Chapter 6
Class Struggle, Political Power, and the Capitalist State

The preceding chapters have attempted to demonstrate the explanatory power of Althusser's concepts of structural causality and ideological interpellation. Chapters 1 and 2 show how structural causality and contradiction provide a new theoretical foundation for the concept of a mode of production and a modernist, non-essentialist explanation of economic determination (in the last instance) and the relative autonomy (specific effectivity) of economic as well as non-economic structures and relations. Chapters 3, 4, and 5 shift their focus to the domain of ideology and the constitution of social subjectivity through processes of psychological identification and subject-centered discursive practices, the bounded yet open-ended structure of practice within a given habitus, and finally the material power of institutions and rituals to (re)produce social subjects "who know their place." This chapter turns to the political instance. In particular, it elaborates the Structural Marxist concept of political power and explains the relative autonomy of the state in contemporary capitalism. We will also situate the "regional" domain of the political into the general framework of structural causality, economic determination, and ideological interpellation. Central to this discussion will be an elaboration of concepts of class and class struggle (terms we have as yet left unspecified) within industrial capitalist societies.

The present chapter is largely concerned with the work of Nicos

Poulantzas, a Greek Communist who lived in political exile in Paris during the sixties and seventies until his suicide in 1979. Although he remained aloof from the politics of Althusser and his circle, Poulantzas was profoundly influenced by Althusser's theoretical work.^[1] In a series of important books written between 1968 and his untimely death—Political Power and Social Classes (1973), Fascism and Dictatorship (1974), Classes in Contemporary Capitalism (1975), The Crisis of the Dictatorships (1976a), and State, Power, Socialism (1978)—Poulantzas produced the basic texts of Structural Marxist political theory. In the following pages, we will examine Poulantzas's important concept of social class as the matrix effect of a mode of production, his interpretation of the political instance as a "nodal point" where the contradictions of a given mode of production are "condensed" or concentrated, his concept of the "power bloc" and the relationship between class power and the state, and finally his brilliant analysis of the capitalist state and class struggle in relation to the internationalization of capitalism in the twentieth century. Poulantzas has had a decisive influence on Marxist political theory, and his wide-ranging work has provided a fruitful point of departure for many empirical and theoretical works. My discussion, however, is necessarily restricted to those aspects of Poulantzas's thought most relevant to our present concerns and to the works of others—Erik Olin Wright, Michel Aglietta, and Göran Therborn—only insofar as they supplement, criticize, and extend Poulantzas's problematic.

Wright: Modes of Determination Within Structural Causality

I begin by defining with greater precision the concept of structural causality, in particular, by specifying qualitative differences in the relations of determination that may exist among the different instances articulated as the complex whole of a social formation. In Class, Crisis, and the State , American sociologist Erik Olin Wright has suggested that structural causality is a global concept that may be broken down into six distinct forms of causality or modes of determination (Wright 1978, 15-25):

1. Structural limitation is a pattern of determination in which some social structure establishes limits within which some other structure or process can

vary and which establishes probabilities for the specific structures or processes that are possible within those limits.

2. Selection constitutes those social mechanisms that concretely determine ranges of outcomes or, in the extreme case, specific outcomes, within a structurally limited range of possibilities. In a sense, selection is a second-order form of limitation: the setting of limits within limits. Wright specifies two complementary forms of selection, positive (mechanisms that determine specific outcomes among those that are possible) and negative (mechanisms that exclude certain possibilities).

3. Reproduction/non-reproduction refers to the facilitation/inhibition of change within one structure by a second. To say that one structure functions to reproduce another implies that the reproducing structure prevents the reproduced structure from changing in certain fundamental ways. This process differs from limitation, Wright maintains, since in the case of limitation, there is no presumption that the determined structure would necessarily change, whereas in the case of reproduction such changes would normally occur. In addition, reproduction/non-reproduction is a variable relation of determination since it implies that the determining structure may become non-reproductive vis-à-vis the determined structure.

4. Limits of functional compatibility specify the mechanisms by which structural forms will be functional with respect to other structural forms and in what ways—in other words, which forms will be reproductive and which non-reproductive.

5. Transformation refers to a mode of determination by which class struggle (practice) directly affects the processes of structural limitation, selection, and reproduction/non-reproduction.

6. Mediation defines a mode of determination in which a given social process shapes the consequences of other social processes. In other words, it is a "contextualizing variable" that shapes the very nature of the relationship between two other variables.

Wright's modes of determination lend greater precision to general concepts such as overdetermination, structural causality, relative autonomy, determination in the last instance, and structure in dominance. Because his terms are capable of greater specification, Wright believes that they are more open to empirical verification as hypotheses and considerably enrich our ability to grasp theoretically a wide variety of concrete social formations. As we have repeatedly insisted, the method of Structural Marxism is not rationalist or idealist; concrete social relations cannot be deduced or read off from abstract concepts. Structural Marxism is neither reductive nor essentialist. It does not dogmatically eliminate the particular by means of the general; rather, it grasps the particular through the general and revises the general in light of the

Figure 1.
The relationship between the economy and the state (Wright 1978, 20).

particular. It is realist and materialist in the sense that structural relationships are held to have an objective, real existence and that we can obtain objective knowledge of them.

Structural relationships have material effectivities that are not only qualitatively distinct but also quantitatively unequal, stratified by their relative determinative significance or dominance. Unfortunately, as Wright explains them, the various modes of determination lack any hierarchical relationship to each other. Because they lack determinative weights themselves, or more accurately because the economic, political, and ideological instances whose articulation they define are devoid of relations of dominance, the modes of determination described by Wright exist in a circular or feedback relationship whereby everything somehow causes everything else. For example, Wright posits the reciprocal relationship between the economy and the state as shown in figure 1.

While one may agree with Wright's description of the relationship between the economy and the state as reciprocal—the economy determining the state by a relation of limitation; the state determining the economy by a relation of reproduction/non-reproduction—it remains unclear which structure, and therefore which relation, is dominant. The problem is compounded as Wright proceeds to add "class struggle" to his model, as shown in figure 2. Now we are faced with the question of the causal priority between relations of transformation and limitation that obtain among "class struggle," the economy, and the state.

Wright attempts to escape the problem by recourse to the "dialectical character of patterns of determination," but this is less an explanation than an admission of confusion: "class struggle, which is itself structurally limited and selected by various social structures, simultaneously

Figure 2.
The relationships among the economy, the state, and class struggle (Wright 1978, 22).

reshapes those structures. The word 'simultaneously' is important in this formulation: social structures do not first structurally limit and select class struggle, after which class struggle transforms those structures. Class struggle is intrinsically a process of transformation of structures, and thus the very process which sets limits on class struggle is at the same time transformed by the struggles so limited" (Wright 1978, 21). The use of "simultaneously" in this passage is obfuscatory and begs the question of structural dominance. The fact that different modes of determination are simultaneously at work is no argument for their equality. Wright also slips from structural to transitive causality and ends up focusing on an endless chain of cause and effect, unable to advance beyond a superficial description of patterns of events. Such a transitive "dialectic" is incapable of discerning the existence of generative structural mechanisms beneath patterns of events, mechanisms whose unequal effectivities determine precisely why certain patterns occur and not others. Finally, Wright's description of class struggle as a process that sets limits "at the same time [that it is] transformed by the struggles so limited" blurs the significant distinction between the irreversible, or intransitive, articulation of structural forces constituting social subjects (and every possible practice) with the interactive, open-ended, or transitive, practices of individuals (invested with and constrained by social powers).

To eliminate the circular, transitive causality implied by Wright's model and transcend the classic hermeneutic fallacy of theoretical hu-

manism (and political voluntarism) that underlies his conception of "class struggle," it is necessary to insist on the hierarchical stratification, in descending order, of the economic, political, and ideological instances. I will not repeat arguments already made regarding the primacy of the economic instance and the tertiary effectivity of the ideological field of institutional apparatuses, social subjects, and the habitus, nor will I anticipate the discussion of the intermediate effectivity of the political instance that is the subject of the present chapter. Here I wish only to clarify the term class struggle , which in its present form obscures rather than illuminates the relationship between structural determinations and the practice of social subjects. In Wright's usage the concept of class struggle is an empty generalization because he excludes the ideological instance and thus the very processes by which social subjects are constituted as class subjects. Wright evades, in effect, the central question of what class struggle actually is in order to assert not only its existence but also its primacy, that is, its transformative effect on political and economic structures.

In posing the question of class struggle, it is necessary to represent the field of practice not, as Wright does, as self-evidently inscribed with economic determinations and political values yet at the same time independent and coequal with respect to political and economic structures, but rather, following Althusser and Poulantzas, as the "matrix effect" of the articulation of economic, political, and ideological structures on the capacity of social subjects to act. That the field of practice is, in the last instance, a field of class practices and class struggle follows from the primacy of economic relations within the articulation of the instances whose matrix effect the field of practice is. In other words, practices are class practices insofar as they are activities of social subjects interpellated by the hierarchical articulation of the economic, political, and ideological instances, as members of what Poulantzas defines as social (as opposed to simply economic) classes. Similarly, the field of practice, what Pierre Bourdieu calls the "social space," is a class field insofar as the structure of positions and powers that constitutes it is the matrix effect of this same articulation. The concept of social class will be discussed at greater length in a moment. I introduce it now to explain why, in figure 3, the concept of the ideological instance, composed of ideological apparatuses and the habitus, is used in place of the term class struggle employed by Wright. This change requires yet another since it is now necessary to add the human agent to the model in

Figure 3.
Modes of determination in contemporary capitalism.

order to distinguish the ideological instance from individual human beings interpellated as social class subjects.

Modes of Determination in Contemporary Capitalism

Figure 3 schematically depicts the articulation of instances characteristic of contemporary capitalism. As in every social formation, the economic function is primary and determinant in the last instance; however, in capitalist modes of production (unlike lineage and feudal modes) the relations and forces of production are formally combined within a distinct complex of economic institutions comprising the eco-

nomic instance. Because the function of economic ownership resides in the economic instance, the latter is the dominant instance of a capitalist mode of production. In figure 3, the dominance of the economic instance is qualitatively expressed as a relation of limitation wherein the economic instance limits the range of possible variations of the political and ideological instances, although it does not directly transform them. Furthermore, because it exercises the ownership function—that is, it controls access to the means of production and the distribution of the social product—the economic instance defines the limits of functional compatibility of the modes of determination exercised by the state on itself—the extent to which the economy can be reproduced or not reproduced (facilitated or inhibited) by the state. Of course, the relations of limitation and the limits of functional compatibility exercised by the economy over the state are not coordinated directly. The absence of direct coordination (reflecting the relative autonomy and uneven development of the social formation) makes it possible, even inevitable, that contradictions between the state and the economy—for example, political practices that are non-reproductive with respect to the economy—will emerge. The economy also exercises an ideological function. By the mode of selection the economic instance determines social subjectivity; it interpellates individuals with respect to their "place" in the forces and relations of production, while the nature and number of available places sets an objective limitation on the practice of interpellated social subjects.

The relative autonomy of the capitalist state is manifested in the formal separation of the political function from the economic instance and from the direct control of the dominant economic class. Although operating within limits imposed by the dominance of the economy (limits that specifically include separate political and economic instances), the capitalist state must have considerable powers not only to reproduce the basic conditions of capitalist accumulation as in the "nightwatchman" state of classical liberalism but also, with the transition to monopoly capitalism, to "regulate" and actively intervene in the economy. Subordinate to the economic instance, the political instance is dominant with respect to the ideological instance. The capitalist state functions as the organizer of the hegemony of the dominant class and class fractions, and by means of its legal code and administrative apparatuses, it limits the practice of human agents to a range of activities compatible with that hegemony. The state determines ideological apparatuses insofar as it declares them legal or illegal, supervises or regulates their

operation, and gives or withholds support (financial and otherwise). By this more specific relation of selection exercised by the state on the various ideological apparatuses, the state indirectly determines the interpellation of social subjects and thus the practice of human agents.

The state directly determines the practice of human agents through the interpellative effects of its own apparatuses, particularly the public school system, where individuals become "good citizens" and patriotic members of the "nation." However, like the economy, the political instance has a polarizing as well as an integrative influence on social subjectivity. While the economy polarizes agents around relations of exploitation and cooperation (ownership of means of production, control over organizational positions and skills, and so forth), politics polarizes social subjects around relations of representation and domination. These bipolar and contradictory interpellations may be specified, following Poulantzas, as "the people" and "the power bloc." Struggles for political power take place in terms of "rights" and "justice" and thus assume a "popular-democratic" rather than an economic-class character.

The ideological instance and the habitus limit the human agent by their objective existence as the social space within which all practice must occur, and select the human agent as a social subject with a unique, often contradictory combination of roles, identities, and capacities. The relations of determination exercised by the economy, the state, and the ideological apparatuses on the human agent are no more directly coordinated with the relations of transformation emanating from the latter than was the case with the relations between the economy and the state. This condition is precisely what makes transformation possible. The human agent—the interpellated effect of the contradictory articulation of instances—is the social subject of all practice, whether antagonistic or non-antagonistic with respect to the dominant mode of production. Thus, the mode of determination of the human agent is always and exclusively transformation.

Economic and political structures define positions or places within social space, and they interpellate human agents on the basis of the positions and places they occupy. However, economic and political relations do not completely define the social space, nor do they exclusively define concrete human individuals who are the agents of all political and economic practice. This tension means, on the one hand, that the effectivity of the ideological instance with respect to the individual agent cannot be reduced to economic or political forms. Although subordinate to both the economic and political instances, the ideological

instance not only mediates relations among itself, the economy, and the state—it is only through the values, meanings, and strategies formed within the habitus that the economy and the state can function at all—but it also interpellates social subjects who are more than simply citizens or workers. On the other hand, the ideological field is limited and selected by the dominance of economic and political structures. As a result, the capacity of a social subject to act within the fields of economic, political, or ideological relations is always already structured insofar as its condition of existence is the intransitive articulation of the instances and their distinct and unequal effectivities. Given the primacy of the economic instance within the social formation, the practice of social subjects is always already a class-structured practice.

To prevent a serious misunderstanding, I should point out that each of the instances depicted in figure 3 (and thus their various modes of determination) is assigned its place and function by the historical matrix of the structure as a whole. In other words, the intransitive effectivity of the articulation of the instances, what Althusser calls the "matrix effect of a mode of production," determines the transitive effectivities, the distinct and unequal modes of determination of the economic, political, and ideological structures. Similarly, the articulation of the instances, as internalized by social subjects and manifested as social space (the field of practice), defines all possible practices and gives to them a determinate class-based character. It is Poulantzas's original achievement to have identified this indirect, intransitive, and overdetermined class relationship and to have given it a name: social class. Poulantzas has also clarified an important distinction between economic and social class relations about which we must be absolutely clear. Economic class relations, defined by the forces and relations of production, exert direct or transitive determinations on the political and ideological instances (as depicted by the arrows in figure 3); social class relations, defined by the historical matrix, exert an indirect or intransitive effectivity that is always already there with respect to economic (as well as political and ideological) relations as their assigned place and function within the complex whole (depicted in figure 3 as the relative dominance of the instances and the qualitative character of their modes of determination). Thus economic class relations are always already social class relations, but it is as economic (social) class relations that they exert their dominance over political and ideological (social class) relations in capitalist modes of production.

Similarly, by means of the matrix effect of the articulation of in-

stances, the capitalist state is always already a class state insofar as it has been assigned a place and a function by an articulation dominated by the capitalist mode of production. It acts as a class state because its apparatuses are limited by the primacy of the economy. It remains a class state to the extent that social class relationships, actualized within the social space and the habitus by social class subjects, remain polarized around the objective economic class interests of the dominant class(es) and not the exploited class(es)—the extent, in other words, to which the ideological discourse of the ruling class is hegemonic within the field of ideological interpellations, thereby producing agents whose practice reproduces the existing relations of production and the existing structure of the state. The state's ability to fulfill its structural function of reproducing the "rule" of the ruling classes depends on its ability to minimize the contradictions between domination and execution within the state itself and mask their interpellative effects, that is, the distinction between the power bloc and the people. Although the capitalist state is not the source of the power it wields (power has its ultimate source in the forces and relations of production), it has acquired, by means of its separation from the economy and the resulting ideological consequences of this separation, the exclusive right to power. The power of the state is particularly difficult to undermine not simply because it has acquired its own coercive apparatuses and monopolized the means of violence but also because its relative autonomy effectively conceals the economic taproot of political power behind a mask of popular-democratic forms.

The modes of determination depicted in figure 3 are provisional. While the economic, political, and ideological instances—combined within a single institutional framework (a lineage) or taking separate institutional forms (a state, an enterprise)—and their hierarchy are constant social facts, their modalities are presented here as hypotheses, not axioms. I do not seek to limit dogmatically the modes of determination to just the six forms described by Wright, nor to imply that these forms might not be revised considerably in light of substantive variations of the economic, political, and ideological instances in different modes of production and social formations. Here I have sought only to demonstrate how political practice and the capitalist state might be explained in terms of economic determination without being reduced thereby to reflections of the economy.

Poulantzas: The Specificity of the Political Instance

In Political Power and Social Classes , completed on the eve of the events of May 1968, Poulantzas sets out to develop a "regional theory" of the political instance in capitalist social formations. In particular, he seeks to resolve one of the most perplexing theoretical problems of Marxist social theory: how the capitalist state can be both a class state yet at the same time formally distinct from the economy and the direct control of the capitalist class—in other words, how class power, the product of a vast network of social struggles that are in the last instance economically determined, is concentrated in and deployed by the political apparatuses of a capitalist state that is typically popular-democratic. The key to Poulantzas's solution to this important and difficult problem, indeed, the central insight on which his entire body of work is built, is a creative application of Althusser's concepts of structural causality, overdetermination, dominance, and economic determination in the last instance, reformulated by Poulantzas as the concept of the "matrix effect" of a mode of production, "the articulation of social instances on the basis of a dominant mode of production" (Poulantzas 1973, 15).

By the term mode of production , Poulantzas refers essentially to what I have previously defined as an extended mode of production, that is, not only the forces and relations of production proper (what I have called a restricted concept of a mode of production) but also those political and ideological relations indirectly determined by the economic instance and essential for the reproduction of the latter. "By mode of production," Poulantzas explains, "we shall designate not what is generally marked out as the economic (i.e., relations of production in the strict sense), but a specific combination of various structures and practices which, in combination, appear as so many instances or levels, i.e., as so many regional structures of this mode. . . . The type of unity which characterizes a mode of production is that of a complex whole dominated, in the last instance, by the economic. The term determination will be reserved for this dominance in the last instance" (Poulantzas 1973, 13-14).

Although he does not himself use the term extended mode of production , Poulantzas clearly distinguishes the structure of the forces and relations of production and their specific effectivity (the transitive mode of determination exercised by the economy on the political and ideo-

logical instances) from the combined or overdetermined effectivity of the articulation of the economic, political, and ideological instances peculiar to the capitalist mode of production (the matrix effect that constitutes the intransitive determination by the structured whole on the specific effectivity of each instance). Moreover, both the forces and relations of production proper to the economic instance and their matrix effect on the social whole are necessary for an adequate concept of a mode of production. "What distinguishes one mode of production from another and consequently specifies a mode of production is the particular form of articulation maintained by its levels: this articulation is henceforth referred to by the term matrix of a mode of production. So to give a strict definition of a mode of production is to lay bare the particular way in which determination in the last instance by the economic is reflected inside that mode of production: this reflection delimits the index of dominance and overdetermination of this mode" (Poulantzas 1973, 15).

Thus when Poulantzas refers to the concept of the political as a "regional instance" that is "specifically political," he is speaking not only of the relative autonomy of the political, that is, what distinguishes political practices and apparatuses from those that are economic or ideological, but also, and with even greater emphasis, of the "place and function [of the political] in the particular combination which specifies [the] mode of production" (Poulantzas 1973, 17). Because the matrix effect of a mode of production "governs the constitution of its regional instances" and "defines the extension and the limits " of each regional instance (Poulantzas 1973, 17), Poulantzas's well-known emphasis on the "relative autonomy" of the state in capitalist social formations—the formal separation of political from economic apparatuses characteristic of capitalism—must be understood not in terms of the independence of the state from the capitalist mode of production but rather as the particular place and function assigned to the state by the structure of a capitalist mode of production.^[2]

The specificity of the political instance is defined by Poulantzas in terms of its "global function" of maintaining (or transforming) the unity of a social formation. "The state has the particular function of constituting the factor of cohesion between the levels of a social formation." The state is "a factor of 'order' or 'organizational principle' of a formation," Poulantzas maintains, "in the sense of the cohesion of the ensemble of the levels of a complex unity, and as the regulating factor of its global equilibrium as a system" (Poulantzas 1973, 44-45). The

state is able to exercise this function (which I have further broken down into the modes of determination of reproduction/non-reproduction, selection, and limitation) by virtue of the fact that the political instance is the central "nodal point" where power is "condensed," that is, formally concentrated and legitimized, and the site from which power is deployed throughout the entire social formation.

While the state is the site where power is formally concentrated, it is not the source of the power that flows through it. Poulantzas defines power not in terms of a function located in a particular institutional apparatus but rather as a relation of forces, a structured force field that is the product of social struggles overdetermined, via the matrix effect, by economic class relations' and practices. All social struggles, and thus all power relationships, have a class character that is both explicit and implicit, directly and indirectly determined by the mode of production. The condensation of power in the state apparatuses in no way alters the class nature of power and its economic taproot. Even though the state cannot be said to have any power of its own, Poulantzas insists that political power is "primordial" insofar as the institutionalized power of the state "constitutes the field" where the unity of the formation is either maintained or transformed.

The fact that power originates in social struggles overdetermined by economic class relations and practices means that the condensation of power in the political apparatuses of the state cannot be a neutral process. Whatever the degree of formal separation between political and economic apparatuses, in class societies state power is always class power. In a given mode of production the state may exercise apparently representative or technical economic, administrative, judicial, and legislative functions, but these "local" functions are always overdetermined by the matrix effect and by the "global" function of the political instance, which organizes and maintains the unity of the social formation. Because cohesion always reflects the political interests of the dominant class, this global political function marks the class character of the state in all class societies, including those dominated by capitalist modes of production:

the state's economic or ideological functions correspond to the political interests of the dominant class and constitute political functions, not simply in those cases where there is a direct and obvious relation between (a) the organization of labor and education and (b) the political domination of a class, but also where the object of these functions is the maintenance of the unity of a formation, inside which this class is the politically dominant class.

It is to the extent that the prime object of these functions is the maintenance of this unity that they correspond to the political interests of the dominant class. (Poulantzas 1973, 54)

That they constitute the point where power is condensed and deployed, as well as the site where the cohesion of social formation is maintained or transformed, is another way Poulantzas expresses the fact that the political apparatuses of the state reflect the articulation of a social formation as "the nodal point where the contradictions of the various levels of a formation are condensed in the complex relations governed by overdetermination and by their dislocation and uneven development" (Poulantzas 1973, 41). Because the state constitutes the "strategic point where the various contradictions fuse," political practice, whose object is to maintain or transform the unity of a formation, must have the political structures of the state as its point of impact and specific strategic objective: "the specificity of political practice depends on its having state power as its objective" (Poulantzas 1973, 43). Given the place and function of the political instance, political practice must be considered "the motive force of history."

[The political instance is] the place in which we find reflected the index of dominance and overdetermination which characterizes a formation or one of its stages or phases. The state is also the place in which we can decipher the unity and articulation of a formation's structures. . . . It is from this relation between the state as the cohesive factor of a formation's unity and the state as the place in which the various contradictions of the instances are condensed that we can decipher the problem of the relation between politics and history. This relation designates the structure of the political both as the specific level of a formation and as the place in which the transformations occur: it designates the political struggle as the "motive power of history" having as its objective the state, the place in which the contradictions of instances (dislocated in their own time sequences) are condensed. (Poulantzas 1973, 45)

Social Class, Social Relations, and Class Struggle

The concept of the matrix effect of a mode of production, when applied to the concept of class struggle, yields a promising solution to yet another long-standing problem of Marxist social theory, the relation between class determination and class practice. In working out his concept of class struggle, Poulantzas seeks an alternative to subjectivist and objectivist concepts of class. In opposition to subjectivist views, which

define class by the consciousness of the agents as members of a class, Poulantzas conceptualizes class as an objective field of structural relations or "class determinations," not as a subjective field of consciously chosen "class positions." While class determinations are held to "fix the horizon of the class struggle," Poulantzas rejects the notion of a rigid and precise correspondence between objective class interests and subjective class positions. Therefore, in opposition to objectivist views that equate class with places in the forces and relations of production in order to reduce consciousness to a reflection of class interests, Poulantzas insists on a concept of class that acknowledges and explains discrepancies between objective economic determinations and conjunctural political and ideological positions.

Poulantzas recognizes a hierarchy of dominance within the capitalist mode of production—the primacy of the economy and the secondary and tertiary effectivities of the political and ideological instances. However, he insists that classes must be defined in terms of the social formation as a complex whole, the "ensemble of structures," and not in terms of the economy alone: social classes are not merely economic classes.

Marx's analysis of social classes never refers simply to economic structures . . . but always to the ensemble of structures . . . and to the relations which are maintained there by the different levels. Let us anticipate and say that everything happens as if social classes were the result of an ensemble of structures and of their relations, firstly at the economic level, secondly at the political level and thirdly at the ideological level. A social class can be identified either at the economic level, at the political level, or at the ideological level, and can thus be located with regard to a particular instance. Yet to define a class as such and to grasp it in its concept, we must refer to the ensemble of levels of which it is the effect. (Poulantzas 1973, 64)

By virtue of the matrix effect, economic class relations are never purely economic, and political and ideological relations are always invested with a social class character. It is helpful to remember that the adjective "social" is always used by Poulantzas to distinguish the in-transitive effectivity of the matrix effect of a mode of production from the transitive effectivity exercised by any particular instance on another. The matrix effect of the mode of production in the field of social relations is referred to by Poulantzas as the social division of labor and is organized by relations of social class. According to Poulantzas, "social classes do not present themselves as the effect of one particular structural level on another structural level: i.e., as the effect of the economic

structure on the political or ideological structures; hence they do not manifest themselves inside the structure, but entirely as the global effect of the structures in the field of social relations which, in class societies, themselves involve the distribution of agents/supports to social classes" (Poulantzas 1973, 64).

Like the economic, political, and ideological instances, the interpellated social subject is also a product of the matrix effect in the sense that individual agents identify with and internalize the complex whole as a structured system of roles and relations. Poulantzas refers to this global system of relations as social relations (as opposed to economic, political, or ideological structures and relations), and he insists that it cannot be represented within the social formation because it is always already there, manifested or actualized as the structure of social space, the habitus, and the consciousness of social subjects.

Social class is a concept which shows the effects of the ensemble of structures, of the matrix of a mode of production or of a social formation on the agents which constitute its supports: this concept reveals the effects of the global structure in the field of social relations . In this sense, if [social] class is indeed a concept, it does not designate a reality which can be placed in the structures: it designates the effect of an ensemble of given structures, an ensemble which determines social relations as class relations. . . . There is no theoretical homogeneity between the concept of class, denoting social relations, and those concepts denoting structures. (Poulantzas 1973, 67-68)

To put the same thing a bit differently, Poulantzas is saying that a double order of class determination is at work: a global, historical, indirect, and intransitive determination and a regional, contemporary, direct, and transitive determination. The intransitive effectivity of a capitalist social formation, that is, social class relations or the "social division of labor," has indirectly interpellated agents as social subjects of capitalism even as the transitive effectivities of economic, political, or ideological structures of the social formation interpellate them as, say, proletarians, citizens of Texas, and wives and mothers. The regional instances of capitalism define places and relations that are objectively real and whose material effectivity is selective with respect to the interpellation of social subjects and limiting with respect to their practice. However, the regional political, ideological, and economic structures always already bear the mark of the complex whole that has assigned them a place and a function; they are thus always already capitalist social relations and never only just political, ideological, and economic relations.

We clearly distinguish [regional] structures and social relations . . . the latter designating the distribution of supports [interpellated subjects] to social classes. These two fields are covered respectively, at the economic level, by the concept of relations of production . . . and by that of social relations of production. . . . [W]e can also speak in all strictness of political "social" relations and of ideological "social" relations. These social relations, as class relations, isolated here with respect to the instances of the political and the ideological, manifest themselves as the effect of the political and ideological structures on social relations. (Poulantzas 1973, 65-66)

For Poulantzas, class struggle is the global effect, in the field of social relations, of the contradictions and uneven development of the social formation as a whole. As the constitution of social classes is related not to the effectivity of the economic instance alone but to the matrix effect of the ensemble of the instances, within which economic determinations are primary, so the field of social relations, structured by the ensemble, assigns a social class function and power to each and every practice. While economic class relations exercise direct determinations on political and ideological relations, Poulantzas calls our attention to the even more pervasive influence of the indirect determination of the matrix effect and social class. Directly, but even more significantly, indirectly, the social space is a field of class struggle: "the organization of instances in economic, political and ideological levels is reflected, in social relations, in economic, political and ideological class practices and in 'struggle' between the practices of the various classes. Since social relations are a field-cum-effect structured from the system of structures, the levels of class struggle are related in the same kind of way as the instances of the matrix" (Poulantzas 1973, 69).

Social classes are necessarily antagonistic. Social space, reflecting the articulation of economic, political, and ideological structures and relations, constitutes a unity, the field of class struggle, but a unity that is dislocated by relations of dominance and subordination determined, in the last instance, by economic structures and relations.

Social relations consist of class practices in which social classes are placed in oppositions: social classes can be conceived only as class practices, these practices existing in oppositions which in their unity, constitute the field of the class struggle . . . . Class practices can be analyzed only as conflicting practices in the field of class struggle, composed of relations of opposition, i.e., of relations of contradiction in the simplest sense of the term. At every level the relation of conflict of the practices of different classes, "struggle" between classes, the very existence of classes themselves, are the effect of the relations of the structures, the form assumed by the contradictions of the structures in social relations. They define at every level fundamental connec-

tions of domination and subordination of classes (class practices) which exist as particular contradictions. (Poulantzas 1973, 86)

Pertinent Effects, Class Power, and the Predominance of Politics

By means of his concepts of social class and social relations, Poulantzas has demonstrated how class struggle exists in political and ideological relations typically viewed as unrelated to class conflict or at best related to it by historical accident (elective affinity). He has also reformulated the traditional Marxist thesis of primacy of class relations without slipping into reflectionism or essentialism. However, the fact that economic relations are always already social class relations does not vitiate the fact that it is as economic class relations that they exert their modes of determination within the social formation. It is in this sense that Poulantzas does acknowledge a kind of reflection or "presence" of the economic within the other instances. The presence of economic class interests may be discerned in what Poulantzas calls the pertinent effects of economic relations on political and ideological structures and relations.

The "bearers" of pertinent effects are non-class relations and functions—for example, state bureaucracy, academia, media journalism—which Poulantzas defines as social categories. The concepts of social categories and pertinent effects make it possible to distinguish, within political or ideological fields, between relations and functions that are generically political or ideological and those that "reflect" the presence of economic class interests within the political or ideological instances. Pertinent effects reflect the primacy of economic modes of determination and their externality, the fact that the political or ideological structure in question would not in and of itself produce these effects. "The reflection of the place in the process of production on the other levels constitutes a new element which cannot be inserted in the typical framework which these levels would present without this element. This element thus transforms the limits of the levels of structures or of class struggle at which it is reflected by 'pertinent effects,' and it cannot be inserted in a simple variation of these limits" (Poulantzas 1973, 79).

Poulantzas also distinguishes between economic classes and class fractions . Whereas economic classes are distinguished by the predominant relation of exploiter-exploited, class fractions are defined as "sub-strata" of classes distinguishable by a combination of secondary economic criteria (for example, the size or nature of enterprises) and the

different political and ideological effects produced by such combinations. In other words, secondary economic distinctions produce class fractions precisely insofar as they generate pertinent effects. The capacity of agents acting as members of class fractions to produce and reproduce pertinent effects constitutes class relations of power and class fractions as "social forces."

Since all social relations are overdetermined by the matrix effect of the dominant mode of production, all power is indirectly class power. The classic Weberian definition of power is thus redefined by Poulantzas as "the capacity of a social class to realize its specific objective interests" (Poulantzas 1973, 104). Economic class relations are relations of power, Poulantzas argues, not in the sense that one is the foundation of the other but in the sense that they are constituted in a "homogeneous field," the field of social relations or simply the class struggle. Moreover, as the specific effectivity of each instance is organized by the matrix effect of the mode of production that gives it a place and a function, so power is, in the last instance, an effect of the intransitive complex whole of the social formation and not of the transitive effectivity of any particular instance intervening in the field of another.

Class relations are no more the foundation of power relations than power relations are the foundation of class relations. Just as the concept of class points to the effects of the ensemble of the levels of the structure on the supports, so the concept of power specifies the effects of the ensemble of these levels on the relations between social classes in struggle. It points to the effects of the structure on the relations of conflict between the practices of the various classes in "struggle. " In other words, power is not located in the levels of structures, but is an effect of the ensemble of these levels, while at the same time characterizing each of the levels of the class struggle. (Poulantzas 1973, 99-100)

The class power deployed by social subjects engaged in economic, political, or ideological practice is always already structured by the intransitive matrix effect of the mode of production such that the concept of power cannot be applied to any one level of the structure. Even in the case of state power, Poulantzas insists that the indirect, social class structure of power precedes and determines its direct deployment. "When we speak for example of state power, we cannot mean by it the mode of the state's articulation and intervention at the other levels of the structure; we can only mean the power of a determinate class to whose interest (rather than those of other social classes) the state corresponds" (Poulantzas 1973, 100).

If the foundation of power is intransitive, power is nevertheless

"globally distributed" throughout the social formation such that its deployment is always transitive. Each application of power directly or indirectly reflects the opposition of class interests and, therefore, an intervention in the field of class struggle. Because power always designates the capacity of a social class or class fraction to realize its objective interests, class power is always relational or differential: it can be defined only with respect to other relations of force constituting the class struggle. "The degree of effective power of a class depends directly upon the degree of power of other classes, in the framework of the determination of class practices in the limits set by the practices of the other classes. Strictly speaking, power is identical with these limits to the second degree" (Poulantzas 1973, 108).

Poulantzas provides a compelling alternative to both "voluntarist" and "reflectionist" views of political practice. Given that the state is the place where class struggle is "concentrated and reflected," Poulantzas argues that political struggles have a predominant place and function within the field of social relations. However, the predominance of the political is itself determined by the primacy of the economic as refracted through the matrix effect.

To the extent that the political superstructure is the overdetermining level of the levels of the structure by concentrating their contradictions and by reflecting their relation, the political class struggle is the overdetermining level of the field of class struggles (i.e., social relations), concentrating their contradictions and reflecting the relations of the other levels of class struggle. This is so to the extent that the political superstructure of the state has the function of being the cohesive factor in a formation, and to the extent that the objective of political class struggle is the state. (Poulantzas 1973, 76-77)

At this point it is perhaps useful to summarize Poulantzas's brilliant but admittedly difficult argument. Poulantzas, following Althusser, argues for economic determination "in the last instance," the primacy of the economic function and economic relations that is always already there in the matrix effect of the complex whole. The primacy of economic modes of determination within the ensemble is no more a function of the economy alone than the condensation of power in the political apparatuses of the state is a function of politics alone. With respect to political power, Poulantzas argues not for the primacy but the predominance of politics. Power is not so much monopolized by the state as assigned to it by the matrix effect of the mode of production. Poulantzas insists not only that all power is relational but also that the field of

power, the social space, is always already a field of social class struggle. Given the central place and function of the political instance with respect to the concentration and deployment of power, struggles for political power are not only social class struggles: they are the predominant form of social class struggle. At the same time, because the predominance of political practice within the field of social class struggles is determined by the matrix effect, conventional political practice tends to reproduce existing class relations, not bring them into question. Because the state is not the source of the power it deploys, Poulantzas's insistence on the predominance of political practice cannot be equated with political voluntarism. If the contradictions between the political and the economic instance become critical, it can only be as a result, in the last instance, of developments within the economy. No less than its conventional counterpart, revolutionary politics is the art of the possible, and even revolutionary practice, which must have state power as its objective, can never succeed by means of state power alone.

The Relative Autonomy of the Capitalist Class State

Poulantzas also seeks to explain how the capitalist state directly serves the interests of the capitalist class while being formally separated from the economy and from the direct control of the capitalists. Poulantzas's concept of the capitalist state is premised on its relative autonomy with respect to the economy and the fact that "it is this autonomy which, as a constant invariant, regulates the variations of intervention and non-intervention of the political in the economic, and of the economic in the political" (Poulantzas 1973, 143). The general concept of a class state, Poulantzas reminds us, does not require that the state be the direct instrument of the dominant class but only that it legitimize and reproduce the conditions and relations of domination and exploitation by which the ruling class is constituted. In capitalist social formations these conditions are defined by the existence of surplus value and the irresistible impetus to accumulate it that occurs when private property exists and labor power is completely commodified. Furthermore, in capitalist social formations, the members of the dominant class are in economic competition with each other, and their competing interests render them incapable of governing directly or with unanimity. Their only common

interests, Poulantzas concludes, are that the exploited class be politically fragmented and that the existence of propertyless laborers "free" to sell their labor power be perpetuated.

Poulantzas argues that the peculiar characteristics of the capitalist mode of production do not require a state that directly represents the economic interests of the ruling classes; rather, they require a state that represents their political interest. However democratic a capitalist state may appear to be, Poulantzas maintains that it always functions as "the dominant class's political power center, the organizing agent of their political struggles" (Poulantzas 1973, 190). The state accomplishes this function by redefining agents of production, distributed in classes, as political subjects, distributed as individuals. The result is an effect of "individual isolation" that is then projected back, via the legal system, from the political realm into the economy to mask the existence of class relationships. The capitalist state is both the source and guarantor of the "rights" of isolated political subjects and thus of its own function of representing the unity of these isolated relations, that is, the body politic of "the people" and "the nation." In other words, "the state represents the unity of an isolation which because of the role played by the ideological, is largely its own effect. This double function of isolating [individuals] and representing [their] unity is reflected in the internal contradictions in the structure of the state" (Poulantzas 1973, 134).

By means of its isolation effect on class struggles, the capitalist state provides the dominant classes with a unique mechanism, the national-popular state, capable of constituting their political interests as general interests and organizing their hegemonic power over the masses:

By its effect of isolation on the socio-economic relations, this state presents itself as the strictly political public unity of the particular, private, economic antagonisms of the ensemble of society. The institutionalized power of the capitalist state presents its own unity in its relations to socio-economic relations (the economic class struggle) in so far as it represents the unity of the people/nation, composed of agents set up, as subjects, as individual/political persons, i.e., in so far as it represents the political unity of an economic isolation which is its own effect. On the level of the relations of the state to the political class struggle, this leads to a result which seems paradoxical but which in fact constitutes the "secret" of the national-popular class state: the institutionalized power of the capitalist class state presents its own class unity , precisely in so far as it can pose as a national-popular state, i.e., as a state which does not represent the power of one or several determinate classes, but which represents the power of the political unity of private agents. (Poulantzas 1973, 276)

While it is clear enough how the capitalist state, by means of its national-popular form, functions to disorganize the dominated classes by denying their existence as classes and relating itself to them as their representative, it is less clear, at least to many of Poulantzas's commentators, how the capitalist state necessarily acts in the political interests of the dominant class.^[3] Such critics contend that the very concept of the relative autonomy of the state precludes a necessary or "reflectionist" relationship between the state and the dominant class. Poulantzas has responded that his critics have misunderstood him by mistakenly emphasizing the notion of the relative autonomy of the state while ignoring the concept of the matrix effect and the view of class power that accompanies it. The autonomy of the capitalist state, Poulantzas insists, is relative to its place and function in the ensemble of instances, and the power it deploys is only the "condensation" of struggles between the different social classes. At the same time, however, the indirect and intransitive determinations of the matrix effect on the political instance make it impossible to say that the dominant classes directly control the actions of the state. Thus the autonomy of the state must be viewed "as a 'resultant' [of the] relations of power between classes within a capitalist formation—it being perfectly clear that the capitalist state has its own institutional specificity (separation of the political and the economic) which renders it irreducible to an immediate and direct expression of the strict 'economic corporate' (Gramsci) of this or that class or fraction of the power bloc, and that it must represent the political unity of the power bloc under the hegemony of a class or fraction of a class" (Poulantzas 1976, 73).

The capitalist state is a class state because it is the condensation of social class relations and social class power. It is in this sense that the very existence of state power necessarily corresponds to the interests of the hegemonic class or class fraction. With respect to the ruling classes, the capitalist state is a class state insofar as it organizes their class powers into a political unity and insofar as it creates and maintains their political hegemony over the dominated classes. "With regard to the dominant classes and fractions, the capitalist state presents an intrinsic unity , combined with its relative [specific] autonomy, not because it is the tool of an already politically unified class, but precisely because it is the unifying factor of the power bloc. . . . The unity of state power is, in the last analysis, to be found in the state's particular relation to the hegemonic class or fraction, i.e., in the fact of the univocal corre-

spondence of the state to the specific interests of that class or fraction " (Poulantzas 1973, 300-301).

Poulantzas defines the "participation of several classes and class fractions in political domination" by the concept of a power bloc . There are two noteworthy things about this concept. First, Poulantzas defines the power bloc as a "contradictory unity of dominant classes or fractions" whose interests are antagonistic rather than monolithic. Second, the power bloc is dominated internally by a hegemonic class or fraction that politically polarizes the economic interests of the other classes or fractions of the bloc in order to establish its own economic interest as the least common denominator in the political field and to make itself the representative of the general common interest of the power bloc as a whole. From this privileged position within the power bloc, the hegemonic class or fraction indirectly reproduces its own privileged place within the relations of economic exploitation and political domination pari passu with the exploitation and domination of the masses by the power bloc as a whole.

Poulantzas defines the hegemony of the dominant class or fraction in terms of "class alliances" and the "unity of state power." Following Gramsci, he maintains that the hegemonic class or fraction exercises class leadership and holds political power: political power is neither shared nor distributed within the power bloc. The non-hegemonic elements of the power bloc are "incapable (through their own organizational means) of transforming their specific interests into the political interest which would polarize the interests of the other classes and fractions of the power bloc" (Poulantzas 1973, 297-98). Because the classes and fractions that make up the power bloc have contradictory interests, they cannot "raise themselves" to the level of a political "unity," where power is shared among the various classes and fractions, but only to the level of a political "alliance," where power is held as a unity by the hegemonic class or fraction. In short, the relations between the various classes or fractions of the power bloc "cannot consist of a sharing out of institutionalized political power, such that the hegemonic class or fraction simply possesses a more important share than the others" (Poulantzas 1973, 297).

By means of the concept of the power bloc, Poulantzas has established the class basis of the capitalist state. However, his description of hegemony and the political power of the hegemonic class or fraction comes very close to eliminating the relative autonomy of the state altogether. By referring to the state as the "univocal expression of the dom-

inant fraction within the power bloc," for example, Poulantzas seems to have reduced the specific effectivity of the state to the pertinent effects of economic class. In Political Power and Social Classes , everything of importance seems to occur outside the state and, perhaps even more significant, independently of the practices of the exploited and dominated classes. This latter point is clearly brought out in the distinction Poulantzas draws between political practice (a concept pertaining to the power bloc) and the political scene (a concept pertaining to the state apparatuses, that is, political parties and the state bureaucracy).

The concept of the power bloc is so important for Poulantzas that for all practical purposes its practices exhaust the field of state power: "the concept of the power bloc is related to the political level and covers the field of political practices , in so far as this field concentrates within itself and reflects the articulation of the ensemble of instances and levels of class struggle in a determinate state" (Poulantzas 1973, 234). The power bloc exists outside the state, yet its pertinent effects seem to determine the very structure of the state; the two are so strongly identified that the dissolution or transformation of the power bloc necessarily entails the transformation of the form of the state itself. The power bloc constitutes the field of political practice and political power, while the state proper, that is, the actual state apparatuses and their structured relationships, exists as a distinct yet distinctly secondary field of political representation: "It [the state] is covered by a series of concepts which indicate class relations in parties, situated in that particular space generally described by Marx as the political scene , in which the direct action of classes operate. We can precisely delimit the dislocation between (i) the field of political class practices (the power bloc) in a form of the state and (ii) the representation of classes by parties in a form of a regime" (Poulantzas 1973, 234).

Despite its tendency to devalue the political scene, Poulantzas's notion that the class power of the power bloc limits the particular rhythms of the political scene (political parties and organizations of class representation) has undeniable explanatory power. Among other things, it explains why political regimes may change in capitalist societies without a significant change in the power bloc and perhaps more important, why the dominant fraction of the power bloc need not correspond to the class (or fraction) that is actually the "ruling" class on the political scene (for example, how a petty bourgeois movement occupies the dominant place in the political scene under the Nazi regime while monopoly capital is nevertheless able to organize itself as the new hege-

monic class).^[4] By means of his concept of the power bloc and its pertinent effects on the political scene, Poulantzas is able to explain how agents on the political scene (politicians and bureaucrats) may end up acting in the interests of the power bloc even when such actions run counter to their own objective class interest (if they are not themselves members of the hegemonic fraction).

Having posited the power bloc as separate from the political scene and yet hegemonic within the field of political practice, Poulantzas is able to explain the existence of contradictions between political positions taken at a given conjuncture and objective class interests without slipping into either reflectionism or voluntarism. Because of the way the dominant class or fraction projects its interest as the interest of the power bloc as a whole, because of the pertinent effects exerted by the power bloc within the political scene, and because of the isolating effect and unifying function of the state within capitalist social formations, Poulantzas insists that a distinction be made between class interest (the objective effect of the ensemble of structural determinations) and class position (the conjunctural place defined by a specific instance or practice, in this case the political scene):

A social class, or a fraction or stratum of a class may take up a class position that does not correspond to its interests, which are defined by the class determination that fixes the horizon of the class's struggle. The typical example of this is the labor aristocracy, which in certain conjunctures takes up class positions that are in fact bourgeois. This does not mean, however, that it becomes, in such cases, a part of the bourgeoisie; it remains from the fact of its structural class determination, part of the working class. . . . In other words its class determination is not reducible to its class position. (Poulantzas 1975, 15)

Class Struggles Within the Democratic State

The most noteworthy development in Poulantzas's thinking in the ten years separating Political Power and Social Classes from State, Power, Socialism is his increasing attention to the political scene and to the existence and importance of class struggles within it. In the former work, Poulantzas emphasizes the externality of the power bloc and the realm of political practices with respect to the state apparatuses and the political scene. Poulantzas implies that the state apparatuses are not affected internally by class contradictions: "This [capitalist] state possesses institutions within which the economic existence of classes and the political class struggle are absent" (Poulantzas 1973, 276). In Polit-

ical Power and Social Classes , Poulantzas defines the state bureaucracy in terms of social categories, not social classes, implying that bureaucracies qua bureaucracies are free of class distinctions and class antagonisms. Bureaucracies, he maintains, reflect the pertinent effects of class power, but they lack real power of their own: "so-called 'bureaucratic power' is in fact the mere exercise of the state's functions. . . . [T]he bureaucracy has no class power of its own, nor does it directly exercise the power of the classes to which it belongs" (Poulantzas 1973, 336). Finally, the forms of the political regime are clearly of secondary importance in Poulantzas's initial framework. Shifting power relations between the legislative (parties) and the executive (bureaucracy) reflect shifts within the power bloc, but the predominance of either does not affect the nature of state power.

Poulantzas continues to defend the essence of his initial position in State, Power, Socialism , and there is no basis for claiming that he abandons his earlier problematic in his last work. However, he does come to realize that his initial framework was inadequate insofar as it ignored the impact of the dominated classes on and within the state, failed to recognize the presence of class struggles within the political scene, and emptied the state of any real significance. Poulantzas corrects these deficiencies not by repudiating his previous view but by expanding his concept of the pertinent effects of class relations on the political scene in such a way as to insert class struggle into the very heart of the state's apparatuses:

The state is the condensation of a relationship of forces between classes and class fractions such as these express themselves in a necessarily specific form, within the state itself . In other words, the state is through and through con-stituted-divided by class contradictions. . . . The executive and parliament, the army, the judiciary, the various ministries, regional, municipal and central apparatuses, the ideological apparatuses—all of these, which are themselves divided into distinct circuits, networks and vantage points, are often the pre-eminent representatives of the diverging interests of one or several fractions of the power bloc. (Poulantzas 1978, 132-33)

The excessively monolithic character of hegemony that predominates in Political Power and Social Classes is qualified significantly in State, Power, Socialism . First, within the realm of the power bloc, the leadership of the hegemonic class or fraction is no longer viewed as having the capacity to exclude the pertinent effects of other classes and fractions from the political scene. Class contradictions within the power bloc, Poulantzas now maintains, are also active within the state.

Indeed, they are active to such an extent that the autonomy of the state is "concretely manifested in the diverse contradictory measures that each of these classes and fractions, through its specific presence in the state and the resulting play of contradictions, manages to have integrated into state policy" (Poulantzas 1978, 135). Contradictions within the power bloc "take the form of internal contradictions between, and at the heart of, [the state's] various branches and apparatuses" (Poulantzas 1978, 132-33). Second, the structure and practice of the state no longer correspond as directly and precisely to the hegemony of the power bloc as is the case in Political Power and Social Classes . In a direct reversal of his earlier position, Poulantzas specifically includes popular struggles within the domain of the state and insists that such struggles no longer be viewed as strictly external and oppositional with respect to the state apparatuses: "in reality . . . popular struggles traverse the state from top to bottom and in a mode quite other than penetration of an intrinsic entity from outside" (Poulantzas 1978, 141).

In recognizing the presence of contradictions and popular struggles within the state apparatuses, Poulantzas by no means abandons the thrust of his original position: that state power is unified, that it represents the interest of a power bloc organized under the hegemony of the dominant class fraction, and that it has a specific autonomy vis-à-vis the dominant class.

The state does not constitute a mere assembly of detachable parts: it exhibits an apparatus unity which is . . . related to the fissiparous unity of state power . This finds expression in the fact that its global polity is massively oriented in favour of the hegemonic class or fraction—today in favor of monopoly capital. But the unity of state power is not established through the cohesive will of the bearers of monopoly capital or through their physical hold over the state. Unity-centralization is written into the capitalist state's hierarchic-bureaucratized framework as the effect of the reproduction of the social division of labor within the state. . . . It also arises from the state's structure as the condensation of a relationship of forces, and from the predominance over other classes or fractions of the power bloc that is commanded within the state by the hegemonic class or fraction. (Poulantzas 1978, 136)

Poulantzas continues to insist that the ruling class rules indirectly and structurally by reproducing existing relations of exploitation and domination through state power. It is the structured effectivity of the state—that is, the functions and modes of determination assigned by the ensemble of structures—and not the direct control of the power

bloc over it that "destines" the state to organize and reproduce the hegemony of the dominant class or fraction. The essential condition of this process is the separation of workers from the struggle over the means of production, and Poulantzas continues to maintain that the predominance of the popular-democratic form of the state within capitalist social formations stems from the fact that this form and no other reproduces this separation most successfully. For Poulantzas, as Martin Carnoy acutely observes, "the state is neither just political nor just juridical in the sense that it reproduces or enforces the legal bases of capitalist exchange. Rather it is fundamental to the conditions under which the bourgeoisie can accumulate and control capital, displacing struggle and conflict to the political from the economic sphere" (Carnoy 1984, 112).

For all these reasons the presence of the dominated classes within the state does not imply that they share or participate in political power or that the existing state apparatus may be reformed from within rather than destroyed from without. Indeed, Poulantzas rejects both of these positions in State, Power, Socialism :

The dominated classes exist in the state not by means of apparatuses concentrating a power of their own , but essentially in the forms of opposition to the power of the dominant classes. [They can never] lastingly hold power without a radical transformation of the state [for their subordination is inscribed in] the very material structure of the state, comprising as it does internal mechanisms of reproduction of the domination-subordination relationship: this structure does indeed retain the dominated classes within itself, but it retains them precisely as dominated classes. . . . The action of the popular masses within the state is a necessary condition of its transformation, but is not itself a sufficient condition. (Poulantzas 1978, 142-43)

In State, Power, Socialism , the dominant classes continue to predominate within the field of political power, and the repressive apparatuses of the state remain the ultimate expression of this power. Poulantzas points out that the repressive apparatuses of the capitalist state (the army, the police, and so forth) are relatively unaffected by the democratic or dictatorial nature of the regime; in both cases they perform precisely the same function—undergirding by force the mechanisms of separation by which the capitalist state reproduces the existing forces and relations of production. In all capitalist social formations "state monopolized physical violence permanently underlies the techniques of power and mechanisms of consent: it is inscribed in the web of disci-

plinary and ideological devices; and even when not directly exercised, it shapes the materiality of the social body upon which domination is brought to bear" (Poulantzas 1978, 81).

The Political Deflection of Class Struggle

Poulantzas concedes that the capitalist state, despite its ultimate appeal to violence, works its separation of workers from the struggle for economic power through processes that are not overtly repressive. In State, Power, Socialism , he discusses four of these processes: individualization, the law, the nation, and the division between manual and intellectual labor. The concept of individualization adds little to Poulantzas's earlier notion of the isolation effect. It is the process by which the capitalist state separates individuals from their production-based class identities in order to constitute them as "free and equal" citizen-individuals and unify them again under the aegis of the state. The capitalist state remains, as before, the source and guarantor of the very existence of its citizens and therefore the essential precondition for totalitarianism and democracy alike. There is, however, one very significant modification to his earlier framework: Poulantzas now recognizes the fact of the existence of citizen-individuals as a "decisive limit" to the hegemony of the power bloc and the development of a totalitarian capitalist state. This limitation occurs because the process of individualization produces, sooner or later, irresistible pressures toward representative democracy.

Although the individual-private [sphere] has no intrinsic essence opposing absolute external barriers to the state's power, it limits that power through being one of the privileged modern representations of the class relationship within the state. The nature of this limit is well-known: it is called representative democracy . However truncated by the dominant classes and by the materiality of the state, it still constitutes a mode whereby popular struggles and resistance are inscribed in that materiality; and while not the only limit to the power of the state, it is nevertheless decisive. (Poulantzas 1978, 73)

The idea that the political power of the dominant classes is limited, at least to some extent, by representative democracy reflects a heightened awareness of the contradictions created when the dominated' classes and their struggles are inserted into the political scene. A similar tendency is observable in Poulantzas's discussion of the law . Poulantzas retains his earlier view, which linked law and repression: "law in every state is an integral part of the repressive order and the organization of

violence and . . . there is no fundamental opposition between law and repression in the capitalist state" (Poulantzas 1978, 77). He also insists that the legal separation of economic property from political power serves the interests of the dominant classes by codifying the process of individualization. However, State, Power, Socialism explicitly recognizes the fact that the law, by the mere fact of its existence, opens up a struggle for power within the state—a struggle which, if it cannot be won within the state, at least provides some possibility of limiting the formal exercise of power against the dominated classes.

Modern law does not intervene against violence; rather, it organizes the exercise of violence, taking into account the resistance of the popular masses.

It is precisely through a system of general, abstract and formal rules that law regulates the exercise of power by the state apparatuses, as well as access to these apparatuses themselves. . . . By thus giving order to their mutual relations within the state, it allows a changed balance of forces in the ruling alliance to find expression at state level without provoking upheavals. Capitalist law, as it were, damps down and channels political crises, in such a way that they do not lead to crises of the state itself. . . . However, the capitalist legal system also takes the dominated classes into account in regulating the exercise of power. Faced with working-class struggle on the political plane, law organizes the structure of the compromise equilibrium permanently imposed on the dominant classes by the dominated. . . . [W]e need to stress the fact that this juridical system, these "formal" and "abstract" liberties are also conquests of the popular masses. In this sense and this alone does modern law set the limits of the exercise of power and of intervention by the state apparatuses. (Poulantzas 1978, 91-92)

The nation (as opposed to the state) is a third mechanism for separating political and economic practices within capitalist social formations. The capitalist state actively seeks to establish ideological identifications that are "national." In contrast to the traditional Marxist argument, which views nationalization in terms of the historical process of unifying internal markets to facilitate capitalist economic development, Poulantzas views nationalization as a massive reconstitution of social subjectivity and a process of ideological integration.^[5] The traditional economistic argument is inadequate because it stops short of explaining why the economically determined process of unification takes precisely the form of a nation. For Poulantzas, the development of the nation stems from a social-psychological need to unify new individualistic social relations and new conceptions of space and time

that are separated, controlled, broken down, and rendered discontinuous by the transition to capitalism.

This need for unity is political and economic as well as ideological. The modern nation redefines "inside" and "outside" and imposes a new "historicity" of linear, progressive time to provide a goal and a meaning to human existence fractured and segmented by capitalist relations of production. The dramatic transformation of space and time characteristic of capitalism is not simply a matter of thinking and representation. "In reality . . . transformations of the spatio-temporal matrices refer to the materiality of the social division of labor, of the structure of the state, and of the practices and techniques of capitalist economic, political and ideological power; they are the real substratum of mythical, religious, philosophical or 'experiential' representation of space-time" (Poulantzas 1978, 26). The national state redefines territory and time as part of the process of individualization, realizing the "historical" unity of individuals by the very act of separating them from their real history—older identifications with family, village, religion, and so forth, which are modified or destroyed. The historicity of nationalism is a thus a kind of temporal homology of the law that functions to abstract individuals from their traditional social space and reunify them in terms of the contemporary social space of the nation.

Poulantzas's insights into the social nature of space and time provides yet another perspective on the significance of the nation-state for the capitalist mode of production:

To be sure, the state is not the subject of real history: for this is a process without a subject, the process of the class struggle. But we can now understand why the modern nation-states constitute the focal points and basic moments of that real history. This becomes clearer if we bear in mind that the state establishes the peculiar relationship between history and territory , between the spatial and the temporal matrix. In fact, the modern nation makes possible the intersection of these matrices and thus serves as their point of junction; the capitalist state marks out the frontiers when it constitutes what is within (the people-nation) by homogenizing the before and the after of this enclosure. National unity . . . becomes historicity of a territory and territorialization of a history . . . the markings of a territory become indicators of history that are written into the state. (Poulantzas 1978, 114)

The fourth mechanism by which economic class struggle is deflected by political means is the separation of mental and manual labor . Poulantzas contends that this separation is a direct result of an organic relation between knowledge and power in capitalist societies. Capitalist

relations of production, he argues, separate intellectual work from manual work not out of technological necessity, the commonly accepted rationale, but as a way of permanently keeping the masses at a distance from the centers of decision making. This separation is accomplished by the various state-supported institutional devices that monopolize the transmission and certification of knowledge in such a way that the popular masses are effectively excluded from it and thus from the ranks of the professional middle class. The educational system is, of course, primary in this regard, but Poulantzas also calls our attention to the fact that the division of mental and manual labor is incorporated into all of the apparatuses of state, not simply those associated with education:

It is within the capitalist state that the organic relationship between intellectual labor and political domination, knowledge and power, is realized in the most consummate manner. . . . In their capitalist forms of army, law-courts, administration and police (not to mention the ideological apparatuses) these state apparatuses involve the practical supremacy of a knowledge . . . from which the popular masses are excluded. . . . In fact, these functions are centralized through their specific separation from the masses: in this way, intellectual labor (knowledge-power) is materialized in state apparatuses, while at the other pole, manual labor tends to be concentrated in the popular masses, who are separated and excluded from these organizational functions. It is equally clear that a number of institutions of so-called indirect, representative democracy (political parties, parliament, etc.), in which the relationship between state and masses is expressed, themselves depend on the same mechanism. (Poulantzas 1978, 56)

Therborn: The Organizational Technology of the Capitalist State

More than any of Poulantzas's previous books, State, Power, Socialism focuses on the structured effectivity of the capitalist state and its internal contradictions. However, even in this his last work, Poulantzas fails to provide a concrete analysis of the bureaucratic organizational forms of the state apparatuses, nor does he attempt to explain the mechanisms by which such organizational forms insure representation of the interests of dominant classes sufficient to mask the contradiction between domination and representation within the popular-democratic state. In short, Poulantzas lacks a theory of the state apparatus as a formal organization. Thus Göran Therborn's What Does the Ruling Class Do When It Rules? (1978)—an ambitious comparative study of the feudal, capitalist, and socialist states from the perspective of organizational

theory—fills an important gap within Structural Marxist political theory. Only by studying the state in terms of its organizational tasks, personnel, resources, and organizational technology, Therborn insists, can we lay any claim to an explanation of its class character. "If we conceive of organizations as processes formally structured by specific mechanisms of input, transformation and output, we can relate them directly to the ever advancing social processes of reproduction and change which provide the inputs and receive the outputs. The class character of an organization may then be determined by the way in which the input, transformation, and output processes are traversed and shaped by the class struggle" (Therborn 1978, 38).

While agreeing that Poulantzas correctly identifies the class basis of the capitalist state, Therborn maintains that he understates the contradictory effects of the distinction between class power and political power in capitalist social formations. In contrast to Poulantzas, Therborn emphasizes the "disjuncture" rather than the correspondence resulting from the condensation of class power into state power.

In the historical course of the class struggle, the state apparatuses come to crystallize determinate social relations and thus assume a material existence, efficacy, and inertia which are to a certain extent independent of current state policies and class relations. It follows that, although the variance between state power and the state apparatus is limited by the fact that they express the class relations of the same society, at any given moment significant disjunctures appear between the two. The possibilities of variance are substantially increased by the coexistence within a particular state system of several apparatuses, in which different sets of class relations may have crystallized. (Therborn 1978, 35)

According to Therborn, the national popular state apparatus exists as both the expression of class domination and as the representative of society as a whole executing necessary social tasks. The contradiction between these two functions generates the bounded yet open-ended organizational dynamic specific to the capitalist state. "The new tasks and problems confronting the state basically derive from the changing social totality in which it operates. But the successful organization of class domination in the state apparatus itself generates new problems of government, administration, judicature and repression—problems which call into question the existing organizational forms of domination" (Therborn 1978, 47). Because it cannot be finally resolved within the context of the dominant mode of production and its antagonistic class relationships, the contradiction between domination and representa-

tion constitutes the ongoing motive for change within the apparatus of the capitalist state. Therborn also identifies the presence of contradictions between the four distinct functional apparatuses of the capitalist state (the rule-making or governmental, administrative, judicial, and repressive apparatuses). Each of these apparatuses has its own internal rhythms and contradictions, which create contradictions within the political instance as a unified whole: "It cannot be taken for granted that they [the state apparatuses] share a common class character. . . . Even though the state is, in a fundamental sense, always one, the level of integration of its apparatuses varies considerably" (Therborn 1978, 41).

We cannot follow Therborn's discussion of the inputs, processes of transformation, and outputs of state apparatuses within capitalist (let alone feudal and socialist) social formations in its entirety. However, his concept of "organizational technology" and his discussion of "formats of representation" and "processes of mediation" within capitalist forms of the state are particularly relevant to our present analysis. By the term organizational technology Therborn refers to "a particular technique of getting things done within productive organizations"—the organizational dynamic that governs the handling of tasks, the patterning of personnel, and the use of incoming material resources. While all tasks, personnel, and material resources are products of class struggle (they are described by Therborn as "crystallizations of class relations"), organizational technology is considered the strategic variable because "it is applied in the process of transformation and affects the regulation of all other inputs and outputs" and because it "directly involves institutionalized social relations of command and compliance, leadership and execution" (Therborn 1978, 40-41). Roughly speaking, organizational technology may be said to constitute the "relations of production" within the state apparatus.

Therborn is particularly interested in the relationship between prevailing class relations on the one hand and two sets of relations, "leadership and execution" and "command and compliance," which constitute organizational technology, on the other. Relations of leadership and execution refer to external relations between the state apparatus and the rest of society and denote what Therborn calls a "directive dynamic," that is, "a mode of orientation and a basis of leadership." Relations of command and compliance refer to internal organizational hierarchies, the "mode of activation of the members of the organization, whereby their contribution to its orientation is insured" (Ther-

born 1978, 62). These concepts permit Therborn to develop an interesting contrast between the organizational technologies of feudal and capitalist modes of production and a suggestive account of the emergence and evolution of the latter.

Therborn characterizes the organizational technology of feudal societies as relatively unified and non-contradictory. He sees no fundamental conflict between the directive dynamic of feudal organization, based on aristocratic privilege and personal loyalty, and its seigneurial relations of command and compliance, based on the economic self-sufficiency of the manor and the need to defend it militarily. Bourgeois revolutions, however, split the relatively unified organizational technology of feudal societies into two distinct technologies, each congruent with the two distinctive characteristics of bourgeois class rule: (1) a "liberal-individualist" technology that combines personal freedom and equality with forms of domination inherent in the relations between capital and labor; and (2) a "rational-bureaucratic" technology that separates mental from manual labor and subordinates the latter to the former.

In its initial period, characterized by competitive capitalism and factory despotism, bourgeois organizational technology takes the bifurcated form of a "formal-legal" directive dynamic, signifying the stratified monopolization of intellectual knowledge, and a "representative" dynamic, which adjudicates and mediates the relationship between the state apparatus and the nation. These new directive dynamics engender new modes of command and compliance that Therborn labels "impersonal bureaucracy" and "parliamentary politics." In contrast to Weber, for whom bureaucracy is an ideal-typical combination of specialization, hierarchy, and knowledge, Therborn stresses the class basis of the "formal rationality" of bureaucracies, which always accept as "given" not only the content of the rules they apply but their enforcement as well. "The market sets the rules of bourgeois society and provides the economic constraint for their enforcement, even if ideological socialization proper, and in the last instance coercive violence, are also necessary. This social dynamic is located in the realm of private enterprise and capital accumulation, and it is the common public needs of these that are insured by the 'calculable rules' of the state" (Therborn 1978, 52).

We have already introduced the mechanisms by which parliamentary politics represents the ruling class while mediating the divisions that exist between it, the state apparatuses, and the dominated classes. Ac-

cording to Therborn, the predominant formats of representation in bourgeois parliamentary politics center on "the links of unity-division manifest both between different fractions of the ruling class and between the class of economic agents and its specialized political personnel," while its processes of mediation "concern primarily the strength of the ruled classes" (Therborn 1978, 181). Therborn contends there has been a profound shift in the forms of both representation and mediation within bourgeois organizational technology in the twentieth century, a shift that corresponds to the transition from competitive to monopoly capitalism. During the period of competitive capitalism and parliamentary politics, leadership devolved on politicians who owed their position to personal abilities (although, Therborn hastens to add, to possess any political ability at all, the individuals concerned had to be members of the ruling class, its allies, or its clientele). "The parliamentary politician governed above all by skillful mediation between fellow MP's of his class, each with his idiosyncrasies and immediate economic and social preoccupations: by playing them off against one another, creating heteroclite and shifting coalitions, and by persuading and cajoling with a particular kind of abstract oratory" (Therborn 1978, 53).

Under parliamentary politics, the masses could be either excluded from the legal nation (restricted franchise) or encapsulated by local bosses or notables (patronage system, machine politics, and other systems of "captive populations" controlled by landowners, company towns, and the like). As the pressure of popular struggles intensified, however, and as capital shifted to its monopoly phase, the classical form of parliamentary politics was no longer an adequate instrument of representation or mediation: it had to be supplemented or replaced by an original form of leadership able to take hold of the increasingly (if only partially) emancipated masses and keep them in a position of subordination. This new kind of bourgeois leadership, which Therborn (following Weber) calls "plebiscitary politics," operates by a combination of increased executive autonomy from the parliamentary process and an intense program of political propaganda designed to substitute emotional manipulation and identification for genuine political participation. "By means of mass appeals, the politician's message, and above all his image and attractive personal qualities, are conveyed to the people through public posters, mass-circulation newspapers, loudspeakers, and the television screen" (Therborn 1978, 53).

The transition from parliamentary to plebiscitary politics is accom-

panied by a shift in the other mode of command and compliance characteristic of bourgeois societies, the impersonal bureaucracy: the formal-legal directive dynamic of the latter is gradually replaced by a "substantial-technical" dynamic within a new organizational form that Therborn calls "managerial technocracy."

In the twentieth century and particularly the last few decades, a new mode of organizing the bourgeois state has developed alongside the legal bureaucracy. Like the latter, it is characterized by specialization, impersonality and stratified monopolization of intellectual knowledge by professionals. But it does not rely to the same degree upon calculable rules and fixed hierarchies. We may term this form managerial technocracy . Its rationality is substantive rather than formal; and, instead of juridical knowledge, it promotes technical and scientific expertise, applied with discretion and consideration of actual effects, rather than with calculable legal precision. The state hierarchy [of impersonal bureaucracy] is broken up by ad hoc committees, working parties, and special enquiries. (Therborn 1978, 54)

The new organizational technologies of plebiscitary politics and managerial technocracy emerge above all in connection with the increasingly social character of the productive forces under monopoly capitalism, the rising challenge of the working classes, and the increasing need for state interventions in the economic sphere. It is at this point that Therborn's discussion of organizational technology, mediation, and representation rejoins Poulantzas's analysis of the interrelation of the state and the economy and the shift in that relation from competitive to monopoly capital.

Poulantzas: The Internationalization of Monopoly Capitalism

In Classes in Contemporary Capitalism and State, Power, Socialism , Poulantzas examines the structural transformation of monopoly capitalism since its emergence in the late nineteenth century. He is concerned, above all, with the emergence of a dominant international capitalist mode of production and the effects of internationalization on the class structures and relatively autonomous states of national social formations. Poulantzas maintains that the most important developments within the international capitalist system have been the dissolution of independent national social formations, however economically powerful these may be, and the emergence of new internal relations of domination and dependence that extend across the national boundaries of even the most industrialized countries.

The present stage of monopoly capitalism, Poulantzas insists, is de-

termined by a social division of labor that is global in nature and that has subordinated national capital and national class relations to the requirements of international capital and its extended reproduction. By focusing on the relationship between American capital and the European Economic Community in the seventies, Poulantzas is able to grasp the general process by which a truly integral, multinational capitalist mode of production has come into being. For this reason his analysis of the seventies retains its interest, advancing our knowledge of the painful restructuring of the eighties and the relentless global Gleichschaltung of the nineties. Poulantzas explains the monopoly stage in terms of the extended reproduction of the capitalist mode of production and its twofold tendency first to expand within social formations, where it takes root and establishes its dominance, and second to expand beyond the limits of this formation. While internal and external expansion are interrelated, Poulantzas argues that external expansion, imperialism, becomes increasingly dominant as monopoly capitalism develops. Gradually the importance of commercial capital and the export of commodities, characteristic of imperialism during the stage of competitive capitalism, is supplanted by the importance of industrial capital and the export of capital during the monopoly stage.

During the monopoly stage, imperialism becomes progressively a more integrated structure. It is an "imperialist chain," Poulantzas insists, made up of national formations linked together by relations of dominance and dependency such that each link "reflects the chain as a whole in the specificity of its own social formation" (Poulantzas 1975, 42). As monopoly capitalism develops from a primarily national to a primarily international mode of production, each social formation making up the imperialist chain internalizes the international structure until, eventually, a point is reached where national structures become little more than "a function of the forms that the dominance of the capitalist mode of production at the international level assumes over the other modes and forms of production that exist within a social formation" (Poulantzas 1975, 42-43). From its very inception, the imperialist chain is forged by relations of domination and dependency. Initially a fundamental cleavage separates the imperialist "metropoles" from a subordinate "periphery" of social formations dominated and dependent on imperialism. During the stage of competitive capitalism, the metropoles remain national, largely autonomous, social formations, while an equal relationship of town (industry) and country (agriculture and raw materials) is established between the core and the

periphery. With the transition to monopoly capitalism, however, a process of transformation is initiated in the relationship between metropole and periphery and between the imperialist social formations themselves. As monopoly capitalism becomes progressively more internationalist, it generates new relationships of domination and dependency that restructure the imperialist chain as a whole, undermining the national independence of the metropoles and breaking down the distinction between core and periphery.

What drives the development of the imperialist chain is the uneven development of capitalism within the metropoles. In Classes in Contemporary Capitalism , Poulantzas distinguishes the stages of competitive and monopoly capitalism and delineates the different phases of the latter. By the end of the eighteenth century, at least in the economic heart-lands of Europe, competitive capitalism had largely supplanted the articulation of feudal and capitalist modes of production characteristic of the transition to capitalism. Competitive capitalism is defined by Poulantzas as primarily a national mode of production involving the real subsumption of labor to capital and the dominant as well as determinant place of the economic instance within the social formation. It is characterized by an articulation of petty commodity production and industrial capital, with the latter, in the form of the factory system, dominant.

Production units are relatively simple and decentralized, and the level of the concentration of capital is still quite modest during the competitive stage; as a result, industrial capital exerts certain "conservation" effects on petty production. Extensive exploitation of labor predominates—the major transformation being the restructuring of petty production along the lines of greater regimentation and more efficient division of labor—but the beginnings of exploitation by means of machinery are clearly discernible. Of particular significance, with respect to the distinction Poulantzas wishes to make between competitive and monopoly capitalism, is the "unity" of the plurality of powers of capital during the former stage and their "dissociation," restructuring, and reintegration during the latter stage. Competitive capitalism is characterized by a "coincidence of economic ownership and possession" and by a "unity of the powers deriving from the ownership and possession of the means of production" in the person of the individual capitalist entrepreneur.

By the late nineteenth century the concentration and centralization of industrial capital in the advanced capitalist social formations inau-

gurates the stage of monopoly capitalism. Poulantzas distinguishes three phases in the development of monopoly capitalism: he "transition from competitive to monopoly capitalism" beginning in the later nineteenth century and ending with World War I; the "consolidation of monopoly capital" as a national mode of production and the international contradictions it creates, contradictions that culminate in the Great Depression and World War II; and finally the "present phase," characterized by the birth and development of a multinational mode of production, a phase of global integration beginning with the "Americanization" of Europe after World War II.^[6] The timing of Poulantzas's periodization is less important to us here than the general tendency it describes, namely, the transformation of a predominantly national mode of production to a predominantly international, or, perhaps more accurately, a multinational, mode. Let us briefly summarize Poulantzas's discussion.

The transition to monopoly capital involves the appearance and extension of monopoly capital in the context of an unstable articulation of monopoly and competitive capitalism that is still national in character. It is a stage dominated by the concentration and centralization of capital in the form of holding companies or trusts, but it is also characterized by a "massive dissociation" of the plurality of powers of capital previously concentrated in the person of the individual entrepreneur. The predominance of joint stock companies produces a dissociation of economic and legal ownership, while the appearance of holding companies and trusts introduces a dissociation between powers deriving from economic ownership. Economic ownership concentrates, embracing several production units, yet the relatively distinct powers of economic possession attached to these units are subordinated, not eliminated, and considerable powers of possession remain in the hands of subordinate ownerships. A further dissociation is introduced in the form of professional managers, agents who exercise the powers of capital without actually possessing them. Finally, previously existing forms of production units, extending across diverse and distinct branches and industries, are also preserved. Despite the fact that the exploitation of labor becomes progressively more intensive, the labor processes of individual production units are only formally subordinated to monopoly capital; in actuality they retain a degree of autonomy characteristic of competitive capitalism.

Poulantzas has little to say about the consolidation of monopoly capitalism as the dominant mode of production and the crises of revolu-

tion, counterrevolution, war, and depression produced by its uneven development. This silence is unfortunate, for it is hardly an exaggeration to describe the global crisis of the interwar years as the birth pangs of multinational capitalism. It is also important to recognize the significance of a new monopoly capitalist regime of accumulation, "Fordism," which emerges first in the United States in the early twentieth century and becomes the ideal-typical model of monopoly capitalism after World War II. The path-breaking analysis of Fordism and the consolidation of monopoly capitalism appeared in A Theory of Capitalist Regulation (1976; English translation, 1979), a study of American capitalism from the Civil War to the Carter presidency published two years after Class in Contemporary Capitalism by French economist Michel Aglietta, another former student of Althusser.^[7] According to Aglietta, the decisive change between competitive and monopoly capitalism occurs in the mode of accumulation or method of maximizing surplus value during each stage. Aglietta defines the mode of accumulation under competitive capitalism and during the transition to monopoly capitalism as "extensive": the organization of labor through mechanization is the primary method of maximizing surplus value, while existing patterns of social consumption and "traditional ways of life" are left to persist or dissolve as they will without radical reorganization. With the consolidation of monopoly capitalism, however, there is a shift from extensive to "intensive" accumulation: an entirely new "way of life" is created for the wage-earning class as the totality of time and space, consumption as well as production, is reorganized to maximize surplus value.

The new regime of accumulation, which Aglietta (following Gramsci) calls Fordism , is based on the "synchronization of mass production and mass consumption." More specifically, it is the articulation of a labor process, organized around the semi-automated assembly line, and patterns of consumption, organized by stratified "life-styles," each characterized by individual acquisition of commodities mass-produced for private consumption. The linchpin of Fordism, Aglietta maintains, is a reversal of the "iron law of wages," at least for key segments of the labor force, and a new strategy of gearing wages to productivity and getting them back through a rapidly expanding domestic market. The result, the so-called consumer society, is characterized by a more regular and more rapidly increasing rate of surplus value than its predecessor. The internationalization of Fordism inaugurated and sustained the global economic hegemony of the United States throughout the twentieth century.

The third and contemporary phase of monopoly capitalism is, according to Poulantzas, dominated by the restructuring of global capitalism after World War II under the direction of the United States. The internationalization of Fordism, accelerated by American control over the economic and political destinies of Europe and Japan, has transformed monopoly capitalism into an integrated multinational mode of production that has gradually supplanted predominantly national forms. The spectacular, if temporary, economic boom of the postwar "economic miracle" obscured, at least until the seventies, the less felicitous outcome of this globalization process: a shift of the burden of exploitation back from the peripheral economies to the metropoles themselves. Only with the global recession of the seventies and the economic restructuring of the eighties is it becoming clear that the classic metropole/periphery distinction is dissolving and that multinational corporations have become globally integrated production units maximizing profits in all geographic regions and in all areas of economic activity.^[8] Poulantzas also emphasizes the tendency of multinational monopoly capitalism to reverse the dissociation of the powers of ownership and possession previously introduced by monopoly capital in its national form. Multinational conglomerates replace trusts and holding companies as the dominant organizations and establish single centers of economic ownership of what are, in effect, multinational industrial firms. Concentrated economic ownership entails real economic possession of subsidiaries and the real subsumption of their previously autonomous labor processes. Finally, Poulantzas notes how the new complex production units closely articulate and integrate labor processes divided between various establishments in several countries. Gaps between different levels of economic ownership and different levels of possession are also being closed, and the powers associated with them are concentrating again.

The Internalization of Internationalization

The formation of multinational production units transforms the articulation of multinational monopoly and national competitive capitalism within each of the industrialized social formations. Competitive or non-monopoly capital remains subordinated to monopoly capital as before, but it is now qualitatively transformed and integrated into a multinational mode of production. The closure of the gaps, within monopoly capitalism, between economic ownership and possession and between

the powers deriving from economic ownership produces an ongoing loss of powers of possession by non-monopoly capital. Indeed, Poulantzas goes so far as to maintain that behind the facade of independent ownership, the very boundaries of non-monopoly capital, that is, the forms of its enterprises and production units, are being dissolved progressively.

Non-monopoly capital loses effective control and direction of its labor process as multinational monopoly capital imposes standardization of basic products and norms of work organization. The dependence of non-monopoly capital on monopoly capital is introduced through patents and licenses controlled by the latter, through the subjection of non-monopoly capital to a social division of labor that confines it to sectors with a low level of productivity and inferior technology, and through limited margins for self-financing of non-monopoly capital. The revival of small-scale businesses, from professional services to skilled craft-work and sweatshop manufacturing, during the restructuring of the eighties confirms Poulantzas's views on this point. As multinational monopoly capital seeks to shift the burden of the crisis onto others and retain maximum flexibility for itself, there has been a marked increase in the subcontracting of skilled as well as unskilled tasks and an increased reliance on a temporary, part-time, or subcontracted labor force. Such businesses are neither atavistic nor autonomous; they are marching to the beat of multinational monopoly capitalism and the rhythms of its expanded reproduction. Far from becoming disorganized, as critics of Poulantzas might suggest, capitalism is becoming more organized precisely as Poulantzas predicted. The "creative destruction" being experienced by contemporary capitalism is in actuality a Gleichschaltung reflecting the newly emerged hegemony of multinational capital.^[9]

The internationalization of monopoly capitalism has restructured class relations and political power within the industrial capitalist metropolises themselves. The concentration of effective economic ownership into the hands of monopoly capital and the dependence of non-monopoly capital has not produced serious antagonism between these two fractions of the capitalist class. This result is not really surprising, Poulantzas argues, because there is no class split between these fractions of the bourgeoisie: relations of domination that might divide the class fractions are structured by global relations of exploitation that unite them. The principal contradiction within the multinational mode of production remains the antagonism of capital as a whole and labor as a whole. Instead of a growing antagonism between monopoly and

non-monopoly capital, we find non-monopoly capital declining as a social force. Instead of resisting its subordination to the new global division of labor, non-monopoly capital is internalizing it, becoming an integral part of the "induced reproduction" of the political and economic conditions of its own subordination.

As the dominance of the global economy is forcing an economic restructuring of national capital, so class structures and domestic politics are being reorganized by the dissolution effects of the global economy on the national autonomy of advanced capitalist states. International integration has eliminated national imperialist rivalries to the extent that Poulantzas feels it is no longer appropriate to speak of contradictions between imperialisms but only of contradictions within a single imperialist chain. The international division of labor cuts across the various fractions of capital within each national social formation and redefines older class divisions, rendering terms like foreign and indigenous capital irrelevant. The political result, unevenly developed but unmistakable, has been for international capital to increasingly dominate national politics "from within." Poulantzas argues that the "national" bourgeoisies of the metropoles are being transformed from relatively autonomous class fractions into "domestic" bourgeoisies bound by multiple ties of dependence to an international division of labor and an international concentration of capital that they do not control.

Although they possess their own economic foundation and base of capital accumulation and thus cannot be confused with a comprador class fraction, the domestic bourgeoisies of the First World are becoming integrated with and dependent on international capital and have politically internalized their place and function within the global economy. By means of this process, which I call "the internalization of internationalization," the hegemony of multinational capital within the power blocs of the imperialist powers is organized and the interests of multinational capital projected as the common interests of the bourgeoisie as a whole. National party politics, that is, the political scene within each of the industrial capitalist states, is being rapidly restructured to reflect this change in the power bloc. National parties across the entire political spectrum are suddenly no longer amenable to the class compromises of the Fordist-Keynesian welfare state and instead declare themselves fully prepared to endorse a massive shift of resources from the public sector of the economy to the internationally oriented private sector and equally prepared to tolerate, if not actually endorse, a comprehensive attack on the standard of living of the working class as well as political attacks on its real or imagined allies.

The internalization of the internationalization of monopoly capital explains not only the economic and political restructuring of the advanced capitalist metropolises but also the transformation of dependent development in the Third World as well. In contrast to the dependency theory that predominated in the seventies, Poulantzas insists that the extension and consolidation of monopoly capitalism in the metropoles initiates and advances the dissolution of the "town and country" relationship between metropole and periphery. In other words, capitalist development in the periphery, which begins with the forced implantation of capitalism in pre-capitalist social formations and which is articulated with imperialist capital in a relationship of dependency, nevertheless advances steadily and with growing autonomy within the context of the uneven development of the imperialist chain as a whole. Similarly, national independence movements in the Third World emerge with a national bourgeoisie capable of competing with the comprador bourgeois class fraction for hegemony within the power bloc of colonial politics and capable of allying with the popular classes in an anti-imperialist alliance.

However, the national bourgeoisie in the Third World never approaches a degree of autonomy comparable to that of its counterpart in the imperialist metropoles since its economic base develops within limits set by the international division of labor and dependent development. Thus from the moment of national independence, if not before, this class fraction is hardly distinguishable from a domestic bourgeoisie. Still, the emergence of a domestic bourgeoisie in the periphery is a development of major significance, and not only because it is a step up from comprador status. In conjunction with the transformation of the national bourgeoisie into a domestic bourgeoisie in the First World, the transition from a predominantly national to a predominantly international capitalist mode of production has clearly initiated a process of leveling and integration whose final outcome may be the elimination of the core-periphery distinction altogether.

It is precisely this convergence that justifies Poulantzas's claim that the principal contradiction within the imperialist chain is always between the bourgeoisie as a whole and the working class as a whole. The particular contradictions within the dominant classes and fractions always depend on this principal contradiction as do particular relations of exploitation between capital and any given segment of the working class. Accumulation crises are the direct expression of the ongoing struggles of the working class against exploitation, while economic re-

structuring and its political aftershocks are, in the last instance, simply the response by the bourgeoisie to the struggles of the working class. However much it offends delicate postmodern, post-Marxist sensibilities, Poulantzas is only stating the obvious in reminding us that "the extended reproduction of capital is nothing other than the class struggle, the contradictions within the dominant classes and fractions being only the effects, within the power bloc, of the principal contradiction" (Poulantzas 1975, 107).

If Poulantzas is to be faulted, I believe it is for failing to recognize a relative decline in the power of the United States and for continuing to interpret the internationalization and internalization of monopoly capitalism in the context of a single dominant state. Poulantzas, writing in the seventies, was understandably impressed with the internalization of American hegemony in Western Europe and the global predominance of American capital in the internationalization process. Without attempting to decide the question of American decline here (that is, without attempting to distinguish the current problems of the United States as a national social formation from the massive power of American-based international capital and the enormous significance of the American market for the global capitalist economy), I would suggest that where Poulantzas sees the ongoing predominance of the United States as a nation , it might better speak of the predominance of multinational capital increasingly removed from national identifications and constraints. Such a formulation, it seems to me, is more consistent with Poulantzas's own analysis of global integration and the internationalization of social classes.

This said, Poulantzas is by no means incorrect to continue to insist on the crucial significance of the nation-state for contemporary capitalism. The internationalization of monopoly capital and its political internalization do not mean that the nation-state has been superseded, suppressed, or bypassed. The process of internationalization has hitherto been effected under the dominance of capital still associated with a definite national base, and the national state remains, despite the considerable degree of regional restructuring that is occurring, the relevant unit of both international and domestic politics. The national state is not merely a tool or instrument of the dominant classes to be manipulated at will, nor is the relationship between economic internationalization and political internalization such that every step toward a global economy necessitates a parallel step toward political "supranationalization." Indeed, Poulantzas himself explains why this is not the case.

By including the political and institutional forms of nation-states in systems of interconnections no longer confined to the play of external and mutual pressures among juxtaposed states and capitals, Poulantzas demonstrates how the global economy has dramatically affected these forms. His concepts of a domestic bourgeoisie and induced representation explain how the power bloc and political scene are restructured in such a way that the states themselves take charge of the interests of the now dominant imperialist capital and its development within the national social formation. The process of internalization is not without its own contradictions, however, and it is to these that we now turn.

Monopoly Capitalism and the Interventionist State

The development of monopoly capitalism demands more and more government intervention in the interests of capital. The growing importance of the state is considered so significant by Poulantzas that he describes the shift from a "nightwatchman" to an "interventionist" state as nothing less than a "displacement of dominance" from the economic to the political instance—a profoundly misleading statement since control over the means of production, that is, the function of economic ownership, remains firmly within the economic apparatuses—and as the definitive structural distinction between competitive and monopoly capitalism. Of course, the nightwatchman state of classical liberalism was a capitalist state, and despite the formal separation of the political and economic instances characteristic of competitive capitalism, it performed significant economic functions, among them taxation, factory legislation, customs duties, and the construction of economic infrastructure such as railways. "The capitalist separation of state and economy," Poulantzas maintains, "was never anything other than the specifically capitalist form of the state's presence in the relations of production" (Poulantzas 1978, 167). Still, the role of the liberal state was restricted largely to reproducting the general conditions of the production of surplus value, enforcing the rights of property and the "freedom" to sell labor power, and providing representative government for and by the propertied classes.

With the emergence of monopoly capitalism, however, the liberal nightwatchman state becomes economically and politically dysfunctional. Mass participation in politics becomes unavoidable for regimes seeking social stability, and popular-democratic institutions provide

at least a formal veneer of social legitimacy for rapid, state-promoted capitalist industrialization. The emergence of mammoth industrial enterprises employing intensive production techniques, involving the integration of expensive and complex production units and seeking international markets and materials as well as economic protection and stability at home, means increasing intervention of the state in the process of capitalist accumulation. With the development of Fordism, the interventionist state becomes actively involved in the extended reproduction of capitalism as a "way of life" with a concomitant increase in state intervention at all levels of social existence. The Keynesian welfare state, in other words, evolves directly with the needs of monopoly capital for a Fordist regime of accumulation that coordinates mass production and mass consumption under conditions of social and economic stability.

The internationalization of monopoly capital demands ever-increasing government intervention in the interests of capital, thereby threatening to undermine the formal "independence" of the state from the economy. International capital complicates the role of the interventionist state by rendering internal or "national" relations of domination and dependency that extend across the national boundaries of even the advanced industrial countries. State power, by definition national, no longer corresponds to a capitalist mode of production that has become international in scope, yet the interventionist state remains, now more than ever, necessary for capitalism. The interventionist state must organize the domination of monopoly capital over other modes and forms of production not only nationally, as before, but internationally as well. At the same time, it must manage domestic economic and political contradictions to facilitate the expanded reproduction of monopoly capital and increasingly intervene directly in the economy to manage crises and use public funds to maintain the valorization of capital at an acceptable level.^[10]

The induced reproduction of an international mode of production by means of national political structures introduces an entirely new set of contradictions between the fractions of domestic monopoly capital and imperialist capital, contradictions that must be added to those already existing between domestic monopoly and non-monopoly capital and the various fractions of both. "The national state thus intervenes, in its role as organizer of hegemony, in a domestic field already structured by inter-imperialist contradictions and in which contradictions between the dominant fractions within its social formation are already

internationalized" (Poulantzas 1975, 74-75). The problem of keeping the contradictions of this new articulation in check, however, devolves on the previously existing apparatus of the nation-state, which appears increasingly obsolescent and incapable of managing forces that are beyond its control. The configuration of the power bloc is scarcely located at the national level anymore. Foreign capital does not participate directly as a relatively autonomous social force in the power bloc, yet its presence is overwhelmingly felt via restructured financial markets, deindustrialization, migrations of labor and capital, fiscal crises of government at all levels, increasing regional and urban competition for scarce investments, and above all, in the destruction of the Fordist compromise between labor and capital.

The internalization of the new global mode of production has profound consequences for the political place and function of the capitalist nation-state, consequences made dramatically clear by the accumulation crisis of Fordism and the global economic restructuring it has triggered. As Poulantzas predicted, the contemporary capitalist state is increasingly caught up in growing contradictions between international and national developments. On the one hand, as a class state organizing and reproducing the domination of the bourgeoisie, the capitalist state finds itself intervening more often and more directly in the interests of international capital (as reflected through the interests of its own domestic bourgeoisie and their place in the international division of labor). On the other hand, as a popular-democratic state, it must act in the interest of "the people" as a whole, a task it is increasingly unable to perform because of international economic pressures over which it has little or no control. The result is a new "crisis of democracy" whose resolution Poulantzas fears may lead away from democratic participation, perhaps even representation, toward a more authoritarian, pseudo-democratic regime increasingly dominated by managerial technocracy and plebiscitary politics. The shift from the Fordist to the neo-conservative state in the eighties is scarcely reassuring in this regard.

The interventionist state, whether Fordist-Keynesian or neo-conservative, remains a class state. "The task of the state is to maintain the unity and cohesion of a social formation divided into classes, and it focuses and epitomizes the class contradictions of the whole social formation in such a way as to sanction and legitimize the interests of the dominant classes and fractions as against the other classes of the formation, in a context of world class contradictions" (Poulantzas 1975,

78). However, the function of unity and cohesion is especially difficult (yet all the more necessary) given the crisis of Fordism and the international, national, regional, and urban restructuring it has set in motion. The ideological impact of the restructuring, for example, is truly staggering. Poulantzas insists that the internationalization of capitalism is transforming our very concepts of time and space. "Transformations of the spatio-temporal matrices," he points out, "refer to the materiality of the social division of labor, of the structure of the state, and of the practices and techniques of capitalist economic, political and ideological power; they are the real substratum of mythical, religious, philosophical or 'experiential' representations of space-time" (Poulantzas 1978, 26). Through technological developments in communication and transportation, capitalism is not only creating a global mode of production and an international class struggle but is also creating, simultaneously, a globally homogenized commodity culture and a deep nostalgia for a vanishing individuality. Such developments have potentially serious political implications and require unprecedented intervention by the state in the field of ideological practice as well as in the economy.

The national interventionist state persists precisely because it is neither neutral nor technical. It is the nexus where repressive and ideological power is centered, and it is inconceivable to Poulantzas that its economic and ideological functions could be delegated to a supranational or super-state apparatus without destroying the illusion of representative government which sanctions and legitimizes the entire process. Of course, the persistence of the nation-state and the continuing predominance of politics must not be misconstrued as incompatible with the power of international capital, much less antagonistic to it. The national state has no power of its own; it is assigned a place and function by what is now an international mode of production. The growing power of both the state and multinational capital is perfectly consistent with the extension and reproduction of capitalism. Furthermore, the neo-conservative revolt against "big government" during the eighties, which Poulantzas did not live to see, in no way signifies a reduction in the power of the state. Far from an attack on state power per se, the Reagan-Thatcher effort to "roll back the state" involved a massive application of state power against Fordism, Keynesianism, and the welfare system. Privatization and deregulation are real enough, but they proceed hand in hand with undiminished government intervention directed, as Poulantzas would have expected, by the exigencies of capitalist accumulation and in the interests of international capital. The

interventionist state is not withering away: it is, like capitalism itself, restructuring.

The interventionist state, whether Fordist or neo-conservative, retains its relative autonomy vis-à-vis monopoly capital. Poulantzas rejects the notion that a "fusion" of the state and monopoly capital is taking place. The state, he argues, takes responsibility for the interests of monopoly capital as a whole, but it does not concretely identify itself with any one of its components. Monopoly capital is the dominant class fraction; but it is the bourgeoisie as a whole that is the dominant class. The state serves the long-term interests of the hegemonic fraction, but only in the context of the reproduction of capital as a whole. "Today as always the state plays the role of political unifier of the power bloc and political organizer of the hegemony of monopoly capital within the power bloc which is made up of several fractions of the bourgeois class and divided by internal contradictions. The relation between the state and the monopolies today is no more one of identification and fusion than was the case in the past with other capitalist fractions. The state rather takes special responsibility for the interest of the hegemonic fraction, monopoly capital, in so far as this fraction holds a leading position in the power bloc, as its interests are erected into the political interest of capital as a whole vis-à-vis the dominated classes" (Poulantzas 1975, 157).

The Crisis of Democracy and Authoritarian Statism

In retrospect, Poulantzas was one of a handful of commentators on the Left to grasp how profoundly the internationalization of capitalism was undermining the stability of the national Fordist-Keynesian state. Although he did not provide an economic counterpart to his masterful explanation of the political aspects of the crisis—for that one has to turn to Ernest Mandel's Late Capitalism (1975)—no one understood sooner or better the structural transformations that the global economy has wrought on the class structure and state apparatuses of the advanced capitalist social formations. Although Poulantzas cannot be said to have foreseen the magnitude of the crisis of Fordism, he certainly grasped the crisis of "democracy" that it produced and the contours of the right-wing reaction by which the crisis was to be "resolved." Although his views were formulated before the triumph of Reaganism and Thatcherism, it remains instructive to compare the drift toward "au-

thoritarian statism," which Poulantzas feared, with what has in fact come to pass, a regime that has been aptly described as "authoritarian populism."^[11] The expression "crisis of democracy" has a bit of a hollow ring in the aftermath of the collapse of Stalinism. However, Poulantzas gives the expression a distinct twist and enduring theoretical substance by defining it functionally, by linking the contradictions of democracy as a structured political regime to those of capitalism as an expanding, imperialist mode of production. Furthermore, he gives it enduring political substance by virtue of his rigorous, realistic, some would say pessimistic assessment of the conditions for political class struggle.

Neo-conservatism has indubitably exacerbated the inherent contradiction between domination and execution within the capitalist state. Whatever else the attack on Fordism means, it entails an intensification of the economic activity of the state, which must manage, as best it can, the chaos created by internationalization. The neo-conservative state is still interventionist and still at the very heart of the reproduction of international capital, so much so, in fact, that it is hard to argue with Poulantzas's assessment that "the totality of the operations of the state are currently being reorganized in relation to its economic role" (Poulantzas 1978, 168). Similarly, Poulantzas is not wrong to contend that a political crisis has emerged that strains the structural autonomy of the capitalist state and that "certain major contradictions within the state are now located between its economic role and its role in maintaining order and organizing consent" (Poulantzas 1978, 168). The undiminished thrust toward privatization and deregulation, the reduction of government funding for social services, and the ongoing struggle against organized labor make it increasingly difficult to maintain the fiction of the representative character of the interventionist state. The decreasing separation between the state and the economy revealed by such interventions has resulted in an increasingly politicized attitude toward state activity. "The state has thus been transformed," Poulantzas argues, "from a buffer or safety valve on economic crises into a sounding-box for the reproduction of crises of social relations. . . . The state's subordination to the logic of monopoly capitalist reproduction, which is thus experienced as 'its' inability to respond to the needs of the masses, has never been more flagrant than it is at a time like the present, when the state is intervening in all domains in which these needs present themselves" (Poulantzas 1975, 172).

Poulantzas's contention that the growing contradiction between the state's functions of domination (organizing the hegemony of the power

bloc) and representation (legitimizing the regime by means of popular-democratic forms responsive to "the people") constitutes a structural crisis, a "crisis of democracy," expresses more or less adequately the political aspects of the crisis of Fordism by the mid-seventies. According to Poulantzas, the state's response to these tensions at the very source of its power follows an authoritarian pattern; state control over socio-economic life intensifies and combines with a relative decline of the institutions of political democracy. Within the state apparatuses the bureaucracy-executive gains at the expense of the parliamentary party system, and the decline of parliament parallels a "loosening of ties of representation" between the power bloc and political parties and between the legislative and executive apparatuses as well.

New plebiscitary and authoritarian political forms, which Poulantzas calls "dominant mass parties," subtly replace the classical parliamentary system and legal-formal bureaucracy. These organizations are characterized by their subservience to the executive, and within the state apparatus their major function is to "unify or homogenize the state administration; to control and propel (in the direction of general government policy) the cohesiveness of its various branches and sub-apparatuses" (Poulantzas 1978, 233). Simultaneously, new mechanisms, deployable by the executive alone (media manipulation and foreign policy adventures), displace the "interest-aggregatory" functions of older parliamentary parties. The dominant mass parties operate as transmission belts of the state ideology to the popular masses and as a means of manipulating consent from the electorate through plebiscitary tactics. The sum total of all these developments—state intervention in the economy, crisis of democracy, dominant mass parties—represents a shift, within the limits defining its relative autonomy, of the capitalist state from its democratic toward its authoritarian pole.

Such a political regime, which Poulantzas calls "authoritarian statism," attempts to preserve the formal separation of the political and the economic by progressively undermining its content. This aim is accomplished by several devices: "greater exclusion of the masses from the centers of political decision-making; widening of the distance between citizens and the state apparatus, just when the state is invading the life of society as a whole; an unprecedented degree of state centralism; increased attempts to regiment the masses through 'participation' schemes; in essence, therefore, a sharpening of the authoritarian character of political mechanisms" (Poulantzas 1978, 238). However, because the state remains formally separate from the private core of eco-

nomic power, its policies cannot really touch the causes of the conflicts that it faces. Thus Poulantzas sees the state caught in a trap largely of its own making: "from now on the state can go neither forwards nor backwards. . . . At one and the same time, it is driven to do both too much (crisis-inducing intervention) and too little (being unable to affect the deep causes of crises). The state is constantly oscillating between the two terms of the alternative: withdraw and/or get further involved. It is not an all powerful state . . . but rather a state with its back to the wall and its front poised before a ditch" (Poulantzas 1978, 191).

This, the more gloomy side of Poulantzas's prognosis, corresponds all too closely to the actual course of political restructuring during the eighties. Unfortunately, the more optimistic possibilities he describes have as yet failed to materialize. Poulantzas had hoped that the "crisis of democracy" created by the internationalization of monopoly capitalism might create new, and not necessarily unfavorable, conditions for the political class struggle. Because the class biases of the interventionist state are being more and more exposed, Poulantzas felt that its policies would become politicized and that this increasing political antagonism might weaken rather than strengthen the hegemony of capital. He had hoped that the economic policies of the state might bring into question its popular support, in particular the political alliance between the capitalist bourgeoisie and the "white-collar" middle class. Such a division, Poulantzas realized, would deal an incalculable blow to the legitimacy of the capitalist state and would also have cardinal significance for the state apparatuses themselves by polarizing the higher and subaltern layers of the bureaucratic administration.

However, Poulantzas, as the author of a brilliant analysis of fascism, was very much aware of the authoritarian potential of middle-class populism in the context of economic and political crisis. His thoroughly justified fear of plebiscitary politics was based on a profound assessment of the ideological gulf that separates the middle class (which Poulantzas defines, rather controversially, as a class fraction of the petty bourgeoisie)^[12] from the working class and a grim recognition of the success with which Fordism deepened the structural roots of bourgeois hegemony. By creating and coordinating mass consumption and by directing the economy by means of Keynesian neo-corporatist policies, the Fordist state integrated and de-politicized labor; by internationalizing the class struggle and globalizing the capitalist mode of production, it rendered a national political struggle against the hegemony of multinational capitalism irrelevant.

Poulantzas exposed with uncompromising integrity the structural developments that created the crisis of Fordism and determined the course the crisis would take. Had he lived, he would hardly have been surprised by the plebiscitary politics of the New Right, which have belied his hopes and substantiated his fears. Far from becoming discouraged with the blatant class bias of the capitalist state, the middle class has torn the humanitarian veil from the face of the state in order to identify with it more strongly than ever. Faced with a traumatic loss of economic security and objectively declining life chances, the middle class has reacted savagely, not against the internationalization of capitalism, the real source of the problem, or even against the hegemony of monopoly capital, whose interest the neo-conservative state serves just as assiduously as its Fordist predecessor did, but rather against the working-class majority and their real or imagined political and ideological allies. In the eighties the middle and lower middle classes have emulated their European predecessors of the twenties and thirties and conducted a political and ideological pogrom against all classes and class fractions beneath themselves. If the result is not quite fascism, the family resemblance is unmistakable.