The author, political science professor David Lewis Schaeffer, is writing about John Rawls (welfare state liberal, roughly) and Robert Nozick (libertarian) from a conservative point of view, complaining that the are really too much alike -- which, from his point of view, I guess they are. But what is the point of similarity that he sees?
From a conservative point of view, both advocate too much personal freedom. "Victimless" offenses, like using dangerous drugs or practicing prostitution, would be permitted, if they had their way. (It's actually not obvious that this is true of Rawls, but let it pass.) Where they differ, if at all, is on the issue of distributive justice. Amazingly, Schaeffer claims they agree here too.
As you may know, Rawls advocates the "Difference Principle," which says that inequalities of wealth and income are only to be permitted if they benefit the people who are the least well off. (For instance, they may be necessary incentives to induce talented people to produce these benefits for the least-well-off.) This requirement of benefiting the least well off could justify some sort of coercive redistribution of wealth, from those who have "too much" to those who have "too little." This idea is an example of what Nozick calls an "end-result principle." Regardless of the process by which you acquired what you have, if the result of the process does not fit this principle, you are liable to coercive redistribution.
Nozick's alternative is what he calls a "historical principle." Whether what you have is rightfully yours depends on past history. On his view, the "Entitlement Theory," you are entitled to your holdings if you acquired them in the right way (eg., by purchase and not by theft) from someone who was entitled to them (by this same principle). Since this rule is recursive, it goes back to the beginning of time unless it is qualified somehow. Of course, there have been people who acquired their holdings in the wrong way (eg., by theft). This brings in the Principle of Rectification, which requires that those who were wronged, or their heirs, be compensated for the wrongs they have suffered.
According to Schaeffer, this is where Nozick suddenly morphs into Rawls:
Ironically, however, Nozick himself ultimately acknowledges that his entitlement theory is insufficient to refute demands for a redistributionist state, since it can never be demonstrated that existing holdings derive from an unbroken series of voluntary transfers... Hence, surprisingly, he ends up suggesting that something like Rawls’s difference principle is morally required after all, in the name of “rectification,” on the dubious premise that those currently least-well-off have the highest probability of being descended from previous victims of injustice.In the next paragraph, Schaeffer suddenly upgrades Nozick's "suggesting" this into an "area of agreement with Rawls." Apparently it was something more than a suggestion, according to Schaeffer. Let's look at what Nozick actually says.
Beginning with the paragraph in Anarchy, State, and Utopia (p. 152) in which he introduces the idea of rectification, Nozick acknowledges a number of times that the problem of how this principle is to be specified and applied raises all sorts of difficult questions. Right away, he raises, as an issue he will not try to settle, this question: "How far must one go in wiping clean the historical slate of injustice?" In other words, maybe we should recognize a sort of moral statute of limitations. Maybe, if the injustice from which your holdings derive happened before a certain cut-off date, you owe nothing on their account.
For whatever it might be worth, my own view is that it is virtually inevitable that we do this. The further back in time we go in finding injustices, the more insoluble the difficulties in deciding what sort of compensation would be required to set it right.
In a note on the next page (p. 153 n.), however, Nozick raises the possibility of quite a different approach. Suppose that rectification means making the victim's heirs as well off as they would have been had the injustice not happened. Suppose, further, that there are several different ways things might have turned out. Which one should the rectifier bring about? Here, he says, we might use some end-result principle to choose between these states of affairs. We might bring about the scenario that would create the most happiness overall (utilitarianism), or we might bring about the one that benefits the people who are now worst off (the Difference Principle), and so forth.
Again, much later (pp. 230-231), he plays with yet another possibility. (One thing Nozick had a lot of is ideas.) Maybe, instead of trying to figure out how things might have been if really old injustices had not happened, we could substitute, as "a rough rule of thumb," something like the Difference Principle, on the theory that the least well off people are most likely to be the ones whose ancestors were treated badly (African slaves, dispossessed Indians, etc.). He admits, though, that "this particular example may well be implausible."
This last idea must be the one that Schaeffer is (mis)reporting.
Obviously, Nozick is not agreeing with Rawlsian redistribution, for at least two reasons. For one thing, he is not asserting that any of these three ideas. They are mutually inconsistent, so he can't believe them all, and he has not picked one. Also, even if he were to opt for the last, most Rawlsian-sounding, one, the point of the transfer payments that he would be recommending would be to rectify past wrongs. Nozick repeatedly points out that whether it is redistributive to take from Peter to give to Paul depends on what our reasons are for doing it. If the reason is historical, if we are trying to right past wrongs, then we are not redistributing. You are simply returning things to their real owner. If, on the other hand, regardless of what happened in the past, you just think Peter has "too much," then -- and only then -- you are redistributing. Nozick's reasons, if he were to opt for this third idea -- which he has not done! -- would be non-redistributive.
As I have suggested, I think the obviously best solution here is the simplest one: some sort of statute of limitations. Nothing that even looks Rawlsian about that!