The welfarist's dilemma, once again, runs like this: for a welfarist approach to succeed, the basic minimum must be tied closely to the achievement of one's conception of the good. Otherwise, Nussbaum's argument for a capabilities approach sticks: a welfarist approach to the basic minimum would insist on a moral reason to force individuals to live lives they do not value. WBM successfully avoids this concern: the basic minimum is not construed as the achievement of particular states or functionings in abstraction from conceptions of the good. The basic minimum is the achievement of a valued project.

So far so good. But the second horn of the dilemma looms. To be a valued project, as noted in the last chapter, crucially depends on the attitudes one takes toward this project and its context. But such attitudes are malleable, often in insidious ways. The fact that one values p can be shaped by poor circumstances and a lack of opportunity. If so, one should think that the mere fact that A values a particular project is not enough for A to maintain the basic minimum. A's valuing might be the product of adaptive preferences.

The phenomenon of adaptive preferences clearly poses some problem for WBM. But the problem it poses is ill-understood and under-theorized. In this chapter, I distinguish two categories of adaptive preferences: shallow adaptive preferences, and deep adaptive preferences. Shallow adaptive preferences and deep adaptive preferences, I shall argue, pose very different problems for a welfarist approach to the basic minimum.