Kratsios Wants Rigor—But Where’s the Integrity?

The new OSTP science guidance gets the method right and the politics dangerously wrong

I was sitting at my desk, scrolling through the newly released White House Office of Science and Technology Policy (OSTP) guidance on Trump’s “Gold Standard Science” executive order, and I felt… uneasy. Not because the guidance was overtly hostile to science. It wasn’t. In fact, much of it read as technical, even reasonable. It urged transparency, peer review, reproducibility, and clearer communication of uncertainty. At first glance, it could almost pass as a straightforward, good-faith effort to promote better science in policymaking.

But let’s be clear: the guidance isn’t the problem. At least not entirely. In many ways, it aligns with practices that federal scientists were already following long before the executive order was signed. Open science. Data sharing. Providing clarity about uncertainty and adhering to good peer review practices. These aren’t controversial within the research community—they’ve been core values across agencies for decades. Federal researchers have been steadily working to improve transparency and reproducibility, as this is what good science requires. The guidance, in that sense, attempts to codify what was already underway.

But precisely because these are already well-established practices, the utility of this guidance is questionable.

The guidance doesn’t offer new tools or address systemic weaknesses—it simply restates principles that scientists are already upholding. What it leaves unaddressed are the actual threats to scientific rigor in federal agencies today. Audits from the GAO, the National Academies, and agency-level reviews have consistently shown that misconduct like fabrication or plagiarism is not what typically undermines the quality of federal science. The most significant threats stem from political interference in the scientific process and capacity constraints. When there aren’t enough scientists to carry out the work, or they’re forced to operate under impossible timelines, rigor suffers. This guidance doesn’t acknowledge those realities. It reinforces norms that already exist while ignoring the conditions that jeopardize them. In that sense, it’s not just insufficient. It’s essentially useless.

The more I read, the more that sense of unease settled in. Not because of what the guidance said, but because of what it left out. Though it avoids using the phrase “scientific integrity,” the guidance wades directly into that space, recasting integrity as a matter of technical rigor, rather than political independence. And while there’s little in the document that’s outright objectionable, the genuine concern lies in what’s missing.

There’s no mention of the central aim of scientific integrity policy: protecting science and scientists from political interference. No safeguards for independent scientific communication. No acknowledgment of the chilling effect that top-down control can have on evidence-based work. If this guidance is meant to supplement ongoing integrity efforts, its silence is troubling. But if it’s meant to replace them—as the broader context and rhetoric suggest—it fundamentally misrepresents what scientific integrity is intended to do.

This guidance wasn’t born of a years-long community effort. It wasn’t the product of a bottom-up push for better norms. It comes directly from an executive order signed by a president whose record on science is, at best, strained. And in that context, every line demands a second read.

During the Trump administration from 2017 to 2021, federal scientists were sidelined, reports were edited or suppressed, and scientific evidence was routinely ignored in policymaking. Scientific funding became politicized, and entire fields, such as climate science, environmental health, and reproductive health, were targeted. Those years left a lasting scar on federal science.

Now, we’re watching a familiar story unfold. In just the first five months of the current administration, billions in research funding have been slashed. Offices and departments dedicated to scientific research have been dismantled. Careers have been upended, not because the science was flawed, but because it was inconvenient. The message is clear: science that doesn’t align with political goals is once again under threat.

Worse still, the executive order that prompted this guidance does the opposite of promoting integrity. It calls on political officials—not career scientists—to determine what evidence should be used and how it should be presented. That’s not support for integrity; it’s a formalization of interference. And yet the guidance remains silent. That silence is not neutral. It’s a redirection. It trades in a broad, protective understanding of integrity for a much narrower focus on research practices. Essential practices, yes, but not sufficient on their own to truly ensure rigorous science informs policies.

This narrowing also shows up in how the guidance defines “rigorous” science. It treats experimental replication as the gold standard, with little consideration for the many other ways scientific knowledge is produced across federal agencies. Observational research, longitudinal surveys, and social science studies all contribute critical evidence, but they don’t fit neatly into the mold of “reproducibility.” Their rigor depends on careful sampling, a strong study design, and transparent analysis, rather than whether a study can be repeated under identical conditions. By emphasizing one model of science, the guidance marginalizes others.

What’s also missing is any recognition that science-based policy depends not on individual findings, but on the weight of evidence. It’s not one paper or one dataset that makes a case—it’s the convergence of multiple lines of research over time. That principle has long guided risk assessments, regulatory action, and public health guidance. And yet, it’s entirely absent from this document. That omission creates space for cherry-picking, where isolated results are elevated to support a political agenda while ignoring the broader consensus. And we’re already seeing that pattern. Climate science is being erased from official documents. Environmental health research is being dismissed. Equity-focused findings are being scrubbed. It’s not just about what science is produced, it’s about who decides what counts.

That’s why today’s op-ed from OSTP Director Michael Kratsios in Science defending the executive order was so disheartening. He described the scientific community’s response as “reactionary,” claiming critics were letting politics cloud their judgment. But he made no mention of the administration’s own political interference. No acknowledgment of gutted advisory committees, funding cuts, or retaliatory culture. No recognition that trust is earned, not demanded. When leaders ignore the political conditions shaping science and then ask the scientific community to meet them halfway, it’s not bridge-building. It’s gaslighting.

And yet, I’m not arguing that every line of the guidance is wrong. Federal scientists have long championed transparency, open data, and clear communication. Much of what the guidance outlines is already standard practice within agencies. If implemented in good faith, parts of it could reinforce improvements that scientists have been working toward for years.

But implementation is the question. Who will enforce this guidance? To what end? In whose interest? In this moment—when entire scientific offices are being dismantled, when grant programs are on the chopping block, and when science that contradicts political messaging is being actively suppressed—the context cannot be separated from the content. A seemingly technical document can still be politically dangerous if it redefines core values in a way that makes them easier to undermine.

So no, I’m not ready to throw the whole thing out. But I am paying close attention. Because in science, as in policy, what isn’t said often tells us the most. And right now, what this guidance doesn’t say speaks volumes.

That’s it for today - Thank you so much for reading SciLight!

If you enjoyed today’s post, please like it or share it with others. You can also support the work we do to shine a light on the politicization of science by becoming a paid subscriber!

If you want to share today’s post as a web page with your network, click this button:

If you have suggestions, questions, comments, or want to drop us a line - send it all to scilightsubstack@gmail.com