determinant role of class struggle and to the absence of any 'iron laws of motion of capitalism*. 1 Introduction. The geographical unevenness of capitalist ...
Environment nncJ Plonnlnfj D: Society nnd Spnco, 1904, volumo 2, pngos 329-345
The geographical transfer of value: notes on the spatiality of capitalism1' C Hadjimichalis School of Technology, University of Thossaloniki, University Box 491, Thossaloniki, Grooco Rocolvod 21 April 1983; In revised form 14 March 1984
Abstract. Marxist research on the geographical unovenness of capitalist development remains highly heterogeneous. Although virtually all Marxists agree on certain general concepts, such as capital accumulation, agglomeration in space, spatial division of labor, capital mobility, the role of the state, etc, there is strong disagreement on how these processes operate and on the nature of their impact on various places. In this paper, a critical evaluation of two dominant lines of thought, the autonomous or semiautonomous development thesis and the transfer-ofsurplus-development thesist is presented, A third approach, based on the notion of the geographical transfer of value, an attempt to use the labor theory of value in geographical terms, is also introduced. This view is further elaborated by focusing first on the articulation of production and circulation, second on state and local state intervention in this articulation, and third on the contradiction between equalization and differentiation. Throughout the paper, a special emphasis is given to the determinant role of class struggle and to the absence of any 'iron laws of motion of capitalism*. 1 Introduction The geographical unevenness of capitalist development has many sources. Our inability to distinguish among separate forces operating simultaneously is to a large extent responsible for the present confusion surrounding the question of uneven development over space. In this paper, in which I synthesize some of the ideas contained in a forthcoming book (Hadjimichalis, forthcoming), I aim to challenge the current understanding of the relationship between the Marxian theory of value, development, and space. More specifically, I will try to go beyond immediate appearances of geographical phenomena by investigating their underlying mechanisms and tendencies. To begin the analysis with issues such as capital mobility, agglomeration in space, spatial division of labor, development of productive forces, etc is a necessary, but too elementary, approach. Instead, I will try to use the labor theory of value in geographical terms, in a framework dominated until recently by aspatial categories and processes. In so doing, I propose an alternative framework for possible future research in the field, based on the notion of the geographical transfer of value (GTV). Taken singly, the ideas presented here are not entirely original. Indeed, my work is influenced by, although I am not in agreement with them, theorists such as Gramsci, Lefebvre, Emmanuel, Shaikh, Lipietz, Soja, de Janvry, Cooke, and others. This paper could be described as an effort to make explicit the influence of certain conditions which are 'space-specific' and 'particular to a location' in the production, circulation, and realization of surplus value. I am aware that GTV is or could be a very controversial issue. Its association with the Marxist theory of value and 'surplustransfer theories' (in particular, unequal exchange), and its dialectical focus on sociospatial processes, could make it highly vulnerable to the criticisms of neo-Ricardians, unequal exchange critics, and to those who still approach space as a simple and t An earlier version of this paper was presented at the workshop on 'Capital Restructuring and the Regional Question* in Copenhagen, May 1982, and at the '29th North American Meeting of the Regional Science Association' in Pittsburg, November 1982.
330
C Hadjimichalis
passive container of productive forces. I will try to defend my thesis, first by focusing on the articulation of production and circulation, and second by arguing for the important role of class struggle and the state inscribed in this articulation. 2 Autonomous versus transfer-of-surplus-development theories During the 1960s and 1970s, Marxist research on the geographical unevenness of capitalist development succeeded in establishing itself as an alternative analytical framework to neoclassical and liberal theories. Initially built upon general laws and abstract formulations, approaches 'borrowed' from international development theories, and theories of capital accumulation, geographical analysis using Marxist tools and methodology was and remains highly heterogeneous (Massey, 1978; Soja, 1983). Though virtually all Marxists call upon these processes to explain uneven development between firms, sectors, and regions, there is strong disagreement on how such development operates and on the nature of its impact on various places. With the risk of oversimplifying, I will identify two dominant lines of thought within this heterogeneous intellectual tradition. The first, based on the process of capitalist production, addresses the self-expansion of capital and puts forward the autonomous or semiautonomous development thesis. In this case, different geographical areas (however defined) grow or decline depending on concretely different production processes and types of product manufactured by firms located in these areas, and on different use-value circumstances of capital and labor operating in these places (Harvey, 1975; Walker, 1978). Development is dependent upon the putting together of the necessary conditions for profitable accumulation (mainly via exports). Sectoral composition and dynamics are major causal forces shaping regional differentiation (Markusen, 1983). This line of thought originates from Lenin and Bukharin, who centered their analyses of uneven development on the contradictions of accumulation that exist within certain capitahst areas (Lenin, 1973; Bukharin, 1973). In their view, it is possible for various capitalist sectors to grow relatively autonomously from other sectors: this offers local investment opportunities for capital, but also leads to the accumulation of enormous masses of capital that drive down the rate of profit. As Lenin (1973, pages 73-74) observed, "... surplus capital then seeks 'your areas' external to the capitalist sphere: in these areas profits are usually high, for capital is scarce, the price of land is relatively low, wages are low, raw materials are cheap". This combination of different rates of profit in different areas is also analyzed by Harvey (1975), who uses Marx's concept of the 'annihilation of space by time' to emphasize capital's goal to lower its time of circulation and to speed its self-expansion. In an attempt to use the profit-cycle model, Markusen (1983) argues that super profits in certain firms are a return to capitalists resulting from lower production costs (socially necessary labor time) or from the reduced time a household spends in reproducing labor power. Along similar lines, Lapple and van Hoogstraten (1980) suggest that the 'fundamental causes of unevenness' are the continuous expansion of surplus-value production and the increase of productivity, all occurring within the same firm at the same place. Similarly, Webber (1982) argues that the regional question concerns the spatially uneven development of branches of production, with agglomeration economies providing the immediate cause of such uneven development. Four serious weaknesses are identifiable in the autonomous approach to uneven geographical development. First, although it focuses on accumulation, it denies the importance of circulation and exchange. Rather, production is the driving force for development, and anything else is subordinate to it. Second, and unsurprisingly, regional development is conceived as an outcome of the autonomous growth of local productive forces only, thus denying any relationship between sectors and firms in
Tho fincwjrnphical transfer of value
331
different geographical units in the outside world, Third, because the main focus of abstraction is upon the logic of production and accumulation, there is little reference to the role of class mobilization, political action, and the state, And fourth, there is a tendency to regard everything that occurs in space and time as conducted on behalf of capital, ending either in cyclical arguments or in economic rcductionism. The second approach places major emphasis upon the sphere of capital circulation and exchange, ignoring (in some cases) production relations. Through circulation and exchange, surplus produced by workers in some areas ('the periphery*) is extracted and transferred (via a number of mechanisms) to benefit capitalist development in other sectors and areas ('the core*). This approach is called the tramfer-of-surplmdevelopment thesis (Amin, 1978; Baran and Swcezy, 1970; Emmanuel, 1969; Liossatos, 1979; Lipictz, 1977), Development is said to be dependent on the strength (economic and political) of the core's social classes, which are able to accumulate wealth and power while the periphery is 'drained* via surplus transfer. Bauer (1924) was a direct supporter of the surplus-transfer thesis, arguing (against Lenin and Bukharin) about the appropriation of surplus value from less-developed areas in cases where these areas have less capital* compared to developed areas. Four years later, in 1928, Kuusinen, at the 6th World Congress of the Comintern, defended a thesis that capitalist penetration in the periphery of the world system under the aegis of imperialism was not only a source of surplus extraction to the benefit of the center, but also created a 'bottleneck* to industrialization and resulted in stagnation (Kuusinen et al, 1973). After World War II, these ideas acquired wider theoretical legitimacy both through the work of Baran and Sweezy (1970) and through the 'development-of~underdevelopment*, 'dependency*, and 'core-periphery* schools of thought (Carney et al, 1975; dos Santos, 1970; Frank, 1969). The surplus-transfer concept, however, gained its major theoretical status through the writings of the Greek Marxist economist Emmanuel (1969) in his book on 'Unequal Exchange* and through the long debate that this book sparked off. Emmanuel distinguishes three forms of unequal exchange: first, unequal exchange in the 'broad sense', when only differences in the organic composition of capital exist; second, unequal exchange in the 'narrow sense', with differences only in wages; and third, unequal exchange in the 'strict sense*, with differences both in organic composition and in wages. Amin (1978), Emmanuel (1969), and Mandel (1975) are aware that, within a given social formation, redistributive mechanisms could wipe out the inequality of exchange. This seems, however, a very optimistic expectation, as the research of Lipietz (1977), Liossatos (1979), Ferrao (1982), Perrota (1981), and Bajec (1983) has shown. Interregional polarization develops as extroverted, as opposed to introverted, accumulation is taking place, or as differences between the organic composition of capital and of labor increase. The main problem with these accounts of uneven geographical development is their incomplete abstraction of the determinants of the processes to be explained. First, there is an overemphasis on the sphere of circulation and exchange, contrary to the autonomous development thesis. Second, their prediction of a continuous 'underdevelopment or stagnation' of the periphery through surplus transfer was simply wrong, as international and regional experience has shown. Third, they use an unclear definition of exploitation via external factors and exchange, thus ignoring local social relations and class struggle. This rests upon a mainly technological explanation of the sources of increased productivity and, hence, of wage increases. Fourth, they assume equilibrium in the capital market (internationally and nationally), which implies that capital movements (the more mobile factor) have already resulted in a world or regional equalization of the rate of profit. And fifth, some of the authors, adopting a functionalist categorization between core and periphery, failed to
332
C Hadjimichalis
identify social classes within particular areas, thus arguing indirectly about place exploiting place. It seems, therefore, that although considerable progress has occurred in the theorization of the process of uneven geographical development—because of a deeper understanding of the underlying processes associated with capital accumulation and circulation—that such progress may be undermined by an emerging dogmatism in the two dominant lines of thought. Struggling to be serious and rigorous in their application of Marxist methods, many (on both sides) began to establish certain boundaries beyond which a Leftist analysis could not reach. The process seems to be analogous with what Soja (1980, page 208) has described in his analysis of the sociospatial dialectic: "... instead of sensitively probing the mix of opposition, unity and contradiction which defines the socio-spatial dialectic, attention has too often been drawn to the empty question of which causes which or to endless arguments about preeminence". Instead of focusing upon a fundamental linear idea of capital accumulation—either autonomously or via surplus transfer—I propose a third view, which focuses both upon production and exchange and upon local social classes and the state. Theories which treat exchange as the dominant source of exploitation and the principal factor of development in regions have been found inadequate. However, it has also been seen that a focus on production requirements and accumulation imperatives, to the exclusion of exchange relations, tends to be closely associated with an inappropriate economic reductionism (Cooke, 1983). What must be clearly demonstrated then is how and in what ways production, circulation, and exchange relations connect accumulation to specific localities, and whether a relation could be established among these localities based on the labor theory of value. Before I pass to this level of analysis, a short comment on the labor theory of value and the unity of production and circulation is necessary. 3 A short comment on the labor theory of value and the articulation of production and circulation The debate on whether Marx's labor theory of value (LTV) is necessary in the analysis of capitalism, and whether it is irreconcilable with the actual relations involved, is an old one. It has acquired great popularity through Sraffa's (1960) work and more recently through that of Steedman (1977) and the neo-Ricardian school. Initially confined to a relatively narrow grouping of economists, the controversy about value theory has spread to wider circles on the left. It is beyond the scope of this paper to provide a detailed review of the ongoing debate. Instead I will highlight some of the points relevant to the discussion of the geographical transfer of value (GTV). Theories adopting the Sraffian perspective argue that profits can be considered as a direct consequence of only two factors: the sociotechnical conditions of production and the real wage paid to workers. Other factors of profit have an effect only by virtue of their influence on these two factors. This argument is based on a mathematical analysis of the necessary conditions for formally calculating profits from a set of initial conditions. The categories of the LTV, neo-Ricardians argue, do not enter into this calculation at all (Hodgson, 1981; Steedman, 1977). The critical point of this model, as Wright (1978) maintains, is not only to reduce the theory of profits into a simple two-factor account, but rather to argue that other causes have their effect on profits via real wages and technical conditions. Thus class struggle, political conditions, the state, ideological hegemony, and—as I go on to argue later—space could not have effects on aggregate profits. In accordance with Wright's criticism, Shaikh (1980; 1981) emphasizes the fact that if one's interest in
Tho gooorophicnl transfer of voluo
333
studying profits is limited to calculating profits, then the nco-Ricardian level of abstraction is as far as one needs to go, This is evident, for example, in Scott's (1980) otherwise very interesting attempt to use nco-Ricardian concepts in studying urban land. Scott adopts Stccdman's approach in treating real wages as exogenously determined if the economy is to yield determinate prices and profits, But Scott fails to recognize that problems such as the spatial differentiation of the rate or profit, capital polarization in space, and changes in urban land (location, distribution, etc) are not restricted to technical changes only. If, on the other hand, one's interest is in understanding the social determinants (or structural limitations, according to Wright (1981)) of profits, one needs to move to a higher level of abstraction. This is precisely what the Marxist model of profit determination attempts, using the LTV. As Lipietz (1981) recently proved, the concept of value, including the magnitude of value, illuminates the whole qualitative and quantitative analysis of price relations, uncovering relationships where the neoRicardians see merely discrepancies. An important problem, which is directly related to my discussion of GTV, is the failure of neo-Ricardians to make the distinction between value and realized value. This rests on their inability to recognize the difference between abstract labor and socially necessary labor (compare Himmelweit and Mohum, 1978). Marx (1976; 1981), on the other hand, is explicit about this difference in Capital (Volumes 1 and 3), where he makes a clear distinction between use-values produced for direct use and converted into commodities only when exchanged, and use-values produced for exchange and hence produced as commodities (Shaikh, 1981). Indeed, exchange is a process in which the different labor-times involved in producing these use-values actually confront each other and are eventually articulated into a social and spatial division of labor through the medium of money prices. But can we say that the mass of surplus value gets bigger or smaller depending only on relative money prices? Shaikh (1981, page 300) answers this as follows: "... consider a crisis in which so little of the social product is sold that profit is actually negative (this is a recurrent real phenomenon in capitalism). Are we then to say that even though workers were exploited and a surplus product produced surplus-value is itself negative? If we are not allowed recourse to the distinction between value produced and value realised, then of course surplus-value is no longer connected to any rate of exploitation at all. It is merely an epiphenomenon of circulation. And so what begins as a tactical capitulation to the neo-Ricardians turns into a rout". Shaikh's interpretation points directly to the articulation of production and circulation. Such an articulation derives from the fact that, after the act of production, surplus value is embodied in the value of commodities that must circulate (that is, be sold) for surplus value to be realized as money and for capital to grow. In this competitive context, the necessary relation between production and circulation is also contradictory: circulation is the negation of production, and production is the negation of circulation (de Janvry, 1.981; Lipietz, 1983). The essence of economic crises is to a large extent found in this contradictory relationship. The class nature of capitalism implies the existence of interclass and intraclass conflicts regarding the production and distribution of economic surplus and, in particular, the determination of profit made by capitalists, versus their costs. Moved by the profit motive and class conflicts, capital is thus driven to expand in the sphere of production, and simultaneously encounters a barrier in the sphere of circulation (Harvey, 1983). The articulation and contradiction between production and circulation results in the periodic emergence of unsold commodities, the selling of commodities below their value, or an increase in circulation time.
334
C Hadjimichalis
Thus the concept of value cannot be seen as a production theory, as in the neoRicardian interpretations, since in the absence of circulation—that is, of sale and exchange—there is no creation of value at all. The transformation of private into social labor occurs only through exchange (de Vroey, 1981). On the other hand, however, this is not a pure circulation-exchange theory, because once there is a sale, and thus a creation of value, the magnitude of value depends on the average social conditions of production prevailing at the place of production. Exchange creates value, but social production determines the magnitude of value. 4 The state and the local state In this section, I present a condensation of a short theoretical analysis of the relationship between the state and capitalist spatial development. The parts of the argument which are exposed here focus on the particular role of state intervention, mainly in the spheres of circulation and reproduction. I pointed out in section 2 that one major inadequacy of autonomous and transferof-surplus-development theories was their limited concern with the state and the local state, when the latter exists as part of an elected administration system. In a number of theories, the state is viewed as an instrument in the hands of the capitalist class. It seeks to manipulate workers' resistance, capital demands, natural resources, etc through specific policies to the benefit of capital (Miliband, 1974; Toft Jensen, 1982). Close to this instrumentalist view is the state-monopoly-capital thesis (Boccara, 1971; Cherpakov, 1969). In this, the modern capitalist state, after the evolution of some internal laws, has passed from the competitive to the monopoly stage of the economy and is now totally dependent on big monopoly capital. Central and local state policies are mainly oriented to secure high rates of profit for monpoly capital, both nationally and internationally (Damette, 1980; Lojkine, 1977). Without adopting the simplicity of instrumentalist and state-monopoly-capital theories, capital-logic theories view the state as a 'fictitious collective capitalist', the principal role of which is to respond to capitalist interests (Alvater, 1973). The state is mainly responsible for the material reproduction of the means of reproduction in order to secure social cohesiveness and to promote international expansion for national capital. Contrary to the previous theories, capital-logic theories allow for a certain autonomy of the local state derived from its conflict-diversification function performed on behalf of the central state (Cooke, 1983; Dear and Clark, 1981; Hirsch, 1981). But they share similar problems with instrumentalist and state-monopoly-capital theories: they read social action through purely economic categories. These views have been criticized for their inadequate theorization of the class character of the capitalist state and for their functionalist and reductionist vision of capitalist development (Cooke, 1983; Poulantzas, 1978; Urry, 1981). As such, they require a holistic concept of civil society to overcome their tendency to reduce political relations and actions to economic relations and decisions. The idea that civil society has a specificity of its own which assists the structuration of political action in particular localities is first found in the work of Gramsci (1971; 1973; 1975). Gramsci elaborated a new conception of hegemony and civil society and was able to break with permanent duahsm and economism, expressing thereby an appropriate emphasis on the political. In his emphasis on the 'ensemble of relations', Gramsci's critique of what he called 'absolute historicism' focused attention on the political, ideological, and cultural problems of capitalism and upon the role of the state. These problems were concretized in time and space, in history and geography —the indispensable conjunctural framework, or the material basis upon which revolutionary strategy had to rely. As Paggi (1979) argues, Gramsci represents both a continuity and a rupture with Leninism. On the one hand, he extends Lenin's
Thu ooorjniphicnl tninnfor of value
335
major argument that history does not proceed via a simple functional logic and that every articulation and radical change depends on a political intervention by the revolutionary subject. On the other hand, he does not confine the primacy of the political to revolutionary conjunctures only, but makes it the articulatory principle of every social situation. It is precisely this Gramscian notion of the political as the articulatory principle that lies at the heart of the present problem of uneven geographical development in most European countries. Such a conjuncture is better theorized in the so-called 4 class theories' and nco-Gramscian theories of the state (Dulong, 1976; Glucksmann, 1979; Laclau, 1977; Offe, 1974; Poulantzas, 1976; 1978). In these two approaches, the local state is not conceived as a simple agent of the central state, subject to its mandate and sharing its burden of administration. The local state may at certain times experience autonomy, but is constrained by the limits which local and 'national' class relations both impose (Dulong, 1976). The state, in the latest of Poulantzas's (1979) interviews, is itself a relation, and (to be more precise) the condensation of class relations. It is neither an 'entity' on its own, nor totally a part of modern capitalist society; it is a field of class struggle, a product and a determinant of contradictory social and spatial relations. In the neo-Gramscian tradition, further emphasis is given to the role of ideological hegemony as a means of securing the cohesivencss of the capitalist system (Laclau and Mouffc, 1981). In these theories, the local state could become separate from the central state when the former is not reproducing capitalist hegemony. Dominant classes must secure the acceptance of the existing mode of exploitation by dominated classes through a continuous displacement—in time and space—of popular demands. These demands could produce a series of crises, which could vary from particular crises (such as a legitimation crisis, a representation crisis, a regional crisis, etc) to the more general crisis of the state itself (Poulantzas, 1976). These two theories, and especially the neo-Gramscian view, provide us with an important distinction between the economic, the political, and local civil society. It is precisely the difficulty of making the separation, and, at the same time, the connection among these three categories, that relates the local state with capitalist development. As I discussed in the previous section, the unity between production and circulation (or exchange) rests upon the nature of the commodity form, which only releases the surplus value it contains when that value has been realized on the market. The nature of the relationship between labor and capital is one of exchange: surplus value is 'injected' by workers into the commodity at the point of production, but labor receives less than this in the form of wages. The sphere of production relations per se is an area into which the state, either central or local, has the greatest difficulty in intervening. When it does so, it is always a reflection of class struggle, and usually takes the form of an indirect intervention, mainly via legislation. The state and the local state, however, do respond directly to pressure from capital and/or labor in the sphere of circulation and reproduction, of which exchange is a part. The clearest examples of this are those aspects of policy which can be broadly included under headings such as regional incentives to capital and labor, agency assistance, zoning, means of collective consumption, and physical infrastructure. Certain implications of these spatially differentiated components of circulation and reproduction are considered in the next section. For the moment, it is necessary to emphasize that the growing state intervention in circulation and reproduction changes the whole process of indirect regulation performed by the LTV. This requires a modern conceptualization of the LTV to account for cases in which the state (through price regulation, norms of exchange, or special trade legislation) could weaken or strengthen the 'pure' LTV (Wright, 1981).
336
C Hadjimichaiis
Though the commodity system remains anarchic—there is no explicit and direct allocation of social labor and of the means of production—a certain social cohesion is required to ensure the reproduction of capitalist relations. This occurs through a posteriori norms via the LTV and through contradictory policies introduced by the state and the local state. 5 The contradiction of equalization and differentiation During the process of accumulation and competition for higher profits, all capitalists do not succeed equally; the cost of failure is also unequally distributed among different sectors, regions, and localities. Different sectors and different firms within the same sector, located in the same or separate localities, have different rates of profit. These differences arise from certain conditions of capitalist development which are unevenly distributed across sectors and regions and which can be distinguished in three major categories: (1) conditions which affect the sphere of production, that is, the production or extraction of surplus value; (2) conditions which affect the sphere of circulation or exchange,.that is, the realization of surplus value; and (3) conditions which affect the reproduction of dominant social relations and in particular the reproduction of the labor force. I have already discussed the intervening role of the state and the local state, particularly in the sphere of circulation and reproduction. These conditions are part of the so-called 'total capital' available for each individual capitalist, which is to a large extent 'space-specific' and 'particular to a location'. Some of these conditions constitute what Harvey (1981) calls a 'spatial-fix', such as workers' housing, urban and regional infrastructure, social facilities, and other means of collective consumption. Every individual capitalist enterprise thus tends, through competition and state intervention, to improve its own 'internal' dynamic in terms of technology, labor process, restructuring, etc; it also improves its own 'external' dynamic in terms of these localized conditions, so that—at least temporarily—it enjoys higher than average social profits. In addition, collectively as a class, capitalists form national and regional associations and, through the state and local authorities, intervene in issues of common interest. Although these fundamental concepts and processes are quite familiar, their contradictory implications within the geographical setting have seldom been explored. One reason for this may be that very often radical analysis has narrowly focused on one aspect of the problem only. I will argue here, however, that such a partial viewpoint mystifies and obscures tendencies and contradictions that permit the production of surplus value to diverge from its geographical distribution (Harvey, 1983). Such an approach requires a fresh look at these localized conditions, not as simple attributes of place, but as 'active moments' in the geographical organization of capitalism. The improvement of the whole tissue of antagonistic conditions in all sectors and regions of a social formation is the driving force behind capitalist development and expansion, and it is achieved through competition and state intervention. Through these improvements, general conditions of social production across regions and sectors tend to equalize towards a social average, producing at the same time an homogeneous space. As Marx (1981) observed in Capital (Volume 3), profits tend also to approximate the social average; in other words, a tendency towards equalization of the rates of profit is in operation. This tendency, however, creates problems for capitalist development. The very tendency towards profit equalization and spatial homogenization minimizes capitalist expansion, since no excess profits can be obtained with equal rates of profit in all sectors and all regions. No more accumulation will be possible, except that necessary
Tho geographical transfer of vnluo
337
for demographic growth, and this itself would be modified in its own turn by the impact of the severe economic stagnation that would ensue (Emmanuel, 1969). Furthermore, the equalization tendency threatens the reproduction of capitalist social relations, which are founded on the unequal distribution of the social product, both between capitalists and workers and among capitalists themselves. A very effective counterbalance to the equalization tendency has been a combination of class strategies, by individual capitalists and the state, towards differentiation across sectors and regions of the very conditions described earlier, For example, the introduction by an individual firm of technical innovation could differentiate it from others and restore its profitability; a managerial strategy to increase wages in a certain firm or region above the national agreement with the state creates 'barriers of entry* to other less-powerful firms in the same sector, or pushes others to cease production; an allocation of public investments to specific regions increases their competitiveness in attracting private capital vis-iVvis other regions; and the introduction by the state of a special land-use legislation or the construction of a new highway system assigns specific and differentiated roles to every territory, so that a parccllizcd and fragmented regional space is produced (Lcfcbvrc, 1974; Lipietz, 1977; Poulantzas, 1978). These observations help to clarify another tendency, namely that towards differentiation of the rate of profit. The organization of space is a very important element in the tendency towards differentiation. Some spatial differentiation is inevitable in any mode of production; it is derived, on the one hand, from the simple friction of distance, and, on the other, from the basic principle of agglomeration versus dispersal in human spatial organization. These transhistorical spatial characteristics have been used by capital and the state to produce a disarticulated and fragmented regional space based on the relative mobility of capital, the relative immobility of labor, and the territoriality of certain general conditions of social production. As Harvey (1983, page 441) observed: "The homogeneity towards which the law of value tends contains its own negation in increasing regional inequalities. All kinds of opportunities then arise for competition and unequal exchange between regions .... The result is a chaos of confused and distorted motions towards both homogeneity and regional differentiation". Taken together within the framework of uneven geographical development, these contradictory relations can be summarized and focused around the principal contradiction of spatial differentiation and spatial equalization in capitalist development. This contradiction is specific to the capitalist mode of production and central to capitalism's spatial problematic. Whether we refer to the internationalization of capital within a world system, to regional inequalities within a particular social formation, or to the political economy of urban space, uneven regional development arises from these two simultaneous and opposing tendencies, and acts back upon them in terms both of the logic and strategy of capitalist accumulation and of the logic and strategy of working-class struggle. The broad significance of the contradiction has been elaborated by a number of Marxists. Palloix (1975) has been perhaps the most explicit analyst of equalization and differentiation on an international and regional scale, showing the connection of the contradiction with Lenin's law of uneven development'. Emmanuel (1969) and Mandel (1975) emphasize its economic implications, and Lefebvre (1974; 1976) puts his emphasis on the spatial dimension of the contradiction, especially at the urban scale. Iipietz (1980) discusses the 'perpetual struggle' between 'inherited' and 'projected' space in regional development. Finally, Poulantzas (1978) identifies the linked pairs of paradoxical oppositions, such as fragmentation versus homogenization, division versus unification, and parcellization versus structuring in connection with the key role of the state.
338
C Hadjimichalis
The combination of differentiation and equalization is not some dialectical sleight of hand designed to absolve those who have been criticized for leaving unresolved the question of whether capitalism increases or decreases regional inequalities, and whether it homogenizes or differentiates in its spatial extension (Soja, 1981). Capitalism can do both, and can do both simultaneously. But this happens not as the result of some inner logic of capital, or of some laws of motion of the capitalist mode of production', but as the result of an intervention by a political agent producing changes in the character of the dominant social relations of production, exchange, and consumption (Dulong, 1978; Poulantzas, 1978). This criticism applies to Harvey's (1983) economistic emphasis in The Limits to Capital, in which politics, ideology, class struggle, and the state are almost absent from the discussion. The movement towards equalization or differentiation is not a matter of free choice by capital and the state. It is the outcome of class struggle between capital and labor over the production of surplus value, and of competition between capitalists over the distribution of surplus value. The reproduction of the general social conditions is an 'active moment' within the overall accumulation process, and is bound up with the perpetual motion in the spatial configuration of capitalist social relations. Thus the reproduction of these conditions becomes a political necessity via the intervention of the central state and the local government (Folin, 1981; Lojkine, 1977; Preteceille, 1981). This is why issues such as capital restructuring, regional and urban decline, and 'the regional crisis' are part, not only of the general crisis of capital, but also of the crisis of the state in virtually every European capitalist social formation. 6 The geographical transfer of value It is now possible to return to a discussion of the geographical transfer of value (GTV) and to 'locate' it within the framework of capitalist social relations and the contradiction inherent in the equalization-differentiation debate. GTV can be defined as the tendency or process through which the value produced for exchange by workers at one locality: (a) is realized by the capitalists who have employed these workers, but profits have been reinvested elsewhere; and/or (b) is at least partially realized by the capitalists who have employed these workers and is added as excess profits to other capitalists in other localities. The first case could be referred to as direct and/or derived GTV, whereas the second could be called indirect and/or originating GTV (Hadjimichalis, 1980; Soja, 1981). In both cases, however, profits have been added to the accumulation process enjoyed by other capitalists in other areas. A number of clarifications are necessary. To begin with, GTV differs from the autonomous and transfer-of-surplus-development theses because, first, it takes production and circulation as a dialectical unity; second, it is not an automatic process, but it is conditional, that is, its operation is dependent on certain conditions which are analyzed below; third, it is not the only or even the major factor influencing development, but one among many; and fourth, its relative weight depends on a concrete time-space conjuncture. Another important clarification is the definition of 'locality'. 'Local' refers to the importance of local variations in action, consciousness, and conditions, with a labor market playing the crucial role (Cooke, 1983; Duncan and Goodwin, 1982). 'Locality' is not a predefined sociospatial unit, but it must result as an effect from the analysis of local labor market characteristics and boundaries (Kayser, 1970). Finally, GTV is not the simple geographical component of the sectoral transfer of value analyzed by Marx. If this were the case, we would end with a contextual view of abstract space and the argument would be a cyclical one. The peculiarity of GTV
UuMioofjniphiail tmnnfnr of vnlun
339
derives from its connection with concrete space materialised in sociospatial relations which are particular to a place. This connection with material space is evident in three important characteristics: the local class structure, the local labor market, and local general conditions of social production. These three characteristics, differentiated or equalized among localities, are the preconditions for GTV to occur, and they are controlled or produced by the central and/or the local state. In the following, I shall consider how these characteristics could be classified as local* (most of them are considered as 'national') and how they are related to the concept of GTV. I begin by noting that historical changes in contemporary capitalist relations have increased the importance of local class structures (Duncan and Goodwin, 1982; Urry, 1981). The accelerated derealization of economic relationships has been followed by an increasing 'poiiticization' of economic activity through conflict affecting local struggles in class, gender, or ethnic composition. Of major importance here is the increased role of the state and of large national and multinational expenditure and employment patterns unevenly distributed amongst various localities. These changes may serve as a 'material basis' to raise workers' consciousness, but may also serve to raise territorially based class alliances (Harvey, 1983). Such territorial class alliances are mainly issue-oriented; they are mobilized either in defence of the local social reproduction process (for example, to protect jobs and state expenditure) or to improve it vis-a-vis other localities (for instance, to attract a new plant). The more territorial class alliances are organised to gain positive agreements with capital and the state (along with rising productivity), the better the relative position of a particular locality in the national and international division of labor. Thus, having been 'eliminated' as an economic structure, the locality reappears with another meaning, and on a different scale, as the framework of a new form of exploitation (Damette, 1980). The connection between labor-market theories and local development has been widely applied in Italy during the 1960s and 1970s (Arcangeli et al, 1980; Garofoli, 1974; Magniani, 1978), and later in the neo-dualism thesis (Mingione, 1978). It has also attracted major interest among English-speaking scholars (Cooke, 1983; Friedman, 1977; Massey and Meegan, 1982; Urry, 1981). The basic idea is that the development process produces, and in turn reproduces, a spatially diverse division of labor. Some areas display a concentration of prestigious well-paid jobs for which a substantial level of qualifications is a prerequisite; others tend to display the inverse of these qualities. Once established, this spatial division of labor changes over time, undergoing a permanent dynamic process of recomposition in relation to the conflict between accumulation imperatives and worker resistance. The internationalization of capital and the growing intervention of central and local state create a fragmented local labor market able to provide the necessary stability and loyalty to corporate decisions (Friedman, 1977; Garofoli, 1983). Since labor power is produced and reproduced outside the factory, the relationship between capital and labor becomes one of double-dependence. It also creates a paradox: in an era of hypermobility and telecommunication, labor markets tend to be more and more decomposed and fragmented at the local level, while labor is willing less and less to move outside the locality for jobs (Garofoli, 1983). Finally, of major local importance are the general conditions of social production, to which Marx (1973, page 33) alludes in a famous passage in Grundrisse. By general conditions, I mean those services, infrastructure, and goods produced directly or indirectly by the central and local state (Folin, 1981; Lojkine, 1977). The provision of these general conditions is highly uneven over space and is the basis for the differential attractiveness of various localities to capital and labor. The kind, quantity, and quality of these general conditions increase the local rate of surplus
340
C Hadjimichalis
value and/or the rate of profit. This is achieved by increasing the productivity of labor, or by reducing the cost of the reproduction of labor power, or by reducing the turnover time of capital (O'Connor, 1973). In other words, the state tends to socialize the cost of production and to lower the cost of circulation (Folin, 1981). The whole process is contradictory, however, because these conditions are produced as public expenditures and not as commodities. At the same time, these expenditures act as a 'counter tendency' to the fall in the rate of profit, thus increasing the competitiveness of certain localities vis-a-vis others. Thus far, I have argued that important changes in local class structure, local labor markets, and general conditions of social production will produce significant variations among different localities. My focus on these issues served also to emphasize the unity between production and circulation, and the particular importance of spacetime structures for the realization of surplus value. In the following, I will consider how such variations may give rise to GTV. Marx defined the 'realization of surplus value' in terms of the successful movement of capital through the three phases of circulation: value appears first as money (when capitalists are acting as buyers); second it appears as labor force (when capitalists are acting as organizers of production); and, third it appears as a material commodity (when capitalists are acting as sellers). These transformations are not taking place at the same time and within the same territory. This means that realization of surplus value could take place within or outside the area where it has been produced, and in short or extended periods of time. I have to emphasize that neither value nor surplus value are created in circulation, because in this process commodities are merely exchanged, not created. It is in circulation, however, that the value magnitudes take their money forms: value, the form of money-price; and surplus value, the form of money-profit (Morishima, 1973). The sum of profit for individual capitalists located in a particular locality is calculated at the end of the third movement. It forms a 'localized accumulation', thus representing the spatially differentiated impulses towards capitalist expansion. Through the operation of equalization or differentiation, localized accumulation is dependent on local class structure, the local labor market, and the relative development of the general conditions of social production. The greater the mass of surplus value produced, the greater the proportion of surplus value realized locally, as well as in other regions; and the greater the mass of excess profit, the faster will be the whole process of capital turnover. Broadly speaking, GTV occurs in direct and indirect ways. Direct or derived GTV occurs when certain visible actions by social agents (usually, but not exclusively, outside the donor region) intervene so that part of the surplus value produced locally is transferred elsewhere. The mediation by these agents involves a series of actions, which vary historically from direct violence, war, and plunder to more sophisticated policies such as state taxation policies, price policy, public transfer payments, multinational profit repatriation, and transfer-pricing. The common denominator behind all these means of direct GTV is their mediation through institutions, especially through the state in combination with industrial and finance capital. In this respect it is important to note the material basis for an immediate link of derived or direct GTV both to the role of the state and to class struggle over the production of regional space. Indirect or originating GTV operates 'indirectly' through the antagonistic capitalist market. The problem here is associated with the transformation of values into prices of production, and then to actual prices in the market. This is known as the 'transformation problem', a much-debated and problematic issue in Marxist political economy (Elson, 1979). It reflects the differences which exist between the value content of commodities and their relative exchange prices. It should be noted,
Tho Oftoomphleo! tronnfor of vnluo
341
however, that the operation of the capitalist market was never perfect or selfregulated, To some extent, indirect GTV could be located under the same general concept of direct GTV, especially in connection with the growing economic function of the state. Thus the distinction between direct and indirect GTV is a methodological one and has no political implications, as I attempt to demonstrate below. In indirect GTV, the transformation from direct prices to prices of production does not involve any real change for the system as a whole (Morishima, 1973; Shaikh, 1980). The total mass of commodities, and the various portions of it going to each class, remains the same as before. By the same token, so do the sum of values and the sum of surplus value. All that changes is the manner in which given production relations are manifested in circulation (Shaikh, 1980), What the transformation brings about is a different division of the total pool of surplus value among individual capitalists having their firms in different or in the same regions. With the direct prices, each capitalist realizes an amount of money-profit equivalent to the surplus value contained in the commodities being sold. With prices of production, each regional sector's money-profit is no longer proportional to its surplus value produced locally; since the sum of values (and hence the total surplus value produced in the social formation as a whole) is still the same as before, the above change of form has the effect of redistributing surplus value from one sphere of production to another, and from one locality to another. As dc Vroey (1981) and Lipictz (1983) have shown, the prices of all commodities produced in various sectors and regions of a national market (taken together) must be equivalent to their collective value, but the price of an individual commodity is not necessarily equal to its value. The tendency towards equalization of the rate of profit does not equalize the price of commodities. On the contrary, the average rate of profit and the market prices permit a sector or firm which operates in a certain region under more favorable technical, social, and locational conditions to realizeaside from its own surplus value—part of the nonrealized surplus value of other sectors and regions. As Marx states (1981, page 829): "... it is true in fact that the average prices of commodities differ from their value, thus from the labour realized in them, and the average profit of a particular capital differs from the surplus value which this capital has extracted from the labourers employed by it". And he also states (1969, page 30), describing the transfer of value: "It is therefore wrong to say that competition among capitalists brings about a general rate of profit by equalizing the prices of commodities to their values. On the contrary it does so by converting the values of the commodities into average prices, in which a part of the surplus value is transferred from one commodity to another and from one capitalist to another capitalist". These value transfers could be better understood in terms of circulation profits, circulation losses, and losses of value (de Vroey, 1981). Profits and losses in circulation result from the lack of full realization of the norms of exchange. They occur in different localities when there is a situation of a seller having power, derived from a local strength (described earlier) in terms of class structure, labor market, and general conditions of production, to impose a premium on the equilibrium price. The specific matter of these transfers is that globally they are cancelled out, since the circulation profits of some are, by definition, the losses of others; their sum amounts to zero. As Marx (1981, page 168) pointed out: "... although in selling their commodities the capitalists of the various spheres of production recover the value of the capital consumed in their production, they do not secure the surplus value, and consequently the profit, created in their own sphere by the production of these commodities ....
C Hadjimichalis
... the various capitalists are just so many stock holders in a stock company in which the shares of profit are uniformly divided per 100, so that profits differ in the case of the individual capitalists only in accordance with the amount of capital invested by each". Similarly, losses of value arise from two sources: lack of sale or devalorization of the means of production (Harvey, 1983). The first results from mistaken choices of production and bad management in circulation. The second results from a sudden drop of value of the means of production. In contrast to losses in circulation, losses of value always represent waste in the allocation of the social labor of a commodity system. In this case, GTV does not occur; it is the 'price' paid for the way social cohesion is formed in a particular society. Thus the total mass of profit available for accumulation in a given locality at a certain time will consist of two positive and two negative components: first, profit stemming from surplus value extracted and realized locally or elsewhere, plus interclass profit or circulation to the benefit of capital; second, profits will be eliminated as a result of direct and indirect GTV, and/or because of losses of value. These joint effects form the basis for an analysis of individual sectoral and geographical differentials in profitability, the motive for capitalist competition and expansion; individual capitalists struggle to sell at or above the social average, while producing at a local cost which must be lower than the social average. 7 Concluding comment These observations clearly deserve further elucidation and interpretation. The main point to be made, however, is twofold. First, contradictory tendencies towards equalization and differentiation operate throughout the hierarchical structure of uneven geographical development, affecting the reproduction of dominant social relations at multiple scales. This in turn provides the material basis for GTV to occur under certain conditions, shaped by potentially conflicting strategies of equalization and differentiation, preservation and destruction, fragmentation and homogenization. Second, GTV is taking place in the sphere of circulation and exchange, but its magnitude and direction (positive or negative) is determined in the sphere of production. If GTV exists, it is not an automatic process, an outcome of some internal law of capital accumulation; it requires the direct and/or indirect intervention of a political agent. This is different from the sectoral transfer of value, which is supposed to act through imperfect competition and the capitalist market. Thus if GTV is influencing uneven geographical development, it is expressed primarily as the outcome of a spatially differentiated accumulation process in which the development of some regions is always at the expense of others. What this means in part is that GTV is the result of particular historical processes, rather than a specific process in itself. This should not be interpreted, however, to suggest that GTV (as an 'outcome' of uneven capitalist development) can be separated from the dynamic of capitalist development as a 'process'. Clearly the aim was to show how uneven geographical development both shapes and is shaped by GTV. To a significant degree, the. subject of GTV has only recently been revealed as part of a much broader process of demystifying the spatiality of capitalism. Just as other Marxists used the commodity form and its demystification to unveil the inner workings of capital in general, so too it may be possible to use an interpretation of uneven geographical development based on GTV both to comprehend and to transform capitalism.
Tho cjoonrophicnl irnmfor of vnluo
343
Acknowledgements, I am grateful to all who made helpful comments on various occasions, especially Dina Vafou, Ed Soja, Alain Lipietz, Peter Maskcll, Ray Hudson, Viggo Plum, and Honrlk Toft Jensen, I also found the comments of Michael Dear and two anonymous referees very constructive. None of these individuals are responsible for the final product. References Alvater E, 1973, "Notes on some problems of state intcrvcntionisni" Kapitalistate issue 1/2, pp 18-31 Amin S, 1978 Ww Law of Value and Historical Materialism (Monthly Review Press, New York) Arcangcli F, Bozzaga C, Groglio S, 1980, "Patterns of peripheral development in Italian regions, 1964-77'* paper presented to the 19th European Congress of the Regional Science Association, London; copy available from C Hadjimichalis Bajcc J, 1983, "Regional disparities in Yugoslavia*' in Tltc Crises of the European Regions Eds D Seers, K Ostrtim (Macmillan, London) pp 47-61 Baran P, Swcezy P, 1970 Monopoly Capital (Penguin Books, Harmondsworth, Middx) Bauer 0, 1924 Die Nationalitatenfragcn und die Sozialdemokratie (Vienna) Boeeara P, 1971 Traite dfEconomic Marxiste: Le Capitalisme Monopollste de VEtat (Editions Soelalcs, Paris) Bukharin N, 1973 Imperialism and World Economy (Monthly Review Press, New York) Carney J, Hudson J, Ivc G, Lewis J, 1975, "Regional underdevelopment in late capitalism: a study of the North East of England" in London Papers in Regional Science 6: Theory and Practice in Regional Science Ed. I Masscr (Pion, London) pp 11-29 Chcrpakov V, 1969 State Monopoly Capitalism (Progress Publishers, Moscow) Cooke P, 1983 Theories of Planning and Spatial Development (Hutchinson, London) Damcttc F, 1980, "The regional framework of monopoly exploitation: new trends*' in Regions in Oisis Eds J Carney, R Hudson, J Lewis (Croom Helm, Bcckenham, Kent) pp 76-91 Dear M, Clark G L, 1981, "Dimensions of local state autonomy" Environment and Planning A 13 1277-1294 De Janvry A, 1981 77i
