The Trump administration has recently proposed placing limits on the science that the Environmental Protection Agency can use in formulating public health regulations. The “Strengthening Transparency in Regulatory Science” rule would prohibit the agency from relying on findings relevant to agency decisions unless the data underlying those studies are made “publicly available in a manner sufficient for independent validation and analysis.” On its face, the rule seems to be a welcome safeguard for promoting good governance: Who could be against more transparency and independent checks on influential research?

Indeed, in this moment of political polarization and “alternative facts,” transparency seems like a tonic with the potential to foster agreement by shining light on what is verifiable and what is not. In science, transparency has emerged as a watchword of responsible practice and a strategy for addressing the “replication crisis.” In government, transparency has been a cardinal virtue since the 1967 passage of the Freedom of Information Act, with its hope that “sunshine” could serve as a disinfectant against corruption.

It’s hard to argue against sunshine. And this is precisely why some smoke-stained interests — including the tobacco and fossil fuel industries — have long sought to take advantage of transparency’s sterling reputation to advance an anti-regulatory agenda under its banner. The Trump administration’s EPA proposal borrows from this murky tradition.

This “weaponization of transparency” can be traced to the 1990s and the tobacco industry’s efforts to poke holes in the science linking smoking to disease, particularly the dangers associated with secondhand smoke (also called “environmental tobacco smoke” or ETS).

During the 1950s, a series of powerful epidemiological studies funded by the American Cancer Society and the British government identified a strong link between smoking and lung cancer — a disease that had been all but unknown before the rise of the cigarette in the early 20th century. Tobacco companies quickly sought to undermine this evidence by “creating doubt about the health charge,” as one internal industry memo later disclosed in state litigation put it.

Such arguments became difficult to sustain as the evidence piled up. After the surgeon general declared unequivocally that smoking caused lung cancer in 1964, the tobacco companies re-centered their efforts on building a defense against the charge that nonsmokers were being harmed by ambient cigarette smoke. It was much more difficult for scientists to demonstrate conclusively that secondhand smoke caused disease, but by the 1980s, emerging research was convincing government authorities that smoke-filled rooms posed a danger to bystanders. Internally, the tobacco companies acknowledged this reality, but nevertheless stuck with their self-described strategy of “throwing up a smoke screen” on the science.

It would be the EPA that delivered perhaps the decisive punch against secondhand smoke. In 1992, the agency issued an exhaustive risk assessment that depicted the human toll in hard numbers: an estimated 3,000 lung cancer deaths annually in U.S. nonsmokers and some 150,000 to 300,000 annual cases of respiratory infections in infants. Most powerfully, the EPA declared secondhand smoke to be a “Group A” carcinogen — a known cause of cancer.

It was a disaster for the tobacco industry, which responded with two strategies. First, the cigarette companies searched for allies. “The credibility of EPA is defeatable, but not on the basis of ETS alone. It must be part of a larger mosaic that concentrates all of the EPA’s enemies against it at one time,” observed a 1993 industry memo. The second strategy involved gaining access to the raw data behind key studies. Tobacco firms knew from experience in lawsuits that they could hire statistical experts to reanalyze the data behind epidemiological findings so that they coughed up more industry-friendly conclusions.

The result of these strategies: an initiative called the Advancement of Sound Science Coalition, an organization dedicated to educating the public about the perils of “junk science.” Outwardly, the industry-backed coalition portrayed itself as a “grassroots-based, not-for-profit watchdog group” aimed at elevating the standards of science-based policy. Internally, however, tobacco strategists noted their hope that the campaign would link warnings about secondhand smoke “to junk science in [the] public’s mind” so that being around cigarettes was no longer “seen as a significant health risk.” In addition, industry consultants reached out to the scientists behind the most influential study that had implicated secondhand smoke as a cause of cancer to request the raw data. However, the scientists declined to share it, citing concerns that their data “not be distorted by the economic interests of other parties who analyze them.”

At the time, the fossil fuel industries were in a similar predicament: The scientists behind an influential study that was poised to shape EPA pollution standards declined to grant industry researchers access to their raw data. Tobacco strategists seized the opportunity to “Develop a Coalition to Support Data Access as a Means to Stop EPA Regulation” — provided precautions were taken to ensure that their efforts were “not visible.” (So much for transparency.) In the end, the industry coalition managed to convince allies on Capitol Hill to quietly slip riders into two different gargantuan federal appropriations bills. The result was the Data Access Act and the Data Quality Act, laws that required public access to data produced in federally funded studies and established a procedure to allow people to request corrections to information put out by the government. No hearings were held on either provision, and the laws did not pertain to industry-funded science. (Again: So much for transparency.)

Today’s EPA reform proposals echo these age-old tactics. From 2014 to 2017, Republicans in Congress worked (unsuccessfully) to advance various “pro-transparency” bills, including the “Secret Science Reform Act,” that would have prohibited the EPA from developing regulations based on science that was “not transparent or reproducible.” In suggesting that existing EPA decisions are reliant upon “secret science,” opponents of regulation are piggybacking upon the compelling language of the open-data movement. Now, the “secret science” legislation has mutated into a proposed rule from the agency itself, under the leadership of former coal lobbyist and EPA Administrator Andrew Wheeler.

The reality is that these “pro-transparency” initiatives actually seem to be a transparent effort to hamstring the agency from making rules to protect population health — a Trojan Horse through which other aims are pursued. David Michaels, former head of the Occupational Safety and Health Administration, has warned that past industry-linked open-data efforts have slowed agency activities by consuming scarce resources. A recent statement from the editors of six leading scientific journals expressed alarm that the proposal might be used as a tool for “suppressing” evidence, potentiating a public health “catastrophe.”

In theory, data access supports a democratic approach to the evaluation of scientific findings — promoting rigor and accountability and bringing us closer to the best policy outcomes. But in practice, if poorly implemented, it may perversely increase industry capture of regulatory processes, as resource-rich special interests exploit imperfections in the evidence to impose administrative delay. If the EPA is prevented from relying on epidemiological studies for which data cannot be released because they contain private medical information, it will magnify the uncertainties inherent in the science of public health.

This is the great challenge confronting agencies like the EPA: Science, by its nature, is always incomplete and fraught with uncertainties, whereas policymaking requires certain and timely action. In the face of emerging dangers such as tobacco smoke or climate change, decisions cannot be put off indefinitely while studies are subjected to endless re-analysis and replications. Instead, public health agencies must sometimes take action to reduce risks before definitive evidence is available. Demands for “transparency” from actors with no track record of interest in genuinely sound science must be regarded with great skepticism.