japanese “imitations” were competitive responses, whose character dictated to their states more centralised and less “liberal” political frameworks than existed in the states they were supposedly “imitating”. In any particular branch of capitalist production, some capitals are “ahead” and others “behind”; the “backward” experience the law of value as a coercive force dictating different methods of “catching up” from those adopted by the already “advanced” capitals. So too with the relations between nation-states. Latecomers and laggards, to succeed, seem to require a greater degree of centralisation than their more advanced rivals.

If we asked, what forces move the states of Western Europe towards “planning” and “nationalisation” in the present period, our answer would have to include the massive weight of US capital in the world economy. The shift towards national “planning”, a marked feature of West European states since the late 1950s, is not an imitative response to the USA, where such processes have gone least far. It is a competitive response, an attempt to mobilise capital resources for competition in the world market.

In H&P’s article, though, the world market appears only as an occasional afterthought (as p. 93). It is not central to their analysis of the state and its relation to capital, but is just one of the “myriad extraneous circumstances which affect the way the crisis presents itself, and may provide escape routes for particular national capitals”. But, what is an “escape route” for one capital is thereby a source of major disruptive pressure for another. Overseas investment by US multinationals like IBM may be, for them, an “escape route”: it’s hardly been that for the British domestic computers industry.

When H&P very properly suggest that state intervention does not do away with the anarchy of capitalism, they propose only that the inherent antagonisms of the market, where capitals meet as “hostile brothers”, reproduce themselves within the state. They certainly do. But they also reproduce themselves between states, who also meet as “hostile brothers” whether in the IMF vaults or in the United Nations soup queue. The internal forms assumed by nation-states are shaped, not only by domestic class struggle, but also by the movement of capital on the world stage, a movement that includes themselves.

The clear implication of my argument is that there is no abstract, theoretical reason why history should not have happened as it has: no reason, that is, why particular nation-states should not have become “capital personified”. “State” and “capital” can become a unity. Indeed, there are structural aspects to the state’s form which suit it to this purpose: it has aspects of a homology of form with every individual capital, being doubly determined by the capital relation as anarchy-plus-despotism. In most Marxist writing on the state, it is the second determination – the repressive aspect of the state – which is stressed. To that extent, Marxist understanding of the state has tended to continual one-sidedness.

When Marx wrote Capital, any tendency to the assimilation of “state” and “capital” was still in its infancy. Yet in Anti-Dühring (1878) Engels was already discussing such developments. In any case, Capital itself stands incomplete in two respects vital to the present discussion: the planned volumes on the state and the world market seem never to have been even sketched out. What Marx did give us, however, was a clear account of the real historical tendency to concentration and centralisation of capital into the joint-stock company form and – by mere extension – into the form of “state capital”. [7]

The development of state capital is a predominantly 20th century phenomenon, its emergence within world capitalism being characteristically uneven. At one extreme are those states where the means of production are more or less totally centralised into one national-state capital, whose development is only comprehensible in terms of their competitive interactions with the rest of world capital, and not simply in terms of their domestic class relations considered in isolation. [8] At another pole stand the nation-states of western capitalism, around which are formed more or less articulated and enlarged blocks of capital, which are either the direct legal property of the states (nationalised industries, etc.) or are increasingly effectively state property (given their dependence on the state for investment capital, orders, etc.). Capitalism in the last quarter of the 20th century can no longer be described as if “capitalism” were synonymous with personal private property. The erosion of the apparent neutrality of the state, to which H&P refer, is a real erosion: crucial to it are all those processes, including capital concentration, which press in the direction of a fusion of state and capital.

Any such fusion of state and capital, achieved either partially or totally, in no sense abolishes the conditions of capitalist production, nor resolves its contradictions. There are of course ideologues of late capitalist development who conceive that socialism is precisely national, single-state capital unified and concentrated into one block, a view of socialism shared of course by the bourgeoisie. The view of socialism as equivalent to single-state nationalisation, the equation of capitalist nationalisation with the Marxist idea of the socialisation of production, is widespread still within the workers’ movement. That such a conception of socialism should continue with such strength, despite all the experiences of the working class in this century, is witness to the under-theorised condition of Marxist state theory. It is, seemingly, a characteristic of the 57 varieties of reformism that their conception of socialism stops at the local frontier posts. If the critique of political economy is to advance its attack on the capitalist state form, we must critically discuss the conceptions of “state” and “capital” that we’ve received, half unconsciously, from previous generations. I’m aware that this Note raises more problems than it solves. Nonetheless, today more than ever we should remember that the working class has a world to win.


Colin Barker lectures in sociology at the Department of Social Science, Manchester Polytechnic, Aytoun Street, Manchester 1.

1. My thanks to the following for very helpful comments on an earlier draft: Claudia von Braunmühl, Simon Clarke, Ian Cough, Peter Green, John Harrison, Bob Jessop, Doreen Massey, John Ure. They will agree that my responsibility is total.

2. It is very odd what an “absence” there is in Marxist theorising on the state in this respect. To my knowledge, the only Marxist who has explicitly called attention to the matter is von Braunmühl (1974, 1976, 1977). Even Marxists whose other theoretical positions ought to have sensitised them to this aspect of the capitalist state seem not to remark on it. For example, Martin Shaw (1974) produced a very useful critical review of some problems in the development of the Marxist theory of the state without even mentioning the state’s national form: yet his view, that Russia and Eastern European countries are state capitalist formations, depends theoretically on exactly this conception.

3. There is a very useful discussion of this in Banaji, 1977.

4. E.g. Marx, 1973, p. 421: “Since value forms the foundation of capital, and since it therefore necessarily exists only through exchange for counter-value, it thus necessarily repels itself from itself. A universal capital, one without alien capitals confronting it, with which it exchanges – and, from the present standpoint, nothing confronts it but wage-labourers or itself – is therefore a non-thing. The reciprocal compulsion between capitals is already contained in capital as realised exchange-value.”

5. E.g. Marx, 1976, p. 477: “... in the society where the capitalist mode of production prevails, anarchy in the social division of labour and despotism in the manufacturing division of labour mutually condition each other ...”

6. In the case of Elmar Altvater, who is extremely explicit in theorising a conceptual gap between “capital” and “state”, the problem is rooted in a misreading and misinterpretation of Engels’ Anti-Dühring. Engels, who admittedly did not theorise his point fully, wrote of the possibility of the centralisation and concentration of capital reaching the state itself, and gave a clear affirmative answer to the question, would such a national state capital still be capital? (Engels, 1959, p. 384). The modern state, he explained, is “essentially ... the ideal personification of the total national capital”. Altvater does not notice that Engels is talking of a total national capital, and reads him as referring to the absurd notion of a completely centralised “capital in general”. Had Engels meant any such thing, he would – as Altvater suggests – have been uttering a nonsense. But he wasn’t: Altvater was misreading him, seemingly because he himself forgot that national limits do not coincide with capitalist limits. See Altvater, 1973, esp. p. 99; Barker, 1977.

7. It is sometimes argued, against the validity of the very concept of state capital itself, that capitalist production rests on a particular form of private property relations. Far from disputing the point, I would only point out that “private property” is not necessarily limited to “personal property”. The private conduct of surplus-value production is in no sense logically opposed to nation-state capitalist production. “Private” means simply “not-social”, that is, not under the collective will and control of the associated producers. Whether capitalist production is carried out under the aegis of personal owners, churches, joint-stock companies, trusts or baboon colonies, or nation-states within a world market, is an important but secondary question. To treat it as the primary question is to reduce the concept of the social relations of production to the vulgar bourgeois sphere of concern with relations of distribution, and to add another “Marxist” head to the Hydra of reformism.

8. The conception of Russian society as a “state capitalist” formation, which cannot be defended here for reasons of space, rests essentially on the view that – from the inception of the First Five Year Plan and the forced collectivisation of agriculture – the Russian revolution must be regarded as decisively “lost”. Thus, far from being viewed with Trotsky as a “degenerated workers’ state”, or with others as a “transitional formation”, or a “bureaucratic collectivism”, or a “statist socialism”, Russia is seen as a single state capital, whose form of economic and social development under Stalin and his successors was conditioned above all by the accumulation needs of competition with the rest of world capitalism. If the theory – extended also to Eastern Europe and other “socialist” countries – has been left in a rather undeveloped state, I find it still the most promising approach to the notorious “Russian question”. The most easily available accounts are to be found in Cliff, 1970 and Harman, 1974.


Altvater, Elmar, 1973, Notes on some problems of state interventionism, Kapitalistate, 1.

Banaji, Jairus, 1977, Modes of production in a materialist conception of history, Capital and Class, 3.

Barker, Colin, 1977, On Altvater, mimeo, CSE Manchester.

von Braunmühl, Claudia, 1974, Kapitalakkumulation im Weltzusammenhang. Zum methodischen Ansatz einer Analyse des bürgerhichen Nationalstaates, Gesellschaft, 1, Frankfurt.

von Braunmühl, Claudia, 1976, Die nationalstaatliche Organisiertheit der bürgerlichen Gesellschaft: Ansatz zu einer historischen und systematischen Untersuchung, Gesellschaft, 8/9, Frankfurt.

von Braunmühl, Claudia, 1977, The nation state and the world market: on the national organisation of bourgeois society in John Holloway and Sol Picciotto, eds., The State and Capital: a Marxist Debate, Edward Arnold, forthcoming.

Cliff, Tony, 1970, State Capitalism in Russia, Pluto.

Engels, Frederick, 1959, Anti-Dühring, Moscow.

Harman, Chris, 1974, Bureaucracy and Revolution in Eastern Europe, Pluto.

Holloway, John and Picciotto, Sol, 1977, Capital, Crisis and the State, Capital and Class, 2.

Marx, Karl, 1973, Grundrisse, Penguin.

Marx, Karl, 1976, Capital, Vol. I, Penguin.

Shaw, Martin, 1974, The theory of the state and politics: a central paradox of Marxism, Economy and Society, 3.4.

Colin Barker Archive   |   Trotskyist Writers’ Index  |   ETOL Main Page

Last updated: 7 February 2019