Document Type


Publication Title

Arizona State Law Journal

Publication Date



In a series of recent writings, Paul Robinson has defended “empirical desert” as the way of deriving distributive principles for determining who should be punished and by how much. Desert is, of course, an idea with a long history, and its precise role in criminal law has been much debated. In addressing various criticisms of desert in criminal law, Robinson distinguishes empirical desert from what he calls “deontological desert” and “vengeful desert.” Robinson’s strategy, which I call “divide and deflect,” fights off various objections traditionally leveled against the use of desert in criminal law by arguing that most of those objections may be valid for “deontological” and “vengeful” desert but are not applicable to “empirical desert.” So, for instance, “vengeful desert” can be too harsh, may be based on anger and hatred, and give only vague guidance, and “deontological desert” judgments may be too contested to be useful for policy makers, but, Robinson claims, empirical desert suffers from no such defects. Robinson’s core claims – about the need to tie desert assessments close to ordinary intuitions and the substantial crime control benefits to be derived when the state can successfully command respect as a moral authority – are valid. However, Robinson’s “empirical desert” needs what he calls “vengeful desert” and “deontological desert” to succeed, and his attempts to make his proposal resistant to the usual anti-retributivist objections by jettisoning the latter two may in the end hurt his project more than help.