Eight reasons the public health and scientific communities oppose EPA’s continued push with its controversial “transparency rule”

The Environmental Protection Agency (EPA) is pursuing its controversial 2018 “transparency rule” with a supplement expected as early as this month. Garnering a whopping 600,000 comments on its initial release in April 2018, the Strengthening Transparency in Regulatory Science proposed rule requires EPA actions be based only on science for which the underlying data and code are publicly available.

The scientific and public health community was quick to react to the news with opinion pieces about why EPA should not proceed:

• Researchers must unite against US environment agency’s attack on scientific evidence, editorial, Nature, November 20, 2019.
• Getting the EPA back on track, Marie Lynn Miranda, guest editorial, Science, December 6, 2019.
• Joint statement on EPA proposed rule and public availability of data (2019), editors-in-chief of five major scientific journals, Science, December 6, 2019.
• The EPA’s proposed ‘transparency rule’ will harm health, safety, and the environment, leaders in open and reproducible science practices, opinion piece, STAT News, December 9.

A previously scheduled congressional hearing of the House science committee, Strengthening transparency or silencing science? The future of science in EPA rulemaking, took place two days after the November 11 New York Times article reporting on a leaked draft of the supplemental rule. All of the five non-governmental witnesses expressed concern if not opposition to the transparency rule in their testimony. More recently, the EPA’s own science advisory board, in a 12/31/19 document “finds that key considerations that should inform the Proposed Rule have been omitted from the proposal or presented without analysis, and certain key terms and implementation issues have not been adequately defined or described.” They also express “concerns about the scientific and technical challenges and feasibility of implementing some requirements of the Proposed Rule.”

Most if not all the pieces listed above call for action by community in opposition. In addition, without exception, they support making the scientific process as open as transparent as possible. 

The opposition from the public health and scientific communities—both in the last month and in 2018, including from the American Statistical Association (ASA)—generally fall into one of the eight following categories:

1.     Proposed rule offered under false pretenses: Many say the EPA public justification for the proposed rule—to increase transparency in science and public access to scientific data underlying EPA decisions—is not the actual intent. The editors-in-chief of five major scientific journals wrote, their original statement “reflected alarm that the proposal’s push for ‘transparency’ would be used as a mechanism for suppressing the use of relevant scientific evidence in policy-making, including public-health regulations.” The leaders in open and reproducible science practices stated, the EPA’s proposed rule “allegedly aims to strengthen transparency in regulatory science suggests that science is broken. It isn’t.” The Nature editorial asks several questions about EPA proposing the rule, including why “EPA’s leaders choose to override their own science advisers.” They addressed the questions by writing, “answers might be found by considering the rule in the context of the wider actions of the administration of President Donald Trump on the environment so far” and noting this Administration’s cancelling the Clean Power Plan, withdrawing from the Paris climate agreement, weakening fuel-efficiency standards, and cutting back on environmental research. Two Washington Post op-ed authors refer to EPA’s transparency effort as a “weaponization of transparency”. They say the EPA borrowed from the tobacco and fossil fuel industries work taking “advantage of transparency’s sterling reputation to advance an anti-regulatory agenda under its banner.”

Dr. Todd Sherer, CEO of the Michael J. Fox Foundation for Parkinson’s Research, testified as follows to Congress:

"As overall justification for the proposed rule, the EPA claims it is following the accepted practice of many science organizations, including many scientific journals. However, we believe this is a misleading claim. Major journals in the field only require data be made confidentially available to other researchers for the purposes of reproducing or extending analysis. No major journal requires scientists to publish raw data to the public in all cases."

Dr. Mary Rice, chair of the American Thoracic Society Environmental Health Policy Committee stated this point in the following words:

"Instead of promoting transparency in regulatory science, this rule would decrease transparency, by giving the EPA administrator unchecked authority to pick and choose which research studies will inform policies that affect the health of the U.S. population."

This point was not part of ASA’s response to the 2018 EPA transparency rule. However, It is congruent with my experience with the House bills widely believed to be the precursor to the EPA proposed rule: the Secret Science Reform Act of 2014, the Secret Science Reform Act of 2015, and the HONEST Act (2017). Each of the three bills passed the House behind the leadership of then House science committee chair, Rep. Lamar Smith of Texas. As the first bill worked its way to full House consideration in fall 2014, the ASA and the greater scientific community—led by AAAS—expressed their strong concerns. Despite the scientific community’s reaction, the committee was quick to reintroduce a nearly identical bill in early 2015 for the new Congress without meaningful engagement with the scientific community. If the bill’s purpose is to strengthen the science underlying EPA decisions, it is reasonable to expect the committee to work with the scientific community to do so. This is for the same reason we would expect decisions to improve the U.S. military or the country’s medical practice be made with the input of military leaders or doctors, respectively.

2.     Proposed Rule excludes use of best science available and hampers the use of evidence: This point is present in nearly all the expressions of opposition from the scientific and public health communities. Dr. Linda Birnbaum, former director of National Institute of Environmental Health Sciences and the National Toxicology Program, said that, by requiring EPA to support actions with only science for which the underlying data is publicly available, the proposed rule “will prevent EPA from using the best available science in making policy”, and “will practically lead to the elimination of science from decision making.” In her Science guest editorial, Rice University Statistics Professor Marie Lynn Miranda concurred, “The rule could eliminate many public health studies from consideration.” David Allison, Dean of Public Health at Indiana University (IU) and representing the National Academies Committee on Reproducibility and Replicability in Science, agreed “that if reproducibility … becomes the sole and essential criterion for inclusions of data, then yes, such a requirement could risk that sound science would be excluded from EPA environmental and public health regulations.” Brian Nosek, the executive director for the Center for Open Science, testified the EPA proposed rule “had the negative quality of suggesting that evidence failing to meet these principles should not be used in policymaking” and concluded, “[t]his approach would degrade the quality of policymaking.” 

The editors-in-chief stated anything less than the most relevant science informing public policy “will harm decision-making that claims to protect our health.” The authors of the Nature editorial wrote the following:

"Many of the data that underpin public-health and environmental studies include information about people who will not have consented to disclosing their confidential data, including where they live; their travel habits; their age and gender identity; and the state of their health. Many such data were integral to the Six Cities study, published in 1993 by what was then the Harvard School of Public Health in Boston, Massachusetts… The results of the Six Cities study led directly to the imposition of life-saving limits on fine particulate matter from emissions. But this research would have been inadmissible under the EPA’s proposed rule."

Besides the private sector science just noted, science done internationally and for which the EPA has no authority to provide the underlying data could also be excluded.

3.     Proposed rule gives too much power to EPA administrator and introduces potential bias in the rulemaking process: Scherer of the Michael J. Fox Foundation for Parkinson’s Research makes the following point in this category:

"this proposed rule also grants too much power to one individual. As written, the proposal grants the EPA administrator broad authority to exclude individual studies. This could have wide-reaching impact depending on the preference of the administrator at the time and allows the administrator to overrule scientists regarding their own research. Allowing politically appointed officials to make decisions about whether a study qualifies for an exception is dangerous. The administrator already has broad authority to decide what action to take on an item. It should not have the power to hide evidence that does not support the action.

Dr. Rice of the American Thoracic Society (ATS) also made this point: “The ATS is concerned that delegating the discretionary authority solely to the Administrator would grant excessive authority to one person without accountability to the public. Allowing the Administrator to pick research in this way is secretive, and flies in the face of any transparent ethical process.”

Dr. Miranda of Rice University noted the possible introduction of bias through vested interests:

"The proposed rule claims that additional analysis of raw data and models will improve science. Who will do this analysis? Most likely, vested interests will finance work slanted toward a particular outcome, rather than undertake scientific inquiry without an agenda. For example, lead paint industry defense attorneys have attributed children's neurological deficits to landlord neglect and parental failure."

The ASA wrote on bias in its 2018 comments:

"Bias is likely to be introduced through the process of identifying what science can and cannot be used because there is not a bright line for what studies this proposed rule would include or exclude. Specifically, the requirement to make a decision for what constitutes “high quality studies” is problematic because the term is not defined in the proposed rule nor are operational criteria for its definition specified… The exclusion of otherwise valid and reliable research from consideration would also often result in regulatory analyses that suffer from biased selection and inadequate statistical power.

"The proposed rule’s statement, “EPA shall conduct independent peer review on all pivotal regulatory science used to justify regulatory decisions,” also risks the introduction of potential bias through the selection of the peer reviewers.”

IU Dean Allison also discussed the threat of bias in the rule:

"The likely quality of the outcomes as a result of the proposed rule would also depend upon the extent to which the request that underlying data be transparent and the studies be reproducible be implemented flexibly and in an unbiased manner or inflexibly or in a biased manner. Were a rule to be implemented that provided strong encouragement and incentives for making science reproducible and transparent, that would be good. In contrast, if such a rule became dicto simpliciter and a sole arbiter of whether information could be included, that would be bad. Certainly, the current EPA rule contains many situations in which exceptions can be made. That is wise. Yet what is unclear to me is whether the rule is necessary at all and, if it is valuable, how these exceptions will be adjudicated and whether the process of making them will lead to excessive use of time, excessive exclusion of studies, and potential bias in terms of which studies and datasets ultimately are allowed to be included."

4.     Individuals’ confidential data could be exposed: The EPA’s attempts to make data publicly available by anonymizing data could end up exposing confidential data, as noted by Dr. Scherer of the Michael J. Fox Foundation for Parkinson’s Research:

"The EPA’s rule puts these individuals at great risk of having their Parkinson’s or other diagnoses exposed. Such exposure could, for example, result in unfair job loss, which then causes loss of income, insurance, and other supports necessary to maintain quality of life."

Dr. Rice of the American Thoracic Society (ATS) also made this point, stating,

"EPA’s proposal to make the underlying data of policy-relevant studies fully available to the public in de-identified form risks disseminating the sensitive information about health problems, deaths and addresses of study participants. The long-term consequences of such a data breach could be devastating, not only for the study participants whose private health information is leaked, but also for the future of environmental health research."

Dr. Miranda of Rice University opened her Science guest editorial with a similar warning:

"Your information will be kept confidential, and the lessons learned from your participation will serve society—those are the promises made by researchers to participants in studies designed to inform environmental policies, from clean water and air to chemical exposure limits. The United States Environmental Protection Agency (EPA) may well break this fundamental pact next year, putting the agency at odds with its very mission “to protect human health and the environment.”"

 The reason for such a data breach risk is that any assurances to protect individuals’ confidentiality while anonymizing data for public release is likely over-promising. Rice’s Miranda wrote the following:

"In epidemiological and clinical studies, people provide information—their medical histories, behaviors, education, employment, and other personal details—under the condition that it will not be shared and their privacy will be protected. Anonymizing data is already difficult, if not impossible. With geographically referenced data, a capable programmer can leverage machine learning and brute computational strength to determine the location, and subsequently the identity, of a study participant. Similarly, facial recognition software has been applied to images reconstructed from cranial scans to identify study participants." 

The ASA in its 2018 comments, acknowledged “important advances have been made in the last decade in enabling protection of the confidentiality of data protections” but went on to write,

"However, advances in information technology, proliferation of linkable auxiliary data, and other factors make it increasingly difficult and sometimes impossible to ensure that confidential data are protected. A “one-size-fits-all” approach for access to regulatory data does not work."

 5.     Rescinding regulations may have detrimental health and environment effects: Numerous individuals and groups noted the point that the EPA proposed rule could harm public health and the environment. The leaders in open and reproducible science practices made the warning in this way: “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.” Scherer of the Michael J. Fox Foundation for Parkinson’s Research concluded the proposed rule “will compromise [EPA’s] mission to protect human health”, after explaining in detail how EPA protects public health, particularly as it relates to Parkinson’s disease. The authors of the Nature editorial concluded their piece with these sentences: “The EPA was created to protect the nation’s environment. As it approaches its 50th birthday next year, it must not be allowed to continue on a course of action that will weaken its ability to fulfil that role.” The editors-in-chief stated warned, “foundational science from years past—research on air quality and asthma, for example, or water quality and human health—could be deemed by the EPA to be insufficient for informing our most significant public health issues. That would be a catastrophe.”

The ASA’s 2018 comments made this point as follows:

"If some rules and regulations are rescinded due to failure to comply with the data requirements, health and environmental damage may occur. These rules and regulations were subject to public comment and were promulgated to promote human health and environmental protection. For example, according to the EPA, the Clean Air Act will prevent as many as 230,000 early deaths in 2020.[1] Rescinding such rules would result in loss of such health and environment benefits, which should be fully considered in a cost-benefit analysis. The failure to adopt new rules because of the faults in this proposed rule poses a similar risk to health and the environment."

6.     New studies/science may be discouraged: Several groups and individuals have pointed out the deleterious impacts the proposed EPA rule could have on science. Dr. Linda Birnbaum, former director of National Institute of Environmental Health Sciences and the National Toxicology Program, testified, “the impact of EPA’s proposed transparency rule will make it not only more difficult for human studies to be conducted ethically, but in many cases will make it impossible to use any information collected, not only prospectively, but looking back at the vast treasure trove of existing investigations.” Dr. Miranda of Rice University stated that re-identification, “will simply discourage people from participating in future health studies.” She further wrote, “Moreover, successfully recruiting and retaining participants depends on trusting relationships built on meaningful and sustained interaction between researchers and participants, especially with disadvantaged populations who are underrepresented in research. The EPA rule assumes that people will consent to their data residing in a repository where decisions about data use are made by persons unknown to them.”

The testimony of Dr. Scherer of the Michael J. Fox Foundation for Parkinson’s Research included this section:

"Chilling impacts to science

"If the EPA’s rule takes effect, it could introduce selection bias that may slow studies and alter results, and thereby affect regulatory decisions. Large-scale population studies rely on many people — often numbering in the thousands — to reveal sensitive or private information. These studies may have difficulty recruiting or retaining volunteers if the researchers are required to make de-identified data publicly available, as some may be more hesitant to share their information. Those who are willing to participate may be different from others, which could introduce confounding variables and bias that may question the study’s results."

 7.     Costs and Benefits are not fully considered: This point is well by Dr. Scherer of the Michael J. Fox Foundation for Parkinson’s Research:

"Even if there was an acceptable way to mask personal data while maintaining enough information to comply with the rule, costs of such anonymizing are unnecessary and expensive. When Texas Congressman Lamar Smith’s Honest and Open New EPA Science Treatment Act of 2017 (Honest Act) — a bill with content very similar to the currently proposed rule — was under consideration, the Congressional Budget Office estimated it could cost up to several million dollars a year to comply."

The ASA’s 2018 comments made this point as follows: “Cost is cited over a dozen times in the proposed rule, yet no explicit cost benefit analysis is provided to help understand how the benefits compare to the costs. A proposed rule of this scope should be informed by a cost benefit analysis that helps demonstrate that the benefits of its adoption outweigh the costs imposed.”

8.     Mechanism already exists: Dr. Rice of the American Thoracic Society (ATS) made this point as follows:

"An independent EPA-funded resource already exists to resolve scientific disputes

"The Health Effects Institute (HEI) is a research group funded equally by the motor vehicle industry and the U.S. EPA that has played a key role in resolving disputes about pivotal environmental health research that informs EPA regulation. For example, in July of 2000, the HEI conducted a re-analysis of two early air pollution studies: the Harvard Six Cities Study and the American Cancer Society Study, on the link between particulate matter pollution and mortality. The re-analysis was conducted by a team of independent scientists, and overseen by a broad board of stakeholders, and confirmed the findings of the original studies. In addition, to assess differences in industry versus academic/NCI analyses regarding risk estimates for diesel exhaust as a carcinogen, HEI held a public workshop and convened an independent panel of scientists. HEI then issued a report verifying the original findings in that scientific controversy, too.5 These past interventions by HEI to independently verify the data analysis for studies on controversial scientific issues provide a generalizable model to address the challenge of privacy vs. transparency in evaluating scientific research directly relevant to the regulatory process. Using the HEI to vet such key results is a practical and proven approach to address concerns raised about transparency, without compromising health data privacy of study participants."

Do these concerns resonate with how you believe the EPA should be fulfilling its mission using science? Assuming so, I echo the call of the authors above for a vocal response from the scientific and public health communities, and others. Besides responding to the expected call for public comments in the coming month(s), people should speak out through social media (including LinkedIn), op-eds, letters to the editor, and communications to Congress. I suggest using the hashtag, #ScienceIntegrity. If you have concerns other than those listed above, please share them with me, either by email or in the comment space of this article.