Is the Department of the Interior’s Scientific Integrity Office Protecting Scientific Integrity?

Posted by Anne LeHuray, March 16, 2015

Federal agencies disseminate vast quantities of information that is not subject to public notice or comment, but often has the effect of regulation. A good example is the Food and Drug Administration’s (FDA) decades-long recommendation to reduce dietary salt for better health. For over 40 years, Americans struggled to reduce salt intake, only to learn in 2013, after a comprehensive review commissioned by the Centers for Disease Control that, in fact, there is no evidence that reducing salt improves health outcomes. Indeed, the review suggested that more salt in the diet might be a good thing. The salt industry knew the FDA recommendation was based on inaccurate information, and spent years attempting to get the FDA to make corrections. One legal finding was summarized in an article by Catherine Meshkin for the Federalist Society thusly:

The court upheld the rights of federal agencies to have the final word on the quality of their facts, figures, and research used in rulemaking and other decisions. The court determined that the [Data Quality Act] does not create a legal right to information or information correctness and referred complaints back to the agency.

No right to information or information correctness? No checks on the adequacy of science underlying rulemaking or policy decisions? While begging the question “who controls the controllers,” unchecked data undermines public trust as well as the undeniable principle that sound policy should be based on sound information. Federal agencies often are under enormous political pressure to interpret data in a way that supports an ideology. In the face of such pressure, the temptation to weigh evidence with a thumb on the scale might become irresistible. Without a meaningful right to challenge agency information, what recourse does the public have to make sure that agencies are making decisions based on defensible data? Turns out the Obama Administration has created an administrative mechanism that might be useful for challenging incorrect information – the scientific integrity initiative.

Shortly after his first inauguration, on March 9, 2009, President Obama issued a Presidential Memorandum on Scientific Integrity emphasizing the importance of science in guiding Administration decisions and the importance of ensuring that the public trusts the science behind those decisions. At the time, the memorandum was viewed by many as making good on an election-year talking point about political interference with government scientists, which candidate Obama claimed was a feature of his predecessor’s administration. The Memorandum gave the Office of Science Technology and Policy (OSTP) responsibility for implementing the policy throughout the Executive Branch. Without much fanfare, OSTP directed all departments and agencies to craft scientific integrity procedures appropriate for the particular mission of each. The overarching goal was to prevent political interference in scientific processes, but OSTP also set other laudable goals including ensuring independent peer review, expanded transparency, and free flow of data to promote the “accurate presentation of scientific and technological information.” In February 2011, the Department of the Interior (DOI) was the first to announce its new policy and to establish a departmental Office of Science Quality and Integrity (OSQI).

Four years in, how is DOI’s scientific integrity policy being used? Twenty-seven cases [note: DOI has renovated it’s web site – closed OSQI cases can now be found here] have been closed to date. As a complainant, I can attest that DOI is as hush-hush about in-progress cases as any secret national security tribunal, and is not much more forthcoming about closed cases. That said, some information has come out about two high profile closed cases that involve commercial interests: the XL Pipeline and the Drakes Bay Oyster Company.

Of the 27 cases closed so far, OSQI has found only two complaints to have merit. Strikingly, both involved allegations of misconduct that were alleged to benefit commercial interests. A few details about one of these complaints (case ESO-S0000328) are available. An article by E & E reporter Emily Yehle summarized how two supervisors at the Fish and Wildlife Service (FWS) were found to have engaged in scientific misconduct in generating a habitat map.

The habitat map was part of a preliminary assessment of the proposed Keystone XL pipeline’s effect on the endangered beetle. [The two supervisors] used flawed methods to create a smaller — and consequently more industry-friendly — range for the insect. It was never used in any Keystone XL assessment, due to the misconduct finding.

It appears the range of the endangered beetle was decreased by not following guidance that extends modeled ranges to political boundaries (county lines) rather than restricting ranges to ecological boundaries. A scientific integrity complaint was filed by a FWS employee. The OSQI convened a panel of experts drawn from internal DOI organizations. Using only internal experts kept DOI in complete control. If experts from outside DOI had been used, the more open and transparent Federal Advisory Committee Act (FACA) process would have been required. The panel identified flaws with the map and underlying modeling and the map was immediately withdrawn from the FWS website. Nevertheless, the two supervisors continued to tout the map’s use and even published a subsequent article about it in a scientific journal. The panel was highly critical of the actions of the two FWS supervisors, of the internal FWS peer review process, and recommended that the FWS formally retract the published paper. The panel and DOI’s Inspector General acted expeditiously on the internally generated complaint to ensure that the smaller habitat range map was not used in evaluations of the route of the XL Pipeline.

Much more detail is available about a second high profile situation. The Drakes Bay Oyster Company alleged misconduct in preparation of DOI reports that concluded closing the oyster farm was supported by disturbances of the local seal population (case ESO-S0000379). The farm, established in 1935, was shut down at the end of 2014. An article in Newsweek relates the details of the misconduct.

To the bewilderment and eventual outrage of [farm owner] Lunny’s advocates in California and Washington, D.C.—U.S. Senator Dianne Feinstein chief among them—the DOI and its National Park Service spent much of the past decade using scientifically unsound, and at times bizarre, tactics to prove the oyster farm had to go. “The Park Service has falsified and misrepresented data, hidden science and even promoted employees who knew about the falsehoods, all in an effort to advance a predetermined outcome against the oyster farm,” Feinstein wrote to then-secretary of the interior Ken Salazar in March 2012. “It is my belief that the case against Drakes Bay Oyster Company is deceptive and potentially fraudulent.”…“I firmly believe that renewal of the permit is the only way for the Park Service to send an unmistakable signal that the [Obama] administration’s commitment to scientific integrity is real,” Feinstein told Salazar.

The National Academy of Science (NAS), at Senator Feinstein’s request, conducted an external, independent review. The NAS report

…concluded that Park Service scientists, in setting out to prove the farm was causing environmental harm, had “selectively presented, over-interpreted, or misrepresented the available scientific information” and “exaggerated the negative and overlooked potentially beneficial effects of the oyster culture operation.”

Bowing to pressure from environmental groups, Secretary Salazar declined to renew the farm’s lease simply because the lease agreement was up, without relying on the flawed NPS science report. Amazingly, in spite of the findings of selective use of data, misrepresentation and exaggeration by no less a body than the NAS, DOI’s OSQI, again relying on a panel composed of DOI employees, “found no merit to the allegations.” The expert whose findings were ignored or misrepresented in the NPS/USGS report continues to tell his story (here and here).

How can DOI’s OSQI not be suspected of White Hat Bias in finding “no merit” in the case where DOI agencies were found by an independent review body — the NAS — to, through the selective use of data, identify harm where none existed? Did White Hat Bias also play a role in OSQI finding “merit” in allegations of misconduct in the setting the boundaries of a habitat map based on the actual range of an endangered insect rather than a wider area where the beetle seems not to have been observed? In the oyster farm case, the real world impact was a closed business and jobs lost for no discernable environmental good. The finding in the case of the habitat range of the endangered beetle was only a small factor in the XL Pipeline debate, but OSQI again came down on the side of the environmental lobby.

It’s still early in implementation of DOI’s scientific integrity policy and these are just two cases. Perhaps recent revisions to the policy have been made to be more protective of scientific integrity. Maybe, at long last there will be a viable mechanism for working toward not only a preference for, but a right to use reproducible science in formulating government environmental policy. So far, though, the signs are less than auspicious about whether, using Senator Feinstein’s words, the DOI’s “commitment to scientific integrity is real.”

One thought on “Is the Department of the Interior’s Scientific Integrity Office Protecting Scientific Integrity?

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s