OMB continues to push agencies on risk assessment

After a stunning rebuke in January by the National Academies of Science, the White House released today another government-wide policy statement on risk assessment.

The new memorandum to agencies, signed by Susan Dudley, recess-appointed administrator of the Office of Information and Regulatory Affairs (OIRA) in the White House Office of Management and Budget, and Sharon Hays of the White House Office of Science and Technology Policy, is presented as an update of a 1995 memorandum by Clinton administration OIRA chief Sally Katzen.

This memorandum is OIRA’s second recent attempt at a government-wide pronouncement on risk assessment.  In January 2006, OIRA published a proposed set of guidelines to govern all risk assessments. The proposal outlined one-size-fits-all technical standards for all federal agencies to use when conducting risk assessments as well as other scientific documents that are not generally considered “risk assessments.” The OMB guidelines would have applied to risk assessments conducted in preparation for issuing or revising health, safety and environmental rules, as well as important scientific investigations such as the National Toxicology Program’s “Report on Carcinogens,” widely considered the gold standard for identifying substances that are carcinogenic.

OMB submitted the proposal to the public for comment and to the National Research Council for peer review.  The NRC came back in January 2007 with an unexpectedly harsh rebuke of the OIRA proposal, emphasizing “the likely drain on agency resources, the extended time necessary to complete risk assessments that are undertaken, and the highly likely disruptive effect on many agencies” and echoing problems raised by public interest groups, including written comments submitted jointly by Public Citizen and OMB Watch.

The new memorandum reaffirms what OIRA had previously told the press after the NRC report, that it was withdrawing the proposed risk assessment bulletin.  Instead, it reiterates the text of a 1995 memorandum stating general principles for risk assessment, and it supplements the 1995 text with additional comments.

Although the new memorandum couches its language in precatory terms, it nonetheless sends clear signals to the agencies about OIRA’s expectations for “influential” risk assessments (a term borrowed from the controversial 2002 Data Quality Act guidelines):

• Reiterates OIRA control over risk assessments.The memorandum declares that the new guidelines for “significant guidance documents,” which include full notice and comment for “economically significant” guidance as well as review by senior agency officials and OIRA, apply to “influential” risk assessments. [Memo at 13] It also reiterates that other controversial Bush administration policies, including the 2004 peer review guidelines and 2002 Data Quality Act guidelines, govern important assessments.
  
  The most stunning development is the declaration that the new policies for guidance documents apply to "influential risk assessments." Risk assessments are scientific evaluations that apply expert judgment to the weight of the evidence to bridge the gap between what is known and what is not yet known in order to help risk managers craft appropriate public policy. The scientific and regulatory policy communities are clear that risk assessment must be rigidly separated from risk management, with the assessment remaining removed from political considerations. Applying the new guidance document policies will allow industry junk science to delay policymaking and OIRA, an arm of the political offices of the White House, to politicize risk assessment.



• Reiterates OIRA preference for risk ranges.The memorandum continues to emphasize mushy risk ranges at the expense of point estimates.In many contexts, such as deciding whether to order a Superfund cleanup, risk managers must make their decisions based on numbers provided by risk assessors.In the case of a Superfund cleanup, for example, risk managers need to know if the risk of harm to the most exposed individual is minimal (less than one in one million), definitely actionable (greater than one in one thousand), or in a discretionary zone.Providing some large range of value would be meaningless, especially given OIRA’s preference for quantitative expression of uncertainty to splay out the ranges to vaster and vaster dimension.



• Reiterates OIRA call for risk/risk comparisons.The memorandum once again expresses OIRA’s intention to have agencies express risks in terms of other, more familiar risks in the course of conducting risk assessment.Dudley’s predecessor John Graham, who drafted the 2006 proposed risk assessment bulletin, built a career before his appointment at OIRA minimizing risks in the press by comparing fatality risks from exposure to hazardous substances, for example, to the risk of being struck by lightning.In Graham’s hands, risk/risk comparison was a vehicle for anti-regulatory propaganda.



• Reiterates OIRA call for “reproducibility.”Risk assessments are the expression of scientific judgment rather than scientific experiment.Risk assessment is the application of scientific judgment and technical expertise to bridge what is known and what is not known in order to provide assessments that aid agencies in making policy decisions.Risk assessment, which is the application of judgment and expertise, cannot be expected to be “reproducible” in the ways that a scientific experiment can.