There’s a common argument that libertarians make against the idea of social or distributive justice. The argument, made by both Robert Nozick and Friedrich Hayek, purports to show not merely that the idea of distributive justice advocated by left-liberals like John Rawls is immoral, but that it is conceptually confused. Asking whether the distribution of wealth in a society is just is like asking whether the color blue is heavy, or whether a stone is moral.
I don’t think this argument is entirely without merit, but I also don’t think it shows that libertarians should reject the idea of social justice. In this post, I’ll set out the argument and explain how it succeeds, and how it falls short.
Here’s what Nozick has to say about the concept of distributive justice:
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. Into this process of distributing shares some error may have crept… [But] [t]here is no central distribution, no person or group entitled to control all the resources, jointly deciding how they are to be doled out. What each person gets, he gets from others who give to him in exchange for something, or as a gift…There is no more a distributing or distribution of shares than there is a distribution of mates in a society in which persons choose whom they shall marry.
– Anarchy, State, and Utopia, pp. 149-150.
Friedrich Hayek is even harder on the idea:
I must turn now against an abuse of the word [justice] which threatens to destroy the conception of law which made it the safeguard of individual freedom…‘Social’ justice (or sometimes ‘economic’ justice) came to be regarded as an attribute by which the ‘actions’ of society, or the ‘treatment’ of individuals and groups by society, ought to possess. As primitive thinking usually does when first noticing some regular processes, the results of the spontaneous ordering of the market were interpreted as if some thinking being deliberately directed them, or as if the particular benefits or harm different persons derived from them were determined by deliberate acts of will, and could therefore be guided by moral rules…It is a sign of the immaturity of our minds…
See the respective texts for the full argument. Basically, though, I think it can be reconstructed as follows:
- A distribution of holdings is only unjust if it results from the unjust (deliberate) actions of individuals.
- There is no central distributor responsible for the distribution of holdings in a free society.
- Therefore, the distribution of holdings in society cannot be unjust.
- The distribution of holdings in a society arises from the aggregation of all our ordinary decisions about what talents to develop, where to live, what to purchase, and so on.
- Except for special cases specifically identified as unjust by a theory of justice in transfer (e.g. theft, fraud), these ordinary decisions are obviously not unjust.
- Therefore, the distribution that arises from them cannot be unjust.
To summarize, for the distribution of holdings to be unjust it would have to be the result either of (a) the unjust distribution of a central distributor, or (b) the unjust actions of dispersed individuals. (a) won’t work because the central distributor doesn’t exist, and (b) won’t work because the ordinary actions of individuals that generate the overall distribution of holdings are obviously not unjust.[1. This is an undefended intuition in Nozick, but it is a pretty unproblematic one in my opinion. It is also an intuition that plays a key role in the recently–much–discussed Wilt Chamberlain argument. That argument draws much of its force from the idea that it is obviously not unjust for individuals to give Wilt Chamberlain a little bit of their (justly held) money to watch him play. In the same way, there would be something odd about a theory of justice that condemned ordinary cases of giving money to one’s children, receiving wages from one’s labor, and so on.] Hence the distribution of holdings cannot possibly be unjust.
There is, I think, an important insight in this reasoning. Inequalities that are the product of spontaneous orders are different, morally speaking, from inequalities that are produced deliberately by a central distributor. If our government assigned mates to everyone in the country except 10% of people, those among the unfortunate 10% would have reason to feel that they were not treated with equal respect as citizens. The same would not be true, I take it, if mates were selected as they are now – by mutual consent. Someone who can’t find anyone else who wants to marry them might feel justly disappointed, and perhaps even disrespected by particular other persons. But even if there is a kind of disrespect involved, it is a disrespect that is qualitatively different from the kind of insult to public standing involved in the first scenario.[2. David Schmidtz draws out the connection between inequality and respect in his “Equal Respect and Equal Shares,” a modified version of which appears in his Elements of Justice.]
There is, nevertheless, a flaw in the argument. That flaw appears in the first premise of the second argument. Yes, the distribution of holdings in a free society is determined partly by the countless ordinary decisions of innumerable individuals. But it is also a product of the social and legal rules that govern and structure those decisions: rules that determine the contours of property rights and contracts, that determine whether there will be a social safety net and what form it will take, that determine the extent, limit, and uses of taxation, and so on. These rules might be just or they might be unjust. If they are unjust, we could (as Judith Shklar pointed out) intervene to change them, even if we did not deliberately create them.
Now, one way – not the only way – of determining which rules are just is to think about the effects that different sets of rules will likely have on the pattern of distributions within a society. We might not know how any particular individual will fare under regime X vs. regime Y, but we can be reasonably confident about certain “pattern predictions”: that low-skilled workers will suffer under a regime that includes robust minimum wages, that fewer people will receive medical care if it is not subsided by tax revenue, and so on. And we can use this information in constructing a theory of social justice. We can have a theory of social justice, in other words, that holds that a given set of legal rules is just only if its expected effects on certain social groups meets a certain standard. We could say, like Rawls, that the legal rules (or basic structure more generally) must work to the maximum advantage of the least well-off. Or we could say something else – that it must make everyone better off than they would be in the state of nature, or than they would be under a situation in which all resources were distributed equally, etc.
Which of these standards is the right one to apply to the evaluation of social and legal rules is a question that can be resolved only by substantive philosophical debate. My point here is not to defend any particular standard, though I suspect that any reasonable standard will have at least a prioritarian or sufficientarian element to it. I do not even claim to have given an argument here that any such standard ought, in the final analysis, to play a role in evaluating the justice of legal and social rules. My goal here is a more limited one: to show that such a standard can capture much of what is good about the idea of social justice without falling into the errors that Hayek and Nozick rightly criticize. What I am arguing is that we needn’t suppose that stuff is doled out by a central distributor to believe that justice has something to say about the rules we set up to govern how stuff is doled out by dispersed economic agents. And we needn’t denigrate the importance of individual liberty and spontaneous order to believe that justice requires that legal and social rules be designed so that they can be reasonably expected to work to the advantage of certain social groups – perhaps especially the poor and vulnerable.
Understood in this way, there is nothing inherently un-libertarian in the idea of social justice. What I have defended here is a moral standard that tells us when a set of legal rules is morally justified and when it is not, on the basis of the effects that set of legal rules can be expected to have on various social groups. But even if the standard says “a set of legal rules is just only if it maximizes the well-being of the least well-off,” this doesn’t directly entail that government has to engage in redistributive taxation, finance public education, or any of that. Whether it should do so depends on whether doing so actually, in fact, maximizes the position of the least well-off. A theory of social justice, as I’ve described it, tells us only what ends a set of legal rules ought to try to achieve. It is therefore dependent on the insights of economics and other social sciences to tell us how to go about achieving those ends.
And, actually, if you read them closely, it looks an awful lot like both Hayek and Nozick do endorse a theory of social justice of the sort I have described. Remember, Nozick’s endorsement of free markets and private property is not only qualified by a modified version of the Lockean Proviso, it partially depends for its justification on that Proviso. For Nozick, roughly, a libertarian legal system that allows for permanent bequeathable property is justified only to the extent that that system makes no one worse off than they would have been in the state of nature.[3. Nozick alternates between applying the Lockean Proviso to particular acts of appropriation and applying it to the general system that licenses particular acts of appropriation. For an example of the latter use, see p. 177 of ASU.] If private property and free markets really did lead to the oppression of the poor and vulnerable, then private property and free markets would not, for Nozick, be morally justifiable. This is why he finds it necessary (on page 177) to appeal to standard free-market economic reasoning to show that they do not.
Hayek’s case is more complicated, and I fear that I will not be able to do it justice here. But readers who believe that Hayek completely dismissed the idea of social justice (and they have ample textual support for this belief!), should remember Hayek’s remark about Rawls’ Theory of Justice that the difference between Rawls’ account of justice and his own are “more verbal than substantive” and that he and Rawls “agree on what is to me the essential point.” Perhaps, as Jeremy Waldron suggests, Hayek was simply confused in his reading of Rawls. But I think we should not be so quick to write off Hayek as a sloppy reader, especially when a more charitable interpretation is readily available.[4. Some illuminating comments, and a bit of backpeddling on Hayek’s part, can be found in this excerpt from a conversation between Hayek and James Buchanan. Gosh I wish this clip was longer!] Readers who are interested in a thorough and carefully argued defense of such an alternative interpretation should consult the work that John Tomasi has done on this topic. You can get a glimpse of it here. But you’ll have to wait for the much fuller version presented in chapter 5 of his Free Market Fairness, due out in 2012 from Princeton University Press.