Judges Should Be Discerning Consensus, Not Evaluating Scientific Expertise


One of the most constructive critiques of the Daubert admissibility regime is Professor Edward Cheng’s recent proposal for a new Consensus Rule in the Federal Rules of Evidence. Rejecting the notion that judges and juries have the capacity to evaluate scientific expertise, Cheng’s proposal would eliminate Daubert hearings—and judicial gatekeeping concerning expert testimony—and require judges and juries, in their verdicts, to follow consensus in the relevant scientific community. Significantly, Cheng argues that judges and juries would have an easier time identifying consensus than they have in deciding between experts who disagree.

We find Cheng’s emphasis on consensus compelling, and we support it. We agree that judges and juries are typically not capable of making technical decisions based on conflicting scientific testimony, and we support Cheng’s proposal to leave that expert decision in the hands of the appropriate community of experts—the scientific community rather than the legal community. We differ from Cheng insofar as we are skeptical that consensus would be easy for lay jurors to identify. Cheng is correct that identifying consensus in a scientific field (e.g., smoking is unhealthy) is based in the everyday, ubiquitous expertise of lay persons—it is a social decision, not a technical decision—such that consensus is easier for non-specialists to identify than discerning scientific truth in a specialized field. Scientists, however, in legal contexts, may not agree on what counts as consensus, and just as lay jurors do not possess the type of expertise that would allow them to act as peer-reviewers of scientific experts, they do not have the experience needed to discern what counts as consensus where experts disagree. We believe, however, that a consensus regime that places more of the responsibility in the hands of judges would be more successful.

Judges have more experience than lay juries in listening to and evaluating experts and the communities they represent. Jurors are almost always novices, whereas judges, who undergo significant trial experiences and judicial training prior to becoming a judge, are constantly gaining experience in their profession. While judges may not have the deep understanding that specialized experts have, specifically regarding what qualifies as consensus in their field, their experience likely enables most judges to enhance their ubiquitous expertise in judging people. Through repeated experience with expert witnesses, judges could learn to identify reliable expert informants on the issue of consensus. Indeed, many courts already emphasize the importance of consensus in their expert evaluations. Consequently, we recommend that in cases involving scientific or other technical issues, a judge should hold a pre-trial, Daubert-style hearing (alongside a new Consensus Rule), during which a judge would hear arguments about any existing consensus (or lack thereof) in the relevant field.

Like Justice Blackmun in Daubert, we recommend a set of factors to guide judges in this new task. If a judge determines that consensus exists, the jury would be instructed to follow it. If no consensus exists, the current Daubert framework for admissibility of expert witnesses should remain in effect. Despite the problematic fallibility of juries when they choose between opposing experts, their choices should be limited to those experts who were subject to some level of admissibility gatekeeping. In this way, our proposal maintains some current standards while combining the best parts of Cheng’s proposal with recent insights from sociological studies of expertise.




Courts | Evidence | Judges | Law | Science and Technology Law