I liked this summary of the debate on desert from Lindsay Beyerstein.
Wilkinson claims to have found a conflict between common sense morality and Rawlsian theory. If so, this undercuts Rawls' claim to have codified common sense justice. Wilkinson argues that instrumentalism doesn't really explain our intuition that a hard worker deserves her reward, though it may explain our intuition that it would be expeditious to give it to her.
The instrumentalist position needs to be supplemented with a non-metaphysical theory of desert. It turns out that a contractual/procedural theory of desert explains our intuitions just as well. We don't have to argue desert in terms of free will and moral responsibility. Sometimes promises beget desert. Our society wisely promises people that they will be rewarded if they work hard and contribute a lot. So, justice demands that we make good on that promise by rewarding the high achievers. Instrumentalism explains why it is a good idea to make that promise.
I think this is pretty good summary of my argument. And I'm glad to see DeLong copied it on his blog. (Thanks Lindsay!) However, I don't think our intuitions about desert are necessarily rooted in the social practice of promising, although people obviously do deserve things in virtue of promises and contracts. I think that I can deserve thanks from my friend in virtue of having done him a favor, or deserve love in virtue of the love I have given. Anyway, I've claimed that anti-Rawlsian intuitions about desert run deep part in our moral psychology. (I see that Lindsay is involved in experimental moral psychology, so maybe she can test this!) The argument that it's utility promoting, or instrumental to some other end, to treat people as if they actually deserve things raises the question of why this practice is utility promoting or instrumentally useful. My argument is that treating people as if they deserve things promotes utility because the practice aligns itself with their moral self-conception — their reflective judgment that they do deserve things. A practice or set of social principles that failed to respect this self-conception will be confronted with resistance and non-compliance, and will tend to be self-undermining. Now, the way I see it, if a practice based on people's moral self-conception turns out actually to make people better off on the whole, then that just shows that our moral self-conception in this regard is justified, and establishes the moral facts of the matter. If we think we deserve things, and our acting on that conviction tends to make us all better off, then we really do deserve things. That is, then desert claims have real normative teeth. If vulgar consequentialists, like DeLong, buys the pragmatic argument for respecting desert claims, then he shouldn't be skeptical about the existence or authority of desert.