Publication Date



Behavioral and neuroscientific research provides a relatively clear window into the timing of developmental maturity from adolescence to early adulthood. We know with considerable confidence that, on average, sixteen-year-olds are less developmentally mature than nineteen-year-olds, who are less developmentally mature than twenty-three-year-olds, who are less developmentally mature than twenty-six-year-olds. However, in the context of a given case, the question presented might be whether a particular seventeen-year-old defendant convicted of murder is “developmentally mature enough” that a sentence of life without parole can be constitutionally imposed on him or her. While developmental maturity can be accurately measured in group data, it cannot be assessed in individuals with confidence. This fact is an instance of a fundamental disconnect that occurs at the intersection of science and law between what scientists study and what courts ordinarily need to know. Scientists typically study phenomena at the group or population level, whereas courts usually need to determine whether a particular case is an instance of some known phenomenon. This is called the group to individual (G2i) problem. Although the G2i problem cannot be fully resolved, it can be managed by using the base-rate data available in the research literature to set the burden of proof. Setting the burden of proof is a classic mechanism for allocating the risks of making a mistake. Two factors in particular inform judgments about allocating risk of error, with the first being the likelihood or frequency of the fact in question and the second being the costs associated with the error. The rarer the fact and the larger the cost of a mistake, the greater the burden of proof should be. The latter factor, the costs associated with error, lies behind the traditional burdens of proof of preponderance of evidence and proof beyond a reasonable doubt in civil and criminal cases, respectively. In contrast, while the former factor, the frequency of the fact in question, is used regularly in areas of applied science, it has generally not informed allocations of burdens of proof in court. This Article sets forth a framework of shifting burdens of proof grounded in the research literature that can be employed to allocate the risk of error when assessing development rity in the sentencing of offenders across the age spectrum.

Document Type


Publication Title

William and Mary Law Review