Prospects for prioritarianism

There are three versions of the priority view on the market whose underlying philosophical motivation is quite different. This articles compares them, and concludes that only one of them is defensible. Acknowledgements This article is one of a series for which John Broome has been a constant source of encouragement and criticism. Much of this stimulus for this article came from a long discussion of a related article with Wlodek Rabinowicz, and sections 11 and 12 are the product of correspondence with him about an earlier version of this article, though he may disagree with my conclusions. To them both I owe more than the usual debt of gratitude. Support for this article was provided by AHRB Research Leave and by leave granted by the Philosophy Department at Bristol. To the AHRB and my colleagues, I am duly grateful. 1 Prospects for prioritarianism The priority view acquired its name from the influential work of Parfit (1991), and although the form of the position had been defended in Weirich (1983) and was known to economists much earlier--it is an explicit target of the groundbreaking work of Harsanyi (1955), for example--it was Parfit’s work that let to it becoming a popular and widely discussed position in moral philosophy. But what the priority view actually says, and whether it is so much as meaningful, let alone correct, has remained obscure. The most powerful challenge to it is in Broome (1991). Broome’s interpersonal addition theorem is a reformulation of a theorem of Harsanyi (1955), and he uses it as the basis for a complex and powerful challenge to the priority view. He concludes that the priority view is worse than false; it is more or less meaningless. With only minor qualifications, Rabinowicz (2002) and McCarthy (2003a) independently conclude that the inferences Broome draws from the interpersonal addition theorem are correct, and that prospects for prioritarianism are better if it is formulated outside the framework of the interpersonal addition theorem. They suggest that prioritarians are therefore well advised to try to deny one of the premises of the interpersonal addition theorem. But they recommend denying different premises, leading to different prioritarianisms. There are therefore at least three forms of prioritarianism on the market, and one goal of this article is to exhibit them all together. But the main goal is to try to assess their viability. I say that a theory is viable to say that it is one which a reasonably sensible person could occupy given the current state of theoretical understanding. All parties regard the position Broome’s prioritarian (BP) occupies as not viable, and while Rabinowicz and McCarthy regard the positions their prioritarians (RP and MP) occupy as viable, I doubt they see themselves as arguing that their prioritarianisms are correct. Rabinowicz (personal communication) informs me that he was more concerned to see how the informal position of Parfit (1991) should best be formulated in response to Broome’s challenge, and McCarthy (2003a, c) claims that whether MP is correct depends upon questions about the personal point of view we are a long way from understanding. Thus the focus here will be on the meaning and viability of the three prioritarianisms, not on whether they are correct. 1 The interpersonal addition theorem 2 It is pointless to try state the arguments without the language of decision theory, and I begin with a quick rehearsal. The bearers of uncertainty are states of nature while outcomes contain everything that matters to people. A prospect is a set of state of nature-outcome pairs, where exactly one state of nature along with its associated outcome must obtain. Throughout