The Judge as Gatekeeper

Definition

In Daubert v. Merrell Dow Pharmaceuticals (1993), the Supreme Court held that Federal Rule of Evidence 702 gives trial judges not just the power but the obligation to screen proposed expert testimony before it reaches a jury, checking that it is "not only relevant, but reliable." This replaced the older Frye "general acceptance" test with a flexible inquiry: judges ask whether the underlying theory or technique can be (and has been) tested, has been peer-reviewed or published, has a known or potential error rate, and is generally accepted in its field — a nonexclusive list, not a checklist to be mechanically applied.

In the Book

The manual traces the "Daubert trilogy" in detail. In Daubert itself, the Court reversed and remanded a case about whether the drug Bendectin caused birth defects, declining to apply its own new test to the record. In General Electric v. Joiner (1997), the Court held that appellate review of a district judge's admissibility ruling gets only "abuse of discretion" deference — even when, as there, exclusion of the plaintiff's causation experts was "outcome determinative" and led straight to summary judgment against a lung-cancer plaintiff who had been exposed to PCBs. The Court also held that a judge may reject expert testimony connected to the underlying data "only by the ipse dixit of the expert" when there is "too great an analytical gap between the data and the opinion." Kumho Tire v. Carmichael (1999) then extended the gatekeeping duty beyond hard science to all expert testimony, including an engineer's tire-failure analysis based on experience rather than laboratory method. David Goodstein's companion chapter, "How Science Works," reframes the same four factors from a scientist's side and notes Chief Justice Rehnquist's own admission that he was unsure what "falsifiability" even meant as a legal standard — underscoring that Daubert asks judges, not scientists, to make a scientific-validity call with legal consequences.

Why It Matters

Gatekeeping names a structural role that recurs whenever one system must admit or reject inputs produced by a different system with different standards of proof: a non-expert evaluator (a judge, a manager, an editor) must decide whether specialized work product is trustworthy enough to act on, using proxies for rigor (was it tested, published, error-rated, accepted by peers) rather than re-deriving the underlying expertise themselves. The same tension — screening for method without re-litigating conclusions, and doing so with a flexible standard instead of a bright-line rule — shows up anywhere quality control sits between a technical producer and a non-technical decision-maker.