Abstract

Rehearing Daubert on remand from the Supreme Court, Judge Kozinski introduced a fifth "Daubert factor" of his own: that expert testimony is based on "litigation-driven science" is an indication that it is unreliable. This article explores the role this factor has played in courts' handling of scientific testimony, clears up an ambiguity in "litigation-driven" and some uncertainties in "reliable," and assesses the reasons courts have given for reading such research with suspicion. This analysis reveals that research that is litigation-driven in the stronger of the two senses distinguished is inherently less likely to be evidentially reliable; but also that it is so hard to determine whether research is litigation-driven in this strong sense that this new Daubert factor is not as helpful as Judge Kozinski imagined.