EPA’s Scientific Integrity Policy is unscientific, lacks integrity

The Environmental Protection Agency recently released a draft update to its “Scientific Integrity Policy,” which aims to ensure the agency’s science-based decisions and research are impartial, empirically-derived, transparent, and based on sound data analysis and objective scientific evidence. The update comes in response to a White House framework from earlier this year that provides guidance to federal agencies on how to develop and implement scientific integrity policies.

Unfortunately, the EPA’s new draft policy contains a number of significant flaws that are likely to undermine scientific integrity rather than reinforce it, primarily by promoting partisan ideology over the pursuit of truth.

One area of concern relates to the overly-broad definitions contained in the document. For instance, the draft defines “scientist” as “anyone who collects, generates, uses, or evaluates scientific data, environmental information, analyses, or products.” This could refer to virtually anyone, from professional researchers to casual readers of science news.

The ambiguous language opens the door to inconsistent and arbitrary application of the EPA’s scientific integrity procedures. By essentially treating all career staff as unquestionable scientists, the policy is likely to discourage future administrations from conducting proper oversight, framing their involvement as “political” and career staff activities as “scientific.”

EPA staff are known to be overwhelmingly politically liberal. Thus, the policy conveniently shields their assumptions and decisions from reconsideration. This is despite the fact that many political appointees also have scientific backgrounds.

EPA’s update also incorporates factors like “diversity, equity and inclusion” and “indigenous knowledge” into the scientific process. While in principle there is nothing wrong with agency scientists gathering information from a diverse array of sources, what ultimately matters is the validity of knowledge itself, not the characteristics of those who share it. By promoting certain forms of knowledge as specially-valuable based on arbitrary social categories, EPA risks depriving its decisions of local knowledge beyond those types singled out for special consideration.

The draft also states that economic analyses like cost-benefit analysis are scientific documents, and that estimates in those analyses should not be changed “based on internal or external policy or political concerns.” However, weighing tradeoffs inevitably requires value judgments beyond just technical number-crunching. For example, assumptions about whose benefits and costs to consider, or how much relative weight to assign to those values, often reflect policy choices.

By prohibiting reconsideration of these analytical assumptions, the draft scientific integrity policy will allow unelected career staff to bake their own ideological views into the analysis, essentially binding political appointees to their judgments.

The prohibition on making changes based on external political factors would also seem to rule out any updates to economic analysis based on concerns raised during the public commenting process or the regulatory review process overseen by the Office of Management and Budget. The plain fact is that it is desirable for political officials to weigh in on policy decisions, given their proximity to voters.

In addition to what it does say, the EPA’s update is also notable for what it doesn’t say. The draft lacks provisions on how to guarantee independent validation of the studies and models EPA uses to justify its regulations.

Moreover, some of the agency’s most controversial rules are supported by studies whose data are not publicly available. The agency could do more to guarantee transparency in such cases. Otherwise, the public will have to trust the agency’s decisions on blind faith.

Already EPA actions taken during the Biden administration appear inconsistent with the principles it claims to espouse with respect to scientific integrity. Take the EPA’s recent update to its “social cost of carbon dioxide” estimate. The EPA had an opportunity to take a balanced review of the scientific literature and to make its methods transparent. Instead, the agency ignored scientific literature inconsistent with its preferred policy solutions, and its opaque modelling assumptions led to counterintuitive and unexplained results.

Similarly, though the new policy spends considerable time discussing ways to shield agency advisory boards from politics, upon taking office the Biden EPA took the unprecedented step of removing committee members from these boards who were appointed during the Trump administration. This is exactly the kind of aggressive political intimidation the EPA claims it now opposes.

Far from “restoring” trust at the agency, the EPA’s draft policy politicizes science. The broad and unclear language, the shielding of staff policy assumptions from questioning, and the insertion of arbitrary social justice criteria, all work to inject partisan ideology further into scientific practices at the agency.

To the extent possible, science and policy should be kept separate. But to achieve this, EPA needs to recognize that much what the agency currently calls science is little more than politics dressed up as science.

(To read my full comment on the EPA’s draft Scientific Integrity Policy, signed by a coalition of 20 groups, go here.)