The American Chemistry Council (ACC), a trade association that represents chemical industry interests and is heavily connected to the plastics industry, filed a Request for Correction Monday on the EPA's Chemical Action Plan for Bisphenol A (BPA). The request, filed under a provision of the Data Quality Act (also referred to as the Information Quality Act), is truly astonishing and bears noting. In addition to standard requests that EPA statements be toned down or removed due to conflicting studies, ACC makes several requests that EPA remove statements that are included not as “ knowledge such as facts or data,” but policy statements that reflect EPA’s intent to manage exposure to BPA.
ACC requests in several places that references to a Canadian risk assessment of BPA be deleted because the Canadian assessment was informed by the precautionary principle:
Any reliance in the Canadian assessment to support EPA’s conclusion that … action is warranted is not appropriate under the Guidelines; EPA cannot blindly rely on Canada’s screening risk assessment (particularly as Canada’s assessment incorporates the precautionary principle), but must perform its own weight-of-the-evidence assessment. (p. 9)
And ACC later objects to the use of a “conservative” model:
EPA cannot rely on the highly conservative E-FAST2 modeling of BPA releases in the 2007 TRI to estimate the amount of BPA in drinking water or surface water when there exists sic peer-reviewed assessments of BPA in groundwater and drinking water… (p. 18, citations omitted)
The Toxic Substances Control Act (TSCA) gives EPA the discretion to be precautionary in its approach to chemical management. TSCA Section 5 states that in making a determination about acting on toxic substances, the EPA administrator “shall consider all relevant aspects of the risk, as determined by the Administrator in the Administrator’s discretion." It is therefore inappropriate for ACC to state that EPA cannot make a legitimate policy decision because the structure of U.S. law does not contain the same explicit requirement to be precautionary.
ACC attempts to coerce EPA into retracting these policy statements using some DQA slight-of-hand, by presenting a list of some challenges that are based solely on the science and some challenges to EPA’s policy decisions. ACC then charges EPA with changing its action plan to reflect the request for correction, or if it cannot accomplish this “the Action Plan should be withdrawn entirely.”
ACC’s tactic in using the DQA to challenge not just specific disputed data points, but EPA’s approach to managing BPA echoes recommendations of John Graham, head of the Office of Information and Regulatory Affairs during most of the Bush administration. At least for now, the Obama administration has not expanded the scope of the DQA to include statements on policy.
The ACC’s request for correction is particularly troubling in the wake of the DC Circuit court's decision this March in Prime Time v. Vilsack. The court ruled that Prime Time International Compandy had a right to judicial review of USDA’s calculation of its market share, but did not support Prime Time’s assertion that USDA had violated the DQA by not responding to its request for correction. We've previously discussed how DQA petitions are not judicially reviewable, an interpretation that was supported in 2006 by the Fourth Circuit. But Jim Tozzi and his cohort at the industry-funded Center for Regulatory Effectiveness have argued that Prime Time may open the door for judicial review of petitions filed under the DQA. And while the D.C. Circuit’s decision to deny a petition from the Department of Justice to clarify that its decision did not create judicial review, it is not yet clear how the Prime Time decision will affect judicial review under the DQA.
If DQA requests were judicially reviewable, a request like ACC’s recent submission could have disastrous effects on agency decision making. The judicial review process would have to make extremely careful determinations about what is “information” for the purposes of the DQA and separate “facts and data” from policy determinations of the agencies. Judges, who are likely neither experts on the science underpinning DQA petitions nor the policy determinations made by agencies, would be vested with the ability to make determinations that could effectively result in industry dictating “corrections” to an agency’s policy determination. Such a process would be tremendously disruptive to agency decision making and provide industry with an undue advantage in influencing not just regulatory action, but agency risk management and policy determinations.
Another likely consequence of such a situation would be an even greater contraction of information provided by agencies. The threat of litigation looming over not just “facts and data” but policy determinations of the agency as well – before the agency even takes a regulatory action – would surely have a chilling effect on agencies’ willingness to make information available.