More on Expert Witness Sequestration

Our post on witness sequestration has drawn two responses. One, from a reader who requests anonymity, is as follows:

I agree that sequestering of expert witnesses is generally not sensible (though if they were to find a really nice hotel, I might be convinced otherwise). My only counter-example would have to be something along the lines of the following:

The expert witness writes a report stating that because EvilPharma put out drugs A and B, one million people developed a disease from taking one or more of those drugs. In trial, it comes out that EvilPharma had received glowing test results about A but bad results on B, and put both drugs on the market anyway. Assume that everyone feels that the jury is much more likely to find liability for the effects of B than for the effects of A. The expert witness is then asked in trial whether there was a way to look at his/her data to distinguish between the effects of A and B, and claims that if you squint hard enough and turn one of the graphs on its side, it's clear that A (if a defendant's expert) or B (if a plaintiff's expert) caused at least 999,999 of the one million cases of disease.

Now, one might say this can be avoided by cutting off the answer with some voir dire on whether the expert's pre-trial report and deposition separated out the effects of A and B, but the expert may claim that all he/she is doing is highlighting something that was already in the data presented. I would think that ideally, the judge would tell all the experts to separate out the effects of drugs A and B without any clue as to which side would favor which split. However, that might mean that the expert could never meet with counsel who might smile or make a funny face in planning the revised direct. So, while I'm not convinced that sequestration would be practical, I could imagine it being useful in theory.

The third paragraph of this counterexample is imbued, perhaps, with a certain wistful longing for a world in which experts opine from behind some Rawlsian veil of ignorance. It may be granted that objective, unbiased testimony might be more forthcoming in such a world. But that world is very distant from our current adversary system. Still, even if dreams of any notionally better world were put aside, it would remain to address the underlying scenario described in the counterexample's second paragraph, where factual evidence emerges at trial that undercuts an expert’s assumptions or analysis. In that situation, there may be something to be said for testing the expert’s credibility by surprising the expert with the information and watching the expert squirm in response.

Of course, if that thinking were applied more generally, we might also want to stop providing experts in advance with much of any information about the issues in controversy. The Rawlsian longing expressed in the counterexample’s third paragraph, that is, may be more integral to the counterexample than it initially seemed. If there is an insuperable practical problem with a system where the jury is permitted to watch the expert’s voyage of discovery, it is probably is that such voyages are often unprofitable if navigated improvisationally, without maps or charts or sextants. It takes a lot of time and thought to analyze (e.g.) whether a drug causes a disease. The price of allowing experts that leisure (and of requiring pretrial disclosure of their finished work) is tolerating a world where the expert gets most of the information beforehand – even if that tends to deprive juries of some opportunities to see perspiration beading up on expert brows.

But then again, maybe the thinking doesn’t need to be applied more generally. Maybe leaving some room for surprise, in judicious quantities, is a helpful way of pulling back the curtain and letting the jury get a glimpse of the real wizard, pulling nervously at the levers. The question might be, does cross-examination already present sufficient opportunity for that? And if so, then is the increment of informative unpredictability that might be gained from sequestering experts worth the loss in testimonial coherence that may come from making experts testify without having heard any of the evidence?

A second response comes from C.E. Petit. We’ll discuss it in our next post.

Update 3/18/06: Our anonymous reader responds:

I'll see your Rawlsian longing and raise you a Heisenberg uncertainty. You do bring up an interesting point that the more we get to see the expert surprised, the less well-considered the expert's analysis would have to be.

One partial solution would be to have more court-appointed experts who, while they might still have their own biases, would at least be able to consult with either neither set of counsel or else counsel for both sides. Another potential solution might be to give each side a limited number of pre-trial computation-intensive hypotheticals that it could ask the other side's expert, presumably with the questioner footing the bill for any additional work. (And coming up with and approving an acceptable budget would be a nightmare.) Thus, in the case of the two drugs, the hypothetical would be to assign relative harm to each drug. The expert would then submit a responsive report and if he or she came up with a different answer at trial after the evidence is presented, the responsive report could be used for impeachment. This may involve tipping one's hand to the other side before trial, but I'm sure that lawyers are well-acquainted with making judgments about whether such a strategy is worthwhile.

Anyway, thanks for raising the interesting question. Deep down, I do agree with you that because experts' testimony is supposed to be based on other facts in the case, as opposed to the testimony of a fact witness in most situations, sequestration of experts is generally not a good idea.