During his first week in office, President Biden signed a memorandum on scientific integrity that launched a process to improve federal agencies’ scientific integrity policies. Although work on scientific integrity policies began during the Obama administration, the Trump administration’s  misrepresentation, distortion, and suppression of evidence on topics from climate change to COVID-19 made it particularly clear why strong policies to protect scientific integrity are needed. With the release of a new Framework for Federal Scientific Integrity Policies and Practices, the Biden administration has taken an important step toward establishing such policies.

The process that Biden laid out in his memorandum relies heavily on the White House Office of Science and Technology Policy (OSTP) and groups that it convenes. In June of 2021, OSTP requested information about the effectiveness of existing scientific integrity policies and good practices that federal agencies could adopt to improve scientific integrity. In January of 2022, the Scientific Integrity Fast-Track Action Committee of OSTP’s National Science and Technology Council published the Protecting the Integrity of Government Science report, which described ways in which existing policies were inadequate and identified promising practices. Where I and others who opined on the report felt it fell short was in describing specific incidents in which agencies or an administration violated scientific integrity — because to build a strong defense against potential attacks, looking at past attacks is instructive.

That report was only an interim step, though. We were looking ahead to a framework from OSTP that would guide agencies in the work of establishing and enforcing strong scientific integrity policies. In writing about the report, I expressed a hope that the framework would “provide a strong framework with concrete enforcement mechanisms to safeguard scientific integrity at federal agencies.”

OSTP has now released the Framework for Strengthening Federal Scientific Integrity Policies and Practices, from the Scientific Integrity Framework Interagency Working Group. I’m happy to say that in my view, it does provide a strong framework, although the enforcement mechanisms aren’t as concrete as what I hoped to see. As agencies use this resource to strengthen existing scientific integrity policies and develop new ones, I hope they’ll create robust provisions in some of the areas where the working group gave fewer specifics.

Good: SI definition and model policy

The framework contains several elements that will make it a helpful resource for agencies. First of all, the 2022 report noted that the federal government lacked a consistent definition of scientific integrity, so the framework provides one:

Scientific integrity is the adherence to professional practices, ethical behavior, and the principles of honesty and objectivity when conducting, managing, using the results of, and communicating about science and scientific activities. Inclusivity, transparency, and protection from inappropriate influence are hallmarks of scientific integrity.

Much of the framework document is devoted to something that several advocacy groups pushed for: a model scientific integrity policy that agencies can adopt and tailor to their needs. It makes strong statements about several areas where we saw problems during the Trump administration (and, in some cases, administrations that preceded it). Specifics include the following; each of these statements is preceded by “It is the policy of the [agency] to”:

• “Prohibit political interference or inappropriate influence in the funding, design, proposal, conduct, review, management, evaluation, or reporting of scientific activities and the use of scientific information.” This would apply to situations like the abrupt early cessation of Teen Pregnancy Prevention Program grants by the HHS Office of Adolescent Health or research related to fossil fuel extraction by the Department of the Interior as well as research conducted within agencies.
• “Ensure that scientific findings and products are not suppressed, delayed or altered for political purposes and are not subjected to inappropriate influence.” This would prohibit the kind of interference we saw from Trump appointee Michael Caputo into publications in the CDC’s flagship Morbidity and Mortality Weekly Reports, or the Trump administration’s suppression of the PFAS/PFOA draft risk assessment.
• “Permit, and even encourage, agency scientists to participate in communications with the media regarding their scientific activities and areas of scientific expertise. In communicating with the media, scientists are strongly encouraged to seek advice from [AGENCY] trained career communications experts.” The distinction between encouraging scientists to seek advice from communications experts vs requiring them to do so is important. During the Trump administration, employees of CDC and the US Geological Survey were told they needed to get permission before speaking with reporters. This issue crosses administrations; journalists have been raising concerns about public information officers blocking their access to experts for many years.
• “Ensure the quality, accuracy, and transparency of scientific information used to support policy and decision making including: … Reflect scientific information appropriately and accurately and ensure that it is free of misinformation…” I’d hope this kind of prohibition would prevent things like EPA’s refusal to include highly relevant research about health effects from its assessment of the pesticide chlorpyrifos.
• “Protect those individuals who report allegations of compromised scientific integrity in good faith, as well as those agency employees and other covered entities alleged to have compromised scientific integrity in the absence of a finding that the individual compromised scientific integrity, from prohibited personnel practices (as defined in 5 USC 2302(b)).” Rick Bright, whose whistleblower complaint states that he was reassigned after objecting to HHS moves to direct COVID-19 funds to unproven treatments that lacked scientific vetting, would probably like to see agencies do a better job of this.
• “Encouraging attendance and presentation of research at professional meetings including workshops, conferences and symposia.” During the Trump administration, EPA blocked scientists from giving planned conference presentations on climate change, while the Bureau of Land Management, Department of Energy, and US Geological Survey slashed employees’ attendance at major scientific meetings.

It’s a bit harder to see how this model policy would prevent two of the most egregious actions we saw from EPA under the Trump administration: a rule that used the pretext of advancing transparency to disregard high-quality research showing harms from regulated substances (i.e., the science-restriction rule, which I wrote about here, here, and here) and a policy of excluding EPA grantees from the agency’s federal advisory committees. Federal courts eventually scrapped both the science-restriction rule and the purge of EPA grantees from advisory committees, but I expect a future administration hostile to scientific integrity to try similar tactics.

Someone concerned about a future science-restriction rule could argue that “Reflect scientific information appropriately and accurately” means reflecting it *completely* and not adopting a policy that will lead to ignoring important studies. I’d feel more reassured if scientific integrity policies were to state explicitly that research can only be excluded from policy decisions for relevant reasons, such as characteristics that cast doubt on the reliability of its conclusions.

The model policy contains a whole section on federal advisory committees, which includes this: “The selection of members to serve on a scientific or technical FAC should be based on expertise, knowledge, and contribution to the relevant subject area.” Again, someone concerned about the exclusion of agency-funded scientists from FAC consideration could argue that using grantee status as a criterion isn’t allowed because it’s not part of this list, but I’d prefer to see an explicit prohibition against it.

Good: Equity in workplaces

In addition to the items above, the model policy includes the following (also preceded by “It is the policy of the [agency] to”):

Promote diversity, equity, inclusion, and accessibility in the scientific workforce and to create safe workspaces that are free from harassment and discrimination. Support scientists and researchers including, but not limited to, Black, Latino, and Indigenous and Native American persons, Asian Americans and Pacific Islanders and other persons of color; members of religious minorities; lesbian, gay, bisexual, transgender, and queer (LGBTQI+) persons; persons with disabilities; persons who live in rural areas; and persons otherwise adversely affected by persistent poverty or inequality; and advance the equitable delivery of Federal programs.

And the “Related Policies that Can Intersect with Scientific Integrity” section includes this:

Diversity, Equity, Inclusion, and Accessibility (DEIA) in Addressing and Strengthening Scientific Integrity and the Disproportional Impact of Scientific Integrity Policy Violations on Underrepresented Groups. Policies, practices, and agency culture to promote diversity, equity, inclusion, and accessibility in the scientific workforce and Federal workforce at large and to create safe workspaces that are free from harassment and discrimination are foundational for achieving a culture of scientific integrity. Because of existing power structures, racism, sexism, discrimination and other forms of bias in the workplace, scientific integrity and DEIA policies may intersect in many places. Similarly, scientific integrity entails greater transparency into research processes and policy-making outcomes. The agency will review and address potential scientific integrity policy violations that have a disproportionate impact on underrepresented groups or weaken the equitable delivery of agency programs.

It’s good to see the working group — and agencies that adopt this model policy — acknowledge that scientific integrity violations can have disproportionate impacts on groups that already face discrimination and various forms of structural disadvantage. And creating workplaces free from harassment and discrimination, while actively supporting researchers whose characteristics mean they have likely faced such barriers, will both create more equitable job opportunities and help our government attract a highly qualified scientific workforce.

The model policy also includes sections on protecting whistleblowers and on monitoring and reporting about scientific integrity activities and outcomes. Outside of the model policy, the framework gives an extensive list of metrics (plus sample measurement methods) for evaluating those activities and outcomes, and it identifies the items agencies must report to OSTP so it can assess agency progress.

Still to Come: Specifics on Consequences

To truly shape the behavior of employees and contractors, scientific integrity policies need strong enforcement mechanisms. The model policy states, “Violations of scientific integrity policies shall be taken as seriously as violations of government ethics rules and must come with appropriate consequences.”

The model policy assigns implementation roles and responsibilities to various individuals, starting with the agency head, who “[e]nsures all supervisors and managers comply with the scientific integrity policy and ensure accountability for those who do not” and “[s]upports and respects the scientific integrity official’s independence, recommendations and designation of and agency compliance with corrective scientific actions when violations of this policy are substantiated.”

Agencies have already been directed to designate Scientific Integrity Officials and Chief Science Officers, who must be career federal employees. The CSO “[e]nsures agencies establish as necessary clear administrative actions for substantiated violations of scientific integrity policies, designating responsibility for each aspect of accountability.” The SIO is responsible for overseeing implementation of scientific integrity policies, which includes receiving questions, concerns, and allegations and conducting an initial assessment of allegations. A note to agencies also suggests, “Agencies may want to consider including a description of the circumstances under which the Scientific Integrity Official and other scientific integrity points of contact may be removed from these roles.” The model policy’s “Ensuring Accountability” sections states that in consultation with the Office of General Counsel, it is the policy of the agency to:

Ensure that the Scientific Integrity Official, together with the other Agency scientific integrity officials, as applicable, drafts procedures to respond to allegations of compromised scientific integrity in a timely, objective, and thorough manner. These procedures shall include the following steps: an initial assessment and review, a fact-finding process, an agency adjudication or determination including description of remedies and preventative measures to safeguard the science, an appeals process, follow-up to track implementation of remedies, and reporting.

Remedies are important whenever scientific integrity violations occur; so are consequences. To prevent future wrongdoing, those found to have violated scientific integrity policy must also face meaningful sanctions. This is where I wish the model policy gave more specifics: What happens when allegations of scientific integrity violations are substantiated and the person found to have violated the policy has exhausted their appeals? Agencies will, of course, face constraints on the kinds of consequences they can impose, and will rightly be thinking of inevitable legal challenges from individuals who face sanctions. This will likely be one of the hardest elements of scientific integrity policies for agencies to get right. Having Congress pass a Scientific Integrity Act that allows for sanctioning those found to have violated scientific integrity policies would put agencies on firmer ground.

Next steps

A memo from OSTP Director Arati Prabhakar sets out a timeline for agencies: After the public posting of the framework, they have 60 days to submit new or updated draft scientific integrity policies to OSTP and the Subcommittee on Scientific Integrity for review. Those entities have another 60 days to review agencies’ drafts, using the assessment elements from the framework. Within another 30 days, agencies “are encouraged to publicly post a draft policy for review and comment” and “should provide an opportunity for public input on their scientific integrity policies and practices.” Final policies are due 90 days after that.

The framework was released on January 12, so that make March 13 the due date for draft policies, May 12 the deadline for the OSTP/Subcommittee review, June 11 the date by which agencies “should” provide an opportunity for public comment, and September 9 the due date for final policies. The model policy states that the agency will complete the development of procedures — “addressing scientific integrity concerns, handling differing scientific opinions, clearance of scientific products, scientific communications, authorship and attribution and other topics as needed” – within one year of the policy’s release, so that could mean waiting until September of 2024 to see what kinds of consequences agencies will establish in order to discourage scientific integrity violations. I’m eager to see what agencies commit to doing to advance scientific integrity across the federal government.