OSHA deserves praise for asking commenters submitting scientific information to disclose their funding sources.
As one of the most widely-reprinted New Yorker cartoons reminds us, “on the Internet, nobody knows you’re a dog.” Anyone can “publish” anything on the Web anonymously, under an alias, or even using someone else’s identity. On the other hand, if you want to publish scientific results or interpretations in a peer-reviewed journal, you are now expected to submit them under your real name, disclose who may have paid you to conduct the research or interpret the findings, and probably also reveal any important relationships or experiences that may call into question your objectivity. Should the standards for disclosure about the scientific information submitted to government agencies for use in expensive but life-saving regulations resemble the standards for scientific journals?
Recently, the U.S. Occupational Safety and Health Administration (OSHA) made the first tentative suggestion that participation in regulatory science should be more like academic publication and less like the Blogosphere. Judging from the reaction of some prominent commenters, you’d think OSHA had suggested something bizarre and troubling. As a scientist and (relevant and non-ironic disclosure) former OSHA official, I find the pushback depressing and only wonder why it took so long for someone in government to suggest that the provenance of regulatory science should not be concealed or misrepresented. Of course, encouraging such disclosure is an entirely separate matter from what the government should do with the disclosed information. I would actually caution against regulators using information about funding sources when designing and defending their rules. Indeed, I do not believe that legally they can rely on such information at all.
The latest controversy began last fall when OSHA published a proposed regulation to lower occupational exposure to crystalline silica, a respiratory hazard and carcinogen that more than two million U.S. workers are exposed to on the job. The OSHA proposal requested that commenters might voluntarily provide information on the funding sources behind each scientific report or piece of testimony that was paid for by a party who might be affected by the results of the rulemaking and also to disclose the extent to which such affected parties reviewed or edited the submissions. Several weeks later, 16 Republican Senators sent a letter to OSHA’s head, Dr. David Michaels, asking him to rescind the information request on the grounds that it “could result in dissuading stakeholders from even submitting comments.” The Senators also suggested that OSHA intended to “prejudge the substance of those comments” that were accompanied by disclosures of industry funding or editing. I heard similar objections last month from several former U.S. Environmental Protection Agency (EPA) officials, who criticized OSHA’s information request as infringing on First Amendment rights.
It is tempting to dismiss the Senators’ concern about a chilling effect of voluntary disclosure, given that for the past 17 years the U.S. House (though not the Senate) has required all witnesses at hearings to submit (now for online posting) a “truth in testimony” form that discloses any parties the witness is representing and lists any federal grants the witness has received.
But some of the signers of the Senate letter have relevant personal experience trying to dissuade certain stakeholders from participating fully in science-policy debates. They certainly seem to regard some kinds of funding as a corrosive influence and revelations about certain sources of funding as a potentially powerful weapon against the investigator. For example, in 2010 Senator David Vitter (R-LA, one of the signers of the letter to Michaels) requested that the U.S. Government Accountability Office investigate how much money various non-governmental organizations had received in legal judgments against the federal government, describing this as “taxpayer money…being spent to fund lawsuits by these radical environmentalist groups.”
More well-known is a series of letters (including one from Representative Joe Barton, R-TX) to Dr. Michael Mann, a climatologist then at the University of Virginia, in which lawmakers demanded, among other things, information on “all financial support you [Mann] have received related to your research” and copies of all agreements he entered into with funders. In response to what they apparently regarded as a “fishing expedition” rather than a simple request, the American Association for the Advancement of Science’s Board of Directors issued a statement decrying “attacks on researchers that question their personal and professional integrity.”
In this connection, it is worth noting that the journal Nature wrote an editorial last month defending OSHA’s request, concluding that “transparency is the best defence against the purchase of undue influence by those with the most financial clout.”
But the juxtaposition of the 2013 Senate letter with previous attempts to ferret out financial influence is not the only irony in this debate. Just as we should be able to distinguish simple and legitimate questions about funding from vendettas disguised as oversight, so too we must acknowledge that executive agencies need to be above suspicion in their own use of influence. At least at the time an agency is considering a proposed regulation, the agency should be listening more and advocating less. Perhaps Senator Michael Enzi (R-WY, one of the signers of the 2013 letter) was recalling his own indignation at OSHA from 2001, when he revealed that the agency had paid 20 contractors $10,000 each to testify in favor of its ergonomics regulation. At the time, he remarked that “[OSHA] practiced these people, which also made sure the testimony they were giving was the testimony OSHA wanted given…[and] they paid those witnesses to tear apart the testimony of other folks who were testifying, at their own expense.”
On balance, I see no evidence that today’s OSHA intends to use the information commenters may voluntarily disclose in any of the aggressively dismissive ways that some of the concerned Senators have themselves used similar information in the past. (Nor is there any evidence even suggesting that OSHA will ever again fund commenters to take its side in a controversy or ghost-write their submissions.)
I think there are two inescapable facts about the relationship between money and scientific commentary. First, sometimes the money follows a scholar, rather than vice versa. I was asked repeatedly some years ago to testify against a Presidential nominee to a key regulatory post who had received a large portfolio of industry funding. But I declined to do so because I was confident that this person’s views long predated any funding. Second, when funding does buy a result, the only remedy is disclosure. I remain troubled by the manufacturer of a particular neurotoxic and carcinogenic solvent continuing to rely on a 2002 recommendation that workers could be safely exposed to 75 parts per million (ppm), when most other risk assessments suggest that 1 ppm or less is needed to provide even minimal protection. However, at about the same time that the director of the small consulting firm was producing the 75 ppm recommendation with funding from the manufacturer, he was testifying elsewhere that the science showed that 5 ppm was necessary for worker safety. It is these kinds of “controlled experiments” where funding seems to dictate the result, even more than the statistical correlations often found between funding source and positive/negative findings, that validate OSHA’s concern about the perils of undisclosed financial and editorial arrangements.
Ultimately, the main reason I feel comfortable supporting OSHA in this matter, and would not worry at all if other regulatory agencies should follow suit, is a simple one: unlike private citizens or elected officials, a regulatory agency like OSHA is simply not in a position to impugn researchers based on their financial ties. Agencies have to follow the rules of administrative procedure, which guard zealously against arbitrary and capricious treatment of scientific information and those who proffer it. “I don’t like your funding” is simply not a permissible (or a strategically sensible) response to a critical comment submitted to the rulemaking docket. OSHA instead has to rebut the comment on factual grounds, and it risks judicial invalidation of its rule if its rebuttal is churlish.
Of course, in the end, opponents of OSHA should also take heed: disclosing who paid for or ghost-wrote a comment doesn’t immunize it against criticism. When my colleagues and I at OSHA rejected certain comments as without scientific merit, we were not “prejudging the substance.” We were simply judging it, which is the agency’s right and its responsibility.
