Cornell Law Review

A child makes an out-of-court statement accusing an adult of abuse. That statement is important proof, but it also presents serious reliability concerns. When deciding whether it is sufficiently reliable to be admitted, should a court consider whether the child’s statement is corroborated—whether, for example, there is medical evidence of abuse? More broadly, should courts consider corroboration when deciding whether evidence is reliable enough to be admitted at trial? Judges, rule makers, and scholars have taken significantly divergent approaches to this question and come to different conclusions.

This Article argues that there is a key problem with using corroboration to evaluate admissibility. Corroborated evidence is, indeed, more likely to be reliable than uncorroborated evidence. But that does not mean that corroboration is always a proper admissibility consideration. In fact, if the corroboration simply proves the same material fact as the corroborated evidence, using corroboration to determine the admissibility of the evidence can impede rational truth-seeking through a mechanism this Article calls “structural confirmation bias.”

What should courts and rule makers do, then, when a category of evidence is critically important but also presents troubling reliability concerns? This Article offers a theoretical framework for thinking about corroboration that can help rule makers and judges craft and apply corroboration rules. It first argues that when that type of evidence will typically be introduced by the party with the burden of proof, it is better to require corroboration to sustain a verdict than to require corroboration to admit the potentially unreliable evidence. However, when that type of evidence may be introduced by either party, courts should consider only corroboration that does not trigger “structural confirmation bias.”

To read this Article, please click here: Theorizing Corroboration