A proposed rule by the Environmental Protection Agency (EPA) that allegedly aims to strengthen transparency in regulatory science suggests that science is broken. It isn’t.
We know it works because we can see the life-saving transplant technologies, hurricane forecasts, new medications, pest-resistant crops, and countless other breakthroughs that exist because of science. This discipline isn’t perfect, but it is the best tool available to safeguard the planet and its people.
Last year, the EPA proposed a rule requiring that scientists disclose all raw data before any study conclusions would be considered. The rule, titled “Strengthening Transparency in Regulatory Science,” would apply retroactively to regulations already in place. It would make it harder to enact new regulations, because many studies from the past rely on personal medical information that was collected under confidentiality agreements and include consensus from reports that may not have shared all of the data according in ways compliant with the proposed rule.
The EPA has received nearly 600,000 comments on the proposed rule.
The House Committee on Space, Science, and Technology recently held hearings on the transparency rule and, in a bipartisan show of support, advanced out of committee the Scientific Integrity Act. This legislation would require all federal agencies that conduct or use science to set up systems to address — and better yet, prevent — attacks on science, like the EPA’s proposed rule. Please contact your representatives in Congress to support the Scientific Integrity Act.
Scientists like us have been raising our voices in opposition to a rule rooted in the misconception that science leaps to conclusions based on weak evidence in single papers. In reality, science works because it produces ideas that are continually tested under tremendous critical scrutiny over long periods of time.
We are writing not just for ourselves but on behalf of other leaders in open and reproducible science practices: Todd Vision, professor and founder of the Dryad Digital Repository; Laurie Goodman, editor-in-chief of the open-access journal GigaScience; Elizabeth Iorns, CEO of Science Exchange and the codirector of the Reproducibility Initiative; Carly Strasser, director of alliances and data strategy at the Fred Hutchinson Cancer Research Center; and Mark Johnston, editor-in-chief of the journal Genetics and former chair of the department of biochemistry and molecular genetics at the University of Colorado.
We are a group of scientists dedicated to accelerating discovery and improving the rigor, transparency, and reproducibility of research. So it may be surprising that we are writing in opposition, as others have done, to a rule whose title would lead one to think it’s good for science. We oppose it because it won’t do what it alleges to do: increase transparency or improve the health and safety of people or the environment.
Various stakeholders in science — including researchers, funders, publishers, private organizations, and various public groups — are constantly working to maintain and improve the integrity of scientific research. The proposed EPA rule may seem on the surface to support that goal. But it is instead a misguided proposal that undercuts the ongoing efforts of the scientific community in favor of dangerous vetting by a regulatory agency that will not have access to the best information.
Moreover, the Office of Management and Budget estimated that complying with the transparency rule would initially cost $250 million, and then between $1 million and $100 million a year, making it an expensive boondoggle that reduces the quality of information available to EPA policymakers.
Independent validation and replication of the research on which significant regulatory actions are based is a worthwhile goal. But that wouldn’t happen under the EPA’s rule. Research journals are increasingly implementing data availability policies that, unlike this rule, don’t impose onerous burdens on researchers and EPA administrators and that respect the need to protect the privacy of human subjects when sensitive public health data are involved.
In an editorial in PLOS Medicine, Stanford’s John Ioannidis writes, “If the proposed rule is approved, science will be practically eliminated from all [EPA] decision-making processes. Regulation would then depend uniquely on opinion and whim.” This would be devastating to the EPA and would harm Americans.
The EPA was established in the Nixon administration after the Love Canal environmental disaster, and was strengthened in the Reagan era. This supposedly non-partisan organization must rely on the best available research and minimal political interference to devise regulations that balance safety with vigorous commercial growth. When the EPA comes under political pressure, the consequences are severe, as when former EPA head Christie Whitman incorrectly testified about post-9/11 air quality in lower Manhattan. With the EPA responsible for regulations such as the Food Quality Protection Act, which in 1999 finally removed dangerous organophosphate pesticides from our food, political interference can cost lives.
The attention given lately to the challenges in repeating and extending results of research reported in individual published papers has given rise to new tools and practices to further strengthen the scientific endeavor. Over time, given the self-correcting nature of research, science uncovers the truth. While any single result may be difficult to reproduce, with evidence and consensus, novel claims are verified or refuted, and theories are confirmed or modified.
To be sure, there are countless ways to improve the research enterprise and to strengthen the rigor of individual research studies. Government agencies are in a position to play an effective role in encouraging these changes, but the EPA’s proposed transparency rule does nothing of the sort.
Reducing the EPA’s reliance on scientific knowledge will undermine science and have a catastrophic impact on the health of the country and its citizens.
Lenny Teytelman, Ph.D., is CEO of protocols.io in Berkeley, California. William Gunn, Ph.D., is director of scholarly communications at Elsevier. Joanne Kamens, Ph.D., is executive director of Addgene in Cambridge, Massachusetts.