This on-line version is the pre-copyedited, preprint version. The published version can be found here:
‘On the limits of Limits to Capital’, Antipode, 36 (3), 480-496, 2004.
My contribution to this celebration of Limits to Capital (hereafter Limits or LC) approaches it as a magisterial work that is also a ‘classic’ text. Even though Harvey recently described it, with justified regret, as the least discussed and used of his books (2000), it is a work of massive theoretical ambition and solid accomplishment. As Harvey remarks elsewhere, Limits ‘seeks to integrate the financial (temporal) and geographical (call it global and spatial) aspects to accumulation within the framework of Marx’s overall argument. It attempts to do so in a holistic rather than segmented way. It provides a systematic link between the basic underlying theory … and the expression of those forces on the ground as mediated through uneven geographical developments and financial operations’ (1999: xix). The overall success of this attempt was a major achievement in its time. It proved a defining moment in Harvey’s own intellectual project, driving many of his later theoretical reflections and empirical research. It has inspired a significant body of work on the uneven geographical development of capitalism and the contradictions involved in the built environment and spatial fixes. And it has proved a classic post-disciplinary text with a message that goes well beyond geography.
Harvey’s book is a sustained attempt to develop the basic method, extend the substantive arguments, and overcome some of the theoretical limits of Marx’s classic critique of political economy. Yet Limits to Capital has its own limits and these are often rooted in the limits of Capital itself. Let us recall that the latter is an unfinished project. In the 1857 outline of his future magnum opus, Marx stated his intention to write six ‘books’ (Marx 1973; cf. Harvey 1982: xiv). These were to deal in turn with capital, landed property, wage-labour, the state, foreign trade, and the world market and crises respectively. This order of presentation corresponded to Marx’s method of analysis, which moved from abstract-simple objects to the totality as a concrete-in-thought. Thus the world market and crises would be the ‘rich totality of many definitions and relations’ and could be introduced only after the other elements had been elaborated. Disagreement abounds over the completeness of the first three proposed books (especially the ‘missing book on wage labour’); but most commentators accept that Marx provided little more than sketches and hints about the final three. It is unlikely that Marx abandoned them as unnecessary to his project, however; rather, he seems to have consigned them to an ‘eventual continuation’ (cf. Rosdolsky 1977: 53).
Harvey’s work builds systematically on the first two projected – but never fully completed – books, benefiting from Marx’s earlier publications, rough drafts, and notebooks as well as from the three volumes of Capital. He thereby adds much to a Marxist understanding of the dynamic of capital accumulation; the nature of money; the crucial role of the credit system and interest-bearing capital; the significance of machinery and other forms of fixed capital; the specificity of landed property in capitalism; the crucial function of different forms of ground-rent in the law of value and the equalization of profit, and as a stimulus to competition; the nature of the built environment and transport; the role of the state in regulating the credit system and markets for land; the role of foreign trade in generalizing the logic of capital; and the uneven geographical development of the world market. With his skilful application of Marx’s method to the logic of accumulation, Harvey is particularly illuminating on four issues: (1) the money form and its various contradictions; (2) the credit form, the temporal fix of accumulation, and financial crises; (3) the partial, temporary spatial fixes of accumulation as capital seeks to resolve crises through geographical expansion and uneven geographical development, and switching crises; and (4) the linkages among crisis tendencies, the conflicts between capital in general and individual capitals, class struggle, and competition. In these and other respects, Limits is a major contribution to the Marxist intellectual commons. Reflecting its roots in Capital, however, it is less convincing on wage-labour, the state, and the world market. Marx himself left many issues unresolved here and in this respect, especially concerning wage-labour, the limits of Capital also limit the Limits to Capital.
Harvey provides an excellent and compelling summary of Marx’s method, as outlined in the 1857 ‘Introduction’ and further developed in the Grundrisse and Capital. This involves a movement from abstract to concrete, i.e., the increasing concretization of a given phenomenon (e.g., from commodities in general to labour-power as a unique commodity through the wage relation and the setting of the nominal wage to the formation of the real wage). It also involves a movement from simple to complex, i.e., introducing further dimensions of a given phenomenon. This can be illustrated by the movement in Limits from (a) capital in general, through (b) particular capitals differentiated according to their material role in the circuit of capital, (c) the political overdetermination of the credit system, and (d) the character of finance capital as a power bloc based on the fusion of industrial and money capital, to (e) the competition between different national capitals and their power blocs for control over new markets. This means concepts are never introduced once and for all but are developed, expanded, and refined many times. It follows that, ‘since we cannot possibly have that understanding at the outset, we are forced to use the concepts without knowing precisely what they mean (Harvey 1982: 1-2).
As the spiral of scientific enquiry continues, elements of the ‘real concrete’ are defined with increasing complexity and concreteness. Thus, while Harvey’s ‘first cut’ theory of crisis in terms of the tendency of the rate of profit to fall is relatively abstract-simple, being couched at the level of capital in general, the second and third cut theories are successively more concrete-complex. His particular contribution to concretizing Marxist analysis lies in developing the implications of ‘socially necessary turnover time’; and, to its complexifying, in elaborating the nature of finance capital and the inherent spatiality of accumulation. Given this continual and incomplete spiral movement from abstract-simple to concrete-complex, we should focus on the coherence and explanatory power of concepts and arguments relative to a given stage rather than criticize a theoretical approach just because its movement is not yet complete. This said, it is often appropriate to critique a theory to the extent that later productive development is blocked by the manner in which earlier concepts are presented. This is why, as Marx declared, beginnings are so important in science.
The Missing Book on Wage Labour
Opinions differ on the ‘missing book on wage labour’. For example, Rosdolsky (1977) believes that it was substantially included in Volume One of Capital. In contrast, for Lebowitz (1982), Capital focused one-sidedly on capital’s need for valorization and neglected ‘the worker’s own need for development’ (cf. Balibar 1985). Marx wrote to Engels that he would initially assume that wages are at their minimum so that he could focus on the nature of capital. But he also added that ‘[m]ovements in wages themselves and the rise and fall of the minimum will be considered under wage labor’ (2 April 1858). Removing this assumption would require closer attention to class struggles, their role in setting wages, and workers’ attempts to overcome capital as a barrier to their own development. Although Harvey gives more weight to class struggle in his analysis than Capital did, he nonetheless explores Marx’s account of labour-power less systematically than he does Marx’s analysis of money, credit, and fixed capital. Moreover, while he argues that, ‘[l]ike most of Marx’s key concepts, that of the value of labour power yields up its secrets only at the end of the analysis, not at the beginning’, Harvey mainly adopts the analytical framework of Capital I and does not really indicate how to move beyond it. Volume One treats the commodity simply as a product of labour – its immediate form of appearance – and defines its price in terms of the socially necessary labour time that it embodies. Labour power is apparently seen as a commodity like others. Marx modifies this general approach to commodities in Volume Three, where ‘prices of production’ and the profit form are first introduced. This enables him to distinguish between the value of a commodity (measured in units of abstract time) and its price of production (taking account of the equalization of profit rates across commodities whose production involves different ratios of constant and variable capital). Limits also accepts this crucial distinction. But this raises two serious and related questions regarding labour power: is it really a commodity like any other and is its value established in the same way as other commodities? Harvey is less forthcoming here.
It is in just this context that Diane Elson incisively distinguished between the ‘labour theory of value’ and ‘the value theory of labour’ (1979). Although Harvey recommends her work on this very question, he does not draw out its importance for labour power and the wage form. So let me elaborate its implications. The Marxist ‘labour theory of value’ argues that the value of a commodity produced in the capitalist mode of production (hereafter CMP) is set by the socially necessary labour time required for its production. Now, if the labour theory of value is applied to labour-power because it is regarded as a commodity like any other, the value of labour-power will be set by the value of the commodities required for its expanded social reproduction. Harvey initially endorses this view and, indeed, along with many other commentators, attributes it to Marx (1982: 5). But he soon admits that ‘Marx is not very helpful’ regarding the determination of this bundle of use-values (1982: 48). In contrast, a ‘value theory of labour’ does not attempt to determine the value of labour-power as a commodity but explores the preconditions and effects of capital’s treatment of labour-power as if it were a commodity (Elson 1979: 123). Harvey subscribes to this proposition too (1982: 37). Indeed he supports Sraffa’s important conclusion that ‘[s]ince labour is not a reproducible commodity in the normal sense, the wage rate becomes a variable which has to be determined outside of the technical relations prevailing within the system of commodity production’ (1982: 40). This point emerges even more forcefully in his afterword, when Harvey writes that ‘[t]he crucial commodity for the production of surplus value, labour power, is itself produced and reproduced under social relations over which capitalists have no direct control. … though labour power is a commodity, the labourer is not’ (1982: 447). He also notes that it is odd that Marx did not pay greater attention to this paradox in its multiple dimensions (ibid.). Nonetheless, Harvey does remark much earlier in LC that, because this makes labour-power a distinctive ‘commodity’, its price (the wage) contains, as Marx emphasized, a ‘historical and moral element’ (1982: 46).
This raises the question whether, in addition to having a use-value and an exchange-value, labour-power has a value that is determined by the labour theory of value. The ‘value theory of labour’ denies this because labour-power is a fictitious commodity, not a real commodity. Seen in this light, the wage, the bundle of commodities that it can buy, and the role of non-commodified goods and services (as provided, for example, through domestic labour and/or state-sponsored collective consumption) are determined in the first instance through class struggle and certain capitals’ interest in expanding the market for consumption goods (cf. Marx 1973: 409, cited Harvey 1982: 49). This does not mean that the setting of the wage level – or the exchange value of labour power – is wholly arbitrary. For, insofar as the institutional separation between the economic and political can be maintained, the economic class struggle between capital and proletariat occurs within limits set by the logic of capitalist markets. In addition, capital’s expanded reproduction requires the presence of certain proportionalities between the departments producing capital and consumer goods respectively. This also constrains the historical and moral aspects of capital’s expenditure on variable capital. Harvey seems to endorse this approach as well in noting the need to bring the discussion ‘down to earth by considering the historical processes whereby the standard of living, the value of labour power and the share of variable capital in the total social product are actually regulated’ (1982: 49). This seems to imply that levels of consumption increase or decrease in response to real wages and hence that ‘the value of labor-power has a tendency to adjust to its price — rather than the reverse!’ (Lebowitz 1991: 111). This excludes the application of the labour theory of value to labour-power itself and widens the theoretical scope for class struggle to shape capital accumulation.
Interestingly, Harvey is well aware of the need to go beyond the labour theory of value in his more concrete-complex analyses of other aspects of the circuit of capital. Thus he discusses the indeterminacies in the calculation of the value of machinery and other forms of fixed capital due to differences between historic cost, replacement cost, and current profitability and the extent to which these differences are shaped by class struggle and capitalist competition. He also recognizes the problems in determining the value of money even when it has the form of a real commodity. He notes that ‘money becomes worth what it will buy. The result: the money commodity acquires a dual exchange value – that dictated by its own conditions of production (its “inherent” exchange value), and that dictated by what it will buy (its “reflex” value)’ (1982: 11). The value of money becomes even more problematic when it is uncoupled from gold or other real commodities. Yet Harvey appears content to argue that ‘labour power as a commodity has a two-fold character: it has a use value and an exchange value. The exchange value is set, in accordance with the rules of commodity exchange, by the socially necessary labour time required to reproduce that labour power at a certain standard of living and with a certain capacity to engage in the work process’ (1982: 22). This argument seems to conflate the value and the exchange-value of labour power and to ignore an obvious parallel with the dual exchange value of money, namely, labour power becomes worth what capitalists will pay for it. This issue becomes more complicated still, of course, if one tries to apply to labour power the distinction, introduced in Capital III, between value and price of production. One of Harvey’s key contributions in Limits is his analysis of the credit form as a means to overcome the tension between value and prices of production (see especially 239-82). It would have been equally interesting to explore the role of the labour market in overcoming the tension between the value and exchange-value of labour power. All in all, it seems more sensible to treat labour power as a fictitious commodity with a use-value and an exchange-value (in the form of the wage) but without a value that is determined by the operation of the labour theory of value.
Although Harvey does allude to the problems of determining the value of labour power in Limits to Capital, he glosses over them by claiming that ‘[t]he concept of the value of labour power primarily serves to keep the idea of exploitation in the forefront of the analysis’ (1982: 46). Yet emphasizing this idea risks dehistoricizing the capital relation by focusing on its appropriation of surplus labour rather than on the latter’s form as surplus value (Elson 1979: 116; Postone 1993: 198). It also means neglecting the importance of the wage form as a means of appropriation of surplus value and, a fortiori, underplaying the struggle between capital and labour over the ratio of necessary to surplus labour (Elson 1979: 116; Lebowitz 2003). Confronting these thorny questions would have enabled Harvey to address the antagonism between capital and labour more directly not only regarding circulation but also production and, furthermore, to bring into the heart of his analysis the working class (however defined) as a political as well as economic force. It is just such an alternative departure point that Harvey recommends in his afterword to Limits (1982: 447). This reinforces the parallels with Marx’s Capital. For the principal subject in volumes one to three of Capital is the self-realization of capital and wage-labour figures only insofar as it can be utilized as a concrete force in that self-realization process (cf. Harvey 1982: 114-16). Attempting to give equal weight to wage-labour as a class for itself in this context would have generated difficult, if not insoluble, problems for Marx at this stage in his unfolding of the logic of capital. However, he could well have started from the working class in his missing book on wage labour. Its absence complicates the task of those who would overcome Capital‘s limits as a foundational text for the historical materialist analysis of capital-labour relations.
The Missing Book on the State
The confusions over the labour theory of value and the value theory of labour derive from the limitations of the three published volumes of Capital as well as from the uncertainties created by the absent book on wage labour. The missing book on the state poses different problems. For, while Marx wrote extensively on actually existing forms of state and politics in a wide range of theoretical, historical, and journalistic texts, he did not develop a suitably abstract theory of the form and functions of the capitalist type of state in the CMP. As Harvey (among many others) notes, ‘Marx intended to write a special treatise on the state but never even began the project’ (1978: 268). Likewise, although Limits contains scattered comments on the state, Harvey admits their inadequacies as the basis for a comprehensive account (1982: 449). Yet he did write an essay on state theory in preparation for LC, describing the latter as ‘a book that seems to have taken an interminable time to finish’ (1978: 268). This essay is incisive and, regrettably, Harvey has not subsequently elaborated its insights. He assumes that the state should be viewed, like capital, as a relation and/or as a process of exercising power in and through certain institutional arrangements (1978: 280). In this context, it is a useful abstract category ‘for generalizing about the collectivity of processes whereby power is exercised and for considering that collectivity within the totality of a social formation. But the state is not an appropriate category for describing the actual processes whereby power is exercised. To appeal to the category “the state” as a “moving force” in the course of concrete historical analysis is, in short, to engage in a mystification’ (1978: 280). In short, instead of treating the state as a simple instrument of class power or as a unified rational subject, one must examine its institutional forms, how they shape the political class struggle, and the latter’s transformative impact on the state apparatus. The power of this approach, which is, indeed, thoroughly consistent with Marx’s own approach, can be seen in Gramsci (1971) and Poulantzas (1979).
Harvey’s 1978 essay rests on an intelligent and skilful reading of a good range of Marx’s writings and postwar Marxist state theory. Here, albeit more implicitly than explicitly, he develops both a general, functionalist theory of the class nature of the state in class-divided societies (with the state seen as emerging to control a society split into irreconcilable class antagonisms) and a more specific, form-determined theory of the capitalist type of state (with its specific form and mode of functioning corresponding to basic features of the CMP). Thus the CMP presupposes: (a) the concept of a juridical person; (b) property rights; (c) a common standard of value in exchange, i.e., money; (d) a separation of private interests from social necessities in exchange – with these social necessities being represented in the form of the state (1978: 272-3). Adopting the same dialectical method later deployed in Limits, Harvey shows ‘that Marx’s analysis of the capitalist mode of production can be paralleled at each step by a theoretical derivation of certain minimal state functions: the equality and freedom of exchange must be preserved, property rights must be protected and contracts enforced, mobility preserved, the “anarchistic” and destructive aspects of capitalist competition must be regulated, and the conflicts of interest between fractions of capital must be arbitrated for the “common good” of capital as a whole’ (1978: 275). He also notes the state’s key role in providing ‘public goods’ that would be unprofitable for individual capitals to produce, in crisis management, and in counteracting the tendency of the rate of profit to fall (1978: 274-5). Finally, going beyond its necessary, but narrowly defined, economic functions, Harvey identifies two key political aspects of the capitalist type of state. These comprise, first, the political adequacy of the liberal bourgeois democratic form to the formal requirements of the CMP; and, second, the importance for ruling class legitimacy of exercising – as far as possible – its control through hegemony rather than coercion in a bourgeois society based on ‘individual’, ‘freedom’, ‘equality’, ‘rights’, ‘justice’, etc. (1978: 275-7). This suggests that the liberal democratic state is the most adequate political form corresponding to the basic character of capital as a social relation.
The key themes of this essay also appear in Limits to Capital. Indeed, it concludes with the following general observations:
‘We have considered all these aspects to the modern state in the preceding text. Yet they do not form an adequate basis for a comprehensive theory of the state. Too many elements are left out. The reproduction of the labourer and of labour power, the production and use of knowledge as both a material force in production and as a weapon for domination and ideological control, must all be integrated into the argument. And as we strive to complete this task, two things become apparent. First, the institutions so fundamental to the reproduction of capital (such as the central bank) are to some degree kept quite separate from those that deal with the reproduction of the labourer and labour power. But secondly, some kind of unity has to prevail among diverse institutions, some balance struck, if society as a working whole is to be reproduced. This raises questions about the allocation of powers, of legitimacy, democracy and ideology … Above all, our attention must then focus upon the political struggle for control over the state apparatus and the powers that reside therein. Class struggle is displaced from the point of production into the political arena’ (1982: 449).
In developing these ideas about the state and political struggle, Harvey successfully avoids the more abstruse arguments of the German ‘state derivation’ debate and integrates several of Gramsci’s ideas about the specificity of state power in the era of mass politics. But, for all his interest in the formal adequacy of the capitalist type of state, his arguments betray a residual functionalism. The latter survives because he tends to present the state as a necessary complement or supplement to market forces in reproducing the capital relation. Thus Harvey’s analysis here lacks his usual critical awareness of the limitations of capital’s basic economic forms. He shows little appreciation in the preparatory essay or LC itself of the real extent to which the form of the capitalist type of state (which is undoubtedly ‘internally related’ to the other forms of the capital relation) also problematizes its capacity to perform its allotted functions in capitalist reproduction. This is not necessarily to reject Harvey’s starting point. For the full meaning and significance of concepts only emerges during an endless spiral of further analysis, concretization, and complexification. Accordingly, thus one might argue that his analysis provides an appropriate set of holding concepts that await further refinements that would retrospectively validate and sublate their initial provisional content. However, as suggested above, this is not the case here (for an inspiring alternative attempt to complete the missing book on the state, see Poulantzas 1979).
The Missing Books on Foreign Trade and on the World Market and Crises
Harvey began to develop a theory of foreign trade and the world market in his arguments about spatial fixes and the general dynamics of imperialism. Having incisively analysed how the credit system promoted a provisional, contradictory, and eventually crisis-magnifying ‘temporal fix’ for capital accumulation in his ‘second cut’ at crisis theory, Harvey offers a ‘third cut’ analysis based on capital’s attempts to secure a ‘spatial fix’ to overcome the barriers to accumulation. It is quite logical, following Marx’s scheme in this regard, that the final chapter of LC introduces a ‘third cut’ at crisis theory, the role and the limits of external markets in temporarily resolving capital’s crisis-tendencies, and the core dialectics of imperialism. This chapter develops some key geographical implications of Marx’s claim that the tendency to create the world market is directly given in the concept of capital itself (1973: 408). This affects the operation of all the CMP’s tendencies and laws and also subjects all economic activities (and the places and spaces in which they occur) to the ‘audit’ of the world market. Indeed the tendency for the equalization of profit rates means that the individual capitalist ‘always has the world-market before him, compares, and must constantly compare, his own cost-prices with the market-prices at home, and throughout the world’ (Marx 1966: 336; cf. 1968: 149). But, as Harvey clearly shows, the development of foreign trade, capital exports, and a global proletariat does not lead to global convergence and homogenization. Instead, driven forward by capital’s contradictions as mediated in and through capitalist competition and the class struggle, the world market intensifies uneven development, prompts imperialist rivalries, and even risks global war as ‘the ultimate form of devaluation’. Or, as Marx put it, the world market is the conclusion ‘in which production is posited as a totality together with all its moments, but within which, at the same time, all contradictions come into play. The world market, then, forms the presupposition of the whole as well as its substratum’ (1973: 227-8). Whether or not Marx would have reached the same conclusion about the likely role of global war as the ultimate form of devaluation is debatable. But most of the other arguments in chapter thirteen of Limits are quite consistent with Marx’s general line of argument and thinking.
In particular, Harvey introduces the concept of ‘spatial fix’ to refer to ‘capitalism’s insatiable drive to resolve its inner crisis tendencies by geographical expansion and geographical restructuring’ (Harvey 2001: 24). Harvey examines the contradictions inherent in these repeated attempts to resolve capital’s contradictions as evidenced in the tension between the ‘fixity’ and ‘mobility’ of capital at any given moment and over time. This tension is found within the category of fixed capital itself (e.g., the mutual presupposition of fixed airports and mobile aircraft), within circulating capital (raw materials, semi-finished goods, finished products versus liquid money capital), and within the relation between fixed and circulating capital (e.g., commercial centres and commodity flows). It also unfolds over time. For ‘capital has to build a fixed space (or “landscape”) necessary for its own functioning at a certain point in its history only to have to destroy that space (and devalue much of the capital invested therein) at a later point in order to make way for a new “spatial fix” (openings for fresh accumulation in new spaces and territories) at a later point in its history’ (2001: 25). These new rounds of spatial fix are facilitated, of course, by innovation in the means of communication and transportation (cf. de la Haye 1988); but their particular form also depends on whether capital is seeking a spatial fix to overcome overproduction (new markets), reduction in surplus population, new materials, localized overaccumulation (new investment opportunities), etc.
Moving Beyond The Limits to Capital
The most significant contributions in Limits stem from Harvey’s use of Marx’s own dialectical method to respecify and elaborate the various economic categories and crisis mechanisms in Capital and to reveal their inherently spatio-temporal qualities. Indeed, his work is distinctive because of its strong emphasis on the centrality of the economy of time to the capital relation and on the deeply temporal nature of many of Marx’s economic categories (see also Grossman 1977, Postone 1993, Bensaïd 2002). This is especially clear in his analysis of the credit system and his ‘second cut’ at a Marxist theory of capitalist crisis. At the same time he also presented a highly sophisticated account of the spatial dynamics of the capital relation.
Nonetheless Harvey’s approach to these closely related topics in Limits can be criticized, albeit mildly, on three grounds. First, while he insists that they operate simultaneously, he treats temporal and spatial fixes as distinct. This involves more than the order of presentation – such that the treatment of the two types of fix can later be combined into a single theoretical treatment of spatio-temporal fixes – and does so for two reasons. Thus, whereas Harvey’s early interest in spatial fixes mainly reflects his training as a geographer and his continuing empirical work on cities, his later interest in temporal fixes seems to flow mainly from his growing appreciation of Marx’s critique of political economy and his subsequent recognition of the increasing autonomization of financial capital. These two sets of interests were less well integrated in Limits than in later work. And, in addition, in Limits itself, Harvey treats temporal and spatial fixes as resolving different crisis-tendencies and also argues that spatial fixes displace and defer the contradictions resulting from temporal fixes. He does not explicitly present the latter as displacing or deferring the contradictions of spatial fixes – although in the continuing and contradictory dialectic of accumulation this cannot be ruled out. A more detailed analysis would reveal the spatio-temporal complexities of both these fixes. For the credit mechanism is inextricably spatial as well as temporal insofar as the operation of credit is linked to spatially specific circuits rooted in the tension between national money and international currency; and, even more clearly (especially in Harvey’s own account), the distinction between fixed and circulating capital rests on temporal as well as functional issues. These complexities are clearly implied in Limits but they are not fully explicated in its first edition (for later developments, see Harvey 1985, 1989, 2001, 2003 and, for a critical discussion, Jessop 2004). It is important to note that the temporal fix and the spatial fix are both inherently spatio-temporal and their operation must be linked to specific spatio-temporal matrices (cf. Harvey 1999: xxiv).
Second, Harvey’s discussion of spatial fixes addresses just one of several interrelated economic contradictions. This concerns the alternating forms (or modes of being) of productive capital: as a concrete stock of time- and place-specific assets in the course of being valorized and as abstract value in motion (notably in the form of realized profits available for re-investment). His analyses of these alternating forms are premised on the competitive imperative to reduce socially necessary labour time and socially necessary turnover time and, in this context, he focuses on the dialectic of fixity and mobility in the circuits of capital. This underplays the importance of other economic contradictions (on these, see Jessop 2002: 19-22). Third, as will now be clear, Harvey’s analysis is conducted mainly in value-theoretical terms. This means it is less engaged with the limitations of these economic categories themselves and neglects capital’s crucial extra-economic dimensions.
Many attempts to develop Marx’s analysis of capitalism tend to repeat the errors of orthodox economics: they give primacy to the economy, assume a strict separation of the economic and political, propose objective economic laws that operate behind the backs or above the heads of the producers, treat individuals as crude character masks or bearers of economic forms, assume that the class struggle between capital and labour can be deduced from their economic relations alone, and suggest that the working class must become a revolutionary subject because of objective laws (Krätke 1998a: 123). Harvey avoids such problems. He follows Marx in stressing that capital’s economic laws are historically specific and are mediated through class struggles over the capital relation; he shows a clear understanding of the state’s role in securing the conditions for capital accumulation; and he notes the problems surrounding the development of class-consciousness and class action.
But, equally importantly, Marx regarded the CMP to be political as much as economic (Krätke 1998a: 125). This is clear from Marx’s planned ‘Weiterentwicklung der Theorie’ in 1847, which promised a critique of the political economy of the state concerned with taxes as the essence of the state, economically expressed; and from his later statement of intent that Capital should include a book on the state (Marx 1975: 348; Marx 1973: 227, 264). Economic laws are definitely not un- or apolitical, then, but always profoundly political. This is not surprising. For the elementary categories of the CMP – commodity, money, exchange, wage, capital – cannot be clearly determined without at least implicitly taking account of the distinctive forms of modern politics, the capitalist type of state, and the interstate system (cf. Poulantzas 1979; Rosenberg 1994; Krätke 1998a,b). Capital provides many historical reflections on the state but these do not amount to a complete economic theory of state. Nor should we expect such a theory. For the basic economic forms of the state (taxes, the national money, state credit, state spending, etc.) are also politico-juridical forms; and the state’s economic activities are themselves conducted under the primacy of the political, i.e., the importance of maintaining social cohesion in a class-divided society (Poulantzas 1979). This introduces an inescapable political dimension into an historical materialist critique of capitalism and its reproduction. This is evident not only in the basic forms and activities of the state considered as an institutionally distinct sphere of capitalist social formations but also in the state’s central role in the very constitution of the distinctive economic forms of the capital relation and the organization of the circuits of capital, including production as well as credit (on credit in this regard, see Harvey 1982: 281-2, 306-12, 321). This holds not only for individual states, of course, but also for the interstate system (Rosenberg 1994). Moreover, according to Limits to Capital, once the frontiers for ‘normal’ primitive accumulation are closed in the late 19th century, war between states becomes a new form of primitive accumulation and the ultimate means of devaluation (1982: 445).
Krätke expresses this well in discussing what is political about political economy:
‘one can express the immanent necessity of “politics” for every capitalist economy in a general formula: because the capitalist economy cannot be a “closed” and self-reproducing system, because this economic system cannot itself create some of its necessary “inputs” or elements – such as money, labour power, nature – and because its capacity for self-regulation is systematically limited, it follows that it must always be oriented to “politics”. Only in this way does it become closed and able to function as a political economic system’ (1998b: 153, my translation).
We can best understand what is involved here if we ask why capitalism cannot be a ‘closed’ and self-reproducing system purely on the basis of market relations. The answer surely lies in the indeterminate but antagonistic nature of the capital relation. This has three key aspects. First, there is capitalism’s inherent incapacity to reproduce itself wholly through the value form in a self-expanding logic of commodification. This is linked to the fictitious nature of land, money, and, above all, labour-power as commodities and, in addition, to the further dependence of accumulation on various non-commodity forms of social relations. Hence continued accumulation depends on an unstable and contradictory set of changing extra-economic conditions. Second, more concretly, these problems are reinforced by the various structural contradictions and strategic dilemmas inherent in the capital relation and by their changing articulation and forms of appearance in different accumulation regimes, modes of regulation, and conjunctures. And, third, there are conflicts over the regularization and/or governance of these contradictions and dilemmas through a variable mix of temporal fixes, spatial fixes, spatio-temporal fixes, and institutionalized compromises that help to stabilize in a provisional manner the circuit of capital and the wider social formation (Jessop 2002).
All of this implies that there is no single best way to regularize accumulation. Instead, various accumulation regimes and modes of regulation will develop their own distinctive forms of appearance of capital’s basic contradictions, dilemmas and conflicts and their own fixes and compromises. These will partially compensate for the incompleteness of the pure capital relation and give it a specific dynamic through the articulation of its economic and extra-economic elements. A key role is played here by the imposition of ‘spatio-temporal fixes’ on these economic and extra-economic elements. These operate on different scales to help resolve, partially and provisionally at best, the contradictions and dilemmas inherent in the capital relation by establishing spatial and temporal boundaries within which a relatively durable pattern of ‘structural coherence’ can be secured and by externalizing certain costs of securing this coherence beyond these boundaries. Although Harvey does not employ the notion of ‘spatio-temporal fix’ in Limits (but see Harvey 2003), he does refer to the importance of the specific ‘time-space frameworks’ in which accumulation occurs (1982: 236). And he also notes that the ‘third cut’ crisis theory assumes the co-existence of relatively closed, self-contained regions and more open spaces beyond their borders that offer opportunities for crisis-management or displacement and can be turned, within limits, into their ‘appendages’ (1982: 427). More generally, in his 1999 preface to Limits, he summarizes one of its lessons as follows:
‘Crises have no existence outside the matrix of spatio-temporalities that capitalism itself creates. Crises are as much about reconfiguring the spatio-temporal form of class relations (through all manner of stressful adjustments) as about the internal class contradictions of capitalism specified in some absolute and immutable space and time’ (1999: xiv).
Relevant spatial factors in these matrices mentioned in Limits include place-based social relations, the built environment, land markets, the rural-urban division of labour, urban hierarchies, locational policies, the inevitable territorialization of political power, and attempts to manage uneven geographical development. Harvey also refers to relevant temporal aspects, such as fixed capital and consumpton funds, and, less systematically, to the rhythms of everyday life (including the domestic sphere, individual and collective consumption), social reproduction, and the dynamics of class struggle. The resulting time-space frameworks (or, in my terms, spatio-temporal fixes) are inevitably political as well as economic and have a key role in displacing, deferring, and defusing crisis-tendencies and contradictions. They are also strategically selective, i.e., some classes, class fractions, social categories, or other social forces located within these spatio-temporal boundaries are marginalized, excluded, or subject to coercion. Beyond these boundaries accumulation is more chaotic and anarchic, lacking in structured coherence, and its impact more disruptive and exploitative as particular capitals (or their states) seek to transform external spaces into useful appendages. The overall course of accumulation will depend on the complementarity (or otherwise) of different solutions within the world market and the extent to which the resulting uneven geographical (and temporal) developments provoke increasing opposition and resistance (1982: 427).
In writing this appreciation of Limits to Capital, I have avoided a simple paean of praise for such qualities as Harvey’s fidelity to Marx’s methods of analysis and presentation, the overall clarity of his argument, and his incisive contributions to hitherto problematic or underdeveloped areas of Marxist analysis. These qualities are all now widely recognized, as this special issue of Antipode indicates. Nonetheless Limits is not without its theoretical limitations – as demonstrated by Harvey’s ‘Afterword’ to the first edition and his subsequent attempts to go further. But I have also tried to avoid unrelenting criticism – not because critique is inappropriate in this context but because it could convey the wrong message about this classic text. A classic text is one that may not provide answers that are considered adequate today but that defines the questions to be answered and points towards the solutions. Continued recognition as a ‘classic’ text is not guaranteed. Indeed, ‘in order for a text to achieve the accolade of a classic, it must typically overcome a variety of cultural hurdles; while to survive as one, it must be subjected to continual critical engagement, its concepts reformulated to meet new problems and trials’ (Baehr and O’Brien 1994: 127-8). Many of the questions and answers provided in Limits to Capital are profound but others need further development. Thus the desire of the scholarly community to engage critically with this text once again is a sign of its continuing significance in the light of the continuing uneven development of the historical geography of capitalism, its contradictions, and crisis-tendencies.
The middle way adopted above is to seek the limits of the Limits to Capital in the earlier limits to Capital and to show both how far Harvey managed to move beyond these limits and how far this work was still confined by them. I have suggested that its most basic limitations derive from Harvey’s retention of Capital’s essentially value-theoretical analysis and its one-sided adoption of the viewpoint of capital at the expense of considering the working class as an active subject with its own interests. It would be interesting and important to explore the essential role of the non-value elements that complement and supplement the law of value in reproducing the circuits of capital and the dominance of capital accumulation as a principle of societalization. Rather than challenging the basic Marxian claim that the limits to capital are rooted in the capital relation itself, this would reinforce the point that all attempts to displace or defer capital’s contradictions actually serve to reproduce them elsewhere. Hints of these insights occur in Limits to Capital in, for example, its references to domestic labour, the contradictions of state intervention, and the risks of war inherent in imperialism. But these are not developed in their own terms or reflected in an exploration of spatio-temporal fixes that takes us beyond value-theoretical issues. Interestingly, Harvey’s more recent work has been moving in this direction, illustrating the importance of the continuing spiral movement in theoretical development typical of the best Marxist analyses.
 While Limits was a ‘beginning’ for new rounds of research as well as presenting the results of ten years’ work (Harvey 1982: 446), this article ignores the many fruitful ways in which Harvey has built on and/or moved beyond its arguments. This is a topic for another paper (see, in part, Jessop 2004).
 Harvey himself correctly contrasts Ricardo’s ahistorical labour theory of value with Marx’s analysis of the specificity of labour in capitalism
 Harvey suggests here that Marx holds the bundle constant in order to show that, if the value of that bundle of use-values falls, the value of labour-power can fall without any detriment to workers’ standard of living (1982: 48; cf. Lebowitz 2003: 110-112).
 Given the continuing, spiral development of Marxist analysis, this is not problematic in itself: subsequent moves could well lead to the integration of non-value aspects of spatial fixes provided that such moves are not foreclosed.
 Harvey (2003) introduces ‘spatio-temporal fix’ to help periodize imperialism and explain the overall logic of the latest phase in American economic, political, and military domination of the world system.
Baehr P and O’Brien M (1994) Founders, classics and the concept of a canon. Current Sociology 42 (1):1-151
Balibar E (1985) The notion of class politics in Marx. Rethinking Marxism 1 (2):18-51
Bensaïd D (2002) Marx for Our Times. London: Verso
de la Haye Y (1988) Marx and Engels on the Means of Communication. New York: International General
Elson D (1979) The value theory of labour. In idem (ed) Value: the Representation of Labour in Capitalism (pp 115-180). London: CSE Books
Gramsci A (1971) Selections from the Prison Notebooks. London: Lawrence & Wishart
Grossman H (1977) Marx, classical political economy, and the problem of dynamics. Part One. Capital and Class 2:32-55
Harvey D (1978) The Marxian theory of the state. Antipode, 8 (2):80-89
Harvey D (1982) The Limits to Capital. Oxford: Blackwell
Harvey D (1985) The geopolitics of capitalism. In D Gregory and J Urry (eds) Social Relations and Spatial Structures (pp 128-63). Basingstoke: Macmillan
Harvey D (1989) The Condition of Post-Modernity. Oxford: Blackwell
Harvey D (1999) Introduction to the Verso edition. In The Limits to Capital (pp xiii-xxviii). London: Verso
Harvey D (2000) Reinventing geography. New Left Review (n.s.) 4:75-86
Harvey D (2001) Globalization and the ‘spatial fix’. Geographische Revue 2:23-30
Harvey D (2003) New imperialism: accumulation by dispossession. In L Panitch and C Leys (eds) Socialist Register 2004 (pp. 63-87). London: Merlin
Jessop B (2002) The Future of the Capitalist State. Cambridge, UK: Polity
Jessop B (2004) Spatial fixes, temporal fixes, and spatio-temporal fixes. In N Castree and D. Gregory (eds) David Harvey: a Critical Reader (in press). Oxford: Blackwell
Krätke M (1998a) Wie politisch ist Marx’ politische Ökonomie? (I). Zeitschrift Marxistische Erneuerung 33:114-128
Krätke M (1998b) Wie politisch ist Marx’ politische Ökonomie? (II). Zeitschrift Marxistische Erneuerung 34:146-161
Lebowitz M (1982) The one-sidedness of Capital. Review of Radical Political Economics 14 (4):40-51
Lebowitz M (1991) The significance of Marx’s missing book on wage-labor. Rethinking Marxism 4 (2):105-118
Lebowitz M (2003) Wages. In idem, Beyond Marx (pp 101-119). Basingstoke: Palgrave
Marx K (1966) Capital vol I. London: Lawrence & Wishart
Marx K (1968) Capital vol II. London: Lawrence & Wishart
Marx K (1973) Grundrisse. Harmondsworth: Penguin
Marx K (1975) Capital vol I. Harmondsworth: Penguin
Postone M (1993) Time, Labor, and Social Domination. Cambridge, UK: Cambridge University Press.
Poulantzas N (1979) State, Power, Socialism. London: Verso
Rosenberg J (1994) The Empire of Civil Society. London: Verso
Rosdolsky R (1977) The Making of Marx’s Capital. London: Pluto Press