Modern law on expert testimony insists, as a condition of admissibility, that the asserted expertise be determined by the trial judge to be reliable. Reliability is usually characterized as a dichotomous attribute of evidence, as if expertise were either reliable or unreliable. This article argues that making progress in the development of meaningful and appropriate restrictions on the admissibility of expert testimony requires that we abandon this conceptualization and understand the implications of endorsing a gradational notion of reliability in which evidence can be more or less reliable and in which a comparative assessment of reliability is prominent. Consistent with Supreme Court precedent and available empirical evidence about jury decision-making, this article recommends that, in deciding whether to exclude expert testimony, the court's comparative reliability inquiry should focus on whether more reliable expertise is reasonably available to the proponent, rather than on the question of whether the jury will overvalue the expertise at the offered level of reliability. A rudimentary outline of how this would work is provided.


reliability, experts, Daubert, Kumho Tire, best evidence, deference to expert, FRE 702, Federal Rule of Evidence 702

Publication Date


Document Type


Place of Original Publication

Seton Hall Law Review

Publication Information

34 Seton Hall Law Review 191 (2003)

Included in

Evidence Commons


COinS Dale A. Nance Faculty Bio