How federal regulators use—and abuse—science in the regulatory process has a profound impact on regulated businesses and consumers who purchase their products and services. In addition to the financial impact, every time that an agency forces science and the scientific process to serve its ideological or political agendas, rather than be guided by the neutral data, the public becomes less trusting of government pronouncements based on science. Below are some troubling recent examples of regulatory junk science. The first example demonstrates that protections against junk science do exist in the courtroom. The subsequent three examples reflect the lack of similar protections in the rulemaking and adjudication contexts.
Fourth and Sixth Circuits Slap-down EEOC. For the second time in less than a year, a federal appellate court has rebuked the Equal Employment Opportunity Commission (EEOC) for its use of junk science in accusing an employer of discrimination for conducting criminal background checks in its hiring process. EEOC’s litigation crusade against criminal background checks has faltered since its outset, with federal district court judges in Ohio and Maryland separately dismissing Title VII claims in 2013. Last April, just 20 days after hearing oral argument, the U.S. Court of Appeals for the Sixth Circuit affirmed the Ohio trial judge’s decision in EEOC v. Kaplan. The court found the EEOC’s statistical proof of disparate impact—compiled and presented by expert witness Kevin Murphy, an industrial psychologist—unreliable and “based on a homemade methodology” not generally accepted in the scientific community. A WLF Legal Opinion Letter and a WLF Legal Pulse post, both published last spring, offer more detail on the ruling.
Instead of cutting its losses after the Sixth Circuit decision, the EEOC doubled down in its Fourth Circuit appeal of the Maryland federal district court decision. The agency defended Mr. Murphy’s shoddy, results-oriented studies in EEOC v. Freeman as it did in Kaplan, and again it received a harsh judicial tongue-lashing. Judge Roger Gregory, writing for a unanimous panel, held that an “alarming number of errors and analytical fallacies in Murphy’s report” rendered the expert testimony unreliable under Federal Rule of Evidence 702.
Judge G. Steven Agee authored a concurring opinion to “address my concern with the EEOC’s disappointing litigation conduct.” The “missteps” of EEOC’s “expert” (Judge Agee’s quotation marks) included arbitrarily discounting from his analysis two-and-a-half years of criminal background inquiries that Freeman conducted—1,500 background checks.
Judge Agee added, “These problems would be troubling enough standing alone, but they are even more disquieting in the context of what appears to be a pattern of suspect work from Murphy.” In addition to the Sixth Circuit’s rejection of Murphy’s junk science in Kaplan, the judge cited a number of other instances over a ten-year period where federal courts found the industrial psychologist’s testimony unreliable. The EEOC’s persistent reliance on Murphy’s “expertise” earned it a stern warning from Judge Agee: “The Commission’s conduct in this case suggests that its exercise of vigilance has been lacking. It would serve the agency well in the future to reconsider how it might better discharge the responsibilities delegated to it or face the consequences for failing to do so.”
FDA and Food Ingredient Bans. The Food and Drug Administration (FDA) has proposed to remove partially hydrogenated oils (PHOs) from the list of food items it considers “generally recognized as safe.” In order to reach that conclusion in the midst of a downward trend in PHOs’ use in processed foods, FDA created a hypothetical “high-intake” consumer who “may still consume high levels of trans fat.” That consumer, in FDA’s scenario, would consistently choose “refrigerated biscuits, ready-to-eat frostings, certain brands of frozen pizzas, and certain brands of microwave popcorn.” In other words, FDA consumer is a mash-up of Charlie and the Chocolate Factory‘s Augustus Gloop and Homer Simpson. To add insult to scientific injury, FDA has refused to publicly release the specific details of how it designed this hypothetical, i.e. the foods it used and the exact amounts of trans fat in each. FDA also padded its estimates of consumption of trans fat from PHO by assigning a value of .4 grams of PHO to every food product that can, under FDA regulations, declare zero grams of trans fat on the Nutrition Fact label. Thus, with FDA’s thumb-on-the-scale approach, .0000001 g = .4 g.
EPA and Climate Change Regulations. Businesses and other affected stakeholders lodged a significant number of complaints with the Environmental Protection Agency (EPA) regarding the agency’s use of scientific studies leading to its 2009 “Endangerment Findings” on greenhouse gas emissions. EPA uniformly denied those objections, and the resulting findings inspired new costly regulations. In a February 13, 2015 WLF Working Paper, Revitalizing the Information Quality Act as a Procedural Cure for Unsound Regulatory Science: A Greenhouse Gas Rulemaking Case Study, attorney Lawrence Kogan details EPA’s extensive violations of the Information Quality Act (IQA) committed during the Endangerment Findings process, including unacknowledged conflicts of interest, flawed scientific peer review, and inadequate treatment of requests for correction of data.
Phthalates and the CPSC. Finally, the Consumer Product Safety Commission (CPSC) is stage-managing its scientific inquiries to advance the Commission’s campaign against phthalates, a group of chemicals that help make plastics pliable. A 2008 federal law banned three phthalates and temporarily banned three others pending the results of a Chronic Hazard Advisory Panel (CHAP) convened by CPSC. The CHAP issued its findings late last year, which utilized some very curious science.
For example, as part of estimating current exposure levels to phthalates, the CHAP used a data set from the time period when the permanently-banned phthalates were still in use. Unsurprisingly, that data set showed significant exposure to those phthalates, but much lower levels of exposure to the temporarily banned ones. Rather than follow the data and conclude that the problem was with the permanently banned phthalates (and not the temporarily banned ones), the study used the fact that the permanently-banned phthalates overwhelmingly contributed to cumulative exposure levels to recommend keeping the temporary ban in place too. In other words, the CHAP recommended permanently banning the temporarily-banned phthalates too, even though the data only showed unacceptable exposure to phthalates that have already been permanently banned. And it did so based on cumulative risk assessments that cannot possibly be true any longer in the wake of the permanent ban.
Greater Scrutiny Needed for Regulatory Junk Science. In the 1993 decision Daubert v. Merrell Dow Pharmaceuticals, the Supreme Court confirmed that federal judges must ensure that an expert witness is qualified in the relevant field, the proposed testimony is relevant, and the expert’s methodology is reliable. The EEOC’s enforcement targets in Kaplan and Freeman prevailed because federal judges took their role as scientific reliability gatekeepers seriously. But such standards do not check regulatory junk science outside of court because Daubert and the Federal Rules of Evidence don’t apply to agency rulemaking or adjudication.
WLF has long argued that they should. After all, § 556(d) of the Administrative Procedure Act (APA) dictates that no “sanction may … be imposed” or “rule or order issued” unless “supported by and in accordance with,” inter alia “reliable” evidence. That language has regretfully not prevented agencies from bending, massaging, or completely ignoring science in their decision-making. An Executive Order compelling agencies to apply Daubert-like relevance and reliability standards could positively reinforce that APA provision. We don’t, of course, expect the current Executive Branch to embrace that approach.
In the meantime, we hope that entities negatively affected by regulatory junk science will consider bringing the type of legal action laid out in the WLF Working Paper referenced above. The IQA has heretofore not provided a forceful, substantive check on agencies’ refusal to conduct proper peer reviews, avoid conflicts of interest, or adequately evaluate challenges to flawed government science. The alternative approach our paper suggests, however, may change that.
U.S. Court of Appeals for the Seventh Circuit Judge Richard Posner wrote in a 1996 opinion, Rosen v. Ciba-Geigy Corp., that “Law lags science; it does not lead it.” The same principle applies with equal, if not greater, force in the context of government regulation.
Also published by Forbes.com at WLF’s contributor page