SCIENCE POLICY-WATCHERS now have the President’s Council on Bioethics in their crosshairs. As the council has Bush’s ear on issues such as stem-cell research, many scientists complain that the council’s membership, revised at the end of February, tilts too far to the pro-life, anti-therapeutic cloning right.
But questions of the Bush administration’s scientific integrity have been in the air for several months. In mid-February the Union of Concerned Scientists released a report titled Scientific Integrity in Policymaking: An Investigation into the Bush Administration’s Misuse of Science. Last August Rep. Henry Waxman, ranking Democrat on the House Committee on Government Reform, weighed in with his own report, Politics and Science in the Bush Administration.
Both documents charge that the Bush administration had “manipulated the scientific process . . . generating unprecedented criticism from the scientific community” (as the Waxman report put it). What are we to make of this allegation?
Some of the accusations seem valid. If, as persuasively alleged, the National Cancer Institute website was changed to say that women who undergo abortions have a greater chance of developing breast cancer, then medical reality was indeed distorted. (There is, in fact, no connection.) And if the Centers for Disease Control removed accurate details regarding condom use in preventing HIV and replaced them with information about the virtues of abstinence, this would seem to be a wrongheaded and regrettable purging of health information.
Both reports also question the composition of scientific advisory panels. Specifically, they raise the case of psychologist William R. Miller, an addiction researcher at the University of New Mexico under consideration in 2002 for a slot on the scientific advisory committee of the National Institute on Drug Abuse. I personally recommended Dr. Miller and was dismayed when an HHS “vetting” aide asked me about his views on abortion and whether he had voted for Bush. In my opinion such factors are unrelated to his ability to evaluate research evidence, but Dr. Miller may have given the “wrong” answers because he was not offered the position.
But unmentioned in the critical reports is that views on abortion and voting habits were not the only criteria the Bush White House had in mind. The aide also asked me to recommend researchers who were Hispanic, African American, or female. These traits are similarly irrelevant to analytic abilities but are not ones I have heard Waxman or the Union of Concerned Scientists complain about. They could not do so and still claim with a straight face, as the UCS does, that the Clinton administration applied no litmus tests in filling scientific panels. Theirs were just litmus tests of a different, liberal kind.
It is common knowledge that academic experts in behavioral and public health and the environmental sciences are, on average, left-leaning. After September 11 the first point of the “Guiding Principles for a Public Health Response to Terrorism” put out by the American Public Health Association–the major professional organization representing public health academics–was not surveillance issues in biological warfare but an injunction to “address poverty, social injustice and health disparities that may contribute to the development of terrorism.” Much of the research on social problems, such as teen pregnancy, crime, addiction, “environmental justice,” and “racism” (the latter explicitly designated on solicitations for HHS grant proposals) is conducted by academic psychologists. Their liberal tendencies have been amply documented in professional surveys on party affiliations among academics.
Opportunities for conflict between a conservative administration and left-of-center academics are thus greater than under liberal political leadership. But that doesn’t mean that advisory groups are any more politicized now than before.
In 2001, for example, when I joined the advisory committee at the federal Center for Mental Health Services, I was the only member who did not fall into lockstep agreement with my fellow advisers, all Clinton-era holdovers, that activist groups should be given government funding to pursue the clinically irresponsible mission of discouraging severely ill psychiatric patients from taking their medications.
So it is little coincidence that the reports attacking Bush in an election year come from entities with impeccable liberal credentials. The Union of Concerned Scientists is pro-Kyoto protocol, anti-nuclear weapons, and strongly critical of genetically modified food. Henry Waxman is a liberal icon in the House. Nonetheless, they have shed light on an important issue: how science advisory committees should work.
They should shun litmus tests, strive for balance, make deliberations transparent, and disclose possible conflicts of interest. More important is the need to be concerned about what Michael Crichton calls “consensus science.” In a 2003 lecture at Cal Tech, the physician-anthropologist-filmmaker-science writer cautioned that when scientific uncertainty is especially great–as in differing models of global warming scenarios–agreement can be deceptive: It can simply mean that all scientists chosen for a committee happen to endorse one of several plausible, though ultimately incorrect, models.
In a striking example of consensus substituting for science, Crichton cites the secondhand smoke issue. In 1993 the EPA under Carol Browner announced that secondhand smoke caused 3,000 lung cancer deaths a year. The evidence at the time clearly did not support the claim that secondhand smoke was carcinogenic, yet the EPA rated it as a Class A carcinogen, meaning it was a proven cancer-causing agent in humans. “Eighteen well-respected scientists looked at [our report and] agreed with the conclusion [the EPA] reached,” Browner said. Five years later a federal judge held that the EPA “disregarded information and made findings on selective information.”
NO WONDER Crichton regards consensus science as “extremely pernicious…a way to avoid debate by claiming the matter is already settled.” The key to protecting the advisory process from being hijacked by consensus, he suggests, is to separate the generation of findings from their verification. By having discrete teams of scientists who check each other–one to decide how to gather the data, another to actually gather them, and yet another to analyze them–the advisory process would have more built-in opportunities for self-correction.
The indignation now aimed at the Bush administration for spinning its science or tilting its committees is highly selective. The reality is that other administrations have done so too, just spinning in the other direction. Both are wrong. Scientific panels need to be constructed to ensure that biases of all kinds are, as much as possible, kept out.
Sally Satel is a resident scholar at the American Enterprise Institute.
