1180 13 MARCH 2020 • VOL 367 ISSUE 6483 sciencemag.org SCIENCEPHOTO: ISTOCK.COM/THOMAS DE WEVERW
e want to make sure we get this
right,” Andrew Wheeler, head of
the U.S. Environmental Protection
Agency (EPA), said in September
2019 about a controversial pro-
posal to narrow the kinds of scien-
tific studies the agency would consider when
crafting regulations. Speaking to reporters,
he acknowledged that the initial plan, re-
leased in 2018, had raised numerous ques-
tions about its practical implications and
cost. It had also drawn withering criticism
from many U.S. science and environmental
groups, who saw it as a way to exclude rele-
vant science on the health effects of pollution
(Science, 4 May 2018, p. 472).
Now, EPA has released for public comment
a document that aims to clarify the proposed
rule, dubbed “Strengthening Transparency in
Regulatory Science.” But all is not right, the
critics argue. If anything, they say, the new
version is more troubling, because it greatly
expands the kinds of science that the policy
might put off limits. The rule also proposes
to change how EPA conducts in-house stud-
ies and analyses, and it gives the politically
appointed agency head a greater say in decid-
ing which studies the agency can use.
“The [Trump] administration has doubled
down,” says Thomas Burke, an epidemio-
logist at the Johns Hopkins University Bloom-berg School of Public Health and EPA science
adviser for 2 years until January 2017. “This
has very far reaching impacts ... [and] looks
like it throws an incredible monkey wrench
into the agency.”
The EPA proposal, first floated by former
administrator Scott Pruitt, generally bars
the agency from using studies that have not
made their underlying data and mathemati-
cal models public. Transparency is essential,
the agency says, because it allows anyone to
independently “reanalyze the data and mod-
els and explore the sensitivity of the conclu-
sions to alternative assumptions.”
The original proposal appeared to limit the
restriction to so-called dose-response studies,
used to evaluate risks and set limits on hu-
man exposure to pollutants. In contrast, the
new version applies to all “influential sci-
ence,” even if the agency does not directly
use it to justify regulations. The rule also
proposes requiring EPA’s own researchers
to reconsider long-standing analytical ap-
proaches. For example, rather than rely on
time-tested models that assume exposure to
chemicals or radiation can have some bio-
logical effect at any dose, researchers should
consider whether a chemical has a hard
threshold below which it’s safe.
Even critics say scientific transparency is a
worthy goal, as is periodically revisiting sci-
entific methods. But they fear the true intent
of EPA’s rule is to enable industry groups op-posed to tougher regulation to prevent the
agency from using certain kinds of studies,
particularly epidemiological studies that rest
on confidential health information, and to
entangle regulators in continual reanalyses.
“This rule opens the door to endless delay
and challenging of the science,” Burke says.
One big question is how EPA will handle
past studies based on confidential health
data. Such studies have been key to setting
limits on major air pollutants, including par-
ticulate matter and ozone, that the agency
must review every 5 years. Researchers say
it would be expensive, difficult, and in some
cases impossible, to release their underlying
data and still honor patient confidentiality
agreements. As a result, depending on how
the final rule is worded, opponents of tighter
controls on air pollution and other rules
might be able to challenge EPA’s use of past
studies when it reviews the limits. “This fun-
damentally changes the way the agency can
use science to protect people,” says Gretchen
Goldman of the Union of Concerned Scien-
tists, an advocacy organization.
In its proposal, EPA asks for comment on
an approach to sharing some confidential
health data. This “tiered access” system, mod-
eled on a program at the U.S. Centers for Dis-
ease Control and Prevention (CDC), would
give greater access to researchers who show
they have a bona fide reason for wanting the
data, and agree to protect them. But there is
a difference between data at CDC and EPA,
Goldman says: EPA uses pollution exposure
data that includes people’s locations, which
can make it harder to protect their identities.
An alternate idea, EPA suggests, would be
to allow regulators to consider all studies rel-
evant to a particular regulation, regardless of
their transparency, but to put a greater weight
on those that have released data. But to Linda
Birnbaum, former director of the National
Institute of Environmental Health Sciences,
that approach is “hardly a way to get the best
science on which to base decisions.”
The public will have 30 days to comment
on the proposal after it is published in the
Federal Register. Congress has also asked the
National Academy of Sciences to review the
matter. The agency hopes to finalize the rule
later this year, although it is likely to face a
court challenge. Meanwhile, science and en-
vironmental groups are tracking a similar
scientific transparency policy under devel-
opment at the Department of the Interior,
which sometimes keeps confidential the loca-
tions of endangered species to protect them
from poaching or collecting. Last month, the
department sent its proposal to the White
House for review. jPesticide research often involves confidential health
data, which can be difficult to make publicly available.EPA expands controversial
‘transparency’ plan
Critics blast revised proposal that covers a wider range
of research, saying it will undermine protections
U.S. REGULATORY SCIENCEBy Erik Stokstad“
Published by AAAS