« Dawkins on race | Gene Expression Front Page | HBU »
September 25, 2004

A load of Rawls

Long ago, when I made a post on Heroes and Villains, [000188] someone suggested I should deal with some more recent examples.

I remembered this recently when there was a discussion of John Rawls and his famous Theory of Justice. This prompted me to take another look at the book and set out some thoughts on it.

Hero or villain? Read on to find out…

I’m sure everyone has read Rawls’s book (ho, ho), but here is a brief summary. (Page references are to the Oxford paperback edition, 1973, and numerous reprints).

According to Rawls a society is ‘a cooperative venture for mutual advantage’ (4). A set of principles of ‘justice’ and ‘fairness’ is needed ‘for assigning basic rights and duties and for determining… the proper distribution of the benefits and burdens of social cooperation’ (5). The main ‘goods at the disposition of society’ are ‘rights and liberties, powers and opportunities, income and wealth’ (62, 92) The principles of justice are to be agreed once and for all, and recognised as binding in advance of their application (4, 11, 12, 99, 115). To determine these principles Rawls proposes an extension of the traditional concept of the social contract: ’they are principles that free and rational persons concerned to further their own interests would accept in an initial position of equality as defining the fundamental terms of their association’ (11). In order that this hypothetical ’original position’ should be genuinely equal, Rawls argues that it is necessary that the parties should be ignorant of their social position, class, natural assets and abilities (including intelligence), and anything else which might distinguish their judgement of their own interest from that of any other party (12, 136-142). This is the postulate of the ’veil of ignorance’. Given these assumptions, Rawls argues that rational people placed in this ’original position’ and acting in their own interest would agree on the basic principle that ’all social values - liberty and opportunity, income and wealth, and the bases of self-respect - are to be distributed equally unless an unequal distribution of any, or all, of these values is to everyone’s advantage’ (62). Since the parties are in a state of equal ignorance about their own prospects, they have no reason to accept any inequality of distribution, unless they can be sure that they themselves will gain from such inequality. Rawls supports this conclusion by appealing to the ’maximin’ principle: in a state of uncertainty, it is rational to choose the option for which the worst possible outcome is the best among all the alternatives (152-57). The practical implication of this principle of ‘justice’ is that income and wealth should be redistributed through taxation (278-80) unless and until any remaining inequality improves the position even of the worst-off (for example by providing incentives to economic efficiency and enterprise).


So much for the theory. Here are what I think are the key objections to it:

1. The original social contract theory of Hobbes and Locke is open to many criticisms, but it does have the merit of offering a solution to a real problem: why should we give up the freedom we would enjoy in a state of nature, and submit to laws and government? The social contract theorists replied that the freedom of the ‘state of nature’ would lead, in Hobbes’s terms, to a war of all against all. It is therefore in the interests of everyone to accept limitations on their freedom and agree with each other to submit to laws, to be enforced by an authority stronger than any individual. It need not be supposed that any actual group of individuals (now or in the past) has literally made such an agreement: the point is that it would be in their interest to do so if the need arose. It is always possible that existing institutions could break down and society dissolve into a ‘state of nature’ - and current events in Iraq show that Hobbes’s ‘war of all against all’ is not as far from reality as his critics have supposed. The social contract theory therefore provides an on-going justification for the restriction of freedom involved in submitting to law and government.

The variant of social contract theory offered by Rawls lacks this merit. His question is not ‘why should we give up our freedom?’, but ‘how should we share the benefits of co-operation?’ Unlike the traditional question, this one does not need an answer in terms of a hypothetical social contract. The decision on how to share the benefits of cooperation can be taken by the participants in such cooperation (or by their elected representatives). People are not obliged to enter into social cooperation: they do so because they choose to, on terms which they find acceptable. To a large extent (in modern western societies) they are free to opt out of the mainstream of social and economic life if they wish to. [See Note A for a qualification.] There is no grand initial contract, but a lot of small ones. Entering into cooperation is not a surrender of freedom, but an exercise of freedom. No-one loses by it, so the question why they should do it requires no fancy answer. Rawls’s 600-page treatise is a pursuit of a giant red herring.

[Note A: Admittedly in modern societies people are compelled to pay taxes for public services which they may not want or need. Some people pay taxes far greater than the value of the services they receive. If it is asked why such people are morally obliged to pay their taxes, the short answer is that they are not, though it may be prudent or expedient to do so.]

2. A more specific objection is to the scope of ‘social cooperation’ in Rawls’s theory. He defines it so broadly as to include all the products of the economy (goods and services). Since in Rawls’s account these are part of the benefits of social cooperation, the resulting income and wealth must be distributed in accordance with the principles of ’social justice’. This implies extensive redistributive taxation.

Many critics of Rawls have pointed out that in a free economy goods and services are not produced by ’society’ or the ‘state’: they are produced by the effort and enterprise of individuals and groups of individuals (including corporations, which ultimately represent groups of risk-taking stockholders). The resulting wealth and income should therefore not be treated as a social product, to be shared among all the members of society whether or not they have contributed to it.

I think this objection is broadly correct, though a Rawlsian might reasonably say that part of the value of all goods and services should be attributed to the favourable circumstances provided by social order, property laws, and so on. However, this part of economic output can be paid for by reasonable levels of corporate taxation.

It should be noted that objection (2) is independent of objection (1). It would be possible to accept that Rawls’s argument is sound for social cooperation generally, when properly defined, (contrary to objection (1)), but to deny that economic activity falls within the scope of social cooperation.

3. Another fundamental objection is to the way that Rawls identifies the principles of social justice with the dictates of self-interest in a certain defined situation (the ‘original position‘). In ordinary linguistic usage justice has nothing to do with self-interest. Justice may happen to coincide with self-interest, but they may equally well conflict, and one of the fundamental principles of ’natural justice’, as generally recognised, is that self-interest should be excluded from judicial deliberation: no man can be a judge in his own case. To identify justice with self-interest would be an abuse of language.

It might of course be argued that in the long term the principles of justice are to everybody’s benefit, even if they conflict with self-interest in particular cases. But Rawls’s approach goes beyond this: he does not merely defend the principles of social justice by an appeal to self-interest, he actually defines his principles of justice by reference to self-interest.

This is more than a verbal quibble. The principles of justice are expected by Rawls to be universal. It would be pointless for people to agree to principles of justice in the artificial circumstances of the ’original position’ if they could not be extended to other, more realistic, circumstances. But the dictates of self-interest cannot legitimately be extended in this way. What is in someone’s self-interest in one set of circumstances may be against his interest in another set of circumstances, and if self-interest is the ultimate criterion of rationality, the individual should constantly reassess his values with changing circumstances. By analogy, in Bayesian inference we might make an assumption about a probability in a state of ignorance, but we need not (and should not) continue to apply that assumption when better information is available and we know that the initial assumption was wrong. Rawls assumes that we are bound by the principles of justice that we would agree to in the circumstances of the ‘original position’ (i.e. behind the ‘veil of ignorance‘), but if ’justice’ ultimately reduces to self-interest this assumption is invalid. Rawls gives no reason, that I can see, why anyone should stick to the ’social contract’ as soon as they see that its terms are against their interest.

4. Rawls’s use of the ’veil of ignorance’ has always been one of the most controversial parts of his theory. Why should people base their decisions on a hypothetical and counterfactual state of ignorance, rather than using their actual knowledge of their own abilities and preferences? Rawls‘s defence of this fundamental assumption is remarkably feeble. He argues, first, that the veil of ignorance is necessary if there is to be any unanimous agreement at all on the principles of social distribution (140). But this is a bare assertion, and it is not obviously true. If people are aware of their abilities and preferences, would they not accept a distribution of social goods in accordance with the market value of their own contributions to society, subject perhaps to a safeguard for those whose market value would not provide a subsistence income? In any case, it is not clear that agreement to the social contract needs to be unanimous. If a majority can agree to a set of principles, the minority of dissenters might be allocated territory in proportion to their numbers and told to ’do their own thing’. Indeed, the population might split up into several different groups based on common abilities and interests, each constituting an independent society with its own rules. Arguably this would be more consistent with intuitive concepts of justice and freedom than the Rawlsian approach, which implicitly reduces everyone to the lowest common denominator. (Incidentally, I don’t think Rawls discusses the problems of territory and boundaries. Does his doctrine require redistribution of income and wealth on a global basis? If not, why not?)

The real reason for Rawls’s insistence on the veil of ignorance is to load the dice in favour of ’equality’: the veil of ignorance ’ensures that no one is disadvantaged or disadvantaged in the choice of principles by the outcome of natural chance or the contingency of social circumstances’ (p. 12, see also p. 141). But this is really prejudging the main point at issue, which is whether people are entitled to receive the benefits of their natural gifts. Rawls’s approach simply rules this out by a procedural ploy. The point at issue needs much fuller discussion on its own merits than Rawls gives it (or than I can give it here), but there would certainly be a widespread view that people are entitled to enjoy the benefits of their natural gifts, even though they have not done anything to deserve those gifts. Nature is not fair, but we cannot expect human action to redress all Nature’s inequities.

5. Finally, on a more technical level, even if we were to accept Rawls’s approach to the original position, the veil of ignorance, etc, it is not clear that the maximin principle is appropriate as the basis for individuals to assess their self-interest in the ‘original position‘. The maximin principle has acquired a certain unjustified prestige from its use in game theory, but even in game theory a maximin strategy (pure or mixed) is only the accepted ’solution’ in the special case of zero-sum games for two players, where any departure from the maximin strategy by one player is likely to be to the advantage of his opponent. Attempts to apply the maximin principle to decision theory outside the context of games have not been widely accepted (see Luce and Raiffa, Games and Decisions, chapter 13). The maximin principle implies a risk-averse approach, in which people prefer to give up the possibility of major gains if this involves any possibility of losses, however small. It is not clear that this is ‘rational’, and it is certainly inappropriate if we have some information about the relative probability of the different outcomes. It would not be ’rational’ to give up a high probability of a large gain to avoid a small probability of a small loss, but this is in effect what Rawls’s principle requires.

Rawls’s theory is presented on a very high level of abstraction, but it would be a mistake to suppose that it has had no effect on practical politics. Directly or indirectly it has been enormously influential, especially on the ’social democratic’ left. Even those who have never read Rawls derive some comfort and support from the vague feeling that redistributive taxation (aka state theft) is a matter of ’social justice’: Rawls proved it, didn’t he?

So overall, in my opinion, Rawls is definitely a villain. What I find most objectionable is not the fallacies in his reasoning, but the blatant intellectual dishonesty underlying the whole system. As Rawls himself comes close to admitting (141), the terms of the ‘original position’ and the ‘veil of ignorance’ are deliberately rigged to produce the desired result: namely, a moderate egalitarianism. I can respect those who espouse egalitarianism on overt moral grounds, but not those who try to impose it as a dogma by pretentious and artificial philosophical arguments. Ultimately, of course, moral principles cannot be proved or disproved. [See Note B for a qualification.] We can only try to elucidate their assumptions and implications. People must then make their own choice, influenced by a mixture of innate and acquired factors. Egalitarianism probably has some evolutionary basis, but one of selfishness rather than altruism: in the distribution of goods (such as the produce of a successful hunt) individuals do not want to be cheated out of their own ‘fair‘ share. Experiments have shown that even monkeys have a sense of ’fairness’, and are outraged when rewards are divided unequally. But of course people (if not monkeys) also want to be rewarded in proportion to their ’merits’, and will be dissatisfied if rewards are equal when the merits are not. Rawls’s theory is compatible with the first of these tendencies but not the second.

[Note B: as Hume first pointed out, values cannot be deduced from facts: ’ought’ cannot be derived from ’is’. However, there is an important qualification to this. It is generally recognised that there cannot be a moral obligation to do something that is beyond our power to do, such as a physical or logical impossibility. Since what is possible is a matter of fact, not value, it follows that the absence of an obligation can be derived from facts.]

Posted by David B at 04:03 AM