¹ See, for instance, Parfit, Derek, Equality or Priority?, The Lindley Lecture 1991, Kansas, 1995Google Scholar; Raz, Joseph, The Morality of Freedom, Oxford, 1986, p. 240Google Scholar; and Scheffler, Samuel, The Rejection of Consequentialism, Oxford, 1982, p. 31Google Scholar.

² By ‘utilitarians’, I mean so-called total (act-)utilitarians.

³ Notice that utilitarians cannot rebut this objection by invoking the principle of diminishing marginal utility. According to this principle, an extra monetary unit produces more well-being if it befalls someone who is poor than if it befalls someone who is wealthy. Therefore, it is sometimes argued, the Principle of Utility implies that it is better to give to the poor than to the rich. This argument, however, illegitimately assumes that different people have the same utility functions.

Instead, we may consider a more sophisticated utilitarian argument for monetary equality. If we do not know people's utility functions, we will maximize expected utility, everything else being equal, by distributing money equally (see Brandt, Richard B., A Theory of the Good and the Right, Oxford, 1979, pp. 311–16)Google Scholar. But this is an (indirect) argument for monetary equality, not for an equal distribution of well-being.

⁴ Apparently, Larry Temkin holds this view. That is, he accepts a version of consequentialist (or teleological) egalitarianism according to which well-being is the relevant unit of egalitarian concern. See Temkin, Larry S., Inequality, New York, 1993, p. 10Google Scholar.

⁵ Parfit, , Equality, p. 17Google Scholar.

⁶ Strictly speaking, the move from E to C (or D) could be better for the worse off (in E) in some respects, for instance if in C they are no longer envious of the group that is better off (in E), but not better for them all things considered, since they also suffer a loss in C. The move from E to C would then be better for some people in one respect. To avoid this complication, however, I shall simply assume that in the examples I consider, whenever no one is better off all things considered, no one is better off in any respect.

⁷ Temkin provides a number of ingenious arguments against the Levelling Down Objection; see Temkin, Inequality, ch. 9. I try to rebute his arguments in ‘Egalitarianism and the Levelling Down Objection’, Analysis, lviii (1998)Google Scholar, and in ‘In Defence of the Slogan’, Preference and Value: Preferentialism in Ethics, ed. Rabinowicz, Wlodek Lund, , 1996Google Scholar.

⁸ Parfit formulates this principle slightly differently: ‘Benefiting people matters more the worse off these people are’ (Parfit, , Equality, p. 19)Google Scholar. However, unlike Parfit's, my formulation rather straightforwardly lets natural distributions fall under the scope of the principle. It implies that it is better if, in the natural lottery, a certain sum of benefits is distributed such that everyone is quite well off than if it is distributed such that some are worse off and others better off.

⁹ See, for instance, Heyd, David, Genethics: Moral Issues in the Creation of People, Berkeley, 1992, p. 97Google Scholar. It should be noted, however, that Heyd uses the term ‘potential people’ rather than ‘possible people’.

¹⁰ This view is also defended in Holtug, Nils and Sandøe, Peter, ‘Who Benefits? Why Personal Identity Does Not Matter in a Moral Evaluation of Germ-Line Gene Therapy’, Journal of Applied Philosophy, xiii (1996), 161–4Google Scholar, and more elaborately in Nils Holtug, ‘In Defence of the Slogan’, sect. 6.

¹¹ See, for instance, Heyd, pp. 80–90.

¹² The argument can be made from other theories as well.

¹³ Broome, John, ‘Goodness is Reducible to Betterness: The Evil of Death is the Value of Life’, The Good and the Economical, ed. Koslowski, Peter and Shionoya, Yuichi, Berlin, 1993, p. 77Google Scholar, Parfit, Derek, Reasons and Persons, Oxford, 1984, p. 489Google Scholar.

¹⁴ Of course, this raises the issue of the ontological status of negative facts, but I cannot go into this issue here.

¹⁵ Steinbock, Bonnie, Life Before Birth. The Moral and Legal Status of Embryos and Fetuses, New York, 1992, p. 71Google Scholar.

¹⁶ For a discussion of a similar case, see Derek Parfit, Reasons and Persons, ch. 18.

¹⁷ Interestingly, but implausibly, some philosophers seem prepared to accept not only that one does the child no harm by having it, but also that, everything else being equal, one does not even do anything wrong! For instance, this follows from the ‘generocentric’ view defended in Heyd.

¹⁸ If, along these lines, we claim that the benefits to people that are created when these people are caused to exist fall within the scope of the Priority View (in the sense that this view assigns moral weight to such benefits), this makes the Priority View what Parfit calls a wide person-affecting principle (see Parfit, , Reasons, pp. 396–401)Google Scholar. Wide person-affecting principles should be contrasted with what he calls narrow person-affecting principles. According to narrow person-affecting principles, one outcome can be better than another only in so far as it involves people who exist independently of which outcome comes about and are better off if the former outcome comes about. Therefore, such principles weight only benefits to actual people. In contrast, wide person-affecting principles also weight benefits to possible people. If one outcome brings into existence (happy) people who would not exist if another outcome came about, this may contribute to making the former outcome better since it gives extra benefits to people – the benefits they derive from coming into existence.

¹⁹ For a much more detailed discussion of the proper temporal unit of prioritarian concern, see Kappel, Klemens, ‘Equality, Priority and Time’, Utilitas, ix (1997)Google Scholar.

²⁰ I would like to thank an anonymous referee for making this suggestion.

²¹ According to a person-affecting version of the Principle of Utility, the larger the total sum of benefits to people, the better. It may be worthwhile to make the following observation before I consider the implications of Priority Pluralism. Some prioritarians may want to deny the Value of Existence View. They may want also to assign weight to the well-being that is created when a person comes into existence. This they can do by combining the Priority View and the Principle of Utility, since the latter principle is an impersonal principle (that is, it assigns weight to impersonal increases in well-being). Such prioritarians will hold a view that coincides extensionally with the pluralist view considered in this section.

²² In terms of the graphics of figure 3, this means that in G, p is at a level of well-being where the graph of the strictly concave function is at least as steep as the graph of the linear function.

²³ Hence, the Combined View contradicts what Temkin calls the Intrinsic Aspect View, according to which the value of an outcome depends only on its intrinsic features(see Temkin, Larry S., ‘Intransitivity and the Mere Addition Paradox’, Philosophy and Public Affairs, xvi (1987), 159)Google Scholar.

²⁴ It may be asked why this is a problem. The reason is the following. At least usually, it is held that it is a logical truth that the betterness relation is transitive (see, for instance, Broome, John, Weighing Goods, Oxford, 1991, p. 11)Google Scholar. Therefore, any moral theory that does not satisfy this requirement should be rejected. Nevertheless, some philosophers challenge transitivity. Thus, Temkin argues that many, perhaps all, plausible moral theories generate intransitivities in the betterness relation when applied to the issue of possible people (Temkin, ‘Intransitivity’; see also his ‘A Continuum Argument for Intransitivity’, Philosophy and Public Affairs, XXV (1996))Google Scholar. However, there are clear examples of influential principles that do not generate intransitivities. For instance, the Principle of Utility does not. Nor, for that matter, does Priority Monism. But, of course, Temkin may believe that these views are implausible. Perhaps Temkin's best example of a plausible principle that leads to intransitivity is what he calls the Person-Affecting Principle (PAP), according to which, roughly, an outcome is better (worse) than another if there is someone for whom it is better (worse) and no one for whom it is worse (better), but not vice versa (‘Intransitivity’, 166). Consider the following outcomes in the light of this principle:

PAP implies that J is better than K, since it is better for some (the p-people), and worse for no one. It also implies that K is better than L, since it is better for some (the r-people), and worse for no one. But J is not better than L; rather, L is better than J. Hence, PAP violates the transitivity of the betterness relation (ibid., 169). However, I believe that a person-affecting principle should not be construed along the lines of PAP. This is because PAP assigns no weight to the benefit of coming into existence. For instance, when comparing J and K, PAP assigns no weight to the fact that the r-people will benefit much more in K than the q-people will in J. In my ‘In Defence of the Slogan’, I defend a (wide) person-affecting principle that takes such benefits into account.

²⁵ It may be worth considering a view that is similar to, but nevertheless different from, Priority Pluralism. Consider a restriction of the scope of the Priority View such that this view applies only to so-called same-people choices. Thus, the Priority View should not be invoked at all in different-people choices (this restriction differs from the Combined View, since the Combined View did apply the Priority View (to actual people) in different-people choices – for instance, to p in the comparison of G and I) However, a restriction of the scope of the Priority View may easily lead to new intransitivities. After all, if the Priority View is relevant only in comparisons that consist entirely of actual people, and other principles are relevant when comparing outcomes that also consist of possible people, the same outcome will vary in value when compared to various different outcomes. For a similar point, regarding other principles, see Temkin, , ‘Intransitivity’, 170Google Scholar.

²⁶ Parfit, , Reasons, p. 388Google Scholar.

²⁷ Incidentally, so does the Combined View.

²⁸ For a utilitarian assessment of how much the population should be increased in the world as it is today, see Hare, Richard, ‘Possible People’, Essays on Bioethics, Oxford, 1993Google Scholar.

²⁹ Narveson, Jan, ‘Utilitarianism and New Generations’, Mind, lxxvi (1967), 69–71CrossRefGoogle Scholar. The term ‘the Asymmetry’ was first introduced, I believe, in McMahan, Jefferson, ‘Problems of Population Policy’, Ethics, xcii (1981), 100Google Scholar.

³⁰ Narveson, it should be noted, does not believe that it can benefit or harm a person to come into existence. But the Asymmetry may be combined with either the Value of Existence View or its negation.

³¹ However, the Repugnant Conclusion cannot be avoided by invoking the Asymmetry. The Repugnant Conclusion concerns the value of outcomes, whereas the Asymmetry addresses our reasons for acting.

³² Sprigge, Timothy L. S., ‘Professor Narveson's Utilitarianism’, Inquiry, xi (1968), 338Google Scholar.

³³ A further problem with the Asymmetry is this. Suppose that, sometime in the future, the last few inhabitants on earth can either cause the world to be fully populated again or end the existence of humans by refraining from having children. Let us assume that they will be equally happy themselves, whichever choice they make. They should then refrain from having children because, amongst the billions of people they could cause to exist, there would surely be a few (say one in 100,000) who would be miserable, and while their miserableness would count against their being created, the happiness of the rest would count for nothing (see Sikora, R. I., ‘Is It Wrong to Prevent the Existence of Future Generations?’, Obligations to Future Generations, ed. Sikora, R. I. and Barry, Brian, Philadelphia, 1978, pp. 136–40)Google Scholar.

³⁴ I would like to thank Roger Crisp, Karsten Klint Jensen, Klemens Kappel, Kasper Lippert-Rasmussen, Michael Norup, Ingmar Persson, Thomas Petersen, Jesper Ryberg, Peter Sandøe and an anonymous referee for valuable comments on a previous version of this article. Thanks are also due to the Danish Research Councils for financial support.