Class, politics, and planning: from reductionism to pluralism in Marxist class analysis

Transcription

1 Environment and Planning A, 1990, volume 22, pages Class, politics, and planning: from reductionism to pluralism in Marxist class analysis N P Low School of Environmental Planning, The University of Melbourne, Parkville, Victoria 3052, Australia Received 20 November 1988; in revised form 10 May 1989 Abstract. The fashion for Marxist analysis in planning theory has all but passed without a satisfactory answer being given to the question: what does the theory tell planners to do? This question can only be answered by conceiving of planning as a political activity in the broadest sense, and by examining Marxian political theory which deals with its subject through the perspective of class conflict. The Marxian concept of class, however, is problematic. In this paper recent developments in Marxian class theory are traced and a movement is found away from a reductionist structural perspective towards one which accepts the importance of nonclass as well as class structures, the existence of new classes cutting across the economic categories of capital and labour, and the significance of individual interests and subjectivity, in short a movement towards pluralism. A critique of reductionism is offered, focusing initially on the work of Harvey, and concludes with some tentative answers to the question posed above. Marxist analysis and planning theory Harvey played a foremost role in the resurgence of Marxist urban geography in the 1970s and 1980s and he continues to exert influence as is demonstrated in the works of Smith (1984; 1987a; 1987b), Paterson (1984), King (1989), and Healey et al (1988). He is also a leading exponent in the field of urban studies of the 'capital-theoretical' tradition of Marxist thought (see Jessop, 1982). Urban planners are an obvious audience for such work because their claim to professional status rests to a degree on their understanding of the dynamics of urban development, an understanding to which Harvey has amply contributed. Harvey (1985c) has himself directly addressed the question of the planner's role. However, when we make the step from urban studies to urban planning, or more generally from analysis to action, we encounter difficulties. An adequate theoretical treatment of 'planning' depends on an adequate theoretical treatment of politics and the state. Theoretical connections have to be made between the complex political phenomena of everyday life, of which urban planning is one small but not unimportant part, and the dynamics of class conflict which are said to underlie them. It may be argued, as does Jessop (1982, page xii), that it is not possible "to distil a single coherent, unitary Marxist theory from the various studies that Marx and/or Engels presented concerning the state and political action". Yet if we want to theorize politics (and therefore planning) from a Marxist point of view we must first bridge between the capital-theoretical and the class-theoretical perspectives. This both Jessop and Harvey have endeavoured to do in different ways. Nevertheless, Harvey's efforts remain characterized by the capital-theoretical tendency to reduce all politics to class struggle, and class struggle to the abstract conflict between economic categories. From the point of view of planning, this view of things tends to deny to planners, by virtue of their role, the power to bring about desirable social change. If everything about the state apparatus merely supports and guarantees the exploitative class relationship then planning, as an activity of the state, is not one which can contribute to social justice. In terms of practical political action there is very little to distinguish such a view from the criticisms levelled at planning

2 1092 N P Low from the perspective of Hayekian liberalism. It is true there have been attempts by Marxists to rescue planning from such a dismal conclusion (see Fainstein and Fainstein, 1979), but they carry little theoretical conviction. Planning theory has in the past been subject to abrupt changes in fashion, from proceduralism in the late 1960s and the 1970s to structural Marxism in the late 1970s and early 1980s, a flirtation with the Frankfurt school (Forester, 1980), and a return to a form of empiricism (in the late 1980s) which subordinates structure to the contingency of local politico-economic conditions (for example, Massey, 1984); and now what?: neoliberalism (Houseman, 1982), pragmatism (Hoch, 1984; de Neufville, 1987), a return to procedural theory (Bryson and Roering, 1987; Kaufman and Jacobs, 1987), institutionalism (Faludi, 1987; Healey et al, 1988), radical pluralism (Dahl, 1985; Friedmann, 1987), postmodernism how soon before planning academics fully adopt Bourdieu, Lyotard, and Baudrillard? Beauregard suggests one reason for such changes in fashion: "Within urban and regional planning the theoretical practice of academics is overwhelmed by the practical business of professionals. As a result, planning ideas are buffetted by unstable and unpredictable political and economic forces rather than discussed with scholarly deliberation" (1986, page 172). However, this lets academics off much too lightly, for they themselves, as competitive producers of knowledge, have played a significant role in shifting fashions as they struggle for leadership in their various fields. In order not to succumb to the dictates of external forces or academic fashions, planning theorists should try to solve the theoretical problems posed by one perspective before rushing to embrace another as soon as the first is shown (as is inevitable) to be deficient in some respect. One way of doing this for structural Marxism is to give due consideration to criticisms which have emerged from within the Marxist camp itself. My underlying purpose in this paper, then, is to reinforce a sense of continuity between structural Marxism and the return to a more pluralist view of the politics of planning which has emerged in the late 1980s. In the next section of this paper the reductionist tendency in Harvey's work is explored. In the third section some alternative conceptions of class structure and politics in capitalism are discussed and related to Harvey's work, conceptions which have emerged from the Marxist critique of reductionism. Some implications of this work for urban planning are then briefly discussed. Harvey's class reductionism Hall (1977, page 17) has observed that "One of the most difficult exercises is to 'read'... the laws and motion of capital in terms of the class struggle". On the one hand we have the central power relationship in the capitalist mode of production, the relationship between capital and labour, institutionalized in the 'labour contract'. On the other hand, when we come to describe the historical phenomena of political life in which people come to share certain sorts of values and interpretations of the world (ideologies) which lead to action, we do not find 'capital' and 'labour' or even 'bourgeoisie' and 'proletariat' crisply delineated. The situation is far more complicated. We find many contending groups each with its own definition of its interests specific to the situation. As Hirst puts it: "We encounter state apparatuses, parties (some of which claim a class identification, others not), campaign organisations (Child Poverty Action Group, Stop the '70 Tour, etc.) trade union and employers' organisations, bodies of armed men, demonstrations, riotous mobs etc. but never classes" (1977, page 125). Marx, of course, confronted and dealt with this same problem; but he did so in a way which left much room for interpretation. The economic logic of capitalism is

3 Class, politics, and planning 1093 approached by Marx in a different way from his treatment of politics, the former being raised to a much higher level of abstraction than the latter. Harvey comments "The formal analysis of the capitalist mode of production seeks to unravel the stark logic of capitalism stripped bare of all complicating features. The concepts used presuppose no more than is strictly necessary to the task. But a social formation a particular society as it is constituted at a particular historical moment is much more complex. When Marx writes about actual historical events he uses broader, more numerous and more flexible class categories. In the historical passages in Capital for example, we find the capitalist class treated as one element within the ruling classes in society, while the bourgeoisie means something different again. In the Eighteenth Brumaire of Louis Bonaparte, which is often held up as a model of Marx's historical analysis in action, we find the events in France of analysed in terms of lumpenproletariat, industrial proletariat, a petite bourgeoisie, a capitalist class factionalized into industrialists and financiers, a landed aristocracy and peasant class. All of this is a far cry from the neat two-class analytics laid out in much of Capital" (1982, page 25). Despite the circumspect tone of this passage, which is typical of Harvey's careful treatment of Marx, he is much less respectful of the debates in the 1970s among Marxist scholars on precisely this same matter. In his book Limits to Capital (1982) these debates are relegated to a footnote. Harvey's own treatment of the problem takes four forms which we can label the taxonomic, the causal, the essentialist, and the 'class alliance'. What these approaches have in common is the desire to reconcile human political phenomena with the central engine of capitalism, the relationship between capital and labour. It is rather as if society is being viewed in the same way as the movement of the planets or of particles of matter. Marx is to society as Newton is to matter. The taxonomic approach In this approach observable interest groups are viewed as subcategories of capital or labour. Dealing first with 'the bourgeoisie', Harvey observes that there are 'differentiations' within the bourgeoisie "which have either to be interpreted as fractions or as autonomous classes" (Harvey, 1982, page 74). He describes four subcategories: rentiers who live entirely off interest on their money capital, industrial capitalists who organize production of surplus value, merchant capitalists who circulate commodities, and landlords who live off the rent of land. He says that it does not much matter whether we use the language of classes, strata, or fractions: "what is essential is to recognize the social relationships that must attach to the different forms of distribution [of capital], and to recognise both the unity and diversity that must prevail within the bourgeoisie as a result" (page 74, emphasis added). But of course if the term 'class' is to mean anything at all then the terms used certainly do matter. True, it is essential to recognise how social relationships (organized or unorganized groups) are shaped by the mode of production, and equally how the existence of these groups and their special characteristics react back on the mode of production. Plainly, for example, as Harvey's own work shows, the existence of a stratum of home 'owners' has had an enormous impact in supporting and developing the capitalist mode of production, but we would hardly call home owners a class. Terms like class, fraction, and stratum can help to differentiate between politically significant groups to distinguish whether and in what way they are related to the mode of production. Later in Limits to Capital, turning to the class of labour, Harvey tackles the question of the interests associated with the division of labour: the deskilling of

4 1094 N P Low labour predicted by Marx and the apparently contradictory creation of new sophisticated managerial, conceptual and technical skills (including the skills of the urban planner). To what class do the possessors of these new skills belong? Harvey again insists that the newly skilled be allocated to one of the two classes: "the only question is, then, whether the monopoly powers that attach to such skills are totally absorbed as a power of capital, through the formation of a distinctive faction of the bourgeoisie (the managers and scientists) or whether they can be captured as part of the collective powers of labour" (1982, page 109). Elsewhere he writes of the proletariat and the capitalist class being "fragmented into distinct strata" (1985a, page 113); of the stratification of the working class and the creation of a "social layer of administrators and overseers" (1982, page 31); of a numerous class of industrial and commercial managers (1982, page 258); and of "different factions of capital and labour power" having "different stakes within a territory depending on the nature of the assets they own and the privileges they command" (1982, page 421). The various subcategories, though they may conflict among themselves, nevertheless share a fundamental "community of interest" and can therefore be viewed as capital or as labour (1982, page 74). The urban planner who is a sympathetic reader of Harvey must be asking two questions: to what class do the people belong with whom I regularly interact and whose interests are affected by the decisions I make (or contribute to)? And to which class do I belong? These are perhaps naive questions yet they reflect the less naive underlying question: what is to be done? We scarcely need to be reminded that Marx viewed his theory as a guide to action rather than a neutral explanation of the world. Presumably Harvey would not disagree on this point. The concept of class provides the key to action, for the capitalist class exploits the class of labour. The term has normative content; 'exploitation' is something to be avoided. A taxonomy would supply the necessary normative criteria for action, with two provisos: that the boundaries of classes and fractions can be accurately defined, and that people's interests follow from the class to which they belong. Unfortunately, both these matters are problematic (in ways to be discussed below); none more so than the class position of planners themselves. Are they exploiters or exploited? If planners cannot clearly identify the class to which they belong they cannot know in which direction their own interests lie, let alone how their perceptions of other people's interests might be coloured by their own class position. The taxonomic approach, then, seeks to classify all groups and interests under one of the two major economic categories: capital and labour. Class provides the basic criteria for action. The taxonomy is the theoretical device by which political actors (groups) are related to the causal driving force of capitalism. Note, however, that Harvey sometimes drops the 'r' out of fraction, which suggests that he is after all not happy with the taxonomic connotation of 'fraction' w. The causal approach In his earlier work (1985a) Harvey devised a formula in which social structure is organized into primary and secondary categories. The diverse social configurations observable in day-to-day politics are 'secondary' to the 'primary' capital - labour relationship. These secondary configurations are either 'residual' divisions rooted in forms of life which preceded capitalism or configurations derived from the working out of capitalist contradictions. 'Landlordism' and ethnic underclasses are (1) Marx frequently used the term 'faction' in his political works [for example, The Class Struggles in France (1962a) and the Eighteenth Brumaire of Louis Bonaparte (1962b)].

5 Class, politics, and planning 1095 allocated to the former category; groups which 'crystallize' (to use Harvey's word) around the functions of public and private consumption or are constituted by authority relations in the state or the corporations are allocated to the latter (Harvey, 1985a, pages 114 and 115). In this approach processes are emphasized whereby structures of interest are formed, the process of 'class structuration' (following Giddens, 1979). The language, however, is far from that of Giddens. Where Giddens sets out to show that social structures are continually being constituted and reconstituted, maintained and modified by the action of individuals and groups, Harvey falls back on the abstract language of 'forces'. "Marx focuses our attention on the forces of class structuration" (Harvey, 1985a, page 111). "The power relation between capital and labour may be regarded as the primary force of class structuration in capitalist society" (page 111). Here Harvey seems to be arguing that social configurations are not to be regarded as strict taxonomic categories; indeed it does not really matter whether they are or not. What matters is the mechanism which produces them. Insofar as there is any theoretical content in terms such as 'force', it is the suggestion of the form of causality conceived by the philosopher Hume: the causality of one billiard ball hitting another and causing it to move, or of one cog in a machine pushing against another. A hierarchically ordered taxonomy is thus replaced in Harvey's reasoning by a hierarchically ordered machine of cogs and gears transmitting a centrally produced force, the power of capital over labour. Time is also introduced into the picture such that a past event can result in present effects: there are residues from a 'past' system; present structures are the resultant of certain responses by state and corporate actors to past conditions. Yet even here there is a niggling uncertainty: referring back again to the Eighteenth Brumaire Harvey notes that Marx was suggesting that "class interests (often of a myopic and nonrevolutionary sort) could and did crystallize around forces other than the fundamental power relation between capital and labour" (page 112, emphasis added). The term 'force' in the context in which it is used by Harvey distances the 'cause' of structure from human agency. It implies a "derogation of the lay actor" in Giddens's words (1979, page 71). The message for planners is that they bear no direct responsibility for their acts. They are cogs in a machine whose power comes from a remote source. It is true that the concept 'social forces' is no less valid a superordinate concept in the explanation of the dynamics of a society than 'individual motivation'. But this has been recognized for some time. To refer yet again (Harvey, 1987b, page 386) to Marx's statement about people making their own history but not in circumstances of their own choosing is to restate the obvious. The problem is to theorize the relationship between individual (and group) motivation and social forces. Marx certainly recognized both sides of the coin and began to theorize it, but his work should only be regarded as a beginning. In this respect the work of Michels, Olson, Gramsci, Habermas, Bourdieu, or Giddens are all more helpful than Marx, and none of it need be regarded as inconsistent with a Marxian starting point. Thrift (1987, pages ) is right when he points out (with Berman, 1984, page 23) that intellectuals are powerless to effect social change unless they can touch the subjective consciousness of individuals. People (including planners) are not going to be moved to act if they are portrayed as cogs. The essentialist approach An alternative approach which is even less theoretically specific is to express the matter in terms of 'phenomenon' and 'essence'. Despite what might appear to be the case on the surface, capitalism is 'essentially' a class-based system. In this

6 1096 N P Low there is the language of surface and depth. The surface, Harvey (with Marx) associated with 'exchange relations' in which all the ideological paraphernalia of capitalism liberty, democracy, and the rest seem to apply. However, beneath the surface of exchange relations "entirely different processes go on" (Harvey, 1982, page 28, quoting Marx from Grundrisse 1973, page 248). Later Harvey observes: "The relation between capital and labour has by now become transformed into multiple and conflicting configurations. We have already identified certain features within this process, as capital and labour split into different factions and sometimes reconstitute themselves around some regional alliance. And as soon as we take other aspects of capitalist life into account the formation of a scientific and technical elite, the growth of management functions, of bureaucracy and so on it often becomes almost impossible to discern the single capital-labour relation underneath" (1982, pages 449 and 450). Here Harvey simply acknowledges the problem planners face (as bureaucrats) without providing a solution to it. If there are 'essential' or subsurface mechanisms, how precisely are they connected to the surface phenomena which planners confront and try to manipulate every day in their work? If the nature of the connection cannot be defined, the existence of subsurface phenomena must be of only academic interest to the planner as to any other political actor. Class alliance Harvey's most theoretically sophisticated attempt to come to grips with the relationship between Marx's analytic classes and political phenomena is contained in his essay, first published in 1985, "The place of urban politics in the geography of uneven capitalist development" (Harvey, 1985b). Here he grapples with the frustration with Marx's concepts expressed by Castells (1983) and others (Saunders, 1981; Mollenkopf, 1983). With the notion of the 'structured coherence' of a geographically constrained local economy Harvey begins to get at the idea of a complex of interlocking interests directed towards maintaining the physical and social status quo both as to distributions of power and distributions of assets. This vested interest, extending across class lines, confronts "the capitalist landscape of production" which "lurches between the stabilizing stagnation of monopoly controls and the disruptive dynamism of competitive growth" (Harvey, 1985b, page 139). This landscape will be familiar from Harvey's books and also from the work of Neil Smith (1984). "The coherence", says Harvey, "embraces the standard of living, the qualities and style of life, work satisfactions (or lack thereof) social hierarchies (authority structures in the workplace, status systems of consumption) and a whole set of sociological and psychological attitudes toward working, living, enjoying, entertaining and the like" (Harvey, 1985b, page 140). Harvey writes of 'the coherence' as of an almost tangible object, and there are obvious comparisons to be made here with Gramsci's concept of hegemony. The coherence 'spawns' a distinctive urban politics which is characterized by an unstable class-alliance in each urban region: "If as we have argued, there is an inevitable tendency toward the production of structured coherence of an urban economy, then it follows that everyone has some interest in finding the political means to affect the form that structured coherence takes" (Harvey, 1985b, page 149). Harvey recognizes the systemic unity of the interacting interests in urban politics but he still insists that the components forming this unity are made up of 'factions' of capital and labour, that is to say subdivisions of these 'primary' or 'essential'

7 Class, politics, and planning 1097 elements (pages ): "Land and property owners (including that faction of the working class that has gained access to home ownership), developers and builders, the local state, and those who hold the mortgage and public debt have much more to gain from forging a local alliance to protect their interests and to ward off the threat of localized devaluation than do transient laborers, itinerant salesmen, and peripatetic multinationals" (Harvey, 1985b, page 149). So to which of the two classes do all these groups belong? Does it matter any more? The idea of class alliance moves us towards the concept of 'interest mediation'. There is a remarkable degree of agreement across theoretical positions that interest mediation is a function performed by the state. A long time ago the pluralist Bentley (1935, page 269) wrote that, "All phenomena of government are phenomena of groups pressing against one another, and pushing out new groups and group representatives to mediate the adjustments''' (emphasis added). Giddens (1973, page 173) pointed to "the growth of formalised mechanisms for the mutual exertion of influence", though he was thinking mainly of the expansion of political control over industry. Cawson and Saunders (1983) described a pluralist and a corporatist mode of interest mediation which, they postulated, were each employed in different functional areas of the activities of the state. From a Marxist perspective, Roweis (1983, page 139) characterized urban planning as the "professional mediation of territorial politics". Most recently, Healey et al (1988) have defined in detail the specific institutional forms in which urban planning in Great Britain mediates the conflict of interests in land, and the biases inherent in these forms. Others such as Mollenkopf (1983), Van Til (1984), and Friedmann (1987) have written of the need for new coalitions and new forms of interest mediation. Harvey argues that groups which are subdivisions of capital and labour enter into coalitions to protect their short-run interests. He does not specify the role of urban planning in facilitating the creation of these coalitions, though such a role may be implied. The most interesting implication of 'coherence' as hegemony, however, is that interests may be created in both capital and labour within a single group, in fact within one person. These groups and persons then enter into alliances. It is because they are interdependent in various ways and, therefore, share an interest in both capital and labour, that they can further their interests (albeit short-run) by forming alliances. But this step takes us decisively away from structural Marxism (and into the theoretical domain of pluralism) and Harvey seems loath to take it explicitly. Harvey on planning When Harvey deals 'head on' with the question of planning, the reductionist tendency appears in a strongly functionalist perspective of the state. If political society reduces ultimately and essentially to the domination of labour by capital then the state has to be viewed as the instrument which stabilizes that power structure which is inherent in the mode of production and counteracts the destabilizing effects of class struggle. The state in capitalism 'should': "1. help to stabilize an otherwise rather erratic economic and social system by acting as a 'crisis-manager'; 2. strive to create the conditions for 'balanced growth' and a smooth process of accumulation; 3. contain civil strife and factional struggles by repression (police power), cooptation (buying off politically or economically) or integration (trying to harmonize the demands of warring classes or factions)" (Harvey, 1985c, page 175).

8 1098 N P Low The urban planner occupies "just one niche within the complex of instrumentalities of state power". From this perception Harvey elaborates a picture of the role of planning and its ideology in maintaining stable development, which is so convincing that we are almost led to believe that it is Harvey's own normative position. But no; in the last paragraph he suggests that we 'step aside' and allow a shadow of doubt to cross our mind. If we do this: "We might even come to see that it is the commitment to an alien ideology which chains our thought and understanding in order to legitimate a social practice that preserves, in a deep sense, the domination of capital over labor. Should we reach that conclusion, then we would surely witness a markedly different reconstruction of the planner's world view than we are currently seeing. We might even begin to plan the reconstruction of society, instead of merely planning the ideology of planning" (Harvey, 1985c, page 184). But a Marxist appreciation of the injustice of capitalist society is nowhere near sufficient even to begin to plan the reconstruction of society. At the very least, planning demands that we have a picture of the sort of society we want to move towards, what in particular is blocking progress towards that goal, what political move will have what effect, and of the political actors most likely to initiate action and carry it forward. In short we have to ask what political conditions would lead to planning the reconstruction of society. Hindess (1978) has warned that structural Marxism which reduces classes to 'effects of the structure' has no theory of transition but relies on essentialism and teleology. Planners are not exclusively defined by their position in the state nor is anyone else for that matter. They are subjects both of domination and of struggle and are players in a political game of which structural Marxism by itself gives us too limited an understanding. Here we see that Harvey asks us to believe that the transformation of society will depend on a leap of faith. Of course, ideology (political values) is important to social change, but what is equally important is a more adequate understanding of 'the political' and its relationship to the underlying power structure in capitalism. It is this relationship on which we will concentrate in the remainder of the paper, via a discussion of some of the Marxist work critical of reductionism. The critique of reductionism The Marxist debates of the 1970s on the question of the relationship between class and politics threw up a wide spread of new ideas and positions including a renewed polarization between economic determinism (Althusser, Poulantzas) and the opposite position of 'necessary noncorrespondence' between economic relationships and political struggle (Hindess and Hirst). Within these debates, however, a certain consensus emerged on the need to 'decouple' theoretically at least two elements of structure which, in the reductionist view, are necessarily and causally related: class viewed as 'economic position' and class viewed as a group of people sharing common economic interests, political awareness, and cultural ties (class 'in itself and class 'for itself). Once these structural elements are decoupled, that is, shown to be not necessarily and causally related ('no necessary correspondence') it becomes possible to consider a number of further 'decoupling' propositions. The abandonment of the categories 'capital' and 'labour' as uniquely descriptive of class permits the conceptualization of a variety of distinctive classes, not just 'fractions'. The existence and mode of operation of political actors and the political means by which these actors structure society can be conceptualized independently of class. The existence of nonclass structures which are far from peripheral can be admitted, and even the connection between person and class can be decoupled so that the possibility can

9 Class, politics, and planning 1099 be allowed of a person belonging to more than one class. However, if the whole framework of historical materialism is not to be abandoned, some alternative explanation of the dynamic relationship among these elements must be provided. In this respect the concepts of structuration and overdetermination have taken the place of 'forces' in the analysis. The focus of some of Harvey's most recent work (Harvey, 1985d; 1987a) has shifted from the role of the economic structure of capitalism in the production of space to a concern with the cultural variables which imbue space with meaning. But this work adds little to our understanding of class. Harvey sometimes now writes as though the concept of class were unproblematic and the above 'decoupling' already theoretically accomplished. He mixes use of the term 'class' with terms such as 'strata', 'community', (affluent) 'groups', (low-income) 'populations', and (local) 'interests' (Harvey, 1987a). His earlier favoured terms, 'class fraction' and 'class faction', have disappeared in one paper (1987a), and in another he admits that 'comparisons' between class and 'other social categories' such as 'community, race, ethnicity, and gender' are 'fair enough' (1987b, page 371). Given the theoretically substantial debate about class which has continued within Marxist discourse it would not be surprising if Harvey now felt that his earlier theorization of class was due for reappraisal. In what follows we will consider what a number of Marxist writers have said about the relationship between class as a way of describing economic functions or 'positions' and class as a way of referring to groups of people sharing, at least potentially, political interests. Along the way we will pause from time to time to consider the implications of this work both for Harvey's structural Marxism and for urban planning. The writers discussed have been chosen not only because of their shared concern to connect theoretically the 'economic' with the 'political' in class terms but also because they represent something of a chronological development of antireductionist thought. One cannot of course prove that a particular person or social group belongs to a particular class, in either of the above senses, by applying certain agreed and well established criteria to observed phenomena. There is no agreement on criteria in Marxist thought. The discussion, rather, takes the form of asking what light observation throws on the Marxist theory of society from which criteria are derived. There is a concern to discover or elucidate what Marx and/or Engels (and their authoritative interpreters) really meant. The purpose of doing this is not to claim the authority of holy writ but to discover ways in which the consistency of Marx's theory can be maintained in the light of subsequent thinking, experience, and observation. The Marxian critique of reductionism, then, takes the form of alternative readings of Marx which arguably provide accounts of the phenomena that are consistent with Marx's theory but are more empirically and conceptually plausible than the reductionist reading of the same phenomena. Poulantzas was among the first to decouple political from economic conceptions of class. Although he states that, for theoretical purposes, in defining a mode of production 'in a pure and abstract fashion' we find that modes of production involve two classes the exploiting and exploited he also disconnects this economic abstraction from the concrete reality of actual class societies. First, he says that wage labour by itself does not constitute classes. Wage labour is merely a judicial form and part of the superstructure: "a form of distribution of the social product... although every worker is a wage earner, every wage earner is certainly not a worker for not every worker is engaged in productive labour" (Poulantzas, 1975, page 20). Wage labour is exploitative, but an occupant of a certain class position may be economically exploited and yet function in the economic exploitation and political

10 1100 N P Low domination of another class. Class positions are designated by the economic 'division of labour', by which he means the social relations by means of which economic exploitation actually takes place in a particular class society. Productive labour is labour "corresponding to the relations of production of that mode: it is that labour which gives rise to a specific form of exploitation.... Thus in a capitalist society it is that which (always on the basis of use value) produces exchange value in the form of commodities and so surplus value" (Poulantzas, 1973, page 30). The picture he paints is thus of a system in which each class both plays a particular role in production and dominates subordinate classes. Second, like Harvey, he argues that any actual society will be characterized by alliances between classes. But, unlike Harvey, he denies that the classes which can be found in a particular conjuncture (actual society) are factions or fractions of bourgeoisie and proletariat: "Now in so far as the two fundamental classes are the bourgeoisie and the working class, it is true that in the course of their expanded reproduction other classes tend to polarize around the working class. But this does not mean that as classes they dissolve into an undifferentiated mass: they are still classes with their own specific interest. In other words, the concepts of 'class' and 'people' are not coextensive: according to the conjuncture, a class may or may not form part of the 'people', without this affecting its class nature" (Poulantzas, 1973, page 33). In other words 'classes' as economic positions are not coextensive with, are decoupled from, actual groups of people. Actual groups of people and their interests and perceptions of themselves are shaped by political and ideological factors as well as (and to some degree independently from) the economic division of labour manifested in class positions. This decoupling leads Poulantzas to suppose that certain occupational groups may occupy a 'contradictory class position'. For example, technicians and engineers, according to Poulantzas's economic criteria, are part of the working class because they increasingly contribute directly to the production of commodities. Yet political and ideological criteria assign them to a position of special authority in overseeing the labour process and its despotic organization (page 35). Therefore, to that extent they are part of the bourgeoisie. This contradictory position is not a subjective matter of which class particular technicians and engineers believe or feel themselves to belong to, but is an objective fact of their contradictory position in the economic division of labour. Unlike Poulantzas, Harvey fails to recognize the independent reality of the political and ideological domains in forming actual groups of people with shared interests. Therefore he does not theorize the reality of the contradictoriness of a real interest group such as urban planners. In fact, urban planners belong to a wider grouping of technicians, professionals, and managers whose interests are divided between the abstract economic categories of capital and labour. Interests involve power relationships and values as well as economic relationships. Healey et al (1988) provide evidence of the importance of institutional relationships (the 'political') and access criteria (the 'ideological') in shaping interests in land (the 'economic') in Britain. Comparing the work of these authors with that of Mollenkopf (1983) we can see how different the United States of America is from Great Britain in its land-use politics despite the similarity of the underlying economic relationships. Both these studies illustrate how important are the political and the ideological domains in shaping real material outcomes, for it is in these domains that action occurs to support or change economic relationships. Hall (1977), in a penetrating and careful analysis of 'the political' in Marx's own works, arrives at three conclusions. First, Marx never intended to imply that

11 Class, politics, and planning 1101 political structures (or indeed any observable strata and groupings) necessarily corresponded with the 'objective determinations' of the economic level: "that the terms and contents of the one [political level] are fully given in the conditions and limits of the other [economic]" (Hall, 1977, page 40). The latter merely "establishes the outer limits, the determinations, the horizons within which the forms of the political arise and appear" (page 40). Second, there are certain political forces to which no clear class designation corresponds, though their role and support is pivotal. In the Eighteenth Brumaire (according to Hall) these are the army, the press, the intellectual celebrities, the priests, and the rural population. Third, a political grouping (such as a political party or pressure group) may represent the interests of a particular class through the role it plays on the political stage even though the members of the group may occupy a variety of class positions. A political group is usually a coalition of class interests, but the action the group takes may support or oppose the interests of a class. The example Hall uses is the Social Democratic Party in France (of the Eighteenth Brumaire) which cemented a coalition of proletarian and petty bourgeois interests: "This 'party' has its immediate determinations: it advances, temporarily, the interests of those left aside by the forceful regrouping of bourgeois forces. It has its contradictory internal structure: through their subordination within it the proletariat lose 'their revolutionary point' and gain 'a democratic twist'. Social democracy also has its objective political content: weakening the antagonism between capital and wage labour and transforming it into harmony. It is 'democratic' reform within the limits of bourgeois society. It is in this precise context that Marx warns us about a too reductive conception of political representation. This temporary solution is petty-bourgeois because it advances the narrow interests of that traditional class. Its 'representatives' cannot be analysed in terms of the narrow reduction to their class designation they are not all 'shopkeepers.' The position of this alliance is 'petty-bourgeois' in character because, temporarily, the general resolution to the crisis it proposes and endorses corresponds to the objective limits of the particular material interests and social situation of the petty bourgeois as a class" (Hall, 1977, page 44). If there is no necessary correspondence between economic class positions and organized groups it follows that we must be prepared to account for nonclass structures or structures in which many class positions are represented. Nevertheless, at any given conjuncture the political role of such a group may (though not 'must') represent a particular class interest. Hall's interpretation of Marx makes better sense of the urban planner's daily experience of interests and pressure groups than Harvey's. Such groups may represent a variety of different class positions, not just 'capital' and 'labour'. Nonclass structures may support or oppose the continued domination of that small group whose interests correspond with its class position and which, by virtue of its power to control investment, can strongly influence (if not determine) the economic future of any space smaller than that over which it is allowed to deploy capital. But the concept of class by itself does not provide criteria for dealing with nonclass structures: interests in 'locale' or the physical environment generally, the meaning people give to places, social relationships and objects, forms of domination, social control, cohesion and prejudice (race, gender, ethnicity), institutional structures such as urban planning, and the administration of justice which would be required in any society regardless of its mode of production. However, in the view of another analyst plurality of interests is not just a feature of nonclass structures.

12 1102 N.P Low Wright (1980) builds on Poulantzas's concept of objective structural determination of contradictory class positions. In his analysis of economic class positions he makes use of three polar types: bourgeoisie, proletariat, and petty bourgeoisie. The last of these he regards as more than a mere anomalous remnant of an earlier mode of production. Indeed the three poles emerge from the economic dimensions of domination and subordination within production itself. These dimensions, according to Wright, refer to the control (or lack of control) of money capital, physical capital, and labour [the "actual activity of work within the production process" (Wright, 1980, page 328, footnote 10)]. Contradictory class locations appear because these three dimensions of the social relations of production "need not necessarily coincide perfectly indeed there are systemic forces in capitalist development working against their doing so" (page 330). Even though this conception is intended to designate only class positions, it already produces many more positions than the two class model: nine including all combinations of the dimensions which Wright regards as 'polar' and those he regards as contradictory. In addition, Wright adds categories such as 'partial' control and 'minimal' control over money capital, physical capital, and labour which have the potential to generate yet more class positions. But in his analysis he leaves out what surely must be an important contradictory class position: of those who have at least partial control of some money capital but no control over either physical capital or labour, the small investor class. This class became the target of the ideological strategies of conservative governments in Great Britain and the United States of America in the 1980s. For Harvey, 'class' is axiomatic; that is, it need not be further explained with reference to some higher order concept. This axiomatic use of 'class' is passive and structural. Starting with two fundamental classes, Harvey can do no more than subdivide them. Wright introduced an active and 'political' higher order concept, domination, to explain the nature of class. This strategy enabled him to generate multiple interests from quite simple elements. It becomes possible to ask "How does x dominate yt\ Emphasis on domination, however, moved Wright away from Marx and towards Weber. This is not necessarily a fault. But in response to his critics and in order to maintain the integrity of a specifically Marxist position, Wright has employed an analysis of class in his most recent work based on unequal possession and exploitation (another active political concept). Following Roemer (1982) he explains the observable complexity of class positions by the existence of contradictory modes of production in modern capitalism. Roemer argues that class and exploitation inevitably arise from free choice in a system with unequally distributed assets. Class is not, therefore, dependent on the labour contract (see also Roemer, 1988). Capitalist exploitation, based on inequality in the distribution of the means of production, exists alongside bureaucratic exploitation, based on unequal control of organizational assets, and alongside socialist exploitation, based on unequal control of scarce skills. Wright explains the plurality of middle-class groups primarily by the plurality of their structural locations, that is, their material interests: their possession of assets and their exploitative relations with people in other structural locations. Political phenomena can then be discussed in terms of coalitions and alliances of class locations and contradictory class locations. This reformulation of the 'embarrassing' problem of the middle classes enables Wright to dispense with the need to introduce nonstructural (in the economic sense) concepts to explain the complex phenomena of class structure. However, Wright's new conception of class structure leads him to a radical possibility: that the new middle class and not the proletariat

13 Class, politics, and planning 1103 is today's revolutionary class: "In feudalism, the critical contradictory location is constituted by the bourgeoisie, the rising class of the successor mode of production. Within capitalism, the central contradictory location within exploitation relations is constituted by managers and state bureaucrats. They embody a principle of class organization that is quite distinct from capitalism and that potentially poses an alternative to capitalist relations. This is particularly true for state managers, who unlike corporate managers are less likely to have their careers tightly integrated with the interests of the capitalist class" (Wright, 1984, pages 400 and 402). In this conception there is more than one possible historical transition between capitalism and socialism. If this analysis is correct then urban planners, together with other members of their managerial class, are not mere passive supporters of capitalism but, potentially at least, bearers of a new mode of production. This is not altogether farfetched, for the production of serviced, developable land is already subject to a process which is distinctly different from simple commodity production and exchange. Wright's argument does not deny the existence and possible importance of nonstructural and nonclass political groupings of the sort referred to by Hall (1977). Indeed he observes that it remains problematic whether workers will be formed into a class or into some other sort of collectivity based on religion, ethnicity, region, language, nationality, or trade: "The class structure may define the terrain of material interests upon which attempts at class formation occur, but it does not uniquely determine the outcome of those attempts" (Hall, 1977, page 404). Taking this point only a small step further, we can say that classes as groups of people with political interests are partly, but only partly, shaped by the economic positions which can be deduced from analyses of the power relations inherent in the mode of production. These positions, in all their complexity, are also reciprocally created as a result of actual class and nonclass struggles. As Harvey observes, what we have to take account of in any explanation of the political is time. But even in Marx's day, class struggles between organized groups whose political interests corresponded with their economic position (classes) took place in the context of struggles among groups with no clear 'class' designation. These class and nonclass struggles resulted in agreements, pacts, conventions, rules, norms, and (with reference to the state) policies which created new groupings of political interest and constantly shuffled the boundaries among existing groups. The struggle among these new groups in turn produced new responses (agreements, conventions, rules, etc) which gradually entrenched the position of the dominant class in a hegemonic system. So urban planning practice, even though it contains the elements of an alternative mode of production, can also be viewed as part of a hegemonic system. But a hegemonic system, as opposed to a functionally determinate system, contains contradictions. Przeworski argues that the hegemonic system of institutionalized rules, acted out daily, shapes the perception and expression of interests: "Hegemony becomes constituted when struggles over the realization of material interests become institutionalized in a manner rendering their outcomes to some extent indeterminate with regard to positions which groups occupy within the system of production. It is this kind of organization of social relations which constitutes 'democracy'. Capitalist democracy is a particular form of organization of political relations in which outcomes and conflict are within certain limits uncertain, and in particular, in which these outcomes are not uniquely determined by class positions" (Przeworski, 1980, page 28).

14 N P Low In other words in pluralist democracy certain kinds of struggle are allowed and they take place because the outcome is indeterminate. Therefore the struggle is worthwhile. And because the struggle is worthwhile the participants have a real material interest in winning. The kinds of struggle which are stimulated by the constitution of pluralist democracy also bring into existence real material interests (things worth struggling for) and actual political groups sharing these interests which are independent of class interests but are nonetheless real. At the same time certain kinds of political action are ruled illegitimate and open to being legitimately ignored or suppressed. Capitalist governments around the world have successfully ruled against union picketing and secondary strikes and boycotts, mass demonstrations are routinely described as 'undisciplined rabbles', and in the case of Britain the autonomy of local government as a whole has been drastically curtailed in the interest of suppressing programmes which deviate from central government policy (see Jones, 1986). However Przeworski's analysis is not aimed primarily at explaining the relationship between class and nonclass interests but at demonstrating that 'class compromise' provides the conditions of existence both of the plurality of material interests and of the state. The basis of the compromise is that "workers consent to the perpetuation of profits as an institution in exchange for the prospect of improving their material welfare within some future period" (Przeworski, and Wallerstein, 1982, page 217). The latter prospect is guaranteed by institutions which would permit workers to process their claims with some chance of success: unions, parties, and a relatively autonomous state. In short, "workers consent to capitalism, capitalists consent to democracy" (page 218). In this analysis consent is not a matter of subjective perception nor is compromise a goal. Consent is inherent in action and compromise is an unintended consequence of class struggle: "A compromise holds if and only if it is continually in the best interest of workers and capitalists and only if it is repeatedly remade. A compromise is not a commitment for some indefinite or even limited future; it is an outcome of strategies chosen today which today appear optimal. If the compromise is to last the day after it must be made again the day after, and it will be made only if it constitutes the best solution the day after" (Przeworski and Wallerstein, 1982, pages 218 and 219). Capitalist democracy is precarious precisely because it depends on day-to-day compromise, but it is not, as Marx thought, inherently unstable. Although society may be transformed either into forms of capitalist dictatorship or state socialism, the stability of capitalist democracy (such as it is) rests on the constitution of multiple material interests and the capacity of the state to organize the form of compromise, and on broad electoral support for its particular outcomes. In relating plural conflict of interest to an underlying class process, Przeworski is actually saying something very similar to Harvey when he writes of 'surface' and 'depth' (see above, page 1096). The constant attempts to maintain a balance between the interests of capital and labour involve the state in policies which bring multiple new interests and groups into existence. Some pluralists have also noted how pressure groups tend to develop in response to state policies and programmes (see Olson, 1965, pages ; McConnell, 1967; Zeigler, 1980). However, in such an explanation there is no need to conceive of abstract economic positions as organized groups of people at all, let alone to reduce all organized interests to fractions (or factions) of one or the other economic position. Resnick and Wolff (1987) have taken the concept of 'class process' a step further. Plural conflict of interest is organized and structured by rule and convention.

15 Class, politics, and planning 1105 However, immanent in this plural conflict are structures of economic power, and immanent in these structures is the 'class process': "that in which unpaid surplus labor is pumped out of the direct producers" (Resnick and Wolff, 1987, page 115). Resnick and Wolff regard the class process as one among many distinctive social processes which together provide the conditions of existence of each other; in short, each of the distinctive processes in society is 'overdetermined' by the multiple influence of all the rest. The point here is that these authors explicitly and radically reject the epistemological position which arranges the objects of knowledge in hierarchical taxonomies. Therefore they also reject essentialism (they claim that Marx's epistemology was antiessentialist) and all other tendencies to reduce explanations of society to simple, 'underlying', mechanistic relationships. They do not claim, as many Marxists seem to, that Marxist theory can provide a complete understanding of society and they also locate Marx firmly in the political struggle for human emancipation and against self-perpetuating inequalities of wealth and social standing which constrict and distort the potentialities of human life (Resnick and Wolff, 1987, page 279). Marx's contribution, they say, lay in defining, locating, and connecting the class process to all the other processes constituting the social totality: "Marx's consequent focus on class in his writings did not stem from some spurious claim that the particular process he discovered was more important in shaping the totality than the myriad other processes comprising that totality. Rather Marx's focus on class and hence on economics represented the much more modest self-awareness of someone who knew he had something particular to add to certain forces for social change" (page 279). The working out of the class process produces a number of economic positions. Classes are "subdivisions among people according to the particular position they occupy in the class process. People participate in class processes; they therefore occupy class positions" (page 117). The positions are given not only by the pure extraction and receipt of surplus labour and its products but by their distribution and receipt for services rendered in the process: to the landowner, for example, for the supply of land and accommodation, or to the agent of the state (including the urban planner) who mediates between consumption and production or otherwise regulates the process. Class structure (according to Resnick and Wolff) includes classes and 'subsumed classes', the latter referring to persons occupying the positions of distributors or receivers of the portions of surplus value allocated in order to secure the conditions of existence of the extraction of surplus labour and the accumulation of capital. To apply the capital-theoretical perspective for a moment, we might say that flows of capital in the secondary and tertiary circuits provide the conditions for the creation and reproduction of subsumed classes. The notion of 'fundamental' and 'subsumed' classes, however, has an inescapably hierarchical flavour which conflicts with the author's previously stated position and highlights an epistemological contradiction in Marx's own work. More importantly, the authors (Resnick and Wolff) say, on the one hand, that classes and subsumed classes are groups of people occupying economic positions and, on the other, that people can occupy more than one class position. But they cannot have it both ways. Either groups with shared political interests are defined by their class position or they are not. The authors actually adhere to and subsequently elaborate the latter position. Various class positions are indeed created through the operation of the class process. The economic interest and political organization of some groups in society coincides with the class position their members occupy, but many more political groups are formed from more or less temporary coalitions among people

16 1106 N P Low who have some stake in more than one class position, and also by the operation of nonclass processes. There are social structures and processes which are not class structures and processes but whose working out in a class society nevertheless facilitates the operation of the class process: the extraction of surplus labour from the direct producers and its accumulation in the hands of a small dominant class. Arguably the tendency of this class to invest for future production and accumulation depends less upon the bourgeois values of frugality and thrift (as Weber thought, see Beetham, 1985, page 56) than upon whether the class has more resources than it needs for present luxurious and ostentatious consumption. However, looking back to Przeworski (and Gramsci) the stability of capitalism and the hegemony of capital is assured because divisions in the population have been generated which cut across class lines. Class (and nonclass) struggle has wrung concessions from those who own and control capital. Without such struggle to distribute the surplus more widely than within a single small dominant class the mass potential to 'need' and 'demand' goods and services could never have been tapped. The multiplication of contradictory class positions which, nonetheless, reinforce the class process accounts for the appearance of a 'new middle class' or 'new petty bourgeoisie'. But the new class is not in reality a homogeneous mass but many and varied political and economic interests composed of 'alloys' of class positions. Last, mention must be made of the school of thought in which it is postulated that the processes of circulation, reproduction, and consumption, rather than of production, are responsible for class formation in contemporary capitalism. Thrift and Williams state: "What is clear, then, is that any analysis of class formation as a historically contingent process arising out of reciprocal actions must step outside the social relations of production and venture into social relations of reproduction. Further, these social relations of reproduction cannot be seen in the narrow sense as being just concerned with the reproduction of labour power. Many of them are bent towards reproducing practices which may only tangentially interact with class practices, or, indeed may be divorced from them; culture rears its problematic head" (1987, pages 7-8). Urry is among the leaders of this school of thought and the following discussion will be focused on his work. Urry (1973) began by trying to find in Marx's work a structural explanation for the existence of the middle class. This he found in the proposition that there are two fundamental conditions for the survival of any capitalist system: a mechanism which ensures the realization of profits by maintaining consumption at a higher level than production, and a mechanism for reducing the costs of circulation "and more specifically the need to have efficient administration of the growing capitalist economy" (page 178). But, as he says, this proposition is insufficient by itself to explain why middle classes, in particular, arise to fulfill these functions. In subsequent work (Urry, 1981; Abercrombie and Urry, 1983) Urry develops a theoretical formulation of the process of class formation in the course of political struggle in the spheres of circulation of capital and the reproduction of labour. Together these 'spheres' constitute 'civil society'. Civil society is "that set of social practices outside the state and outside the relations and forces of production in which agents both are constituted as subjects and which presuppose the actions of such subjects first, in the sphere of circulation directly; second, in those social relations within which labour-power is reproduced economically, biologically and culturally; and, third, in the resultant class and democratic forces" (Urry, 1981, page 31).

17 Class, politics, and planning 1107 According to Urry it is erroneous to characterize 'class' as pertaining to abstract economic positions. Classes only exist 'at the level of civil society. That is to say the term 'class' only comes to have meaning in the context of the circulation of capital when buyers and sellers of commodities, including labour, interact with one another. Despite Urry's claim to a position "having more affinities to the Marxist than the Weberian position" (Urry, 1981, page 26) this formulation is a reinterpretation of Weber in Marxist language. Accordingly, Urry has no difficulty demonstrating that many different factors bear on the formation of political interests and political groups; in descending order of importance: spatial organization of labour and residence; sexual division of labour; religion/ethnic /racial 'allocation of subjects'; differentiation on the basis of secondary groups such as trade unions, professional associations, and voluntary groups; and 'general allocation of subjects', a category which is presumably intended to allow for individual distinctions (page 70). The family is at once a refuge from capitalist competitive social relations and one of the noncapitalist forms of production which makes capitalism possible by subjecting women to marginalized labour and the domination of men. Abercrombie and Urry (1983) discuss the distinction between classes as economic positions (class places) and classes as groups of people. The authors reject a radical distinction between 'places' and 'persons' as postulated by Poulantzas, arguing instead that "although classes are constituted as places, class practices are partly determined by the characteristics of persons who fill those places, characteristics determined by a number of determinants outside class places" (page 109). Additionally, both market (exchange relations) and work situations (production relations) determine class places. Thus "the mechanisms of the capitalist mode of production require the functions of conceptualization, control and reproduction to be performed and they produce sets of places comprising specific market and work situations which we call classes" (page 126). So, both a plurality of functions and a plurality of nonclass factors combine to produce a plurality of groups of persons. Urry is really saying that classes are constituted of acting subjects, individuals in effect. The subject's consciousness of her or his political position is given by many social factors including the social relations stemming from the capitalist mode of production and, in particular, the culture and ideology of the society. Urry's analysis works both down to the individual from the social level and up to the social from the individual level. Marrying collectivism with individualism in this way makes it permissable for one to consider that an individual may have interests in both capital and labour simultaneously as well as in other nonclass structures. Therefore, as there is no such thing as a 'pure' individual, there is also no such thing as a pure class. In keeping with Marxist language, however, Urry turns the matter round: "Capitalists and workers are categories of subject produced within civil society and such classes are fissured by the bases of structuration present within civil society. There are no pure classes of capitalists and workers existing outside the forms of differentiation effective within civil society. And further, within civil society there are other 'classes-in-struggle' [a concept of Laclau's] and popular democratic forces and struggles" (Abercrombie and Urry, 1983, page 106). Unfortunately this is confusing. If there are no pure classes, and if classes cannot exist outside the political domain of acting subjects, can it be said that there really are such things as classes at all? The only way out of this conceptual problem is to postulate, via realist epistemology, an underlying 'mechanism' similar to Resnick and Wolff's 'class process' which acts on individuals. However, the mechanism need not be considered 'underlying' or hidden at all. It would be much simpler to say that class is one of the many institutionalized relationships in which

18 1108 N P Low individuals are enmeshed. 'Class' is the term given by Marxist theory to a cluster of very fundamentar social rules structuring work and property relationships. If we take this step we should reserve the term 'class' for that cluster alone. It would therefore be incorrect to talk of 'consumption classes' or indeed of a 'service class' (Abercrombie and Urry, 1983; Thrift, 1987). Rather, we should say that capitalism, in the course of its development, with state 'polyarchy' (Dahl and Lindblom, 1953) and mass consumption, has constituted many different organized groups and social strata, but the fundamental work and property relationships remain unchanged. In all this we have come a very long way from Harvey's reductionist view of class; more than half way not only to Weber but to pluralism. The political role of the planner cannot be fully understood simply as legitimating capitalism and smoothing the process of capitalist exploitation. The role of urban planners as state officials need not be viewed as derivative of the relationship between the bourgeoisie and the proletariat. Indeed Wright (1984) argues that managers and state bureaucrats occupy the pivotal contradictory class location in late capitalism. Because this managerial class represents an alternative mode of exploitation to that of capitalism, its members (and not the proletariat) constitute the revolutionary class. Such a view is obviously debatable, not least because state and private-sector bureaucrats would seem to share little in the way of economic interest. Hoff (1985) points out that the capitalist state apparatus is itself divided on class lines between the 'classical' bureaucracy concerned with the efficient development, management, and control of capitalist economies (the army, the law, public finance, etc) and the 'welfare' bureaucracy, including health, educational and social services, and environmental agencies. This view recalls the earlier arguments of Cawson and Saunders (1983). At all events the political role of the state bureaucrat seems both pivotal and problematic from a nonreductionist class perspective. Marxist analysis and the planner's role Marxist urban analysis seeks to explain how it is that injustice, exploitation, and concommitant poverty, environmental squalor, and degradation enter into the process of urban development. But there is no message in Marxism about how to be a better urban planner. When urban planners reach the conclusion that there is nothing substantial that they can do within the current rules and norms of their particular part of capitalist society to improve urban conditions and solve urban problems, Marxists (for example, see Harvey, 1985c) invite them to step outside the roles given to them by those rules and norms. Such a step may well be demanded by conditions in some of the cities in the USA and Great Britain today. This is not just a failure of planning but a wholesale failure of political structures which are supposed to enable people through legitimate political struggle to improve their conditions of life. Gottdiener (1987) has pointed to the death of democratic politics at the local level in the United States of America. Without intervention by the state, as occurred in the 1960s, people living in extreme poverty have little hope of mobilizing to press demands for improvement. Existing democratic structures simply reflect the biases in society in favour of the powerful. 'Stepping outside' their role does not, of course, mean that planners should give up their jobs, simply that planners should start thinking of the city as citizens do. This extends planners' thinking beyond their job descriptions. Marxism suggests three further steps which the urban planner should take. First, they should become aware of the societal rules which limit the possibility of action and which define and constrain the role of the planner. These rules must become the target for change. Second, they should become aware of the class to which they belong, that is to say look to the people, in the widest sense, who share their values and interests.

19 Class, politics, and planning 1109 Third, they should mobilize for action (in accordance with those values and interests) with members of their class and all available allies, with all the power at their disposal, including the power, however circumscribed, which the institution of urban planning affords. The strong implication of modern Marxist thought (influenced by Gramsci) is that, if society is to be changed, those who wish to change it must themselves be prepared to fight for what they believe in and not wait for some other 'force' or even other actors, such as the working class, to do it for them. Such political action is, of course, always problematic. As Gottdiener observes: "When local political culture has died, it can only be resurrected from outside. All other changes are doomed to be coopted within the encapsulating, totalizing logic of State power. As with the case of mobilization against capital, how one gets outside a globalizing system while being caught within it is a puzzle of some magnitude" (1987, page 268). But when it comes to political action, there is no 'outside'. The only hope for change is to work with whatever political resources exist in the surrounding environment of the political actor. Marxism draws our attention to the importance of periods of crisis in capitalist economies in creating conditions in which structural change involving renegotiation of the rules becomes easier. By reducing classes to economic positions, however, structural Marxists oversimplified this political environment, which is composed of many social and political structures, some organized and active, some latent whose influence is felt only indirectly through the ballot box or through investment decisions. The behaviour of these structures cannot be understood by means of class analysis alone. Placing class labels on politically significant groups does not help one to understand or anticipate their behaviour. Even though all groups may bear some relationship to the class process (the extraction of surplus value and the accumulation of capital in the hands of a dominant class), politically significant groups are not all class fractions or even classes. If the antireductionist arguments are correct we must, at least, differentiate in each particular case those structures which have a strong relationship to the basic 'class process' from those only weakly related or unrelated. Even where the relationship is strong, we must expect to find coalitions of interests within the complex configuration of economic positions. Furthermore, there also exist groupings whose members' interests are divided between different class positions. Many people own a little capital but the value of the little they own is tied inextricably to the fate of the large capitals in the hands of the few. This situation is not just a rhetorical 'icing' of legitimacy on the capitalist cake, it is the cake itself. Thrift and Williams (1987, page 7) say "This is a conclusion to which the Marxist literature has come slowly and painfully over the last twenty yearsclass is not the be all and end all of social and political life." Nonclass and interclass structures play a major part in the central class conflict, though this conflict may no longer be between the bourgeoisie and the proletariat. Urban planners who desire urban improvement need to recognize the texture of local political environments, their relationship to national and global politics, and the relationship between urban planning practice and the mobilization of local interests. In this last respect, Gottdiener draws our attention to the work of Henig (1982) and Crenson (1983). But a wider rereading of the pluralist literature is also essential, for it is within the pluralist 'home domain' of theory that the question of political action has been most discussed (see Alford and Friedland, 1985). By insisting upon Marx's view that the political initiative for change lies with the proletariat and only the proletariat, structural Marxism reduces the incentive for planners and other rawproletarian citizens concerned with social justice and the welfare of cities to act immediately and on behalf of their own values. By conflating

20 1110 N P Low 'human values' with 'false consciousness' structural Marxism robs itself of a powerful means of change, namely the power to hurl back at capitalism its own rhetoric of freedom and democracy. By denying the validity of alternative theoretical frameworks, structural Marxism robs itself both of political allies and of insights into the world of politics. Marxism is not the only theoretical framework which deals with the opacity of the rules which structure social practices: "What those rules should be is the principal question in human affairs; but if we except a few of the most obvious cases, it is one of those which least progress has been made in resolving. No two ages, and scarcely any two countries, have decided it alike; and the decision of one age or country is a wonder to another. Yet the people of any given age and country no more suspect any difficulty in it, than if it were a subject on which mankind had always agreed. The rules which obtain among themselves appear to them selfevident and self-justifying. This all but universal illusion is one of the examples of the magical influence of custom, which is not only, as the proverb says, a second nature but is continually mistaken for the first. The effect of custom, in preventing any misgiving respecting the rules of conduct which mankind impose on one another is all the more complete because the subject is one on which it is not generally considered necessary that reasons should be given, either by one person to others or each to himself." These words could almost have been written by Bourdieu (1977) but were in fact written by Mill (1971, page 69). Pluralist thought has long been conscious of the 'rules of the game' (see, for example, the excellent discussion of pluralism by Waste, 1987). Gottdiener's concept of the two tiers of the organization of the state (the 'first tier' institutionalizing capitalist property relations) is paralleled by Benson's concept of 'deep structures' (Benson, 1975; Benson and Weitzel, 1985). It is, of course, true that modern pluralists have, on the whole, accepted the rules of capitalist democracy as the basis of an ideal polity and have concentrated their analyses on the 'play of power' within the rules. For this they have been widely criticised. But Dahl and Lindblom, arguably among the most sophisticated of pluralist theorists, in 1953 (pages 255, 257, 371, 442, and 481) expressed reservations about the bias of a system in which one part of the public domain (the state) is ruled by 'polyarchy' and the other (the 'corporate sector') is ruled by hierarchy and/or oligarchy. In 1982 and 1985 Dahl returned to the question of bias and now argues that democracy is endangered by the institutional failure to control the power of the leaders of business corporations. Surely here is the basis for dialogue between Marxists and radical pluralists. Last, by drawing attention to the underlying similarity of all capitalist economies, structural Marxism tends (whether intentionally or not) to deflect attention from important differences in capitalist polities. Gottdiener's arguments about the structure of the state in the USA cannot simply be transferred to other capitalist states (as he himself would doubtless agree). He says, for example, "No mechanisms currently exist that can aggregate neighbourhood mobilization of needs into a viable public discourse on the future state of the metropolis" (Gottdiener, 1987, page 283). This is not the case in countries like Australia where most state governments are in effect also elected metropolitan governments (see Low and Power, 1984; Parkin, 1984; Moser and Low, 1986) or in Sweden where local government standards of service delivery are centrally regulated. To take the view that real improvement is impossible unless capitalism, as a whole, is changed is to resign oneself to inaction. The only way capitalism can be changed for the better, as has always been the case, is through political action aimed at partial and gradual change.

21 Class, politics, and planning 1111 Conclusion We began by looking at the different ways in which Harvey tackled the intersection between structural economic relationships and the complex web of political interests, actors, and relationships which together make up the reality of capitalist society. Harvey's approaches to the problem were characterized as 'taxonomic', 'causal', 'essentialist', and 'class alliance'. In each approach Harvey expresses doubts. He says it does not really matter whether we use the language of classes, strata, or fractions. He then drops the 'r' out of fraction to make the less reductionist term 'faction'. He says that class interests can crystallize around 'forces' other than the fundamental power relation between capital and labour. He notes that the relation between capital and labour has today become transformed into multiple and conflicting configurations. He comes close to adopting the Gramscian concept of 'hegemony' in which political interests supporting a mode of production are not reducible to those of capital and labour. Recently he uses the terms 'strata', 'communities', 'groups', 'populations', and 'interests' apparently in preference to 'class fractions' or 'factions'. So perhaps Harvey himself is not a reductionist. But, as we have argued, his theory of class is. Harvey's theoretical strength lies in his ability to relate spatial phenomena to Marx's theory of the circulation of capital. Marx's class theory, however, also leads to a different sort of theoretical connection: between economic and political relationships. In this respect a number of Marxist writers have reached very different conclusions from Harvey's about the way in which Marxist theory can be reconciled with a 'plural' political world. There is no agreement among these writers as to the status of the political. The plurality of political interests is explained by Wright in terms of the plurality of economic relationships, and by Przeworski as the outcome of a class compromise which is continually being reconstituted. Resnick and Wolff portray a plurality of political interests and actions which together 'overdetermine' the system in which the 'class process' can take place. Urry and others argue that classes are constituted by the circulation, reproduction, and consumption functions in a modern capitalist economy. But there is a sort of consensus on the need to decouple the concept of class as 'economic position' from class as 'group of people with shared political interests'. Factors other than economic position intervene to shape interest groups and pressure groups. Therefore there are nonclass interests and these may well be extremely important in the flow of political action. Organized interests act politically in conditions shaped both by economic constraints and by other 'rules of the game' laid down by earlier political action. To speak in these terms, however, is to invoke a realm of discourse commonly thought to be opposed to Marxism: that of pluralism and elite theory. Harvey seems willing to move in this direction but is restrained by the weight of his earlier reductionist conceptions. The theory of urban planning is situated at the intersection between the theories of space (urban) and politics (planning) in the context of society (with all that that term entails). In this paper we have traced a path in Marxian class theory, via a critique of Harvey's conceptions of class, from reductionism to pluralism. If planning rather than the urban system is to be the main theoretical focus, the starting point from a Marxist perspective will be the politics of class. In this respect, however, the future for a radical theory of urban planning practice now appears to lie in dialogue between Marxist, Weberian, and pluralist perspectives rather than within any one of these.

22 1112 N P Low References Abercrombie N, Urry J, 1983 Capital, Labour and the Middle Classes (Allen and Unwin, Hemel Hempstead, Herts) Alford R R, Friedland R, 1985 Powers of Theory: Capitalism, the State and Democracy (Cambridge University Press, Cambridge) Beauregard R, 1986, "Planning practice" Urban Geography Beetham D, 1985 Max Weber and the Theory of Modern Politics second edition (Polity Press, Cambridge) Benson J K, 1975, "The interorganizational network as a political economy" Administrative Science Quarterly 20 (June) Benson J K, Weitzel C, 1985, "Social structure and social practice in interorganizational policy analysis", in Policy Implementation in Federal and Unitary Systems, Questions of Analysis and Design Eds K Hanf, TAJ Toonen (Martinus Nijhoff, Dordrecht) pp Bentley A, 1935 The Process of Government (Principia Press, IL) Berman M, 1984, "The signs in the street: a response to Perry Anderson" New Left Review Bourdieu P, 1977 Outline of a Theory of Practice (Cambridge University Press, Cambridge) Bryson J M, Roering W D, 1987, "Applying private-sector strategic planning in the public sector" Journal of the American Planning Association Castells M, 1983 The City and the Grass Roots (Edward Arnold, London) Cawson A, Saunders P, 1983, "Corporatism, competitive politics and class struggle", in Capital and Politics Ed. R King (Routledge and Kegan Paul, Andover, Hants) pp 8-27 Crenson M, 1983 Neighbourhood Politics (Harvard University Press, Cambridge, MA) Dahl R A, 1982 Dilemmas of Pluralist Democracy, Autonomy versus Control (Yale University Press, New Haven, CT) Dahl R A, 1985 A Preface to Economic Democracy (University of California Press, Berkeley, CA) Dahl R A, Lindblom C E, 1953 Politics, Economics and Welfare (Harper and Row, New York) de Neufville J, 1987, "Knowledge and action, making the link" Journal of Planning Education and Research Fainstein N I, Fainstein S S, 1979, "New debates in urban planning: the impact of Marxist theory within the United States" International Journal of Urban and Regional Research Faludi A, 1987 A Decision-centred View of Environmental Planning (Pergamon Press, Oxford) Forester J, 1980, "Critical theory and planning practice" Journal of the American Association of Planners Friedmann J, 1987 Planning in the Public Domain (Princeton University Press, Princeton, NJ) Giddens A, 1973 The Class Structure of the Advanced Societies (Century Hutchinson, London) Giddens A, 1979 Central Problems in Social Theory (Macmillan, London) Gottdiener M, 1987 The Decline of Urban Politics, Political Theory and the Crisis of the Local State (Sage, Beverly Hills, CA) Hall S, 1977, "The political and the economic in Marx's theory of classes", in Class and Class Structure Ed. A Hunt (Lawrence and Wishart, London) pp Harvey D, 1982 The Limits to Capital (Basil Blackwell, Oxford) Harvey D, 1985a, "Class structure and the theory of residential differentiation", in The Urbanization of Capital Ed. D Harvey (Basil Blackwell, Oxford) pp Harvey D, 1985b, "The place of urban politics in the geography of uneven capitalist development", in The Urbanization of Capital Ed. D Harvey (Basil Blackwell, Oxford) pp Harvey D, 1985c, "On planning the ideology of planning", in The Urbanization of Capital Ed. D Harvey (Basil Blackwell, Oxford) pp Harvey D, 1985d Consciousness and the Urban Experience (Basil Blackwell, Oxford) Harvey D, 1987a, "Flexible accumulation through urbanization: reflections on 'post-modernism' in the American city" Antipode Harvey D, 1987b, "Three myths in search of a reality" Environment and Planning D: Society and Space Healey P, McNamara P, Elson M, Doak A, 1988 The British Planning System in Practice: Land Use Planning and the Mediation of Urban Change (Cambridge University Press, Cambridge) Henig J, 1982 Neighbourhood Mobilization, Redevelopment and Response (Rutgers University Press, New Brunswick, NJ)

23 Class, politics, and planning 1113 Hindess B, 1978, "Classes and politics in Marxist theory", in Power and the State Eds G Littlejohn, B Smart, J Wakeford, N Yuval-Davis (Croom Helm, Andover, Hants) pp Hirst P, 1977, "Economic classes and politics", in Class and Class Structure Ed. A Hunt (Lawrence and Wishart, London) pp Hoch C, 1984, "Doing good and being right, the pragmatic connection in planning theory" Journal of the American Planning Association Hoff J, 1985, "The concept of class and public employees" Acta Sociologica Houseman G L, 1982 City of the Right, Urban Applications of American Conservative Thought (Greenwood Press, Westport, CT) Jessop B, 1982 The Capitalist State, Marxist Theories and Methods (Martin Robertson, Oxford) Jones G W, 1986, "The crisis in British central - local relationships", paper presented at the Conference of the Structure of Government Research Committee of the International Political Science Association, Pittsburgh, 12 December 1986, "Governing in the Midst of Change"; paper obtainable from G W Jones, London School of Economics, Houghton Street, London WC2 Kaufman J L, Jacobs H M, 1987, "A public planning perspective on strategic planning" Journal of the American Planning Association King R J, 1989, "Capital switching and the role of ground rent: 1. Theoretical problems" Environment and Planning A Low N P, Power J M, 1984, "Policy systems in an Australian metropolitan region: political and economic determinants of change in Victoria" Progress in Planning McConnell G, 1967 Private Power and American Democracy (Alfred A Knopf, New York) Marx K, 1962a The Class Struggles in France in Selected Works volume 1 (Foreign Languages Publishing House, Moscow) pp Marx K, 1962b The Eighteenth Brumaire of Louis Bonaparte in Selected Works volume 1 (Foreign Languages Publishing House, Moscow) pp Marx K, 1973 Grundrisse (Penguin Books, Harmondsworth, Middx) Massey D, 1984 Spatial Divisions of Labour (Macmillan, London) Mill J S, 1971 On Liberty (Dent, London) Mollenkopf J, 1983 The Contested City (Princeton University Press, Princeton, NJ) Moser S T, Low N P, 1986, "The central business district of Melbourne and the dispersal and reconcentration of capital" Environment and Planning A Olson M, 1965 The Logic of Collective Action (Harvard University Press, Cambridge, MA) Parkin A, 1984, "The states and the cities: a comparative perspective", in Australian Urban Politics, Critical Perspectives Eds J Halligan, C Paris (Longman Cheshire, Melbourne) pp Paterson J L, 1984 David Harvey's Geography (Croom Helm, Andover, Hants) Poulantzas N, 1973, "On social classes" New Left Review number 78, Poulantzas N, 1975 Classes in Contemporary Capitalism (New Left Books, London) Przeworski A, 1980, "Material bases of consent: economics and politics in hegemonic systems" Political Power and Social Theory Przeworski A, Wallerstein M, 1982, "The structure of class conflict in democratic capitalist societies" American Political Science Review Resnick S A, Wolff R D, 1987 Knowledge and Class, a Marxian Critique of Political Economy (University of Chicago Press, Chicago, IL) Roemer J, 1982 A General Theory of Exploitation and Class (Harvard University Press, Cambridge, MA) Roemer J, 1988 Free to Lose, an Introduction to Marxist Economic Philosophy (Harvard University Press, Cambridge, MA) Roweis S T, 1983, "Urban planning as professional mediation of territorial politics" Environment and Planning D: Society and Space Saunders P, 1981 Social Theory and the Urban Question (Century Hutchinson, London) Smith N, 1984 Uneven Development (Basil Blackwell, Oxford) Smith N, 1987a, "Dangers of the empirical turn: some comments on the CURS initiative" Antipode Smith N, 1987b, "Rascal concepts, minimalizing discourse, and the politics of geography" Environment and Planning D: Society and Space Thrift N, 1987, "Introduction, the geography of late twentieth century class formation", in Class and Space, the Making of Urban Society Eds N Thrift, P Williams (Routledge and Kegan Paul, Andover, Hants) pp