An Effort to Improve Scientific Integrity in the Federal Government
December 2, 2019 by Ramin Skibba - Undark Magazine
In September, after Hurricane Dorian pummeled the Bahamas, it barreled down on Florida, Georgia, and the Carolinas. But while many of these states anticipated evacuation warnings and property damage, President Donald Trump falsely claimed the storm could also endanger Alabama. Leaders of the National Oceanic and Atmospheric Administration then endorsed Trump’s claim, and when employees of the National Weather Service in Birmingham contradicted the claim, Commerce Secretary Wilbur Ross, who oversees NOAA and the NWS, criticized the office and allegedly threatened to fire its staff.
The clash over scientific integrity wasn’t a first for federal scientists in the United States. For instance, the Obama administration pressured scientists to underestimate the size of the BP oil spill in the Gulf Coast in 2010 and NOAA's administrator misleadingly claimed that the majority of the oil was gone in months.
When the federal government interferes with its scientists and manipulates their research and reports, science advocates often clamor for clear scientific integrity policies. Ideally, advocates say, government scientists should follow their research where it leads, talk about it honestly and freely with the press and the public, and release unaltered information about their findings.
A bill introduced earlier this year seeks to move toward that ideal. On March 13, Democratic Representative Paul Tonko of New York, introduced the Scientific Integrity Act in Congress. It passed the House Science Committee on October 17, with six Republicans joining 19 Democrats in voting in favor.
“I think all agencies should have [a] strong scientific integrity policy in place,” Tonko said.
“Science should not be messed with, should not be adulterated or buried or hidden somewhere or mischaracterized.”
The bill — which would require federal agencies that conduct, oversee, or fund scientific research to develop and adopt scientific integrity policies — has to overcome more hurdles before becoming law. The House will need to review and pass it, and then a similar version, sponsored by Democratic Senator Brian Schatz of Hawaii, needs to make it through the Senate. Trump would then either sign or veto a compromise version of the bill. But if Tonko and his allies successfully navigate the bill along that path, it will smooth out and buttress piecemeal scientific integrity policies already adopted by a handful of government agencies.
Tonko says he thinks other representatives will come on board with the bill, if they haven’t already. “We deal with some very complex and technical issues,” he said. “And our policies and our advocacy need to be guided and tempered by science.”
Science integrity policies aren’t new. Still, only half of 18 major U.S. agencies have clear, relevant policies, according to a 2017 report from the Union of Concerned Scientists, a nonprofit advocacy organization. A few agencies, the report argues, such as the National Institutes of Health, have nonexistent or poor policies; others, like NASA and the National Science Foundation, have policies but lack the staff to make them work in practice. And the rules don’t help much if government scientists don’t know how to use them. According to an April report from the Government Accountability Office, the Department of Energy and National Institute of Standards and Technology don’t train employees about relevant integrity policies.
The policies “are uneven, they’re a patchwork of different components, and there’s no one gold standard across the government,” said Michael Halpern, deputy director of the Center for Science and Democracy at the Union of Concerned Scientists, who testified in favor of the bill in a bipartisan hearing of the House Science Committee in July. “These policies can be rescinded or weakened at any time.”
The rules also won’t matter if they aren’t enforced. In May of last year, top Environmental Protection Agency officials appeared to have violated their own policy when the agency delayed the release of a report about the health risks of the common chemical formaldehyde, and the Union of Concerned Scientists and other groups asked its scientific integrity office to investigate the situation. Later in 2018, the Department of the Interior told scientists with the U.S. Geological Survey that they had to get permission before agreeing to interview requests from reporters — a directive employees say represents a departure from prior protocol.
With such threats of censorship, many government scientists simply choose silence. In a 2018 survey from the Union of Concerned Scientists of some 4,200 government researchers, nearly half of respondents from the National Park Service and more than a third from the EPA reported self-censorship. The survey also suggested a decrease in morale at both agencies.
“They may not have even gotten a direct order not to talk about something, but they kind of knew what they were supposed to talk about and what they weren’t supposed to talk about,” Halpern said. “That is one of the other challenges with policies that are vague and not sufficiently enforceable,” he added. “If people don’t know where the line is, they’re gonna stay way back from it.”
Still, other government scientists may speak up outside usual channels as whistleblowers. But whistleblowers take risks when they decide to come forward, and they have limited protections. In 2005, during George W. Bush’s administration, Rick Piltz, then a senior associate in the U.S. Climate Change Science Program, blew the whistle on a Council on Environmental Quality appointee who repeatedly watered down government climate reports and exaggerated uncertainty in climate science. Piltz resigned from his position; the appointee also resigned and took a job at ExxonMobil.
The Trump administration has also seen many climate change-related whistleblowers, including Joel Clement, an Interior Department analyst who in 2017 pointed out climate threats to Alaskan native villages, only to find himself reassigned to an obscure accounting post. More recently, epidemiologist George Luber had led the since-eliminated Climate and Health Program at the Centers for Disease Control and Prevention, but when he criticized the agency for canceling a conference and moving climate funds to other programs, he was barred from speaking publicly and placed on administrative leave in 2018.
Whistleblowers also fared poorly under Obama. While he promised transparency, his government criminally prosecuted more leakers under the Espionage Act than all other administrations combined. In addition to the BP oil spill episode, Obama’s EPA downplayed evidence that fracking can contaminate well water, although the agency reversed its position during his last month in office.
“Both parties’ presidents are guilty of abuses,” said Jason Ross Arnold, a Virginia Commonwealth University political scientist, in a written statement to Undark. “We need a scientific integrity act because government officials have far too often interfered in government scientists’ work to protect political or economic interests, at the cost of American democracy and public health.”
Today, most agencies haven’t completed a whistleblower certification, which would enhance protections. The scientific integrity bill explicitly outlines such protections.
Policymakers aren’t alone in wanting access to solid scientific findings. A majority of Americans say they trust findings more if the data are publicly available, and they also think scientists should be more transparent about potential conflicts of interest with industry, according to an August 2019 survey by the Pew Research Center. But if government scientists don’t have support and independence and their conclusions threaten an industry with a powerful lobby or the reputation of a politician, those conflicts of interest come to the fore. Research can be contentious if it involves, say, Flint-style water pollution, a Katrina-level natural disaster, or gun violence, to name just a few examples. Any time a scientist wants to share that research at a conference or on social media, their freedom to do so may depend on scientific integrity policies.
(An emphasis on transparency can be abused, however, as scientists have pointed out in response to a proposal from the EPA that would limit the type of research the agency uses to determine public health policy.)
If the Scientific Integrity Act is enacted into law, it could encourage a culture of openness, giving scientists the freedom to do and share their work without fear of retaliation, Halpern argues. It would also standardize policies across agencies, Tonko says, and give them the attention they require.
But there’s no guarantee the act will become law. All but two of the 230 sponsors of the House bill and all 14 sponsors of the bill in the Senate are Democrats.
“Without Republican support, it would be tough to see how it gets brought up in the Senate, because it becomes really difficult for any one individual from the other party to sign onto the bill if it looks like it’s partisan,” said Michael Fisher, a senior fellow at the Federation of American Scientists, a nonprofit science policy organization.
The bill’s chances might be bolstered by an October 2019 report from a nonpartisan group at the Brennan Center for Justice at the New York University School of Law. The task force is co-chaired by Christine Todd Whitman, former New Jersey governor and EPA administrator in the George W. Bush administration, and includes another Republican, Mike Castle, a former governor and congressman from Delaware. The report calls for sweeping reforms and makes recommendations that go beyond the Scientific Integrity Act, which mostly covers research, and draws more attention to political appointments, advisory committees, and government data.
Partisanship aside, many agencies will likely welcome this bill, said Wendy Wagner, a University of Texas law professor. “The incentives of an agency as a collective are to get the right answer because they’re going to be blown around by the political winds every four years.”
Still, considering how readily some agencies, but not others, have already adopted scientific integrity policies and how often scientific disputes flare up, agencies may respond differently to a federal scientific integrity law. The bill leaves discretion to the agencies, which would prove helpful in some cases, since the standards and timing to build scientific consensus vary from field to field. But that discretion leaves room for some political interference if the way the policy’s implemented depends on an agency’s leadership. After all, though weather scientists were able to speak up about Hurricane Dorian and Alabama thanks to the agency’s scientific integrity policy, it didn’t prevent retaliation.
The bill lacks teeth on enforcement, argues Dana Gold, senior counsel of the Government Accountability Project, a nonprofit that advocates for whistleblowers. It also lacks a dedicated provision “that would give employees the right to report alleged violations free from reprisal.” The Scientific Integrity Act is a step in the right direction, she added, but whistleblower protections should be stronger.
If the bill stalls in Congress or on President Trump’s desk, the campaign to improve the integrity of government science will continue into the next administration. In that case, Tonko aims to again push the bill. He said he’s optimistic about passing it this time though.
Halpern said he feels the same way. “It’s one of the silver linings of the repeated attacks on science that we’ve seen over the past couple years — that not only are scientists really engaged in standing up for science, but also people across all kinds of different issues, whether it’s public health advocates, or environmentalists, or national security experts,” he said. “People’s eyes have been opened.”
Ramin Skibba (@raminskibba) is an astrophysicist turned science writer and freelance journalist who is based in San Diego. He has written for The Atlantic, Slate, Scientific American, Nature, and Science, among other publications.