As this issue goes to press, the National Academy complex[1] regards itself as under fire after Animal Legal Defense Fund, Inc. v. Shalala[2] found it subject to the Federal Advisory Committee Act (FACA).[3] This is notably reflected in a recent Science editorial, Raiders of the Last Bastion.[4] There, M. R. C. Greenwood, points out how this will interfere with prompt and apolitical advice:
It takes 4 or 5 months to get a charter for a FACA commit-tee.... [T]he government has set a cap on the total number of FACA committees..., making it... impossible to create 400 new committees.... The regulations... also require... appropriate balance... by considering individuals who are affected and interested as well as expert, so that political considerations become a factor in committee membership.Streamlining and facilitating creation of Academy committees seems warranted, but openness and, to the extent possible, balance seem to offer the most reliable control for biases that cannot and probably should not be eliminated.
Greenwood acknowledges that "FACA's aim to make the advisory process more open" is "reasonable and important" and notes that the Academy has opened information-gathering. He then gives two reasons for closed deliberations: To prevent funding sources from exerting pressure on committee members and to ensure that reports are not made public prior to review satisfying "the Academy's standards of evidence." Neither is compelling.
As recently emphasized by Carl Cranor,[5] merely selecting a "standard of evidence" is inherently normative (or "political" in a narrow sense). Such issues should be openly addressed.[6] Moreover, as noted elsewhere in this issue of Risk:[7]
Compared with the period before about 1970, it is unlikely that any scientific or technical group continues to be politically detached. Most are intensely involved through an elaborate system of consultancy, advisory committees and participation as legal expert witnesses.Some have made a case that nothing is purely objective,[8] but, to the extent that this is true, the inherence of normative bias does not compel undifferentiated treatment. The key questions are: How to categorize types of bias, and what are appropriate controls in various circumstances? For example, in contrast with fact, outcome or party bias, Judge J. Skelly Wright found policy ("political" in the broadest sense) bias acceptable, even desirable, in the context of agency rule making:[9]
The appellees have a right to a fair and open proceeding; that right includes access to an impartial decisionmaker. Impartial, however, does not mean uninformed, unthinking, or inarticulate.Greenwood ends his editorial with a call to arms: "It is time for those who care about scientific, not political advice, to speak out." He also asks the Supreme Court or Congress to "fix the problem."
This seems to ignore the fact that most, if not all, advice sought by agencies or Congress is implicitly, if not explicitly, "political" -- in the sense Greenwood contemplates. Scientists, ever less apt to be oblivious to effects of their advice personally, professionally or socially, cannot be isolated. Indeed, it seems difficult to support the proposition that, were it humanly possible, those offering scientific advice should be utterly heedless of its implications.
[2] 104 F.3d 424 (D.C. Cir. 1997).
[3] Pub. L. 92-463, Oct. 6, 1972, 86 stat. 770, codified at 5 U.S.C. App. 2 (1996).
[4] 277 Science 163 (July 11, 1997). All quotations infra are taken from this page.
[5] The Normative Nature of Risk Assessment: Features and Possibilities, 8 Risk 123 (1997).
[6] See, e.g., Thomas H. Pigford, Maximum Individual & Vicinity-Average Dose for a Geologic Repository Containing Radioactive Waste, 8 Risk 9 (1997).
[7] Halina S. Brown et al., Reassessing the History of U.S. Hazardous Waste Disposal Policy..., infra 249, 271.
[8] See, e.g., Sheila Jasanoff, Procedural Choices in Regulatory Science, 4 Risk 143, 145 (1993).
[9] Association of National Advertisers, Inc. v. Federal Trade Commission, 627 F.2d 1151, 1174 (D.C. Cir. 1979), cert. den. 447 U.S. 921 (1980) (emphasis added).
