Levelling down revisited

The ‘levelling down objection’, as originally advanced by Derek Parfit, was meant to place critical pressure on ‘telic egalitarianism’, which claims that it is itself bad if some people are worse off than others. As Parfit notes, it appears to follow from telic egalitarianism that the badness of inequality will be removed however equality is restored. So, if we can replace an unequal distribution with an equal distribution by levelling down the position of the better off without improving the position of the worse off, then the new distribution qualifies as good. But that is deeply counterintuitive, according to proponents of the levelling down objection. Egalitarians have been steadily amassing replies to the levelling down objection, and now it is considerably less obvious whether the objection has the deadly force that it was initially taken, by many, to possess. At least three arguments have made a strong impression on the debate. First, the claim that levelling down is ‘better in one respect’ may be a commitment which is modest enough to spare egalitarians from embarrassment. Second, a variety of forms of ‘conditional egalitarianism’ have emerged, which insist that equality is valuable only given the satisfaction of certain conditions that side-step the levelling down objection, while remaining true to the spirit of telic egalitarianism. Third, the goodness or badness of a distribution can perhaps be imported into an account of personal goodness in such a way as to show that the levelling down objection does not apply. Several questions arise. How powerful are these egalitarian lines of response? And, for those who have been persuaded by it, what does the force of the levelling down objection really consist in? Is revulsion at levelling down just the other side of the coin from an enthusiasm for the Pareto principle? And why does levelling down apply specifically to egalitarianism? For any theory of distribution, egalitarian or otherwise, it seems plausible to suppose that there can be circumstances in which an actual distribution will conform more closely to the theory-prescribed distribution if levelling down takes place. Would that make such a theory of distribution vulnerable to the levelling down objection? It is time to revisit these issues to take stock of these arguments, and to think more carefully about the normative force of the levelling down objection. This is what this paper sets out to do.