Since writing about the Wilt Chamberlain example last week, I’ve been revisiting Anarchy, State, and Utopia and thinking about what legitimate criticisms can be leveled against this particular step in Nozick’s argument.  I still think Stephen Metcalf’s complaints are basically frivolous, and his recent response to his critics does little to change my view.  On the question of whether Nozick meaningfully “repudiated” libertarianism, he stresses that some of what Nozick says in the 1989 essay “The Zig Zag of Politics” seems more at odds with libertarianism than the positions he advanced both before and after—notwithstanding his later claim to have considered himself a libertarian “all along.”   I think my last post sketches a plausible fallibilist reading of “Zig Zag” on which the gap between that essay and his other writing is not as great as it might seem at first blush, but I’m not ultimately sure why this is supposed to matter all that much—or why you’d try to make so much of a change of heart that, if it was significant, was also apparently temporary.

As for the Chamberlain example itself, my objection was not that it is not sufficiently “representative” of Nozick’s views, but that it is not an argument to the conclusion Metcalf thinks it is, which means his attempted rebuttal is tantamount to establishing that the Pythagorean Theorem fails to prove Reimann’s Hypothesis. Metcalf halfway seems to acknowledge this, but insists the example is “rigged” in a way designed to “muddle” our intuitions:

Why, if Nozick did not want to game his example, did he choose Wilt? After all, if Sanchez is correct, isn’t the point made just as well with, say, a happy-go-lucky doofus who rides a wave of Internet exuberance and cashes out big, all while adding to the world precisely zero utility? Absent an injustice in each step (the prospectus is accurate, the bankers price the IPO fairly) the resulting gross inequality itself cannot be regarded as unjust. But I didn’t choose Wilt Chamberlain; Nozick chose Wilt Chamberlain. I.E., he wanted to harvest all of the sentimental associations from a historical reality while leaving behind all its real-world complications.

Well, I don’t know why Wilt Chamberlain. Why trolleys? Maybe Nozick was a basketball fan. Maybe it just seemed simpler than something involving stock options, IPOs, and fine points of securities law. Most likely, though, it’s because at the time Nozick was writing, Wilt Chamberlain would have been the most prominent pop cultural example of a celebrity many people thought outrageously overpaid. Even today, critics of inequality will routinely use professional athletes as an example of the absurd disparities market systems permit, as in “What kind of sane economic system lets a grown man make a million dollars playing a game most people consider recreation, while teachers are underpaid?” In the early 1970s, Chamberlain would have been the poster boy for this kind of argument. He was one of the most famous athletes in the country, and had just made news by signing to the Lakers for a controversial and, at the time, unprecedented $250,000 salary, making him the highest paid pro in any sport. That happens to be the exact sum Nozick imagines him earning in his thought experiment, so this is pretty clearly an example “ripped from the headlines” about that controversy. I imagine the choice was motivated in part by the idea of flipping that common complaint around: “You see, I’ll defend even this canonical example of ridiculous inequality.” It’s pretty doubtful, in other words, that many people around Harvard at the time would’ve thought that using Chamberlain to illustrate the point was stacking the deck somehow.

In any event, since the upshot of the thought experiment is to make a point about certain features of patterned (as opposed to historical) theories of distributive justice, rather than to validate in one fell swoop the pattern of holdings in actual capitalist societies, this just doesn’t seem particularly important to me.  The example is, by design “rigged” to be as unobjectionable as possible in terms of the validity of the specific set of transactions described, and given the limited purpose of the thought experiment, this is no dig against it. If you think the result is fair in a case like this, then you think that historical factors can be relevant to assessing the justice of a distribution, and that a patterned view fails to capture the whole truth.

That said—and perhaps in the spirit of the “Ideological Turing Test“—I want to point out some of the ways I think the Wilt Chamberlain Argument does fall short—not, again, as a supposed proof of the morality of real market institutions, but as a theoretical argument against patterned principles of justice. Here’s what a fairer critic might be able to say.

First, Nozick’s claim that patterned principles will require “continuous” intervention with people’s choices to be “continuously” realized does not have much force independently of the argument for thinking that a distribution will be just if it arises from free transactions consistent with a valid Principle of Justice in Transfer (PJX) from a just (according to your favorite patterned theory) starting point. Rawls’ Justice as Fairness is a patterned theory, but Rawls is quite clear that it—and specifically the Difference Principle—applies to the rules and institutions of the basic structure of society, including (inter alia) whatever PJX is enforced. It is not supposed to be an independent criterion for evaluating the justice of a time-slice distribution, except insofar as this may reveal whether the PJX and other institutions have been suitably well crafted in accordance with that principle. 

Moreover, insofar as Nozick supports both rectification of unjust (e.g. fraudulent) transfers, police enforcement of property rights, and taxation (e.g. sales taxes) to support the minimal state, he must also be prepared to countenance “continuous” interference with various kinds of chosen activities. There may be pragmatic issues with the scale or form of intervention proposed, but they’re distinct from the question of whether the intervention is just.

Second, the force of Nozick’s argument relies in large part on being able to grant the patterned theorist a favored starting point, D1, and show how it could be transformed into a distribution highly inconsistent with that pattern by a series of morally unobjectionable steps. For some versions of a patterned theory, this will not quite work. If, for instance, the patterned theorist argues that people are entitled to a share of the stream of social output over time, then no static allocation at a time slice (such as D1) would actually count as satisfying the requirement. More generally, the patterned theorist can object to Nozick’s move from the justice of a distribution to the claim that people therefore have a right to the specific bundle of assets they’re afforded under that distribution. For someone who thinks the just pattern involves rights to shares of social wealth rather than specific concrete holdings, each new child born will be entitled to a share—requiring a transfer from others even though we presumably don’t want to say reproduction involves intrinsic injustice. The advocate of a patterned view can say that transfers to ensure each child a fair starting endowment does not really interfere with anyone’s entitlement to their (previously) just holdings, because what people were entitled to all along was not the specific stuff constituting their previous holdings, but only an equal (or fill in your favorite allocation algorithm) percentage of the total, which they continue to have. Nozick still has the objection that it seems absurd to claim it’s unjust for people to freely do anything with “their” shares that disrupts this delicate equilibrium. But if we’re clear that share rights remain share rights, and don’t translate into rights over the specific assets constituting one’s share at a time slice, the example loses a lot of intuitive force. If I’m entitled to an “equal share” of the floor space at the yoga studio, there’s no mystery why my share might vary over time as more people arrive or the building is enlarged, on top of any gains I might realize from svelte friends offering me portions of their shares until I slim up a bit.

Finally, and perhaps most importantly, the intuitive appeal of the argument rests on the tacit premise that if D1 is just, and the microtransfers that lead from D1 to D2 are individually morally permissible, then the macrodistribution D2 effectively inherits the justice of the microsteps. As Nozick himself observes, of course, one could “shoehorn” a patterned view into the PJX, such that individual steps contributing to the disruption of the macropattern are ruled out (or corrective transfers mandatory), even though the microtransfers would be unobjectionable absent consideration of their contribution to altering the macropattern. No correct PJX, Nozick thinks, will have this form. This seems to beg the question against the patterned theorist in a way that is hard to justify. This is especially so given that, on a plausible libertarian theory, it will be the case that independently permissible actions may be rendered wrongful by virtue of their contribution to an aggregate effect.

Consider, in the spirit of Derek Parfit’s (itself highly unrealistic!) “harmless torturers” thought experiment, certain kinds of pollution. Suppose it is the case that many producers emit, as a byproduct of industrial activity, a variety of chemicals into the atmosphere. Further suppose that for each factory, its individual emissions make no perceptible difference in the surrounding air quality, but the combined effect of all these gasses emitted by thousands of producers mixing produces some catastrophic ecological effect. 

Each producer can claim that his emissions inflict no harm when considered in isolation, and perhaps they even inflict no marginal harm, holding constant the actions of others. (By analogy with voting: Each voter, at least in a national election, can reasonably point out that her vote makes no difference to the outcome of the election. Yet the outcome is, nevertheless, the joint product of all those individual votes.) In conventional libertarian terms, the ecological catastrophe (should it occur) nevertheless clearly constitutes a serious harm to the health and property of others, in a way that may plausibly be said to violate their rights, whether or not that violation can be ascribed to any individual actor. (To avoid complications involving potential “homesteading” of emissions rights, suppose the producers all act simultaneously and without coordination.) 

Insofar as the ecological harm is foreseeable on the basis of solid science, producers who continue to emit willy-nilly, without coordinating to set limits on their individual emissions, can plausibly be said to act wrongly. If this is so, the injustice of the individual producer’s excessive emission will not be a function of its intrinsic effects considered in isolation, but rather of its role in realizing an aggregate “pattern.” The libertarian can, of course, object that unequal macrolevel distributions do not constitute an analogous “harm,” and that the macropattern at D2 does inherit the fairness of the transfers that give rise to it. But he cannot appeal to a general principle to the effect that rules of justice governing individual action may never factor in the role they play in generating an aggregate outcome.

In the unlikely event that’s not more than enough Nozick for you, Matt Yglesias and I had some things to say about Nozick and political philosophy more generally in a recent Blogging Heads dialogue.