Nozick describes Rawls's view as an "undeniably great advance over utilitarianism." From his standpoint that is a reasonable estimate. Rawls's theory is a half-way house between utilitarianism and Nozick's own position. But if having gone half-way with Rawls we are forced by the logic of our position to go all the way with Nozick, it could be that we went wrong when we started out. None of the arguments Nozick uses against Rawls is decisive when invoked against a utilitarian position. Utilitarianism gives a clear and plausible defense not merely of progressive taxation, welfare payments, and other methods of redistribution, but also of the general right of the state to perform useful functions beyond the protection of its citizens from force and fraud. Utilitarianism also provides an argument in defense of the claim behind Williams's argument for equality—that society should, so far as its resources allow, provide for the most important needs of its members.
Nor do we have to go all the way with the utilitarians to be in a position to advocate state-directed redistribution of income. The problem of whether we can accept a utilitarian account of noneconomic rights like the right to freedom of speech or freedom of worship need not be raised here, for Nozick's argument is mainly addressed to economic rights. We can deal with property in a utilitarian manner, rejecting the doctrine of an intrinsic right to property, without necessarily rejecting the idea that there are some intrinsic rights against the state. For the remainder of this discussion, though I shall talk simply of "utilitarianism," it will be this limited economic utilitarianism to which I am referring.
Nozick, aware that utilitarianism is a more fundamental rival to his position than other conceptions of justice, tries to get it out of the way in the first part of the book, when discussing the moral background of his theory. The discussion is sketchy, however, and falls below the level of the later sections. Nowhere is utilitarianism fully and systematically confronted. Nozick mentions some well-known objections but, with one exception, does not pursue the replies that utilitarians have made when these objections have been raised in the past.
The exception is interesting. In opposition to the view, which utilitarians have held, that the only things that are good or bad in themselves are states of consciousness, or conscious experiences (pleasant or happy ones being good, painful or miserable ones bad), Nozick asks us to imagine that we can build an "experience machine" which would give us the satisfactions of a wonderful life—any life we'd like—while we float in a tank with electrodes plugged into our brains.
If we had such a machine, Nozick says, we would choose not to use it—and this shows that things other than experience matter to us. In anticipation of the reply that we would not use the machine because, as good utilitarians, we would be concerned about other people's (and other animals') experiences as well as our own, Nozick makes the further assumption that everyone is able to plug into one of these machines. This means that we cannot give point to our lives by improving the experiences of other beings; the experience machine gives everyone who wants them the best possible experiences anyway. Nevertheless, Nozick says, we would not plug in, and this is because in addition to wanting to have certain experiences we want to do certain things and be a certain sort of person. We desire to live in contact with reality, and this no machine can do for us.
The entire essay is well worth reading.