The Decision-Theoretic Case for Intermediate Criminal Verdicts

Abstract This chapter studies the circumstances under which issuing an intermediate verdict maximises expected value and is, therefore, justified under decision theory. It starts by presenting the well-known decision-theoretic argument for the selection of the standard of proof in a binary verdict system. On the basis of this argument, the chapter identifies the condition for the superiority of an intermediate verdict with respect to acquittal and conviction – or the ‘superiority condition’. To clarify, the chapter shows that, if an intermediate verdict satisfies this condition, there is a probability range such that issuing the verdict when the probability of guilt falls within this range yields higher expected value than acquitting and convicting and is, therefore, justified on decision-theoretic grounds. The chapter draws from the superiority condition – which is a simple mathematical formula – a heuristic for the identification of a concrete superior intermediate verdict. On the basis of this heuristic, the chapter devises an intermediate verdict, called conditional acquittal, that consists in allowing for the retrial of the defendant in the presence of new incriminating evidence with substantial probative value.