Some of you may be following the blogospheric controversy over Stephen Metcalf's Slate article attacking Nozick and libertarianism. The best quick summary of the excellent replies to this really awful piece is here. All I'll say is that when a writer gets me nodding along with Brad DeLong's quite correct criticism, he must be really bad.
But one point that hasn't been made enough is the way in which Metcalf and some of his defenders including Jonathan Chait are abusing and misreading the famous Wilt Chamberlain example in Anarchy, State, and Utopia. The point of that example was NOT to justify all forms of wealth acquisition and not even voluntary ones as Chait seems to think. It's not an exercise in justification, moral or otherwise. The point was to demonstrate a problem with what Nozick called "pattern" theories of justice such as that of Rawls.
Specifically, that subsection of the book is in the chapter on distributive justice and is called "How Liberty Upsets Patterns." What Nozick is up to there is showing that you cannot have BOTH a theory of distributive justice that is based on a specific pattern of wealth distribution as the ideal AND give people liberty over their private property. Nozick asks the reader to assume that their preferred distribution exists. Assume further though that people are free to dispose of some portion of their income. Suppose they all wish to pay Wilt to play basketball. After all of these voluntary transfers take place, the previous distributive pattern will be upset and Wilt will have more wealth than before and others less. By hypothesis, this is a less than ideal pattern.
Nozick's point here is that if you really believe in pattern theories of distributive justice, you must put major constraints on people's liberty to dispose of their incomes or they will consistently "upset" those patterns. You can have liberty or patterns, but not both.
This is another way of making a point that Hayek made in The Constitution of Liberty: if you want equality of outcomes, you have to treat people unequally; if you want to treat people equally, you have to accept inequalities of outcomes. What Nozick does is to generalize this point to ANY pattern or end-state theory of justice.
The Chamberlain example is not a moral defense of any and all voluntary transactions and certainly not of any attempts at acquiring wealth. It is an argument for the way in which his preferred entitlement theory of justice is compatible with liberty in a way that pattern or end-state theories of justice are not. It's an argument against a particular kind of theory of justice on the grounds that such theories will of necessity seriously violate liberty. It is nothing more and nothing less.
In fact, Nozick even tells you what his point is: "The general point illustrated by the Wilt Chamberlain example...is that no end-state principle of distributional patterned principle of justice can be continuously realized without continuous interference with people's lives." (163) Would it have killed the critics to actually look at the text?
It remains what it intended to be: a brilliant response to Rawls, nothing more and nothing less.
So Chait and everyone else can stop trying to ride it where it was not meant to go. It's not a justification of any specific distribution of wealth but a counter-argument to Rawlsian (or other) end-state theories of justice by claiming, rightly in my view, that they are not compatible with liberty.