Given the outsized role that IARC can sometimes take in litigation and regulation, lawyers and judges should pay some attention to, and give some critical thought about, how the Reference Manual on Scientific Evidence addresses IARC’s classifications and evaluations of putative carcinogens.
The Reference Manual is published by the Federal Judicial Center and the National Academies of Science, Engineering and Medicine to remedy the gaps and defects in the knowledge base of judges who must conduct trials and gatekeep expert witness opinion testimony on scientific issues. The third edition of the Reference Manual on Scientific Evidence was published in 2011; a fourth edition was released in the still of the night on the last day of the last month of 2025. Given the imprimatur of its publishers, the Reference Manual can be influential. Whether its influence is deserved or warranted is very much a matter of debate.
Without any citation support or analysis, the third edition of the Reference Manual accorded IARC undeserved honorifics and accolades. Various chapters of the third edition referred to IARC as “well-respected and prestigious,” “well regarded,” and “reputable.”[1]
The late Professor Margaret Berger, in her opening chapter of the third edition, misleadingly characterized IARC’s review processes as “not review[ing] each scientific study individually for whether it reliably supports the causal claim being advocated or opposed.”[2] Berger’s chapter was published after her death, and it cited cases that were decided after her death. The error noted above cannot be fully attributed to her; as with the posthumous case references, other parts of her chapter may have been introduced gratuitously by an overzealous editor. The gist of the mistaken representation of IARC’s process, made without any reference to the Preamble, is, however, consistent with Professor Berger’s lifetime publications.
Berger’s peculiar, erroneous take on the IARC process appears to accord with her personal belief that courts should not look at the validity of individual studies. According to the Preamble, however, IARC’s working groups are very much supposed to review individual epidemiologic studies for whether they reliably report an association, or animal studies for whether they support a causal interpretation (for the animals in the study).[3]
The new, fourth edition of the Reference Manual[4] continues the tradition of giving IARC a higher regard than the evidence would warrant. The new edition’s chapter on toxicology, for instance, carries forward the characterization of IARC as “respected.”[5] Readers will have a difficult time of what to make of the chapter’s approbation, when the same toxicology chapter references IARC in connection with its discussion of risk assessments, even though IARC most decidedly does not conduct risk assessments.[6] Readers of the Manual may ask whether the authors of the Manual have any idea of what they are saying.
The toxicology chapter engages in some special pleading for animal evidence to remain within the scope of evidence for human health effects. In the context of cancer causation, this chapter points to IARC as a source for justifying the use of animal evidence, and even high-exposure, maximum-tolerated dose, rodent studies as evidence for human carcinogenicity.[7] Not surprisingly, the chapter glibly avoids addressing difficult questions about extrapolating from non-primate studies to conclusions of human carcinogenicity.
What is perhaps more unsettling is that the toxicology chapter epistemically trespasses upon the epidemiology chapter in order to settle a grievance of one of the toxicology chapter’s authors.[8] One of the toxicology chapter’s authors, Bernard Goldstein, fellow of the Collegium Ramazzini, was an expert witnesses for plaintiffs in a case that went up to the highest court in New York State, Parker v. Mobil Oil,[9] which affirmed the exclusion of Goldstein’s testimony. Footnote 37 in the toxicology chapter attempts to offer a defense of Goldstein’s litigation opinions in a way that hardly does justice to the decision of the New York Court of Appeals. More disturbing is that the chapter never discloses that Goldstein, one of the chapter’s authors, was a partisan expert witness in the Parker case. At the time Parker was decided, gasoline was an IARC 2B (“possible”) human carcinogen. Within a few years, in what may have been the result of advocacy groups, IARC reconsidered gasoline and bumped it up two categories to group 1, speaking through a working group that appeared to lack balance and credibility.[10]
Turning to the Manual’s epidemiology chapter,[11] we find again an uncritical and unscholarly description of IARC monographs as “well regarded,” and an assertion that courts “generally recognize” IARC monographs as “authoritative.”[12] The authors offer no support for the former claim. In support of their latter assertion, the chapter cites to a rather dubious appellate decision, which affirmed a trial court’s ruling that an expert witness was permitted to reference the IARC 2A classification of glyphosate in his testimony.[13] The case cited by the Manual never held that the IARC monograph at issue was “authoritative”; indeed, the court never used the word authoritative at all. The two law professor authors of the epidemiology chapter know, or should know, that “authoritative” is a term of art in the law of evidence. Federal Rule of Evidence 803(18), for instance, makes statements made in a so-called learned treatise admissible in evidence if “established as a reliable authority.” The glyphosate case cited by the Manual never suggested that there had been a determination that the glyphosate IARC monograph, or the IARC classification of glyphosate, was a reliable authority. There is a fair amount of learned opinion that IARC badly bungled its evaluation of glyphosate.
IARC did, in fact, controversially, classify glyphosate as 2A, or “probably carcinogenic” to humans. “Probably,” as used and defined in the key IARC document, the Preamble,[14] has no quantitative meaning according to IARC. A more accurate translation into common parlance would have us say that IARC classified glyphosate as “very possibly” carcinogenic to humans. The Manual’s chapter on epidemiology failed to cite the actual Preamble. Instead, the authors cited to a poster that fails to give the relevant definitions.[15] This abridgement distorted the presentation of IARC ideas, such as they are, and thwarted critical scrutiny of the IARC process.
The epidemiology chapter affirmatively misrepresents the work of IARC working groups and their classifications by describing IARC as convening working groups and asking them to reach “consensus” on carcinogenicity. In fact, the working group, and its subgroups, treat a bare majority as sufficient.[16] The Manual also describes IARC as employing a “weight-of-the-evidence” approach, a descriptor so subjective and vague as to be unfalsifiable.
There are many reasons that IARC monographs and classifications should not be well regarded. For a fuller discussion of those reasons, see the recent paper published by the Washington Legal Foundation, on IARC’s Precautionary Science: How the WHO Cancer Research Agency Misinforms Regulation and Litigation.[17]
[1] National Academies of Science, Engineering & Medicine and Federal Judicial Center, REFERENCE MANUAL ON SCIENTIFIC EVIDENCE at 20, 564 n.46, 646 (3d ed. 2011).
[2] Margaret A. Berger, The Admissibility of Expert Testimony, in REFERENCE MANUAL, id.
[3] IARC MONOGRAPHS ON THE IDENTIFICATION OF CARCINOGENIC HAZARDS TO HUMANS – PREAMBLE (2019) [cited herein as Preamble], https://perma.cc/LJE4-K7HH
[4] National Academies of Sciences, Engineering, and Medicine & Federal Judicial Center, REFERENCE MANUAL ON SCIENTIFIC EVIDENCE (4th ed. 2025) (cited as RMSE 4th ed.).
[5] David L. Eaton, Bernard D. Goldstein, & Mary Sue Henifin, Reference Guide on Toxicology, 1027, 1045, in RMSE 4th ed.
[6] Id. at 1036 (citing Hardeman v. Monsanto Co., 997 F.3d 941 (9th Cir. 2021) for its affirmance of a lower court’s allowing an expert opinion to rely upon and discuss at trial the IARC classification, not its risk assessment, of glyphosate as a “probable carcinogen”).
[7] Id. at 1044.
[8] Id. at 1044 n. 37.
[9] Parker v. Mobil Oil Corp., 7 N.Y.3d 434, 857 N.E.2d 1114 (2006).
[10] Craig Dillard & John Kalas, “IARC Classifies Gasoline As a Human Carcinogen – Litigation May Follow,” Nelson Mullins (April 15, 2025).
[11] Steve C. Gold, Michael D. Green, Jonathan Chevrier, & Brenda Eskenazi, Reference Guide on Epidemiology 897, in RMSE 4th ed.
[12] Id. at 919 n. 68.
[13] Hardeman v. Monsanto Co., 997 F.3d 941, 967 (9th Cir. 2021), cert. denied, 142 S. Ct. 2834 (2022).
[14] Preamble, https://perma.cc/LJE4-K7HH
[15] Reference Guide on Epidemiology, RMSE 4th ed. at 919 n. 68 (incorrectly citing the Preamble to a URL, https://perma.cc/XXY4-7DKF, which is a one page abridgement of the actual Preamble, devoid of definitions and qualifications.
[16] Reference Guide on Epidemiology, RMSE 4th ed. at 974.
[17] Schachtman, IARC’s Precautionary Science: How the WHO Cancer Research Agency Misinforms Regulation and Litigation, Washington Legal Foundation Monograph (May 2026), available at https://www.wlf.org/wp-content/uploads/2026/05/05-26Schachtman-Monograph.pdf
