Responding to yesteday’s post, Matt Yglesias argues that Stephen Metcalf is still kinda on point because, even if Nozick remained a libertarian on some grounds—maybe pragmatic or consequentialist ones—he nevertheless abandoned the deeper philosophical opposition to redistributive taxation that characterizes Anarchy, State, and Utopia. Matt does, however, back off when I point him to pages 281-282 of Nozick’s final (2001) book Invariances, where Nozick writes:

The different levels of ethics have a different status. The ethics of respect, largely specified by what I have called the core principle, is the part, the one part (I think), that is (that should be) mandatory across all societies. In saying this, I am putting forward a particular normative position: that the further ethical levels are matters of personal choice or personal ideal. Even if these further levels are not mandatory for all societies, some particular society may attempt to make one or another of these further levels [requiring affirmative assistance to others, caring and love, and embodiment of various spiritual ideas] mandatory within it, punishing those members of the society who deviate or fall short. I also believe—this is an additional component of my own position, presented in Anarchy, State, and Utopia—that no society should take this further step. All that any society should (coercively) demand is adherence ot the ethics of respect. The further levels should be matters for a person’s own individual choice and development.

Yglesias actually thinks this view is arguably more extreme than the original A,S,&U position, though since Nozick’s primary concerns in that chapter are metaethical, I’d be chary of inferring too much from his passing remarks here. Still, it’s clear enough that he regards this aspect of the position laid out in A,S,&U as “his position.” In what sense, then, do we think he became “less hardcore,” and in what sense did he regard his earlier views as “seriously inadequate”? While it’s probably a mistake to try to shoehorn too much of what a thinker as notoriously mercurial as Nozick wrote into a single coherent view, given the span of time involved, I think it’s useful to consider the position laid out in Invariances in light of what Nozick says in “The Zig-Zag of Politics,” which is the essay that gave rise to the popular impression that Nozick had wholly shed his libertarian ideas. I’ll return to that momentarily.

First, an observation about Invariances. Here, Nozick spells out a view of morality as an evolved adaptation with a specific function: Enabling ever broader mutual social cooperation. He divides this cleanly into “levels,” but it’s not obvious that the boundaries are so sharp and clean as all that. On his “core principle of ethics,” for instance, Nozick writes:

It is desirable to extend the realm of people who benefit from coordination and cooperation. A group G1 should extend social cooperation to G2, if this can be done in some way that benefits those in G2 and improves (or does not worsen) the situation of the poeple in G1. [….] However, a group is not required to extend cooperative relations to another group with whom it has no interactions, at the cost of lessening the benefits to itself.

Now, pace Nozick, it seems at least possible to me that this principle is compatible with some forms of affirmative assistance. It might, for instance, be the case that the (or a) Pareto optimal scenario for interaction between a Better Off and Worse Off group, on the whole and over the long term involves some level of subsidy from Better Off to Worse Off, in the form of provision for education or insurance against extreme hardship, because the social surplus from the productive cooperation thereby enabled—and in particular the share of the surplus accruing to Better Off—is greater than the amount of the subsidy. (It may also be less expensive than enforcing libertarian rights against members of Badly Off who may feel compelled to turn to crime in the absence of viable opportunities for productive cooperation.) Where this is the case, that sort of limited aid might fall within Nozick’s “core principle” even when it is not motivated by an independent principle requiring aid to others for its own sake, at least if that principle is interpreted broadly. If familiar collective action problems make it difficult to supply an adequate subsidy without compulsory transfer, the initial interaction will not be to the “mutual benefit” of those compelled to provide it, but the system of interactions it enables may be. I don’t say this is the case with respect to any particular system of transfers—that’s an empirical question—but insofar as this scenario is not wildly implausible, it is at any rate subject to reasonable disagreement. What falls within the “level” specified by the “core principle” of interaction to mutual benefit is, in other words, not quite so sharply defined as the system of side-constraint rights spelled out in A,S,&U—and my loose interpretation of it may well be in direct conflict with what he says in response to Rawls in Chapter 7 of that book—even though there are enough obvious affinities between these ideas that Nozick himself is prepared to link his “core principle” with the position of the earlier book.

Turning back to “The Zig-Zag of Politics,” I note that one of its primary concerns has to do with the attitude one ought to have in the context of a liberal democratic society in which many different moral views coexist. Where A,S,&U is concerned with ideal theory and the sort of government that could justly arise from a state of nature, “Zig-Zag” is more concerned with how to deal with the range of divergent positions members of a pluralistic society holds. Here, he effectively takes a fallibilist stance that leads him to embrace the titular Zig-Zag of democratic politics over intransigent insistence on one’s own view as uniquely avoiding injustice, at least within a certain range. In some ways, it prefigures the Rawls of Political Liberalism, who focuses on how to deal with the conflict of diverse reasonable moral conceptions.

We hear echoes of this in Nozick’s language in Invariances. In contrast with the stark boundaries of rights-respecting and rights-violating states found in A,S,&U, we find a discussion of multiple levels of ethics, all of which are equally part of morality, even as Nozick argues that only one level is legitimately enforceable. There is, in other words, a stronger sense of tension between competing values: A state which coercively transfers resources to the poor may be enforcing a real ethical obligation even if that enforcement violates the claim to non-interference embedded in the “ethics of respect,” and even if we think those claims should take priority. That seems to mark an important distinction from simple theft, even if we think the transfer is on net morally bad.

Note also the (perhaps canny) ambiguity in the claim that “some particular society may attempt to make one or another of these further levels mandatory,” though he believes no society “should” do so. Does that “may” express a mere possibility, or a permission, as in “it is not the morally optimal policy, but it is in the set of morally permissible options”? (Would that make libertarianism supererogatory?) One way to construe this, again borrowing Rawlsian jargon, is as suggesting that (some modified version of) the view from A,S,&U is one among many “reasonable conceptions of justice” that might order a liberal society, and in Nozick’s view the most defensible one of that set. This is something analogous to the way Rawls recasts the theory of Justice as Fairness articulated in A Theory of Justice as one of several possible “modules” that may be plugged into the superstructure of Political Liberalism.

Since something close to this is roughly my own view, I’m wary of too-confidently projecting it on a thinker who is, sadly, no longer around to correct the liberties I take with his ideas. Still, it is one (I think appealing) way to square Nozick’s claim that he had become “less hardcore” with the textual evidence that his substantive commitments to a moral (and not merely pragmatic) claim against coercive state redistribution remained intact. (There are also, as David Boaz notes, reports that Nozick had modified his views on the inalienability of certain rights.) On this interpretation, Nozick’s substantive view of libertarian rights remained largely the same, but less “hardcore” insofar as that view is situated within a family of reasonable views that give different weights to competing ethical values, rather than simply in contrast to varying degrees of theft, slavery, and predation.