EvidenceProf Blog

Editor: Colin Miller
Univ. of South Carolina School of Law

Friday, August 2, 2013

Noteworthy Forthcoming Article: A Clash of Studies -- Using Removal Statistics to Assess Daubert

A new study by Andrew W. Jurs and Scott DeVito revisits – and disagrees with - the innovative 2005 study by Edward Cheng and Albert Yoon that examined the impact of Daubert on litigant behavior:  specifically, litigants' likelihood to remove cases from state to federal court.  

Since these studies are based on the idea that litigants might remove cases to federal court to take advantage of Daubert, it would be interesting if litigators could comment (below) on whether they actually do seek removal (or fight removal) based on state/federal variance in expert gatekeeping standards . . . .

Here’s the information on the new study:

The Stricter Standard: An Empirical Assessment of Daubert's Effect on Civil Defendants

Andrew W. Jurs (Drake University Law School) and Scott DeVito (Florida Coastal School of Law)

Catholic University Law Review, Vol. 62, 2013


While Daubert was clear in its rejection of Frye and the substantive standard for expert admissibility, its effect on litigants has been hotly debated. Several studies since 1993 used quantitative analysis through case study analysis and judicial surveys, to measure Daubert’s effect. Yet these methodologies have reached contradictory results. In 2005, Edward Cheng and Albert Yoon offered a revolutionary new approach in their work Does Frye or Daubert Matter? A Study of Scientific Admissibility Standards. They proposed that studying removal of cases from State Court to Federal Court in the period 1990 to 2000 could quantitatively demonstrate Daubert’s true effect. It works because a litigant could, by removing a case to federal court, switch scientific admissibility standards in some circumstances. The aggregate change in behavior of all litigants can therefore be measured. 

We agree that removal rate offers the best hope for assessing the true effect of Daubert, and so in this study we offer our analysis of removal rates using econometric tools never before applied in this area. Our analysis reveals a startling discovery: Daubert is the stricter standard for expert admissibility. Not only does a change removal rate after Daubert clearly demonstrate this result, but it is confirmed through a “shift back” to state courts when the state also adopts Daubert and removal no longer entails a change in standards. Our results directly contrast with Cheng & Yoon’s conclusions, and so we also revisit their study and deconstruct its methodology piece-by-piece. In so doing, we will describe several errors in that study both explaining the different results but also ultimately undermining its validity. 

Ultimately, our research into aggregate case data from real cases demonstrates a new and conclusive finding: Daubert has been the stricter standard.

- - - - - - -

- JB


| Permalink

TrackBack URL for this entry:


Listed below are links to weblogs that reference Noteworthy Forthcoming Article: A Clash of Studies -- Using Removal Statistics to Assess Daubert:


I'm a litigator in the defense bar. I prefer being in federal court for multiple reasons, with Daubert's heightened standards being only one of them. Moreover, my state (Arizona) recently adopted Daubert at the state court level, but I still prefer to remove and do remove at every single opportunity.

Posted by: N | Aug 5, 2013 9:03:56 AM

In my jurisdiction, I almost always remove certain actions if I have a basis to do so. I do analyze how anticipated issues will play out before recommending removal to clients. With that said, I can say that the application of Daubert versus Frye (I am in a Frye jurisdiction) has never been determinative for me and my clients. Rather, there are two primary reasons for removal: the jury pool and the summary judgment standard. Even if my state adopted Daubert, I would still tend to remove at the same rate I do now, all other things remaining equal.

I would like to comment on the use of removal statistics by Cheng and Yoon and Jurs and DeVito. When I began reading the studies, I was confused about the questions being posed. Did they want to measure which standard admitted more evidence? Did they want to measure the impact of the standards on removal rates? Did they want to measure the impact of the standards on actual jury awards or settlements? It appears that they wanted to determine whether Daubert was a “stricter” standard than Frye by measuring how often parties remove cases. In my view, this conflates attorneys' possible perception of Daubert benefits with actual admissibility determination outcomes. But perception is not necessarily reality.

The authors' logic appears to be as follows: if defense attorneys perceive that Daubert is stricter, they will remove from Frye jurisdictions. Furthermore, attorneys' perceptions are based on prior events (successful or unsuccessful admissibility challenges). Therefore, removal data which demonstrates attorney perception of Daubert versus Frye will tell us, indirectly, about admissibility determinations made in other litigation, prior to the removal of the analyzed actions.

But removal-motivating attorney perceptions do not necessarily come from personal prior experiences. They may come, for example, from courtroom gossip, from CLE presenters, and yes, even from law review articles like these. If an attorney's perceptions do come from personal experience, there is a question of quantity: did the removing attorney experience multiple adverse Frye determinations, or one high-stakes one? Furthermore, the motivating perceptions may not even be that Daubert is “stricter”--but that it offers greater opportunity to challenge proffered scientific evidence, since its inquiries are not restricted, as sometimes occurs with Frye, to novel scientific evidence.

Using removal rates to measure the comparative strictness of the two standards does not strike me as a reliable method. Can removal rates show something interesting about litigant response to procedural changes? Certainly. Can they be used in discussing the impact of procedural changes on federal and state caseloads? Sure. I remain unconvinced, however, that removal rates are an accurate way of measuring whether Daubert is actually more stringent than Frye.

Posted by: J. | Aug 26, 2013 10:51:49 AM

Post a comment