¹ For an account of this covert form of discrimination see Warren, Mary Anne, “Secondary Sexism and Quota Hiring,” Philosophy & Public Affairs, vol. 6, no. 3 (1977).Google Scholar

² Although goals are often distinguished from quotas, especially in legal arguments, I shall draw no such distinction here. The important differences between alternative policies lie in their administration, not in the terminology we use to characterize their objectives.

³ See, for example, Thomson, Judith Jarvis, “Preferential Hiring,” Philosophy & Public Affairs, vol. 2, no. 4 (1973).Google Scholar

⁴ The standard objections can be found in Simon, Robert, “Preferential Hiring: A Reply to Judith Jarvis Thomson,” Philosophy & Public Affairs, vol. 3, no. 3 (1974)Google Scholar; Goldman, Alan H., Justice and Reverse Discrimination (Princeton: Princeton University Press, 1979)Google Scholar, ch. 3; and Richards, Janet Radcliffe, The Sceptical Feminist (Harmondsworth: Penguin Books, 1982)Google Scholar, ch. 4.

⁵ For the best treatment of this issue see Friedman, Marilyn A. and May, Larry, “Harming Women as a Group,” Social Theory and Practice, vol. 11, no. 2 (1985).CrossRefGoogle Scholar

⁶ That traditional discrimination is an evil is clearer than why it is. If we say, for instance, that job candidates should always be assessed purely on the basis of merit then we will be barred from taking account of such additional factors as age, nationality, place of residence, seniority, and so on. On the other hand, if we say that the problem lies in assigning some independent weight specifically to gender then we will be forced to conclude that positive sexism is also an evil. A more convincing story will probably condemn traditional discrimination in employment for its assumption of the inferiority of women, or for the way in which it conspires with other oppressive social practices to deny them equal opportunity with men. For a discussion of these issues in the context of race see Dworkin, Ronald, A Matter of Principle (Cambridge: Harvard University Press, 1985)Google Scholar, ch. 14. In Section 3 below, I consider the closely allied question of which rights (if any) are violated by discriminatory practices.

⁷ It is not essential to this argument that one social group is here disadvantaged in order to benefit another. The same group (broadly defined) may be both victim and beneficiary of positive discrimination. I recall reading some years ago of an inner city housing development in New York which consisted of a black majority and a white minority. Most applicants for flats in the development were black, but experience showed that if the percentage of black residents rose above some critical point then the phenomenon known as ‘tipping’ occurred and the whites began to leave, leading in the end to the creation of yet another black slum. In order to save the black residents from this fate, a quota had to be established which gave preference to white applicants. Thus in this case (some) blacks were discriminated against in order to prevent even greater discrimination against (other) blacks.

⁸ For a catalogue of such arguments, see Warren, “Secondary Sexism.”

⁹ In Section 3 below, I consider the claim that the male victims of positive sexism suffer not merely a loss but an injustice, which may not be balanced in this way against the injustices to women which such a policy will avert.

¹⁰ Those who wish to pursue them should consult the recent literature on agent-centered restrictions, especially Scheffler, Samuel, The Rejection of Consequentialism (Oxford: Clarendon Press, 1982)Google Scholar, ch. 4; and Nagel, Thomas, The View from Nowhere (New York and Oxford: Oxford University Press, 1986)Google Scholar, ch. 8.

¹¹ Richards, Sceptical Feminist, pp. 142–143.

¹² See my The Moral Foundation of Rights (Oxford: Clarendon Press, 1987).

¹³ I owe this last suggestion to G. A. Cohen.