The Continuity Test is the principle that a proposed distribution of resources is wrong if it treats someone as disadvantaged when they don't see it that way themselves, for example by offering compensation for features that they do not themselves regard as handicaps. This principle -which is most prominently developed in Ronald Dworkin's defence of his theory of distributive justice -is an attractive one for a liberal to endorse as part of her theory of distributive justice and disadvantage. In this paper, I play out some of its implications, and show that in its basic form the Continuity Test is inconsistent. It relies on a tacit commitment to the protection of autonomy, understood to consist in an agent deciding for herself what is valuable and living her life in accordance with that decision. A contradiction arises when we consider factors which are putatively disadvantaging by dint of threatening individual autonomy construed in this way. I argue that the problem can be resolved by embracing a more explicit commitment to the protection (and perhaps promotion) of individual autonomy. This implies a constrained version of the Continuity Test, thereby salvaging most of the intuitions which lead people to endorse the Test. It also gives us the wherewithal to sketch an interesting and novel theory of distributive justice, with individual autonomy at its core.
IntroductionSuppose we think that it is a proper role for at least some political and social institutions to eliminate or mitigate individual disadvantage, howsoever construed. One question we should ask about this activity is: what ought to be the relation between the judgements of disadvantage encoded in such activity, and the judgements of disadvantage made by the individuals concerned? In particular, is there something wrong with a proposed distribution of resources that treats someone as disadvantaged when they don't see it that way themselves, by offering compensation for features that they do not themselves regard as handicaps?Answering 'Yes' to these questions is, at least at first sight, an attractive stance for a liberal to take. The alternative (that we treat people as disadvantaged even when they don't regard themselves in that light) is unattractive for a number of reasons, especially if -as seems reasonable -we consider only judgements which don't depend on factual errors or manipulation. For one thing, a distributive scheme which doesn't pay attention to individuals' own views in this way seems unjustifiably paternalistic. For another, it looks like it would be controversially perfectionistic: we would be justifying compensation for a feature (deafness, for example) on the grounds of some value judgements (e.g. that deafness is bad, or impairs one's ability to live a good life) which are controversial and rejected by the individuals concerned. Not all liberal political philosophers oppose all paternalism and perfectionism; but anti-paternalism and anti-perfectionism are central to enough liberal thought that it is worth investigating ...