To his credit, in his article, "The Limits of the State: Beyond Statist Approached," Tim Mitchell does indeed go "beyond" the boundaries that have been drawn in most debates among political scientists on this subject, but he leaves us with the question of whether he has gone far enough. After effectively criticizing various writers' attempts to establish a clear division between state and society, Mitchell offers a version of their union that emphasizes the numerous everyday activities that constitute the forms and hierarchies of order and control, which are our political institutions, while simultaneously making these institutions appear separate from and independent of the larger social context. The same practices that give rise to the state are held to be primarily responsible for the widespread misunderstanding of its nature, and in particular of its limits.
But why do we want to understand what the state is, in the first place? Primarily, I would have thought, it is because we want to understand better what the state does. As this suggests, most people already have a rough idea pf what the state does-it makes laws, administers and enforces them, adjudicates disputes on the basis of such laws, and represents people who live under the rule of one state in their dealings with other states. Until greater clarity is attained regarding who and/or what does all this, it is difficult to distinguish the state from such kindred phenomena as the government and the nation, or to understand why setting the boundaries of the state should be such a problem. More important, it is impossible to explain the patterns observed in how the state conducts its work and in the effects that result from it. That is, some groups in the country consistently play a greater role in the higher reaches of the state and benefit to an inordinate degree from most of what the state does. Well, who are they, and why should this be so? And does it have anything to do with the nature of the state as such? These would seem to be reasonable and even obvious questions.
While all the writes Mitchell criticizes are indeed guilty of setting up the state as an independent entity, they do at least provide us with answers to these questions-if it is only to deny the accumulated evidence, though it is usually more, such as attributing the observed bias to the decisions of politicians. Mitchell, on the other hand, offers no answers whatsoever. The Foucault-inspired micro-practices of "organization and articulation" (92) that constitute the state in his account are disembodied, depoliticized, declassed, and, despite a couple of vague references to history, ahistorical. Emptied of all social content, they can tell us nothing about why the state acts as it does.
Yet there is much in Mitchell's essay that suggests that he really knows what the answers are. He says, for example, that the spillover of society into what is ordinarily taken to be the state, their internal relation, has to do with how a "social and political order is maintained" (90). But we never learn who benefits most from maintaining the present social order or how exactly the state helps them. The social order itself is never named. Capitalism is nowhere mentioned in his article.
Further, Mitchell is aware that a major effect of the attempt to remove the state from society is that it reinforces the view that the state and/or officials acting on its behalf are free and independent agents, only occasionally influenced-and then only to a minor degree-by larger social forces. From this, it is but a short step to recognize, as I believe Mitchell must but nowhere makes explicit, that there are few beliefs more important to the health of the status quo than the one which treats state institutions as neutral in the political and economic clashes between different groups, such that each of them has a more or less equal chance of winning the support of these institutions. To believe that the state has special ties to one group, insuring that its major interests always come on top, would cost the state a great deal of its legitimacy and make it far more difficult for it to do whatever it does on behalf of the favored group.
Finally, Mitchell's own examples suggest to us that he knows more than he says. Do the Aramco Oil Company (90) and American banks (90), two of Mitchell's main examples, engage in politics to establish a clear separation between the state and society, or to promote their own money-making interests? The question should answer itself. If in the process of promoting their interests, they make use of means that disguise their integration into the state-chiefly as a way of avoiding democratic accountability-this certainly deserves to be mentioned, but it should not require that we exchange one disguise for another. For while Mitchell isolates a major ideological distortion that emerges from the particular ways in which a favored group helps constitute the state in the very process of using it, he does so in a manner that further mystifies who exactly is acting and why they are acting in these ways.
Why such hesitancy? After criticizing various thinkers for setting the state out from society, Mitchell seems unwilling to investigate the systematic bias that he clearly suspects is present in state practices because he fears falling into the same trap as those he criticizes. To present the state consistently acting as it does due to pressures (whatever their form) located in society to reestablish, he seems to believe, the very separation between state and society that he took such pains to demolish. In short, Mitchell's explanatory framework and commitments keep him from recognizing what, at some other level, he knows to be there.
Is there another way of treating this matter that would avoid being caught up in this dilemma? Mitchell's own approach suggests that there is. The key to this approach involves extending the notion of the state back in time to include the wide variety of practices by which the institutions of the state came into being. Having stretched the boundaries of the state so far, there can be no objection in principle to stretching them a little further in order to include who and/or what are primarily responsible for the aforesaid practices. Mitchell's conception of the state, in other words, could-and, in my view, should-include the group(s) whose special ties to the state give its forms and practices their distinctive character. This would enable us to explain the patterns we observe in the latter's effects without introducing a split between state and society. If the everyday activities of citizens helps explain how the state is constituted, it is only the connection between these activities and the requirements of the favored group(s) that can account for why it is constituted in just these ways at just this time. An adequate conception of the state must contain answers to the questions, "why" and "when," as well as the question "how."
There is no single answer to where the state ends and where society begins, because these are not neighborhoods separated by broad avenues. Rather, the one performs a range of essential functions within the other. The ways and even the forms through which these functions are performed, however, vary considerably depending on both the nature and stage of development of the society in question. They also vary depending on which group has enough power based on its structural position in society to ensure that its interests generally emerge on top, that the requirements rooted in these special interests get taken as society's requirements. Such a group is more accurately referred to as the "ruling class." With changes in society, in its ruling class and in what this class requires to best serve its interests, the functions of the state will also vary and with them the boundaries that establish state limits as well as the very meaning of "state" in our discourse.
Functions imply a system, however, and our quest is complicated by the fact that the functions performed by the state belong to three different, though overlapping, systems. These systems are class society, capitalism, and modern American democratic capitalism. They differ from each other in the degree of generality (or specificity) that mark all their qualities. Which is to say, our society is a class divided society and as such possesses a host of interacting qualities that are similar to those found in other class divided societies during the whole period that they have existed on earth (roughly five to ten thousand years). Ours is also a capitalist society and has many systemic qualities that distinguish it as such, which it shares with all the capitalist societies in the five to six hundred years that this social formation has existed. And, finally, our society is a stage in capitalism that has come to the fore under the special conditions that have developed in America in recent years (say the last twenty to fifty years). What we call "American society," then, is a layered construction composed of sets of internally related qualities that fall on three different levels of generality, indicative of three different time frameworks.
While the qualities that constitute these three systems are often treated together, we can also distinguish among them, and for certain purposes it is essential that we do so. As regards the state, it is clear that what is meant by the ruling class as well as by its requirements varies somewhat as we move from viewing the United States as a class society to viewing it as a capitalist society, to viewing it as modern American democratic capitalism.
At the level of class society, the dominant economic class requires political help in reproducing the conditions that underlie its position of dominance. Such help involves various means of socializing the population, and especially is most oppressed sections, into an acceptance of the status quo and a repressive apparatus-chiefly police, soldiers, and courts-that can be called upon when the former doesn't wholly work. Socialization is aimed primarily at gaining popular acceptance for the institution of private property, for the basis-whether in divine will, birth, or money-of the prevailing division of labor, and for the idea that the special interests of the dominant class are the same as the interests of the nation. The mystification involved in the latter leads Marx to refer to the state during the entire period of class history as an "illusory community" (1942, 74). It is a "community" because it tries to reproduce the conditions underlying the mutual dependence of all people living in a given geographical area. It is "illusory" because it privileges the interests of one class under the guise of serving the interests of everyone. So much is true of the state in all class societies.
But, as we said, ours is also a capitalist class society, and as such it requires political help tailored to the peculiar character and interests of the ruling capitalist class. Perhaps the most distinctive feature of the capitalists as a ruling class is their constant drive to accumulate, to treat wealth primarily as a means to produce still more wealth. While socialization and repression remain essential state functions for capitalism, the nature of accumulation in this period requires certain modifications on the forms through which these functions are carried out. To cite just one example, workers perform their roles in accumulation more effectively if they feel they have had some freedom in choosing a job and at least the possibility of influencing state policy in so far as it affects their job. All of which gives to the production of capitalist ideology, and the institutions devoted to this activity, a much more important place in capitalism-and therefore in the capitalist state-than it has in other class societies and their states.
The centrality of accumulation in capitalism also forces the state to undertake a function that does not exist in other class societies, and that is helping the capitalists realize the value of their products. Given how capitalism works, capitalists simply require more direct political help in their efforts to enrich themselves than did feudal aristocrats or the slave owners of antiquity. They require what amounts to state planning on their behalf-and they get it. While rendered mostly invisible by their ideology, such planning covers everything from investing, to financing, to doing research, to training workers, to buying, to selling the finished products. Most of our tax, trade, labor, and monetary policies (including both laws and their administration) some under the rubric of capitalist planning. The political institutions and practices that serve capitalist interests in these ways are essential parts of the capitalist state.
While Mitchell's examples show corporations and banks only interested in making money-an interest aided by "appearing" private, outside the state-he never examines the system in which this occurs. Consequently, there is no mention of capital accumulation and the kind of help this requires from the state. Had he placed his analysis squarely within a capitalist framework, he would have noticed that the state's appearing separate from society is but one of the ways-though, perhaps, ideologically the most important-in which the state serves the interests of the capitalist ruling class.
Just as ours is a class society and a capitalist society, it is also, as we have said, the modern American democratic capitalist society, and some of what our state is and does pertains to the special interests of the American capitalist class, given its place and role in world capitalism at this stage of its development. If the first set of qualities we examined is shared with all societies within the period of class history, and the second with other capitalist societies for the entire period that they have existed, these qualities apply only to the American democratic capitalist society of recent times. Among the new developments that have had a major impact on the character and practices of the state on this level of generality, three seem to stand out. The enormous increase in the gross national product, together with the intensification of competition between capitalist societies in the last few decades, has led to the state taking on a larger and more direct role in the realization of value, while for good ideological reasons insisting ever more loudly that nothing of the kind is happening. Also, the Cold War has given us the National Security State, including the "Military-Industrial Complex," which has greatly facilitated the capitalists effort to repress internal opposition and bolster their profits by draping both with the mantle of national security. Finally, the development of communications technology has made it possible for the American ruling class to extend the political work of socialization in capitalist ideas and values to the entire globe, transforming the media into an indispensable arm of the state, which is to say into an aspect of the state itself (or what Louis Althusser dubbed an "ideological state apparatus") (1976, 67-126).
As our American capitalist class becomes increasingly international, with more foreign capitalists investing in the United States and more U.S. capitalists investing abroad, the political help required to realize value has led to the development of world institutions such as the IMF, the World Bank, and GATT, whose economic labels disguise their important political functions. They are the beginnings of a world capitalist state, whose origins and development-as with the rise of the national states that occurred earlier-are best understood as political responses to the changing requirements of the ruling capitalist class on the level of generality of modern capitalist society.
The national state in the United States is also losing some of its power to regional political forms, such as the New York Port Authority and the Appalachian Regional Commission, and for similar reasons: the interests of the capitalist class in this period are better served by having certain political functions performed in this way and on this scale. That both world and regional forms of the state are completely unaccountable to their respective constituencies has also contributed to their popularity among the ruling class. Barring a socialist transformation, what we have here is a preview of the next stage of the "American" capitalist state, with the national forms of the state being reduced to performing mainly repressive and socializing functions while most of the help rendered capitalists in their drive to accumulate and maximize profits is provided at the world and regional state levels. How this will affect the national state's ability to perform its repressive and socializing functions remains to be seen.
Viewing political institutions and practices as internally related to the ruling economic class, as equal and necessary aspects of a more broadly conceived notion of the state, viewing such ties as falling on three different levels of generality, does not keep us from recognizing occasions when a set of unusual circumstances allows some government officials, political parties, and even whole institutions to operate in relative autonomy from ruling class imperatives. The point is that this is the exception, and a relative exception at that-meaning short term and for limited purposes-and not the rule, as is the case in most non-Marxist interpretations of the state. Such exceptions are also largely explainable in terms of the special circumstances that apply. These exceptions also become increasingly rare as we move from the more specific level of modern American capitalism to the more general one of capitalism over all to the still more general one of class society, which is to say that it was easier to have a New Deal (the part of it that favored the interests of the workers) than it is to interfere with the structures that underlie capitalist and, more generally, ruling class domination.
Nor do I mean to wholly dismiss the struggle by subaltern classes to use some part of the state-taking advantage of the problems and contradictions that come with the territory-to promote their interests. While the cards are carefully stacked against such efforts, it is sometimes possible to win a small, or temporarily, or to keep a capitalist initiative from succeeding to the extent that it otherwise would. And by reordering power relations even slightly, every victory resets the odds for the major confrontations up ahead, when the problems inherent in the ordinary functioning of the system have reached the explosion point. The larger goal of these confrontations can only be to socialize the means of production and deprive the capitalist ruling class of the basis of its power. Should this succeed, a new ruling class, representing in this case the great majority, would organize the state in a way tat would best serve its special interests, once again altering the boundaries of the state in the process. The internal relations between the state and society-a very changed society and hence a very different state-would remain the same.
If many political scientists treat the state as a kind of "ghost in the machine" (David Easton's not altogether friendly observation, quoted by Mitchell) (91), it is only because the machine acts in ways that belie all attempts at explanation which are restricted to the decisions of the people who sit at the controls. Something more and different has to be introduced to help account for the persistent patterns found in how the machine works. Until this is done, all attempts to kick the state out the door-in the manner of the political systems' theorists, for example-will only result in having it climb back in through the window. Recognizing this, Mitchell sought to highlight the numerous everyday practices that constitute the state in its present form and enable it to function as it does. In so far as these practices are themselves left unaccounted for, however, he has just substituted one ghost for another. It is only by extending the connections that Mitchell started to trace back further in order to take account of the political requirements of our ruling class-grasped as a ruling class in general, as a capitalist ruling class, and as the modern American capitalist ruling class-that the otherwise inexplicable movements of Easton's ghost (and Mitchell's, too) become comprehensible.