The National Association of Scholars (NAS) welcomes the nomination of Congressmen Lee Zeldin to lead the Environmental Protection Agency (EPA), pending Congressional approval. We believe he will provide strong, reforming leadership for the EPA—which is in strong need of reform. The EPA, like the other federal agencies whose powers are informed by scientific research, does not operate according to the best scientific principles of transparency and reproducibility. We strongly urge Congressman Zeldin to make a priority of bringing best scientific practices to the EPA.
Too many activists debate science in relation to the policy outcomes such science might produce. These activists want more or less regulation of particulate matter emissions, or more or less money spent on “climate change,” and they simply subordinate proper scientific procedures to their desired goals. We think far more damage to EPA in practice has been done by progressive activists, who have embedded themselves successfully and with devastating effect within the EPA bureaucracy and throughout academia. To rebuild public trust, the EPA should expel these activists guiding science policy. The new reformers should seek to institute the best scientific procedures, without regard to the policy outcomes such science might recommend.
They should do so because science policy is uniquely vulnerable to the ambitions of radical activists precisely because of the criminally negligent scientific procedures fostered by deliberately lax federal policy. The irreproducibility crisis of modern science, fueled above all by pervasive politicized groupthink and by scientists’ shift to conducting research by means of culpably negligent statistical procedures, produces masses of false positive research results. Activist bureaucrats actively commission these false positive research results in a host of scientific and social scientific disciplines to justify the mass production of illiberal, radical regulations throughout the federal science regulatory agencies. Scientific procedures ought to restrain arbitrary, ideological policymaking, but the corruptions of politicized groupthink and misused statistics instead facilitate it.
The federal government, moreover, is the largest single funder of scientific research in the world, and federal funds not only distort American regulatory policy but also subsidize the wholesale production of irreproducible research in American universities.
NAS has documented precisely how the irreproducibility crisis has distorted government policy in four separate research reports. Our first report, PM2.5 Regulation (2021), focused on irreproducible research in the field of environmental epidemiology, which informs the EPA policies and regulations. The next three reports investigated the distorting effects on government policy of irreproducible research in nutritional epidemiology, which informs the Food and Drug Administration’s policies and regulations; mathematical modeling and public health policy, which informed the debacles of the Centers for Disease Control and Prevention and the National Institutes of Health in their COVID-19 pandemic response; and implicit bias theory, which informs a wide variety of state and local laws. The first report provides particularly strong justification for why the EPA should reform its scientific procedures, but the four reports collectively provide evidence that the federal government’s use of scientific research has been so systematically distorted that it should engage forthwith in systematic reform.
Therefore we urge Congressman Zeldin to institute reforms of EPA that embody the suggestions we made in our Model Science Policy Code. We wrote these as 10 statutes, to apply to every federal agency, but individual agencies can adopt their principles voluntarily. Seven of our ten model bills reform different aspects of federal policy that weaponize the irreproducibility crisis of modern science to justify illiberal regulatory policy. Three of these bills reform regulatory policy: Reproducible Policy Act, Mathematical Modeling Reform Act, and Weight of Evidence Act. Another four reform grants policy: Reproducible Grants Act, Replication Studies Funding Act, Negative Research Act, and Research Integrity Act. Three further bills reform the three other crises of federal science policy: Indirect Costs Act, Science Depoliticization Act, and Science National Interest Act. These three crises are each important, but each can be addressed by relatively simple legislation. All of these statutes can be transformed easily into EPA regulations.
These ten initiatives articulate reforms that should be enacted simply because they are good policy. Of course the EPA should require best existing practices to ensure reproducibility in science the government funds, or uses to inform policy. Of course universities should not be allowed to overcharge the EPA systematically. Of course EPA science grants should not be used to impose identity-group discrimination and DEI ideology on universities. Of course the EPA should work to prevent foreign intellectual espionage in our universities. Yet we forward these initiatives now not just because they are good policy, but because they are necessary. EPA policy in particular, as American science policy in general, has gone dreadfully wrong. These initiatives are the minimum necessary to right the EPA’s course.
In general, EPA regulations should foster transparency and reproducibility. EPA regulations also should require from academics the same level of rigor in their research as they require of corporations. Finally, EPA regulations should foster the rapid de-identification of personal data, so as to remove the confidentiality excuse used by politicized activists to bar full transparency for their research.
We also believe the EPA should not overturn previous regulations arbitrarily as it implements our recommendations. Regulatory stability is an important goal for the Federal government, and indeed for any system of laws and regulations. American enterprises have invested substantial resources in regulatory compliance, and their investments should not casually be set at naught.
Consider the Harvard Six Cities/ACS studies that are cited in support of current PM2.5 regulation. James Enstrom provided strong evidence that the 1995 ACS study only achieved statistical significance by “data gardening,” since not all the data that were available were used in the analysis. If that study is not reproducible, and recent negative studies are, then how should the EPA unwind its regulations?
We suggest a multi-part reform. The EPA should announce that it will cease using the Harvard Six Cities/ACS studies, and similarly irreproducible data sources, by some reasonably near date, unless the underlying data have been made publicly available. As the same time, the EPA should immediately begin to fund a high-priority program to create a new, substitute data set, with born-open, publicly accessible data and built-in de-identification to address any privacy concerns.
These data will then be available for the EPA to use once it ceases using the Harvard Six Cities/ACS studies and similarly irreproducible data sources. If the new data do not justify the regulations, then the regulations can be withdrawn in an orderly manner. If the new data do justify the regulations, then the regulations can be continued. This multi-part reform should maximize reproducibility reforms and regulatory stability.
Similarly crafted multi-part reforms, enacted throughout the EPA’s remit, ought to maximize the twin goods of good science and stable regulation.
We hope Congressman Zeldin will consider such reforms throughout the EPA. Ultimately, the EPA should use the best scientific procedures, no matter the policy results, to assure that all environmental regulations are based upon transparent, reproducible research.
Photo of Lee Zeldin by Gage Skidmore // Flickr // CC BY-SA 2.0
