Can There Be a Neutral Theory of Justice?

Abstract

The publisher's statement about Bruce Ackerman's Social Justice in the Liberal State begins by comparing the book to John Rawls's A Theory of Justice. I I believe that Ackerman's book clearly merits comparison to Rawls's, both in originality and importance. However, I also believe that the two theories require comparison because of their vulnerability to the same fundamental difficulty. There are some obvious similarities between the two works. Both argue for a unique solution to the problem of distributive justice under ideal conditions. Both argue from a kind of impartial decision procedure. For Rawls, this procedure is, of course, the famous device of the original position. For Ackerman, it is the test of under the ideal conditions of a perfect technology of justice -a kind of science fiction situation where there never is any practical difficulty in implementing the substantive conclusions of a neutral dialogue (p. 21).2 Despite these obvious parallels, I would like to explore a connection between the two works of a more general kind. I believe that this connection indicates a common and fundamental difficulty in both works. Both theories attempt to realize, simultaneously, three aspirations: (a) they offer what I will call structural principles ofjustice; (b) they aspire to uniqueness; and (c) they attempt to maintain neutrality, by which they mean a lack of bias toward any particular theories of good. There are some general reasons for believing that these three aspirations are incompatible. My argument will be that one interpretation of c, neutrality, rules out a while the remaining possible interpretations of c can be expected to rule out b. Hence, the attempt to live up to all three aspirations is doomed to failure. I believe that much of the debate of the last decade about Rawls's theory and much of the current debate about Ackerman's theory should be seen in this light. While I will focus on Ackerman's theory in this review essay, I will also attempt to suggest, if only briefly, how the parallel difficulty arises for Rawls. Let me quickly characterize each of these three assumptions and then argue that the third, neutrality, will undermine either the first or second, depending on how it is interpreted. By a structural principle, I mean an ethical criterion that compares states of affairs in terms of payoffs to positions.3 By payoffs, I mean