Concept
Joiner and Appellate Review
Reliability and relevancy
Joiner and Appellate Review
General Electric Co. v. Joiner is one of the three decisions known as the Daubert trilogy, and it answers two questions that Daubert left open. How closely may a trial judge test the link between an expert's data and the expert's opinion? And how closely may an appellate court test what the trial judge did? The Court granted certiorari to settle the second question, held that abuse of discretion is the appropriate standard, and concluded that the district court did not abuse its discretion when it excluded the plaintiff's experts.1 In reaching that result, the Court confirmed that a trial court may exclude an opinion when too great an analytical gap separates the data from the opinion offered.2
The facts of Joiner's case gave both holdings their shape. Joiner linked his cancer to exposure to PCBs and to their derivatives, furans and dioxins. He had smoked for about eight years, both of his parents had smoked, and lung cancer ran in his family, so he was perhaps already at heightened risk of developing the disease eventually. His suit alleged that the exposure “promoted” his cancer, which otherwise would not have developed for many years, if at all.3 With smoking and family history already in the picture, he relied largely on expert witnesses. They testified in depositions that PCBs alone can promote cancer, that furans and dioxins can too, and that Joiner's exposure to all three was likely responsible for his cancer.4
The case then moved through three courts. The district court excluded certain of those experts, and the exclusion was the sole cause of the summary judgment that followed.5 The Court of Appeals suggested that Daubert had somehow altered the general rule of appellate review where a district court excludes scientific evidence, and it reversed. The Supreme Court in turn reversed that judgment.6 The first question, what a judge may do when the data do not reach the opinion, is the subject of the next subsection. The second, how far an appellate court may second-guess that judgment, follows it.
The Analytical Gap
The Daubert inquiry is supposed to focus on the expert's methodology and not on the expert's conclusions.7 That instruction invites an objection. If a court may examine only methods, how can it reject an opinion for failing to follow from the data? Joiner answers that conclusions and methodology are not entirely distinct from one another. Trained experts commonly extrapolate from existing data, but nothing in Daubert or the Federal Rules of Evidence requires a district court to admit opinion evidence that is connected to existing data only by the ipse dixit of the expert. A court “may conclude that there is simply too great an analytical gap between the data and the opinion proffered.”8
The rule has two halves, and each has its own reason. The first half protects extrapolation. Real disputes seldom come with studies of the very person, dose, and setting at issue, so a rule forbidding every step beyond the studied conditions would silence experts on most questions worth asking. The second half limits that freedom. Daubert points the judge toward method because a sound method is the best available stand-in for a trustworthy result, and the stand-in works only if the method actually carries the expert from the data to the opinion. If methodology and conclusion were wholly separate, an expert could attach any conclusion to a list of loosely related studies, call the list a method, and leave the judge unable to object without appearing to decide the merits. Where the only bridge from data to opinion is the expert's assurance, the judge has nothing to evaluate, and the gap inquiry lets the judge say so. The text of Rule 702 points the same way: the testimony must be the product of reliable principles and methods, and the opinion must reflect a reliable application of those principles and methods to the facts of the case.9 Later discussions of fit use the same vocabulary, asking whether the expert has unjustifiably extrapolated from an accepted premise to an unfounded conclusion.10
The studies in Joiner show where the gaps appear. The Court brushed aside the question whether animal studies can ever be a proper foundation for an expert's opinion, because that was not the issue. The issue was whether these experts' opinions were sufficiently supported by the animal studies on which they purported to rely, and the studies were so dissimilar to the facts of the litigation that it was not an abuse of discretion for the district court to reject the reliance.11 The epidemiological studies failed in two further ways. One made no mention of PCBs and was expressly limited to the type of mineral oil involved in that study, so it did not support the experts' opinions. Another involved a PCB-exposed group in Japan that had seen a statistically significant increase in lung cancer deaths, but its subjects had been exposed to numerous potential carcinogens, including toxic rice oil that they had ingested.12 The Court held that the district court could reasonably find the studies insufficient, whether individually or in combination, to support the conclusion that Joiner's exposure to PCBs contributed to his cancer.13 The extrapolation itself drew no penalty. The unsupported extrapolation did.
Justice Stevens did not concur in the judgment, yet he stressed what the holding does not do. The Court had not held that admitting the testimony would have been an abuse of discretion, and nothing in Daubert or the Rules requires a judge to keep an expert's conclusions from the jury when they fit the facts of the case and rest on reliable scientific methodology.14 The disagreement shows why the gap doctrine is hard to apply. The line between methodology and conclusions has been very difficult to locate in practice,15 and commentators who defend the gatekeeping role describe the judge's task as examining the evidence and deciding whether it is valid science.16
Exam stems that carry an expert from high-dose animal studies, from a different disease, or from a differently exposed population to the plaintiff's case are describing a possible gap. Two answers go wrong. One treats Daubert's focus on methodology as a shield and insists that the judge may not compare the data to the conclusion at all, which Joiner forecloses. The other reads Joiner as holding that animal or epidemiological studies can never support a causation opinion, when the Court said the question was whether these experts' opinions were sufficiently supported by the studies they used.17 A court that excludes for a gap is not declaring the conclusion false. It is finding that the proponent has not shown how the data lead to the opinion.
Abuse-of-Discretion Review
Joiner settled the question Daubert had left open: an appellate court reviews a trial court's decision to admit or exclude expert testimony under Daubert for abuse of discretion.19 Under that standard the reviewing court asks whether the trial judge's ruling fell within the bounds the law allows, which is a different question from whether the reviewing court would have ruled the same way. The standard has deep roots. In Spring Co. v. Edgar, decided in 1879, the Court said that where it is very much a matter of the trial court's discretion whether to receive or exclude evidence, the appellate court will not reverse unless the ruling is “manifestly erroneous.”20
The Court of Appeals had tried to depart from that rule for exclusions of scientific evidence, and Daubert did not support it. Daubert did not address the standard of appellate review at all. It held only that the “austere” Frye standard of general acceptance had not been carried over into the Federal Rules of Evidence.21 The Rules therefore let district courts admit a somewhat broader range of scientific testimony than Frye allowed, but they leave in place the trial judge's gatekeeper role in screening that evidence. A court of appeals applying abuse-of-discretion review to such rulings may not categorically distinguish between rulings that allow expert testimony and rulings that disallow it.22 The Court cited its earlier evidentiary decisions, among them Beech Aircraft Corp. v. Rainey and United States v. Abel, for the deference that review demands.23
The symmetry follows from what gatekeeping is. A judge who keeps out an unreliable opinion performs the screening duty as faithfully as one who admits a reliable one, so an appellate standard that scrutinized only the exclusions would pull judges away from the screening that Daubert requires. The Court of Appeals' error was a failure to give the trial court “the deference that is the hallmark of abuse-of-discretion review.” In applying an overly “stringent” review, it overlooked that standard.24 Justice Stevens, who did not concur in the judgment, confirmed the point from the other direction. The very point of the holding, he wrote, is that the abuse-of-discretion standard applies whether the district judge has excluded or admitted the evidence, and the Court had not held that admitting the experts would have been an abuse.25
Joiner also argued that the setting of the ruling called for closer review, because the exclusion produced summary judgment and was therefore “outcome determinative.” The Court rejected the argument. On a motion for summary judgment, disputed issues of fact are resolved against the moving party, but the admissibility of expert testimony is not such an issue of fact, and it remains reviewable for abuse of discretion.26 The admissibility ruling comes before the summary judgment question and decides what evidence will count in the record. The favorable reading of the record that protects the party opposing summary judgment applies to the evidence that passes the gate, and it does not reach the gate itself. The point has real consequences, because a plaintiff's expert is often the only proof of causation. Joiner affirmed a summary judgment under the deferential standard even though the exclusion of the expert was its sole cause.27
Deference is justified because the trial judge manages the record and sees the experts, and the cost is variation: judges may reach different results on similar expertise, and the appellate court may uphold both.
The order of operations in Joiner shows how the standard works. The Court fixed the standard first and then asked whether the district court had abused its discretion, finding no abuse because the experts' reliance on the studies could reasonably be rejected.28 It never asked whether the Supreme Court would have weighed the studies the same way. A stem that stresses an exclusion ending the case, or a trial judge who excluded an opinion the Rules might seem to favor admitting, invites two tempting errors, de novo review and heightened scrutiny. Both fail, because the standard does not change with the direction or the consequence of the ruling.
The same deference reaches the way the judge tests reliability. A trial court has discretion in deciding whether any particular criterion is satisfied and in deciding what criteria to apply in the case before it.29 The Daubert and Kumho standards are ways to show that the requirements of amended Rule 702 are satisfied. Their factors are strongly advised, but other ways to show reliability are available so long as they do not flout Daubert or Kumho, which are flexible by their own terms.30 The later section on the flexible use of the factors builds on this point.