In a series of books and articles Paul Robinson (along with various coauthors) builds a case for “empirical desert” as a basis for deciding who should be punished under the criminal law and how much. Robinson’s case is unconvincing in several respects. First, his attempt to discredit the leading rival, deontological desert, is based on a misconception of the prevailing methodology in deontological ethics. Additionally, by dismissing as practically irrelevant the grounds for justified punishment, Robinson’s analysis fails to take account of the dynamic interplay between social and political values on the one hand and the meaning and justification of punishment on the other. Finally, because Robinson’s pragmatic conception of desert is untethered from our political morality, it provides neither the occasion to reflect critically on our punitive institutions and practices nor the normative resources to reform them.