Matthew D. Adler · September 2009
77 GEO. WASH. L. REV. 1478 (2009)
How should we take account of the interests of future generations? This question has great practical relevance. For example, it is front and center in arguments about global warming policy. Unfortunately, the question is doubly difficult—doubly, because it not merely implicates generic disputes about the structure of moral obligations (disputes between consequentialists and nonconsequentialists, welfarists and nonwelfarists, and so forth), but because the appropriate treatment of future generations implicates distinct problems that are not resolved merely by adopting one or another generic framework.
This Article addresses the appropriate treatment of future generations within the framework of prioritarianism. In prior articles and a forthcoming book, I argue that the “social welfare function” (“SWF”) approach to policy choice provides a systematic, implementable, and theoretically well-grounded method for rendering policy sensitive not only to efficiency, but also to equity—to the fair distribution of well-being. This framework has firm intellectual roots, drawing both from moral philosophy (where the distinction between prioritarian and non-prioritarian approaches to equity has recently been much discussed), and from welfare economics (where the idea of an SWF originates).
My claim, in this Article, is that the appropriate treatment of future generations, within the framework of prioritarianism, is straightforward in the “core case” where the intertemporal population is fixed and finite. Part I of this Article summarizes the prioritarian approach. Part II discusses the core case. Here, I shall argue that the interests of future generations should be given the very same weight as the interests of the present generation. Arguments to the effect that the SWF should incorporate a utility discount factor, reflecting a pure rate of time preference, are misconceived—at least in the core case.
Part III discusses departures from the core case: non-identity problems, variation in the size of populations, and infinite populations. As we shall see, arguments for departing from neutrality between present and future generations become stronger in such cases. In particular, there is a plausible argument that an individual who does not exist in some outcomes should have her interests wholly ignored for purposes of comparing such outcomes to outcomes in which she does exist. As for infinity cases, we shall see that a strong kind of temporal neutrality is logically inconsistent with the principle of Pareto superiority.
These conclusions, however, are much less firm and definitive than the argument for neutrality in the core case. The main aim of Part III is simply to acquaint the reader with the various kinds of problems for the SWF framework that arise once we relax the assumptions of a fixed, finite population. Given the intrinsic difficulty of these problems, and space limitations, I can hardly do more than that.
