Macquarie University
POL264 Modern Political Theory
ROBERT NOZICK: AGAINST DISTRIBUTIVE JUSTICE
Copyright 1996 R.J. Kilcullen

According to Nozick there are three sets of rules of justice, defining:

how things not previously possessed by anyone may be acquired;
how possession may be transferred from one person to another; and
what must be done to rectify injustices arising from violations of (1) and (2).
A distribution is just if it has arisen in accordance with these three sets of rules. See pp. 151-2.
Nozick does not try to specify in detail the rules under the above three headings ('I shall not attempt that task here', p. 153). However, he does give some further information on rules of acquisition; see p. 174ff. He follows John Locke who as Nozick interprets him held that a person has a right (1) to own what he makes, and (2) to appropriate anything not already owned, provided he leaves 'enough and as good' for others - i.e. provided his appropriation leaves them no worse off. (Nozick calls this the 'Lockean proviso'.) It is not clear how Nozick would defend (1) against his own criticisms of Locke (p. 174-5). As for (2), he points out that the proviso cannot reasonably be taken to mean that there can be no worsening of others' opportunities to appropriate; it must mean that in other respects they are no worse off. Nozick raises the question 'No worse than they would be how?' What is the baseline? In Rawls's theory the representative worst-off person must be no worse off than he would be under any other possible arrangement. Nozick rejects this, but does not define another baseline: 'This question of fixing a baseline needs more detailed investigation that we are able to give it here'; p. 177. However, 'whether or not Locke's particular theory of appropriation can be spelled out so as to handle various difficulties, I assume that any adequate theory of Justice in acquisition will contain a proviso similar to the weaker of the ones we have attributed to Locke'; p. 178. Nozick says that the proviso is violated if a person appropriates all of something necessary to life - or purchases it, or combines with the other owners of it, or finds himself the sole owner when other supplies are lost (e.g. when all the other water holes dry up). Nozick refers to the possibility of losing entitlement to something that was originally yours because of developments since, such as the drying up of other waterholes, as the 'historical shadow' of the Lockean proviso; p. 180.

Comparison with Rawls's Theory
Nozick classifies theories of justice as (1) either end-result or historical, and (2) either patterned or unpatterned. The entitlement theory is historical and unpatterned. It does not demand that the distribution resulting from just acquisitions, transfers and rectifications be patterned, i.e. correlated with anything else (such as moral merit, need, usefulness to society); people may be entitled to things got by chance or gift. Any distribution, irrespective of any pattern it may or may not have, is just provided it has the appropriate history, provided it did in fact come about in accordance with the rules of acquisition, transfer and rectification. Rawls's theory on the other hand, is an end-result theory. Choice of principles behind a 'veil of ignorance', must be based on calculations about what people are likely to end up with under the various possible sets of principles - there is no other way of choosing (is there?); p. 202. Therefore if any historical entitlement theory is correct, Rawls's approach is wrong.
Notice that this imposes on Rawls in the job of showing that no possible version of an historical entitlement theory could be correct. He might reply that he intends to stick to his own theory until someone actually produces a correct entitlement theory; Nozick has not, because his theory is merely a sketch with many important details not worked out.

Nozick points out (p. 207 ff) an analogy between his own entitlement theory and the process by which in Rawls's theory the rules of justice are arrived at. Rawls specifies an initial situation and a process of deliberation, and say that whatever rules results from this are the rules of justice; similarly Nozick specifies a process, and says that whatever distribution results is just. 'Each theory specifies starting points and processes of transformation, and each accepts whatever comes out'. But Rawls's process for generating principles cannot generate process principles, but only end-result principles. Nozick says that this is ironic. It presents a dilemma: if processes are 'so great', it is a defect that the process cannot lead to process principles of justice; if processes are not so great, then why should we accept the outcome of Rawls's process? (This is a weak argument. Rawls can claim that his process is 'great' without having to hold that all processes, just because they are processes, are great.) There are many pages of criticism of details of Rawls's argument which we cannot follow here (or even there, sometimes!).

Patterns and Liberty
Others besides Rawls have put forward 'pattern' theories. Nozick advances an objection against all of them: part of ownership is the liberty to give things to other people. If justice consists in the pattern in which goods are distributed, then giving - which changes the pattern - will be unjust. Thus pattern theories do not merely correct the mal-distribution which allegedly happens under an entitlement theory; they also alter the concept of possession. 'The view that holding must be patterned perhaps will seem less plausible when it is seen to have the consequence that people may not choose to do acts that upset the patterning, even with things they legitimately hold'. (See p. 219-20). Note Nozick's concept of ownership: a right to do whatever you choose with what is your own.
No Central Distributor
'Pattern' theories sometimes picture some person or institution faced with the problem of fairly distributing the sum total of good things: they should be distributed in a pattern corresponding to merit, need, etc. But, Nozick maintains, things are never collected into a sum total to be allocated by a central distributing authority. p. 149 The term distributive justice is not a neutral one. Hearing the term distribution most people presume that some thing or mechanism uses some principle or criterion to give out a supply of things... There is no central distributor, no person or group entitled to control all the resources, jointly deciding how they are to be doled out. 'If things fell from heaven like manna, and no one had any special entitlement to any portion of it...' (p. 198), 'there might be a more compelling reason to search for a pattern. But since things come into existence already held (or with agreements already made about how they are to be held), there is no need to search for some pattern for unheld holdings to fit; and since the process whereby holdings actually come into being or are shaped, itself needn't realize any particular pattern, there is no reason to expect any pattern to result... In the non-manna-from-heaven world in which things have to be made or produced or transformed by people, there is no separate process of distribution for a theory of distributions to be a theory of'; p. 219.
No Presumption of Equality
Nozick ask why it is to be assumed that differences between persons are arbitrary unless they can be justified. A central distributor would perhaps be bound to treat all alike unless for good reason, but in a free society distribution results from many localized exchanges between individuals entitled to bestow their holdings as they wish. p. 223.
The Natural Lottery not Unjust
According to Rawls, the veil of ignorance should conceal the distribution of natural talents, because rules reflecting this distribution would not be just. But according to Nozick, it is not true that a person deserves something only if he also deserves whatever he used, including natural talents, to obtain it. 'It needn't be that the foundations underlying desert are themselves deserved, all the way down'; p. 225. 'Whether or not people's natural assets are arbitrary from a moral point of view they are entitled to them, and to what flows from them'; p. 226.
Equal Opportunity not a Right
Possessions that people are entitled to may not be seized to provide equality of opportunity for others. Life is not a race; 'there is no unified race', p. 235. There are individual exchanges, in which the parties do not usually care about desert or handicaps, but simply about what they get in exchange. 'No centralized process judges people's use of the opportunities they had; that is not what the process of social cooperation and exchange are for'; p. 236. (Statements about what institutions are for are always suspect. How do we decide what exchange is for, and anyway why would this impose a norm?)
Justice and Equality
People often note that wealth is unequally distributed, and proceed immediately to discuss how it might be made more equal. But on the entitlement theory one cannot decide whether redistribution is necessary merely by looking at the prevailing pattern of distribution. Whether it just depends on how the distribution came about. If it came about in accordance with the rules of acquisition, transfer and rectification, then it is not unjust, however unequal it may be.
Redistributive Action by Government Unjust
According to Nozick, taxation is equivalent to forced labour. Taking a proportion of earnings is like making a person work a proportion of his working time for another's purposes. It is unjust to force a person to work for another's benefit.
According to Locke (Section 27), a person has a property in himself and in his labour; each person has liberty to decide what he will do (subject to the rights of others), and a right to reap the benefits of his own actions. But tax-financed social welfare programs institute something like ownership by others of people and their actions. The poor have a claim on the actions and products of others, whether or not those others freely entered into relationships that might give rise to such claims, whether or not they voluntarily take these claims upon themselves, in charity or as part of an exchange. It is inconsistent to allow a right to emigrate, when there is no right to stay and opt out; see p. 173.

(Are there any enforceable duties to do things for others? E. g. is there a duty to help in floods or earthquakes? Enforceable duties to help do not imply that actions or persons are owned. Part of ownership is the right to sell; the fact that someone has a duty to help me does not imply that I have a right to sell his help. Parents are not partly owned by their children. Nozick's analogies need to be analysed carefully. )

Rectification of Past Injustice
Nozick does not attempt to work out the rules of rectification. However he says that it is an important task in each society to work out what operable policy best approximates the results of a detailed application of rules of rectification. It is possible that some tax-financed welfare program, or even Rawls's rule of favouring the worst off group, may be justified as a means of rectification, if it can be assumed that the better off are beneficiaries of past injustices, and the worse off victims. He warns that his entitlement theory cannot be used to condemn any particular welfare scheme unless it is clear that it cannot be justified as a means of rectifying past injustices. See p. 231. (Will this warning be heeded, or will the theory be taken as a justification of the existing distributions?).
Nozick's theory of rights
In his account of the possible justification of the state, and in his entitlement theory of justice, Nozick postulates absolute rights - not merely prima facie rights which might be overridden, but boundaries not to be crossed without the free consent of the persons whose rights they are. Other people's rights are constraints upon our actions toward our own goals.
Why must we respect such constraints? When a person makes a thing, or finds it unowned and appropriates it, why must others not use it without his permission - no matter how great their need, no matter how such things are distributed?

Nozick's answer is that such constraints express the inviolability of other persons; a person is not to be used to benefit others - this would not sufficiently respect the fact that he is a separate person, that his is the only life he has. There is no transcendent social whole for the sake of which individuals can be sacrificed, there are only other individuals. See p. 32-3, 50-1. In effect Nozick agrees with Rawls's criticism of Utilitarianism: in adding and subtracting the costs and benefits to all the persons affected, and in allowing benefits to one person to offset costs to others, Utilitarianism does not sufficiently acknowledge the separateness of persons.

Some criticisms of Nozick's theory
From T. Nagel 'Libertarianism without Foundations', Yale Law Journal (K/29/.A4) 85 (1975) p. 136f.
(1) It is risky to argue from small-scale examples to universal principles.

Nozick supposes 'that it is possible to determine what governments may and should do by first asking what individuals, taken a few at a time in isolation from large-scale society, may do, and then applying the resultant principles to all possible circumstances, including those which involve billions of people, complicated political and economic institutions, and thousands of years of history. What is more surprising, he discovers in himself intuitions about the moral requirements on men in a state of nature which he is willing to endorse as universal principles unmodified in their cumulative effects when applied in any circumstances whatever.... It is hard to see how anyone could seriously arrive at firm moral opinions about the universal principles of human conduct without considering what it would be like if they were universally applied, in iterations which might create complex effects of scale'.

(2) Nozick's intuitions are in any case wrong, because they see as absolute some rights that are only prima facie rights.

'The intuition that Nozick discovers in himself is that everyone has an absolute right to be free from coercion, and an absolute right to acquire and dispose of his property - so long as he is not violating the same rights of others and so long as his acquisition of property does not, for example, give him sole title to the formerly public water supply of a desert community. Nozick's intuition is that each person is entitled to his talents and abilities, and to whatever he can make, get, or buy with his own efforts, with the help of others, or with plain luck. He is entitled to keep it or do anything he wants with it, and whomever he gives it to is thereby equally entitled to it. Moreover, anyone is entitled to whatever he ends up with as a result of the indefinite repetition of this process, over however many generations....

'Nozick's moral intuitions seem wrong even on a small scale. He denies that any of the rights he detects may be overridden merely to do good or prevent evil. But even if it is not permissible to murder or maim an innocent person to promote some highly desirable result, the protected rights do not all have the same degree of importance... It is far less plausible to maintain that taking some of an innocent man's property is an impermissible means for the prevention of a serious evil, than it is to maintain that killing him is impermissible. These rights vary in importance and some are not absolute even in the state of nature...

'Rights limit the pursuit of worthwhile ends, but they can also sometimes be overridden if the ends are sufficiently important. The only way to make progress in understanding the nature of individual rights is to investigate their sources and their relations to each other and to the values on whose pursuit they set limits'.

(3) Nozick is wrong in holding that benefit to one person can never offset cost to another: sometimes it can.

To make sense of utilitarianism,

'All one needs is the belief, shared by most people, that it is better for each of 10 people to receive a benefit than for one person to receive it, worse for 10 people to be harmed than for one person to be similarly harmed, better for one person to benefit greatly than for another to benefit slightly, and so forth.... If a choice among such alternatives does not involve the violation of any rights or entitlements, but only the allocation of limited time or resources, then we regard those comparisons as excellent reasons for picking one alternative rather than another. If we can help either 10 people or one person, not included in the 10, and we help the 10, then we can say that rescue of the 10 outweighs the loss of the one, despite the fact that he does not get some overbalancing good from his sacrifice, and his is the only life he has. So for the purpose of comparing possible outcomes of action, where the violation of rights is not in question, it is clear that the distinctness of individuals does not prevent balancing of benefits and harms across persons. If special constraints enter in when a sacrifice is to be imposed on someone as a means to the achievement of a desirable outcome, their source must lie elsewhere. Such constraints should not derive from a principle which also has the consequence that practically nothing can be said about the relative desirability of situations involving numbers of different people.

(4) Nozick is wrong about the basis of rights.

'Furthermore, the source of rights of the general kind Nozick advocates cannot be discovered by concentrating, as he suggests we should, on the meaning of individual human lives and the value of shaping one's own life and forming a general conception of it. Vague as his suggestions are, they all suffer from an error of focus, for they concentrate solely on features of persons that make it bad for certain things to happen to them, and good for them to have the opportunity to do certain things. But rights of the kind that interest Nozick are not rights that certain things not happen to you, or rights to be provided with certain opportunities. Rather they are rights not to be interfered with in certain activities. They give rise to claims not against the world at large, but only against someone who contemplates deliberately violating them The relation between the possessor of the right and the actor, rather than just the intrinsic nature of the possessor and of his life, must enter into the analysis of the right and the explication of its basis.

'Any theory of rights must explain this structural feature, even if it does not follow Nozick in elevating the unimpeded exercise of the will into the supreme principle of morality. It is of the first importance that your right not to be assaulted is not a right that everyone do what is required to ensure that you are not assaulted. It is merely a right not to be assaulted, and it is correlated with other people's duty not to assault you. This cannot be explained simply by the fact that it is bad to be assaulted, which is merely an item in the catalog of values by which the desirability or undesirability of occurrences or sets of occurrences is to be weighed. That assault is disagreeable or bad does not explain why the prohibition of it should serve as a constraint on the pursuit of other values or the avoidance of other harms, even if those other values outweigh the badness of assault in a pure calculation of the relative desirability of possible outcomes. Sometimes one is required to choose the less desirable alternative because to achieve the more desirable one would have to violate a right....

'An explanation of the basis of right would therefore have to concentrate on the actor and his relation to the person he is constrained not to treat in certain ways, even to achieve very desirable ends. And it would have to explore the interaction between those constraints, and the goals whose pursuit they constrain. There is no reason to think that either in personal life or in society the force of every right will be absolute or nearly absolute, i.e. never capable of being overridden by consequential considerations. Rights not to be deliberately killed, injured, tormented, or imprisoned are very powerful and limit the pursuit of any goal. More limited restrictions of liberty of action, restrictions on the use of property, restrictions on contracts, are simply less serious and therefore provide less powerful constraints'.

(5) The distinction between end-result and historical theories of justice is not as sharp as Nozick supposes (if we recognize that some rights are not absolute, but may be subordinated to certain valuable ends).

A political theory might -

'assign society the function of promoting certain goods and preventing certain evils, within limits set by the differing constraints of different individual rights. It would not judge processes and procedures solely by their tendency to produce certain outcomes, nor would it judge outcomes solely by the processes that had produced them. Social institutions and the procedures defining them would be assessed by reference both to their respect for individual rights and liberty, and to their tendency to promote desirable ends like the general welfare.

'Nozick offers a classification of principles of distributive justice into which such a theory does not fit.... Suppose a theory says that a distribution is just if it results from a process governed by rules that reflect (a) the suitability of certain patterns, (b) the desirability of increasing certain good results and decreasing certain evils independently of any pattern, and (c) a respect for individual rights of differing importance, Such a theory will be at bottom neither purely historical nor purely patterned. It will be formally historical, but the "historical" or process criterion will be partially determined by considerations of pattern and considerations of total outcome. Therefore Nozick's concentrated attack on patterned principles and nonhistorical principles provides no reason to think that his alternative is correct'.

Note: 'More specifically, his arguments against Rawls are seriously weakened by a Procrustean attempt to portray Rawls's principle of distributive justice as a nonhistorical or end-result principle. Rawls does not maintain that the justice of a distribution can be determined independently of how it was produced. He believes that its justice depends on the justice of the institutions, including legal institutions defining entitlement, which were involved in its production. These are assessed only partly on the basis of their tendency to promote a certain distributive end-state. Rawls, for example, gives priority to the preservation of individual liberty... [which] certainly restricts the procedures by which a distribution can be justly arrived at.'

(6) On the Wilt Chamberlain example, by which Nozick shows that preservation of a pattern would violate the liberty of property owners to do as they wish with their own: again, he supposes that if there is a right, it is absolute.

'It only seems a problem to Nozick, and a further violation of liberty, because he erroneously interprets the notion of a patterned principle as specifying a distribution of absolute entitlements (like those he believes in) to the wealth or property distributed. But absolute entitlement to property is not what would be allocated to people under a partially egalitarian distribution. Possession would confer the kind of qualified entitlement that exists in a system under which taxes and other conditions are arranged to preserve certain features of the distribution, while permitting choice, use, and exchange of property compatible with it. What someone holds under such a system will not be his property in the unqualified sense of Nozick's system of entitlement'.

further readings
Western Political Quarterly (JA/1/.W4) 39 (1975) p. 177f has a number of articles on Nozick's book. See especially those by Johnson, King and Singer. Johnson's article gives a more extensive account of the content of the book.
Ethics (BJ/1/.16) 87 (1977) p 97f (January number) also has several articles on Nozick. See especially p. 130-141.

L. P. Francis and J. G. Francis, 'Nozick's Theory of Rights: A Critical Assessment' Western Political Quarterly, 39 (1975) p. 634.

N. Cooper, 'Justice and Historical Entitlement', Canadian Journal of Philosophy, 7 (1977), p. 799f.

J. Narveson, 'Anarchy, State and Utopica (sic)'; Dialogue 16 (1977) p. 298f.

Not mentioning Nozick, but useful background: H. J. McCloskey 'Rights - Some Conceptual Issues' Australasian Journal of Philosophy 54 (1976) p. 99f.