The State

Anthony de Jasay, courtesy of the author
Jasay, Anthony de
(1925- )
Display paragraphs in this book containing:
First Pub. Date
Indianapolis, IN: Liberty Fund, Inc.
Pub. Date

1. The Capitalist State

Violence, Obedience, Preference


Preferences for political arrangements depend on people's conception of their good as well as on the arrangements that are supposed to be preferred.


States generally start with somebody's defeat.


"The origin of the state is conquest" and "the origin of the state is the social contract" are not two rival explanations. One deals with the origin of the state in real time, the other with logical deduction. Both can be simultaneously valid. Historical investigation may establish that, to the extent that we can learn about such things, most states trace their pedigree to the defeat of one people by another; more rarely to the ascendancy of a victorious chief and his war gang over his own people; and nearly always to migration. At the same time, widely acceptable axioms will also help "establish" (in a different sense of the word) that rational people, in pursuit of their good, find it advantageous to subject themselves to a monarch, a state. Since these two types of explanation of the state deal in unrelated categories, it is no use trying to relate them or accord priority to one over the other. Nor is it sensible to infer that because states have come into being and flourished, it must have been rational for people who pursued their good to subject themselves to them—otherwise they would have put up more of a fight before doing so.


Consider in this light a well-regarded attempt at reconciling the (historically) violent origin of the state with the rational volition of the subject which underlies the analytical type of ontologies such as the social contract.*5 In this essay, any person living in the state of nature forms an estimate of all future incomes he is likely to get in the state of nature and another estimate for all future incomes he would receive in civil society endowed with a state. The second estimate is taken to be larger than the first. The two estimates are discounted to present value. It takes time to get everybody else round to concluding the social contract that provides the passage from the state of nature to civil society. The high incomes resulting from the creation of the state are, therefore, some way off in the future and the present value of their excess over state-of-nature incomes is small. It may leave insufficient incentive for undertaking the task of getting everybody round to agree to the social contract. On the other hand, a state can be quickly created by violent means. The higher incomes engendered by the existence of the state thus begin to accrue quickly. They do not shrink so much when translated to present value. The comparison of the present value of incomes under a state formed slowly by peaceful negotiation of a social contract, with that of incomes under a state entering society by the short-cut of violence, must favour violence. If so, the income-maximizing rational person can presumably be expected either to welcome the violence done to him by whoever is bringing in the state, or himself resort to violence to organize it. The reader may either take it (though this cannot have been the author's intention) that this is the reason why most states were not created by peaceful negotiation but by violence or that, whatever was the historical cause in any particular case, this theory of rational motivation is at least not inconsistent with it.


Like the contractarian theories before it, this sort of theory invites the careless conclusion that because states have come into being by violence, and flourished, and because it can make sense for people serenely to submit to violence leading to the creation of the state which they desire but cannot manage to achieve, people did welcome state-creating violence after the event. The underlying assumption is that the state, regardless of its peaceful or violent origin, helps people in the pursuit of their good.


Astonishingly, this assumption is hardly ever cast in a more general form, for instance by allowing for algebraic sign. If it were, it should read "the state helps/hinders," with the actual balance of the expression depending on the empirical content of the terms "help" and "hindrance." More informatively, the assumption could be cast in a form like "the state helps/hinders some people, hinders/helps others and leaves the rest unaffected." The affected are helped and hindered in different ways and to different extents. Unless by a fluke the hindered set is empty (i.e. everybody is either helped or left alone), the algebraic sum is a matter of comparisons between the helped and the hindered. Running up against interpersonal comparisons so early is a sign that our reflections are at least headed in the right direction, towards the central questions of political theory.


If ever there were people in the state of nature, and as a matter of repeated historical fact it took violence to impose a state upon them, it seems pertinent to ask, Why does standard political theory regard it as a basic verity that they preferred the state? The question really breaks down into two, one "ex ante" and the other "ex post": (i) Do people in the state of nature prefer it to the state? and (ii) Do people, once in the state, prefer the state of nature to it? These questions very sensibly allow for people's preferences to be related, in some way, to the political environment in which they actually happen to live.*6 However, once they are framed in this way, they are seen to have a peculiar character. When social scientists say that they know that Smith prefers tea to coffee because he just said so, or because he has revealed his preference by taking tea when he could have taken coffee, they deal in objects which are presumed to be both familiar and accessible to Smith. When Smith is talking about his preferences for things he can at best know from hearsay, difficulties begin to arise. They are compounded when he could not possibly translate his avowed preference into a practical act of choice, because some alternatives are simply not feasible. People who live in states have as a rule never experienced the state of nature and vice versa, and have no practical possibility of moving from the one to the other. It is often a historical anachronism and an anthropological absurdity to suppose such movement. On what grounds, then, do people form hypotheses about the relative merits of state and state of nature?*7


It appears that among certain South American Indians (though conceivably elsewhere, too) an increase in the size of the demographic unit is recognized as favouring the likelihood of the creation of a state, possibly because of the changed scale and kind of wars that this entails. A war chief supported by his quasi-professional warrior followers can coerce the rest of the people into durable obedience. In a book by Pierre Clastres which should prominently figure in any bibliography of the social contract,*8 it is reported that the Tupi-Guarani people used to abort this process by swarms of them seceding, going off to distant and fearsome lands on prophet-led flights from the greater dread of subjection, of the state which they identify with evil. The American Indian people studied by Clastres typically live in the state of nature, a condition which has little to do with the level of technical civilization and everything to do with political power. Their chiefs can exhort but not command, and must rely on oratory, prestige and liberal hospitality to get their way. Their prestige depends in part on seldom risking interference in a matter where their exhortation is liable to go unheeded. There is no apparatus among them for enforcing obedience and the Indians would not dream of voluntarily contracting to obey, though they may choose to agree with the chief on a case-by-case basis.


Theirs are, according to Clastres, true affluent societies, easily capable of producing surpluses but choosing not to do so, a two-hour working day being sufficient amply to provide for what they consider adequate subsistence. Though there is little or no production for exchange, there is private property; there could be no private hospitality, no invitations to feasts without it. There is no obvious obstacle to the division of labour and hence to capitalism, but the goods that the division of labour may provide are not prized. Work is held in contempt. Hunting, fighting, story-telling and party-going are preferred to the sort of goods labour could produce. The question is staring us in the face, Is it because of their preferences that the Indians abhor the command-obedience relation inherent in the state, and choose to stay in the state of nature? Or is it living in the state of nature which predisposes them to like, above all else, the tangibles and intangibles that typically go with it?


Marx would no doubt frown at the role tastes and preferences are allowed to play in this way of posing the question, and would presumably decide that subsistence agriculture, gathering and hunting were phenomena of existence, of the "base," while the institutions of the state were those of consciousness, of the "superstructure." It was thus the former which must have determined the latter. Clastres, for one, asserts the contrary.*9 Analytically (as distinct from historically), both views are true in the same sense as "the chicken caused the egg" and "the egg caused the chicken" are both true. My contention here is that preferences for political arrangements of society are to a large extent produced by these very arrangements, so that political institutions are either addictive like some drugs, or allergy-inducing like some others, or both, for they may be one thing for some people and the other for others. If so, theories that people in general (Hobbes, Locke, Rousseau), or the ruling class (Marx, Engels), mount the political arrangements that suit them, need be approached with much mistrust. Conversely, the view (Max Weber's) that historical outcomes are largely unintended, deserves a préjugé favorable as the more promising approximation to many of the relations linking state and subject.

Notes for this chapter

Robert L. Carneiro, "A Theory of the Origin of the State," in J. D. Jennings and E. A. Hoebel (eds), Readings in Anthropology, 3rd edn, 1970.
A more succinct statement of the same point is found in Michael Taylor's excellent Anarchy and Cooperation, 1976, p. 130: "if preferences change as a result of the state itself, then it is not even clear what is meant by the desirability of the state." See also Brian Barry, The Liberal Theory of Justice, 1973, pp. 123-4, for the related argument that since socialization adapts people to their environment, a heterogenous or pluralistic society is unlikely to turn homogenous and vice versa, although "only one generation has to suffer to create orthodoxy (as the absence of Albigensiens in France and Jews in Spain illustrates)."

However, Barry's use of the socialization argument seems to me somewhat lopsided. Must we exclude the possibility that the environment can generate not only positive, but also negative preferences for itself? Enough examples from second-generation socialist countries and even from third-generation Soviet Russia, attest to a virulent allergy to totalitarian ways and a yearning for diversity on the part of some unknown but perhaps not negligible part of the population. In the pluralistic West, there is a parallel yearning for more cohesion of purpose, for moral attitudes, an allergy to admass, to what Daniel Bell calls the "porno-pop culture" and the "psychedelic bazaar."

This is perhaps saying no more than that all societies tend to secrete corrosive elements (though in only some societies do the rulers suppress them). Yet it is not trivial to generalize the "endogenous preference" argument by admitting that social states may generate both likes and dislikes. Otherwise, the endogenous generation of preferences would ceaselessly cement any status quo and historical change would become even more mysterious, incomprehensible and random than it is anyway.

In the luxuriant literature that has sprouted around John Rawls's Theory of Justice, 1972, no objection appears to have been raised against the "original position" on this ground. The participants in the original position are devoid of all knowledge of their particular persons. They do not know whether they are representative white Anglo-Saxon men or representative Red Indian women, tenured philosophers or welfare recipients. They do not even know the age they live in (though this seems hard to reconcile with their knowledge of "political affairs and the principles of economics"). They are induced to seek a "cooperative solution" to their existence (in game-theory terms), which can be summarily interpreted as agreement on a social contract for a just state.

Failing agreement, in leaving the original position they would exit into the state of nature. They seek to avoid this outcome, because they know enough about themselves and the state to prefer it to the state of nature. They know their "life-plans" whose fulfilment depends on command over tangible and intangible "primary goods." They also know that the state, through the "advantages of social cooperation," entails a greater availability of primary goods than the state of nature. In technical language, the participants thus know that they are playing a "positive-sum game" in bargaining for a social contract (which is just in the sense, and only in the sense, that everybody is willing to stick to its terms). This means that if the cooperative solution is reached, more primary goods can be distributed than if it is not.

The comparison of two bundles of primary goods, however, requires indexing, and the weights adopted for the index (for instance, the relative valuation of time off against real income), cannot help but reflect a logically prior preference for a type of society. In other words, people in the original position cannot say that the bundle of primary goods available in the state of nature (containing, for instance, much leisure) is smaller than that available under the state (containing, for instance, many tangible consumption goods) unless they already know that they prefer to live in civil society. Comparison of the state-of-nature bundle and the state bundle presupposes the very preference which it is employed and required to explain.

The state-of-nature bundle of primary goods contains more of the things which people living in the state of nature are used to and have learnt to appreciate. It is, for them, the bigger bundle. The converse is true of the bundle available under conditions of social cooperation. It is the bigger bundle for people who have learnt to like what it contains and not to mind its constraints. But can people in the original position really tell which bundle is bigger?

Pierre Clastres, La société contre l'état, 1974; English translation, Society against the State, 1977.
Ibid., ch. 11.

End of Notes

Return to top