Absolute Certainty and the Death Penalty

It is generally well understood that the proper standard of proof in criminal cases is the familiar beyond a reasonable doubt standard. Recently, however, there have been proposals to create a new standard of proof - such as "beyond all doubt" or "no doubt" - for capital cases as a way of protecting against erroneous convictions. These proposals have become more urgent in light of revelations of serious errors in capital cases in recent years. This paper critiques such proposals. I argue that a higher standard of proof in capital cases can only be coherently justified on consequentialist grounds; and, if consequences are what matters, a higher standard of proof is unlikely to have the desired effect. Decades of empirical evidence show that changes to the wording of jury instructions are likely to have little, if any, impact on how jurors reach their decisions. As an alternative, I propose several other changes that might actually help guard against erroneous convictions in capital cases. First, I suggest either eliminating or altering the present system of "death qualification" of jurors in capital cases, which leads to an artificially low standard of proof in such cases. Second, I suggest several changes to the way in which jurors are instructed. Changing the verbal formulation is not likely to be effective, but instructing jurors at the beginning, rather than the end, of the case in the standard of proof; giving jurors written instructions; and finally, expressing the standard of proof in quantitative terms to ease comprehension all may achieve more satisfactory results.