The California Supreme Court is poised to set standards regarding the foundational scrutiny of expert opinions in a case that promises to have significant implications for product manufacturers.
There is a continuing controversy in the California intermediate courts of appeal over how strictly trial judges should scrutinize the foundation for scientific testimony, arising from the California Supreme Court’s last words on the subject in 1994. In April 2005, the Court granted review of a decision concerning California gatekeeping standards, Lockheed Litigation Cases, 126 Cal.App.4th 271, 23 Cal.Rptr.3d 762 (2005) (“Lockheed II”), No. S132167 (Rev. Granted Apr. 13, 2005), presumably to deliver some longawaited guidance in this critical area. (Disclosure: Alan Lazarus is the author of an amicus brief fi led in this case.)
This may be the case where the Court establishes a coherent system for assuring that all scientifi c opinion testimony offered in California courtrooms meets acceptable standards of scientifi c validity and reliability.
A. Background on California’s Admissibility Standards for Scientifi c Evidence
The importance of the pending Lockheed case can only be appreciated with knowledge of the existing patchwork of judicial decisions attempting to defi ne trial court authority to exclude junk science.
1. The Kelly Rule
In People v. Kelly, 17 Cal.3d 24 (1976), the California Supreme Court adopted the Frye standard, holding that scientifi c evidence predicated on a novel theory, technique or process is not admissible in California courts unless the proponent demonstrates that the predicate is generally accepted as reliable by the relevant scientifi c community.1
The Court held that such a showing is necessary because of the captivating effect that scientifi c evidence may have on a jury. The risk that a jury will get carried away by novel scientifi c evidence that has not been shown to be reliable is a threat to rational decision-making. Thus, Kelly calls for strict scrutiny of novel scientifi c theory for essentially two reasons: (1) the experimental or novel nature of many scientifi c theories, techniques or processes raises concern over their ability to reliably generate valid conclusions; and (2) the trappings of science tend to confer, in the minds of jurors, unwarranted persuasive power on expert scientists with impressive credentials who claim to demonstrate that science supports or compels their opinions.
2. The Screening Framework For Scientifi c Evidence Arising From Leahy
Much of the current controversy revolves around the scope and implications of the California Supreme Court’s decision in People v. Leahy, 8 Cal.4th 587 (1994). In its narrowest sense, Leahy held that the Kelly standards regarding the admissibility of novel expert testimony survived the then-recent U.S. Supreme Court decision in Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579 (1993). The Court affi rmed that the Kelly rule remained viable for cases presenting novel scientifi c theories, processes or techniques and that, unlike Daubert, the proponent must specifi cally demonstrate to the court that the predicate is generally accepted as reliable by the relevant scientifi c community.
Reasoning its way to that conclusion, the Leahy court reviewed Daubert in detail and discussed the relationship between Federal Rule of Evidence 702, as interpreted by Daubert, and the analogous provisions of the California Evidence Code. “[T]he [Daubert] court expounded on the screening functions federal courts should perform under rule 702 to assure that ‘any and all scientifi c testimony or evidence admitted is not only relevant, but reliable.’” Leahy, 8 Cal.4th at 597 (quoting Daubert, 509 U.S. at 590). Comparing Rule 702 to the California Evidence Code, the Leahy Court concluded that “[s]ections 720 and 801, in combination, seem the functional equivalent of Federal Rules of Evidence, rule 702, as discussed in Daubert.” Id. at 598 (emphasis added). However, because the Kelly standard of general acceptance is a more conservative and more stringent approach to admitting evidence based on novel scientifi c theories, processes or techniques, and because Kelly was decided after the adoption of the Evidence Code, Leahy held that principles of stare decisis did not allow the Court to abandon that decision. Id. at 599. Applying Kelly, the Court ultimately found that the horizontal gaze test used by police to determine the defendants’ level of intoxication was a new scientifi c technique requiring compliance with the Kelly standard and proof of its general acceptance by the scientifi c community. Id. at 612.
B. The Issue: What Foundational Screening Is Required When Kelly Does Not Apply?
1. The Broad View of Trial Court Screening Authority
Kelly is limited to issues of novel science, a category of scientifi c evidence that has sometimes been interpreted narrowly. This leaves a broad range of scientifi c opinion testimony unregulated by Kelly’s reliability safeguards, such as when an expert uses generally accepted “techniques” to draw conclusions not supported by the underlying evidence or by application of appropriate scientifi c principles. In such non-Kelly cases, the concerns in Kelly over the signifi - cant impact unreliable scientifi c testimony may have on the jury do not disappear. Employing an established technique does not appreciably dim the scientifi c aura surrounding the expert’s testimony, nor does it guarantee that the substantive (data and reasoning) or procedural (methodology) foundation for the opinion is scientifi cally sound and reliable. By explicitly drawing the parallel between the federal and state rules of evidence, Leahy implicitly confi rmed that section 801, like its federal counterpart, requires trial courts to assure themselves before admission “that any and all scientifi c testimony or evidence is ... reliable.” Id. at 597. See also People v. Mitchell (2003) 110 Cal.App.4th 772, 794 (drawing parallels between California’s Evidence Code and Federal Rule of Evidence 702 in requiring a reliable foundation for scientifi c evidence). Leahy thus strongly suggests that in a non-Kelly case, once apprised of substantial questions concerning the reliability of the factual and analytical bases for an expert’s opinion, trial courts must evaluate the reliability of that foundation. See also People v. Stoll, 49 Cal.3d 1136 at 1140, 1161(1989); People v. Gardeley, 14 Cal.4th 605 at 618 (1996) (holding all expert opinions must be relevant and reliable to be admitted into evidence).
By reaffi rming that the Kelly rule remained viable for cases presenting novel scientifi c theories, processes or techniques, but also interpreting section 801 (along with section 720) to be the functional equivalent of the federal rule that requires the trial court to assure the reliability of the foundations for all scientifi c testimony, the California Supreme Court suggested a logical and workable framework for trial courts to assess the reliability of the foundation of scientifi c testimony: (1) Where the challenged testimony is based on a novel scientifi c technique, the proponent must show both that the predicate technique is generally accepted and that the foundation is otherwise reliable; and (2) when the challenged testimony is based on established scientifi c methods, the proponent is required to show that it rests on a reliable scientifi c foundation. Satisfaction of the general acceptance standard is required for a novel scientifi c technique, but a general review of the reliability of the foundation for the opinion is still required.
This interpretation fi nds support in the text of the Evidence Code provisions. Sections 720, 801 and 803 require the proponent to demonstrate that the opinion is grounded in reliable methods, data and reasoning. Section 801, in pertinent part, limits expert opinion to testimony:
(a) Related to a subject that is suffi ciently beyond common experience that the opinion of an expert would assist the trier of fact; and
(b) Based on matter . . . that is of a type that reasonably may be relied upon by an expert in forming an opinion upon the subject to which his testimony relates . . . .
Evidence Code section 803 provides, in pertinent part, that “[t]he Court may, and upon objection shall, exclude testimony in the form of an opinion that is based in whole or in signifi cant part on matter that is not a proper basis for such an opinion. . . .” (emphasis added). Taken together, sections 801(b) and 803 require that expert testimony have a reliable basis. Many California courts have read these provisions to require trial courts to strictly scrutinize the foundations of expert opinion to determine its reliability, hence its relevance, and its admissibility. See Korsak v. Atlas Hotels, Inc., 2 Cal.App.4th 1516, 1523 (1992); Pacifi c Gas & Elec. Co. v. Zuckerman, 189 Cal.App.3d 1113, 1134, 1136 (1987).
Many California courts, both before and after the granting of review in Lockheed II, have excluded expert testimony they fi nd to be lacking in reliable foundation. In Zuckerman, the court of appeal rejected the opinion of an expert as to the value of a gas storage fi eld, fi nding his approach, data and reasoning fundamentally fl awed. The fl aws included extrapolating an infl ation multiplier for net income from gross fi gures; extrapolating value of the stored gas fi eld from a producing gas fi eld; analogizing to a “comparable transaction” which was not truly comparable; and calculating the value of an underground storage reservoir by proportional reference to the value of the total surface area rather than the storage capacity. The court of appeal concluded that the absence of a reliable and adequate foundation rendered the opinions of the expert speculative and insubstantial. Id. at 1134-1136.
The court rejected the argument that such foundational fl aws go to weight rather than admissibility, fi nding that it impermissibly seeks “to minimize judicial responsibility”: The value of opinion evidence rests not in the conclusion reached but in the factors considered and the reasoning employed . . . . Where an expert bases his conclusion upon assumptions which are not supported by the record, upon matters which are not reasonably relied upon by other experts, or upon factors which are speculative, remote or conjectural, then his conclusion has no evidentiary value. . . . In those circumstances, the expert’s opinion cannot rise to the dignity of substantial evidence. Id. at 1135.
In an earlier opinion in the Lockheed litigation, Lockheed Litigation Cases, 115 Cal.App.4th 558 (2004) (“Lockheed I”) the court rejected plaintiff’s contention that the court’s authority extends only to determining whether the type of evidence relied upon by the expert in forming the opinion is a type of evidence reasonably relied upon by experts. The expert had relied on an epidemiology study, and plaintiff argued that the expert’s testimony was admissible because toxicologists routinely reach causation conclusions based on epidemiology studies. The court disagreed, observing that, “[a]n expert opinion has no value if its basis is unsound,” and that a study may provide a sound basis for one conclusion but not for another opinion. Id. at 564. The court construed Evidence Code section 801 “to mean that the matter relied upon must provide a reasonable basis for the particular opinion offered, and that an expert opinion based on speculation or conjecture is inadmissible.” Id. Accordingly, the court held that it was proper for the trial court to determine that an epidemiology study showing painters exposed to 130 different chemicals had an increased risk of cancer was not proper foundation for an opinion that any single chemical among the 130 caused cancer.
Relying on Lockheed I, the court in Stephen v. Ford Motor Co., 134 Cal.App.4th 1363 (2005) found an expert’s opinion regarding a defective tire lacked foundation. The expert relied on the failure of other tires manufactured by the defendant and amateur photos of the allegedly defective tire taken by plaintiff’s boyfriend. The court found the former was inadequate because the other tire failures were not shown to be suffi ciently similar to support a causal inference. Moreover, inspection of amateur photos to determine the cause of a tire failure was an analytical method neither reasonably reliable nor generally practiced in the fi eld. The court held that these foundational problems did not merely go to the weight of the evidence but rendered the opinions nothing more than speculation, justifying their exclusion. Id. at 1373.
In Geffcken v. D’Andrea, 137 Cal.App.4th 1298 (2006), the court essentially applied the bifurcated foundational screening approach suggested by Leahy, applying Kelly-Frye to the extent the expert relied on novel predicates, and otherwise examining the foundation of the opinion for reliability. The expert opined that plaintiffs experienced various medical complications, including cancer, neurological problems and respiratory problems, from inhaling toxins emitted from molds. The expert based his opinions on air samples confi rming the presence of mold spores at the properties and two different medical tests measuring antibodies produced in the body as a result of exposure to mold (actual living organisms) and mycotoxins (poisons produced by particular molds). Id. at 1302-1303.
The air samples were excluded as overly prejudicial under Evidence Code section 352 because pervasive chain of custody errors invalidated the integrity of the samples. The court found that plaintiffs failed to demonstrate that the antibody tests were generally accepted in the relevant scientifi c community under the Kelly standard. Both tests were performed by only one lab in the United States, reputable medical literature invalidated the tests as proof of causation, and defendants’ experts testifi ed that such tests were not reliable or accepted in the scientifi c community as valid techniques to determine human exposure to mycotoxins. Because each of the bases for the expert’s opinion had been determined unreliable, the opinions amounted to nothing more than speculation and conjecture and the court excluded the expert’s opinions under section 801. Id. at 1311-12.
The court did not characterize its analytical approach or assign Leahy as its model, but the opinion well illustrates the workability, reasonableness and common sense of such a screening structure.
2. The Narrow View of Trial Court Screening Authority
Another line of cases take a narrow view of a trial court’s screening authority. Perhaps the most extreme example is Roberti v. Andy’s Termite & Pest Control, Inc., 113 Cal. App.4th 893 (2003). In Roberti, the court of appeal interpreted Evidence Code section 801 and Leahy in a manner at odds with the language of both. The defendant argued that Kelly applied to expert testimony that low-level exposures to a common pesticide caused plaintiff’s autism, and if it did not, then “the foundational analysis applied in federal courts” should be applied. The court held that Kelly did not apply, and Evidence Code Section 801 did not authorize reliability screening. The court read Leahy as a rejection of Daubert’s “threshold reliability analysis” and determined that threshold reliability screening is “not in keeping with the law in California.” Id. at 905. “Unless and until our Supreme Court determines that the Daubert analysis is applicable in California, we will adhere to the rule of People v. Kelly and its progeny, and refuse to apply a more extensive preliminary admissibility test as in Daubert to expert medical opinion concerning causation.” Id. at 906.
What the Roberti court failed to recognize is that Leahy rejected Daubert solely in the context of novel scientifi c evidence, such as the horizontal gaze test at issue in Leahy, and primarily because it was considered too “liberal” in that context. Nothing in Leahy’s analysis suggests that trial courts should not evaluate the foundation of scientifi c testimony for reliability in some fashion before allowing it into evidence.
Roberti’s departure from other California cases that do screen the foundation of scientifi c testimony illustrates the need for the Supreme Court to clarify the responsibility of California trial courts in those cases where Kelly does not apply. Some courts continue to view Kelly as the only available check on the intrusion of junk science into California courtrooms, and continue to cling to the outdated philosophy that all substantive fl aws in the expert’s foundation merely detract from the weight of the expert’s testimony. Others believe that trial courts are required to assure that scientifi c testimony rests upon a reliable foundation before it can be admitted into evidence.
C. The Stage is Set for Supreme Court Review
– Lockheed Litigation Cases
The Lockheed litigation is a series of multi-plaintiff toxic tort cases arising from occupational chemical exposures at a Southern California aircraft manufacturing facility. In Lockheed I, the court of appeal affi rmed the exclusion of medical causation testimony in a group of 102 wrongful death cases for lack of reliable foundation. Lockheed I, 115 Cal.App.4th at 565. As noted above, the court held that trial courts can reject scientifi c opinion testimony if the data relied upon does not reasonably support the expert’s opinion, i.e., because the expert’s reasoning or analysis was not valid or reliable. Id.
In January 2005, the same court of appeal in the same toxic tort litigation issued a broader opinion, affi rming the trial judge’s exclusion of similar testimony from the same expert, the ubiquitous Dr. Daniel Teitelbaum (the expert who was excluded in General Elec. Co. v. Joiner, 522 U.S. 136 (1999)), in a group of 24 personal injury cases. Lockheed II, 126 Cal.App.4th 271, 23 Cal. Rptr.3d 762. The new opinion addressed a broader array of evidence, and the case therefore generated more specifi c guidance on how trial courts should evaluate medical causation testimony in toxic tort cases.
The trial court had concluded that the materials Teitelbaum cited (multi-solvent epidemiological studies, animal toxicology studies, case reports, government guidance documents and treatises) did not reasonably support his causation opinion. Plaintiffs argued on appeal that the trial court usurped the function of the jury by analyzing and weighing the underlying data. They again argued that once the expert cites materials of a general type that experts in the fi eld customarily rely upon (i.e., epidemiology studies) as support for the opinion, the court may not second guess that reliance or independently analyze the data to evaluate whether it reasonably supports the expert’s conclusion.
The court rejected plaintiffs’ “overly narrow conception of the inquiry and examination” permitted under California law, reiterating that trial judges do have screening responsibilities under California Evidence Code section 801(b). That responsibility includes determining whether the substance of the supporting material provides a reasonable basis for the particular opinion, “or, on the other hand, [whether the cited support] reveals that the opinion is based on a leap of logic, conjecture, or artifi ce.” Id. at 772. The trial court had “carefully heeded the legal limitations on its role” in excluding the opinion testimony. It had also used a permissible approach, considering each category of data and then the data collectively, to determine whether it supported each opinion. The court again held that the expert’s reliance on a study evaluating the effects of exposure to multiple solvents did not provide reasonable support for the conclusion that any particular solvent within the study is capable of causing a disease. Plaintiffs’ argument that all organic solvents have similar properties and toxicities was not supported by the record.
The court agreed with plaintiffs, however, that an epidemiological study need not show a relative risk from exposure that meets or exceeds 2.0 (i.e., a doubling of the risk of the disease in the exposed population) to provide some support for a causation opinion. The court expressed no opinion “as to whether an expert opinion based solely on an epidemiological study showing a relative risk of less than 2.0 can be suffi cient to support a fi nding of causation.” Id. at 775. Although the trial court erred in excluding the study from consideration based on the low relative risk, the error was not prejudicial, because the study of exposure to multiple solvents offered no support for causation opinions as to any particular solvent.
Plaintiffs argued that animal studies, case reports, and toxic registries and treatises generally provide reliable evidence of causation in humans, particularly when combined with other evidence. The court of appeal rejected this categorical approach and affi rmed the trial court’s more discriminating analysis. As to the animal studies, because of differences in toxicity arising from species differences and the high dose exposures typically used in animal experiments, the expert/ proponent must adequately explain why reliance on the particular animal studies is warranted. The court held that the reasonableness of relying on the animal data had not been adequately explained. Id. at 780.
Because case reports are not controlled studies, and frequently contain little or no analysis, they are ordinarily useful only to suggest a hypothesis of causation or association, not to establish one. As plaintiffs failed to demonstrate why the case reports were probative, the court of appeal held that the trial court did not abuse its discretion in fi nding that they were of little value and, even together with other cited evidence, failed to provide a reasonable basis for the causation conclusion. Similarly, statements concerning the toxicity and effects of chemicals at issue, found in the Registry of Toxic Effects of Chemical Substances, treatises, information cards and similar sources, did not reasonably support the expert’s opinion. These statements were not accompanied by supporting data, some did not distinguish transitory from permanent injuries, some considered the effects only possible, some sources failed to disclose the criteria employed for their listings, and some disclaimed any independent evaluation of the data. Because plaintiffs failed to show why the materials were probative, they failed to demonstrate an abuse of discretion in excluding the expert’s opinion. Id. at 781.
The court of appeal’s opinion in Lockheed II provided valuable guidance on key general principles governing the admissibility of scientifi c expert testimony, and on specifi c issues that occasionally arise in toxic tort cases. Now the Supreme Court has agreed, at last, to answer the question posed by the confl icting court of appeal decisions and to clarify the nature and extent of a trial judge’s responsibilities when asked to exclude purportedly unreliable scientifi c evidence.
The Lockheed case presents the California Supreme Court with issues of great importance to product manufacturers plagued with the injustices of junk science. The lower courts have interpreted Leahy to both endorse and prohibit reliability screening of the foundation of scientifi c opinion testimony. This is the Supreme Court’s opportunity to break that impasse and provide valuable guidance on these key questions.