The proposed Secret Science Reform Act of 2014 runs only three pages, and would set a simple rule: The U.S. Environmental Protection Agency cannot take any action based on research for which all underlying data are not publicly available.
The measure, now awaiting a vote on the House floor, embodies a concept—increased scientific transparency—that enjoys wide support in academe. The bill has nevertheless generated an oversized level of consternation and uncertainty in the research community.
Much of the concern stems from the apparent political motivation behind the bill, which is widely seen as aimed more at battling the EPA over air-pollution rules than balancing the goals of scientific transparency with the need to protect research subjects.
“Political controversy may be exploited to move the open-science agenda,” said John P.A. Ioannidis, a professor of medicine at Stanford University who has warned about reproducibility problems in science, “but one has to be cautious about who is doing what and why.”
The Secret Science bill was approved last month on a straight party-line vote by the Republican majority on the House science committee. Its chairman, Rep. Lamar Smith of Texas, said he wanted to stop the EPA from imposing costly air-pollution regulations without being sure about the actual benefit to people.
Mr. Smith complained that the EPA during the Obama administration has repeatedly justified such rules by citing two decades-old studies financed by the National Institutes of Health and led by Harvard University researchers. One, known as the Six Cities Study, tracked local air-pollution levels and the health records of more than 8,000 people in six American cities during the 1970s and 1980s. The second, organized by the American Cancer Society, made similar comparisons with 1.2 million people in all 50 states beginning in 1982.
After accounting for personal characteristics and behaviors such as smoking, both studies found strong correlations between increased levels of air pollution and worsened public health, said a lead author of the Six Cities Study, Frank E. Speizer, a professor of medicine at Harvard University.
Mr. Smith, however, doubts the science and feels its application by EPA has been too ambitious, costing Americans jobs and higher energy prices for insufficient or unproven benefit. As a chief example, he has cited the EPA’s estimate that its proposed limits on ozone will cost taxpayers $90-billion per year, making them the government’s most costly regulations ever.
In response, the Secret Science bill would ban EPA from moving ahead with such regulations until Harvard—and any other university whose scientific findings are cited as the basis of a regulation—turns over all data associated with its research.
“The American people foot the bill for the EPA’s billion-dollar regulations, and they have the right to see the underlying data,” Mr. Smith told his committee just ahead of its Republican-only approval of the bill.
And the study authors have made clear they won’t comply, Dr. Speizer said. Harvard researchers promised their thousands of study participants that their personal information—names, addresses, medical conditions—would be protected, he said. “I’m not going to give out raw data, because that would be unethical,” said Dr. Speizer.
A Complicated Subject
On some level, that kind of stance might be the end of the debate over air-pollution rules. Advocates of the Secret Science bill have made clear that their primary goal is to stop the EPA from moving ahead with air-pollution regulations, and if the bill becomes law and Harvard researchers and other scientists insist on protecting test-subject data, then the EPA would have to come up with some new basis for its rules.
But for many science advocates, it’s not that simple. Those protesting the bill include the American Association for the Advancement of Science, which said the bill’s attempt to set a flat demand for data disclosure—the EPA would need to make public “all scientific and technical information” that is needed “for independent analysis and substantial reproduction of research results"—could cause major problems if enforced literally.
Mr. Smith and other Republicans have suggested a willingness to clarify the demand so that personal information is protected, through statistical techniques that can help mask the identities of individual study participants. Those techniques are commonplace and often sufficient, said Edward P. Richards III, a professor of law at Louisiana State University who studies privacy issues involving grant data.
Still, it’s a complicated subject, Mr. Richards said, with no guarantees that a person determined to figure out the identities of test subjects couldn’t do it. And that kind of sleuthing—such as figuring out a person’s identity based only on an address and a medical condition—is likely to become even easier as technologies and databases grow more sophisticated, he said.
Raising further concern for some advocates of unfettered scientific inquiry is the fact that Mr. Smith, in previous discussions of the matter, cited James E. Enstrom, a former epidemiologist at the University of California at Los Angeles, as a scientist who should have access to the American Cancer Society data. Democrats have complained about the fact that Mr. Enstrom, who was fired by UCLA in 2012, previously conducted tobacco-industry-financed research challenging health-related concerns about secondhand smoke.
Aside from any debate over any particular scientist, said Andrew A. Rosenberg, director of the Center for Science and Democracy at the Union of Concerned Scientists, there’s concern over the idea that Congress might decide what studies EPA uses and what data-security protections EPA should be demanding. “It’s not entirely clear to me that there isn’t any risk” in giving personal data to outside scientists, said Mr. Rosenberg, an affiliate professor of natural resources and the environment at the University of New Hampshire.
Fear of a partisan industry-financed attack shouldn’t exempt any scientist from the need to have his findings validated, said another leading advocate of transparency, Brian A. Nosek, director of the Center for Open Science at the University of Virginia. But the bill’s authors still need to protect the privacy of research subjects, said Mr. Nosek, an associate professor of psychology. And when they conflict, “values of transparency and confidentiality must be considered simultaneously,” he said.
A Lost Opportunity?
The Secret Science bill is not the first attempt by Congress to guide scientific input into policy making. Last year Congress nearly included language from a measure called the Sound Science Act in a bill setting agricultural policy. The language, removed from the final version of the bill after protests, would have allowed virtually endless challenges to the science-based assessments used by governmental agencies.
And back in 1998, the EPA faced a very similar problem to the one posed by the Secret Science bill when Sen. Richard C. Shelby, Republican of Alabama, included an amendment to federal budget legislation that required all data produced under a federal grant award to be made publicly available. The effect of the Shelby amendment was later tempered in the governmental rule-making process that follows enactment of a bill. Among their actions, administration officials required that any disclosures comply with privacy exemptions set out by the federal Freedom of Information Act.
The authors of the Secret Science bill could have been more rigorous in the pursuit of open-science goals by insisting on the full adoption of the Shelby-amendment language, Mr. Richards said, even to the point of overriding the protections of FOIA and the Health Insurance Portability and Accountability Act of 1996, the standard set of rules ensuring patient privacy.
But they did not, he said, suggesting that the Secret Science bill is more about a political fight with EPA than a wider concern about scientific transparency. And without such a political motive to demand their attention, it’s hard to see what else might get lawmakers to seriously tackle that broader subject, he said.
Mr. Ioannidis, who published a 2005 analysis in PLOS Medicine concluding that most published research findings appeared to be substantially false, also sees a lost opportunity to improve transparency and reliability in science.
The Secret Science bill appears to serve the interests of the fossil-fuel industry, Mr. Ioannidis said. “It is unfortunate,” he said, “that in other situations where sharing of raw data would serve the public primarily—such as the availability of all raw data from clinical trials of drugs—no such bill exists.”
