By Gretchen Goldman, Union of Concerned Scientists
Yesterday the Trump administration issued a supplemental notice to a draft rule long feared by many in the public health and scientific communities as a major disruption to science-based policy as we know it. Purported to be about transparency at the US Environmental Protection Agency (EPA), the proposal—in reality—would prevent the EPA from using the best available science to protect public health and the environment, as its mission holds.
The proposal is similar in substance to a version that leaked to the New York Times last November, with some new twists. Here are five initial takeaways.
1. The EPA’s use of science will be even more severely restricted
The supplemental proposal expands what was already a sweeping proposed rule. The new provisions clarify that the rule now applies to all “influential science” at the agency (not just science used in regulatory efforts) and that it applies to all data and models (not just dose-response models).
It also expands the definitions of data and reanalyze to be bigger and vaguer. In the draft rule, there were open questions about how extensive the rule would apply. With this supplemental notice, it is now crystal clear that the policy stands to fundamentally alter how the EPA uses science. From pesticide regulation to air pollution protections, chemical safety rules to water quality standards, the rule will hamstring the ability of the agency to protect people from environmental threats.
2. The restrictions are more complicated and confusing, but still limit science
The EPA offers two approaches for dealing with data that can’t be made public and both are problematic. First, they offer a tiered-access approach by which data that cannot be made public can be selectively shared for independent validation. There are many issues with this approach including legal restrictions, costs, and capacity; but most immediately, it is unclear what data would even fit this category. All the original confidentiality concerns about this rule still apply even if you are sharing with a select number of people. You cannot share personal health data even if you explain that you are only sharing it with a few trusted friends. Providing a tiered-access option doesn’t address the fundamental concern here: If the relevant health data and other confidential data sources must be kept private, the EPA won’t be able to use it.
The alternative approach offered is a weighted approach where the EPA won’t be prevented from using science that relies on nonpublic data, but those studies will be down-weighted compared to studies with entirely public data. This approach would unfairly and arbitrarily devalue scientific work that relies on things like personally identifying health data. At an agency charged with protecting public health, studies involving health data are especially crucial. Many of those crucial health studies will be needlessly downgraded in EPA decision-making under this proposal.
3. The goal of the rule is to question the science
Notably, in the new proposal, the Trump administration shows its hand. This was never about promoting transparency; it was always about finding ways for those outside of the scientific community to undercut the science supporting EPA decisions. We of course know this from tobacco industry documents from the 1990s, but this proposal makes clear that goal is about allowing non-scientists to cherry-pick the science. The supplemental proposal states that the approaches offered “…can often allow stakeholders to reanalyze the data and models and explore the sensitivity of the conclusions to alternative assumptions…” (emphasis mine).
In what world would stakeholders need to conduct sensitivity analyses on data? It is extraordinarily odd for the EPA to issue a rule for stakeholders to conduct scientific reanalysis studies. Scientists, of course, do conduct sensitivity analyses and other approaches for testing the robustness of their results, but this work is rightfully undertaken in the process of peer-review, by EPA experts, and through the agency’s science advisors. The EPA is not waiting for members of the public to do deep dives into testing the reproducibility of studies. This would waste everyone’s time and resources. The sentence in the proposal reveals their intention: Those drafting the proposal appear to want data to be more accessible so that stakeholders (like regulated industry, for example) can call into question legitimate scientific studies by cherry-picking data.
4. The administration doesn’t know the difference between the EPA and the CDC
The administration seems to believe that the EPA can simply mimic data sharing efforts of the Centers for Disease Control and Prevention (CDC) but this is unworkable for several reasons. To start, there would be tremendous cost and infrastructure needs that must be met to build and maintain the kind of data repository necessary for all the data and studies used by the EPA. These costs, too, are passed to those seeking to access the data. It can cost researchers thousands of dollars to gain access to the CDC’s Research Data Center, which the rule cites as a model. The EPA claims to be concerned with “stakeholder input” but it is unlikely that stakeholders like community groups and public health advocates will have the resources to access the data. But you know who will? Regulated industries.
5. The administrator still has broad authority to decide when the rule applies
Concerningly, the agency has not walked back the aggressive provision allowing the EPA administrator to decide when the rule applies on a case-b- case basis. This is still arbitrary and illogical, but most most alarmingly, it makes the EPA’s use of science wholly political.
We cannot let this happen
Expect more details about what this rule means for EPA’s ability to protect public health and the environment. Stay engaged. This month is the time to submit public comments about the harm that this rule is bound to impose on science and on public health.
Join me in sounding the alarms.