Protecting the Integrity of Federal Science
Until minimum elements of a scientific integrity policy become law, the American public must rely on each successive administration to ensure that federal decisionmaking is based on rigorous and independent research.
In May 2020 Rick Bright, the scientist directing the Biomedical Advanced Research and Development Authority at the Department of Health and Human Services, filed a whistleblower complaint alleging he faced retaliation. That retaliation, the complaint claimed, came in the form of being transferred to the National Institutes of Health after he “insisted on scientifically-vetted proposals, and … pushed for a more aggressive agency response to COVID-19.”
That same month, the inspector general of the Environmental Protection Agency (EPA) published an internal survey from 2018 about the agency’s scientific integrity policy. Although a majority of EPA employees said they were satisfied overall with the implementation of the policy, 59% were dissatisfied with the agency’s culture of scientific integrity. In addition, of those employees with a basis to judge, 21% (368) expressed concern over reporting of their research findings without alteration or suppression. Nearly 400 employees said they had experienced, but did not report, potential violations of the policy due to fear of retaliation and the belief that reporting would make no difference. They were not outliers: 42% of the EPA employees surveyed who were involved in science said they would not feel comfortable reporting a potential violation of scientific integrity.
Although breaches of scientific integrity during the Trump administration were widely reported, concerns about the issue go beyond one administration, or one party. In response to concerns about scientific integrity under the George W. Bush administration, the Obama administration directed attention to the issue. But also under Obama, a judge ruled that actions by then secretary of health and human services Kathleen Sebelius to revoke a decision by the Food and Drug Administration on contraception access was “politically motivated, scientifically unjustified, and contrary to agency precedent.” And integrity concerns at EPA’s Office of Pollution Prevention and Toxics and Office of Pesticide Programs have continued in the current administration.
President Biden’s Memorandum on Restoring Trust in Government through Scientific Integrity and Evidence-Based Policymaking, which he signed during his first week in office, builds on the last time that federal science integrity received presidential attention. The Obama administration undertook similar efforts, beginning with a 2009 presidential memorandum assigning John Holdren, the director of the White House Office of Science and Technology Policy (OSTP), with “ensuring the highest level of integrity in all aspects of the executive branch’s involvement with scientific and technological processes.” Holdren followed up in 2010 with some general guidance but left the details up to the agencies.
President Biden’s memorandum is broader in scope and more detailed. It includes more agencies and applies to all federal employees and contractors. The president directed agencies to designate a scientific integrity official (SIO) to oversee implementation and improvement of scientific integrity policies. The memo also directed OSTP to convene a government-wide Task Force on Scientific Integrity to conduct the first ever comprehensive assessment of scientific integrity policy and practices in the federal government, and to develop a framework for regular assessment and iterative improvement of agency scientific integrity policies and practices.
Although the task force’s January 2022 report, “Protecting the Integrity of Government Science,” overall constitutes a good start in addressing its charge, it does not go far enough in critical areas. Moreover, the task force still needs to develop a framework, and OSTP recently put out a request for information to receive input on what should be included.
There are several steps the administration should further take to strengthen federal scientific integrity, and the framework—if developed properly—can be a valuable tool. In the long run, federal science will only be protected if senior leaders are held accountable and all employees feel comfortable filing reports of possible violations. However, for federal scientific integrity policies and practices to really work, rather than being dependent on the attitude and commitment of each successive administration, they must have the full force of law behind them.
A Close Look at “Protecting the Integrity of Government Science”
Members of the Task Force on Scientific Integrity were given three main tasks. First, they were told to assess whether existing scientific integrity policies in federal agencies prevented political interference. Second, they were asked to analyze any instances in which scientific integrity policies have not been followed or enforced. And third, they were tasked with identifying effective policies and practices that protect scientists’ independence. The number of agencies included was vast, as evidenced by the task force’s membership: 57 representatives from 29 agencies.
The task force report’s findings on the first task—assessing whether existing scientific integrity policies in federal agencies prevent political interference—highlight the need for agencies to strengthen their scientific integrity policies to deter political interference. The report thoroughly and comprehensively documents the types of scientific integrity problems that can occur in conducting scientific research, managing science, communicating research results, and making agency policy decisions. After reviewing the full range of problems, the task force concluded that although agencies were well equipped to deal with research misconduct and security breaches resulting from the actions of individual scientists, they lacked procedures to address scientific integrity violations by senior leaders.
Regarding the second objective of analyzing instances in which scientific integrity policies were not followed or enforced, the report falls far short. The task force focused on only two examples, writing that it was “neither charged nor equipped to address new allegations of scientific integrity violations.” The report does not include any of the highly publicized examples of political interference in communicating the public health threats from COVID-19 and climate change. Instead it highlights political interference in the Hurricane Dorian forecast by the National Oceanic and Atmospheric Administration and in the decennial census questions by the Census Bureau. By focusing only on these two examples, the report fails to convey the magnitude or seriousness of instances when science was censored, manipulated, and hindered. Compiling and analyzing a complete record of scientific integrity violations is a critical step toward recommending procedures that will prevent such violations from happening in the future. Appendix C of the report provides an excellent list of the ways scientific integrity policies can be violated, and the task force should have highlighted at least one example for each type of violation.
For the final task of identifying effective policies and practices, although the task force provided general recommendations, these do not go far enough in specifying either a minimally acceptable scientific integrity policy or appropriate enforcement of such a policy. The report rightly states that a one-size-fits-all scientific integrity policy won’t work: policies must be tailored to the specific mission of each agency while adhering to common principles. This is true, but there are certainly minimum standards that could be laid out. With regard to enforcement, the report recommends that “violations of scientific integrity be considered on par with violations of government ethics, with comparable consequences,” but it provides no details on the consequences or who should decide or implement those consequences. And there are no recommendations on how to address violations from senior leaders, even though the task force report identifies this as the most critical problem.
Further Strengthening Federal Scientific Integrity
After the report, the task force’s next step will be to develop a framework that must, according to the presidential memorandum, “inform and support the regular assessment and iterative improvement of agency scientific-integrity policies and practices, to support the Director and OSTP in ensuring that agencies adhere to the principles of scientific integrity.” The framework represents an opportunity to flesh out the details—in particular, to strike the appropriate balance between government-wide and agency-specific definitions and policies, and to include enforcement policies with teeth.
Defining terms is key to ensuring common understanding, and the federal government needs to adopt a standard definition of scientific integrity. The task force, in its report, noted this, stating that “some, but not all, agencies provide definitions of scientific integrity in their scientific integrity policies. These definitions vary across agencies and would benefit from greater harmonization. The Task Force intends to produce a definition of scientific integrity for adoption by Federal agencies as it develops a framework for assessing scientific integrity policies.”
The task force should consider adopting the definition that the Climate Science Legal Defense Fund (CSLDF) uses in its model scientific integrity policy for agencies, universities, and other research institutions. The CSLDF defines scientific integrity as “a condition that exists when scientific research and publication of the results thereof adheres to the accepted standards, conduct, and professional values of the relevant scientific community. Adherence to these standards is designed, insofar as possible, to ensure objectivity, clarity, reproducibility, and utility of science and scholarly activities and assessments and to prevent bias, fabrication, falsification, plagiarism, outside interference, censorship, and inadequate procedural and information security.”
To go along with this standard definition, the framework should present a default scientific integrity policy that would be in effect until an agency replaced it with its own customized policy approved by OSTP. This default policy should include the minimum elements of an approvable policy: the standardized scientific integrity definition, a process for reporting problems to someone with the authority to investigate, a range of specific consequences for those found in violation of the policy, and identification of who should decide those consequences.
This default policy should also apply to all interagency work done by groups, such as the US Global Change Research Program, to ensure consistency for participating agencies. To promote development of customized policies, the framework should describe how to enhance these minimum elements to develop exemplary scientific integrity policies and practices. The task force should identify best practices from current agency policies as recommended enhancements and commit to a periodic evaluation of their effectiveness.
The other critical area is enforcement. Scientific integrity violations are a misuse of critical, costly government resources. They can disrupt policymaking and could ultimately harm people. The framework should specify serious consequences for scientific integrity violations. Assigning roles and responsibilities is key—for reporting as well as for who investigates, decides, and implements the consequences for each type of violation. The framework should clearly delineate the responsibilities of the SIO, chief science officers, inspectors general, general counsels, and agency managers.
The framework should also explain how whistleblower protections can be used for scientific integrity complaints if an agency’s scientific integrity process fails to investigate and remedy allegations of violations. The Whistleblower Protection Act protects federal employees who disclose illegal activities or instances of fraud, waste, and abuse in the federal government. Violations of scientific integrity involving intimidation or coercion, obstruction and interference, and immunity from consequences are also violations of the act. Before scientific integrity policies were established, employees could only rely on whistleblower protections to address such violations, and many feared taking on that protracted legal process. Agencies must make their scientific integrity policies an effective alternative to the whistleblower route. This requires developing a culture that encourages employees to report possible violations without fear of retaliation. It also requires an expeditious process to investigate those reports and approaches for correcting the scientific record and holding accountable those who knowingly violated scientific integrity policies.
One particularly difficult area is allegations of scientific integrity violations by high level officials, which should be handled outside the agency involved. The framework should call for the establishment of an interagency Scientific Integrity Council formed by all the SIOs to assess, investigate, and adjudicate such allegations.
Although a federal scientific integrity framework is a necessary step forward, even if it is perfectly designed, its implementation will depend on the commitment of the executive branch. In the long run, federal science will only be protected if policies and practices are backed up by the full force of law. One approach that could give these policies permanent standing is H.R.849, the Scientific Integrity Act, which would require agencies that fund, conduct, or oversee scientific research to adopt and enforce scientific integrity policies with specific protections for scientists. Until minimum elements of a scientific integrity policy become law, the American public must rely on each successive administration to ensure that federal decisionmaking is based on rigorous and independent research that is free from suppression, manipulation, and other interference.