In a comment below, Julian reminds me of Hayek’s approving mention of Rawls’s method for thinking about principles of justice. This led me to look it up again. It’s in Law, Legislation, and Liberty, Volume 2: The Mirage of Social Justice, p. 100:
Before leaving the subject [of ‘social’ or distributive justice, ideas that Hayek finds muddled and misleading] I want to point out once more that the recognition that in such combinations as ‘social’, ‘economic’, ‘distributive’ or ‘retributive’ justice the term ‘justice’ is wholly empty should not lead us to throw the baby out with the bath water. Not only as the basis of the legal rules of just conduct is the justice which the courts of justice administer exceedingly important; there unquestionably also exists a genuine problem of justice in connection with the deliberate design of political institutions, the problem to which Professor John Rawls has recently devoted an important book. [Vol 2 of LL&L was published in 1976.] The fact that I regret and regard as confusing is merely that in this connection he employs the term ‘social justice’. But I have no basic quarrel with an author who, before he proceeds to that problem, acknowledges that the task of selecting specific systems or distributions of desired things as just must be ‘abandoned as mistaken in principle, and it is, in any case, not capable of a definite answer. Rather the principles of justice define the crucial constraints which institutions and joint activities must satisfy if persons engaging in them are to have no complaints about them. If these constraints are satisfied, the resulting distribution, whatever it is, may be accepted as just (or at least not unjust).’ This is more or less what I have been trying to argue in this chapter.
The Rawls quote is from his essay “Constitutional Liberty and the Concept of Justice” in Nomos IV, Justice, 1963, p. 102. Hayek follows up the citation in his as always illuminating footnotes thus:
. . . where the passage quoted is preceded by the statement that ‘It is the system of institutions which has to be judged and judged from a general point of view.’ I am not aware that Professor Rawls’ later more widely read work A Theory of Justice (Harvard, 1971) contains a comparatively clear statement of the main point, which may explain why this work seems often, but as it appears to me wrongly, top have been interpret support to socialist demands, e.g., by Daniel Bell, ‘On Meritocracy and Equality’, Public Interest, Autumn 1972, p. 72, who describes Rawls’ theory as ‘the most comprehensive effort in modern philosophy to justify a socialistic ethic.’
Hayek is right. Rawls, at least the early Rawls, seems to me to be nowhere near a socialist or even a continental-style social democrat. He strikes me as a fairly traditional classical liberal working to reconcile the contemporary game-theoretic conception of rational choice with traditional contractarian theorizing and Kant.
Hayek somewhat earlier cites another passage from the same Rawls Nomos essay (which I really must read!) that casts Rawls in a similar classical liberal light. Here’s Rawls:
If one assumes that law and government effectively act to keep markets competitive, resources fully employed, property and wealth widely distributed over time, and maintains a reasonable social minimum, then, if there is equality of opportunity, the resulting distribution will be just or at least not unjust. It will have resulted from the workings of a just system . . . a social minimum is simply a form of rational insurance and prudence.
Here, as in other places, Rawls seems to characterize the social minimum, the safety net, as the price the wealthier must prudently pay for the stability of the order from which they benefit. At other times, Rawls makes more of the idea of fairness-as-reciprocity as a core aspect of the sense of justice, and argues that the motivation to pay into the treasury to fund the safety net is not merely prudential, but is rooted in a more distinctively moral motivation. I think this is one of many instances where Rawls’s good, clear rat choice contractarianism comes into tension with his penchant for Kantian moral psychology.
My total baseless conjecture is that Rawls’s students and some colleagues, who were far to Rawls’s left politically, slowly drew him further left toward a more euro-style social democrat point of view. I’m told that Rawls was a model of intellectual openness, not at all dogmatic, which would make him especially prone to reciprocal influence from his interlocutors. Early Rawls’s quasi-positivist naturalism and classical contractarianism was discouraged by his milieu while his Kantianism, and especially his Kantian moral psychology, was encouraged. So he abandons the idea that the theory of justice is a part of the theory of rational choice.
Rawls’s students seem to be far more vehement in their opposition to Humean moral psychology than in their interest in the contractarian conception of a social order. They seem more interested in coming up with an arguments to the effect that we’re obliged to pay taxes even if we don’t want to (the claims of justice are rationally INESCAPABLE, dammit!) than in exploring the general nature of a theory that could justify something like taxation in terms of its role in maintaining a viable social order. They seem to have co-opted a lot of Rawls’s language, and methodological apparatus, but left most of his core contractarian logic behind.
So here’s my unanswerable question–get out your possible world telescopes: What does the late Rawls look like in a world in which he had folks like Nozick, Lomasky, and Schmidtz as his prominent students, rather than folks like, say, Scanlon, Cohen, Estlund, and Korsgaard? Does he look more like the Rawls that Hayek sees and likes?