Dead salmon litter the mouth of the Klamath River near Klamath, Calif., after federal authorities diverted water.
		AP file, 2002

And then there are the politicians.

Political battles surrounding scientific questions are nothing new, but a new debate has emerged in recent years over "sound science," a phrase used by both presidents Clinton and Bush to describe the basis of their administrations' regulatory decisions.

Not a term used by scientists, sound science has come to mean new rules for determining what kind of scientific evidence can be used to shape regulations.

Five environmental bills now before the 108th Congress have sound-science provisions. The phrase has previously appeared in legislation dealing with endangered species, logging on public lands, public health and other topics.

Among lawmakers, sound science has come to mean a preference for scientific data based on real observations — rather than models or expert judgment — that have been heavily peer-reviewed by outside scientists.

Among critics, sound science has come to mean the selective use of such data to justify a certain agenda.

'Objectivity and utility'

That's too cynical, says risk analysis expert John Graham, the Bush administration's point man on sound science. "Two key ideas are objectivity and utility."

Graham heads the Office of Information and Regulatory Affairs at the Office of Management and Budget. Sound science, he says, means "it is important that (scientific) claims, whether for or against a hypothesis, should be replicated before they drive public policy."

He has championed a "cost-effectiveness" analysis approach to regulation in which scientists estimate the probable effects, with 0% to 100% likelihood, of different approaches to solving a problem. Decision makers should rationally weigh how much risk they can afford based on those numbers. His own work has involved auto safety, for example, weighing the cost of requiring cars to have airbags against the number of lives saved.

When it comes to sound science, "we know it when we see it," says Mike Catanzaro, a spokesman for the Senate Committee on Environment and Public Works. Without sound science, he says, the committee sees "an irrational approach, first regulations and then science," practiced by such agencies as the Environmental Protection Agency and the Department of the Interior.

Critics take a dimmer view. "Sound science is just a political buzzword used to denigrate somebody else's best science in a conflicted area," says science and public policy expert Rustum Roy of Penn State. Risk analysis can easily be distorted, he says, either by scientists or politicians backed by special interests. "Distrust anyone who uses the phrase. Sound science cannot be encompassed in a sound bite," Roy says.

In December, Democrats on the House Committee on Resources attacked the idea in a report called Weird Science: The Interior Department's Manipulation of Science for Political Purposes.

How sound the science?

Debate over the role of science in public policy has intensified greatly in the past few years:

• Part of the battle over oil drilling in the Arctic National Wildlife Refuge, a National Academy of Sciences report found that drilling both helped and hurt the wilderness. Displeased, Sen. Ted Stevens, R-Alaska, suggested that Congress curtail payments for such reports.
• A 2001 "sound science" decision by the EPA to postpone a drop in the amount of arsenic allowed in drinking water was rescinded after public outcry.
• Graham has piled up two "Clean Air Villain of the Month" awards from the Clean Air Trust for his role in the EPA's decision to lower its estimate of the benefits of clean-air controls. "He has promoted a view that isn't sound science at all — he is simply wielding a political ax aimed at helping corporate polluters avoid pollution controls," the environmental group contends.

"Science is the centerpiece. It tells us how to accomplish our objectives without spending our resources fruitlessly," says University of Colorado freshwater scientist William Lewis. As chair of the National Academy of Sciences' Committee on Endangered and Threatened Fishes, Lewis sits in the middle of one sound-science debate — the Klamath River showdown.

On the Klamath

Endangered coho salmon need certain water levels to survive, and other fish are guaranteed to Native American tribes under federal agreement. Upstream, farmers need the water for crops. In the past, farmers have forced open canals and called for an end to Endangered Species Act protection for the fish.

Last year, Lewis' committee released an "interim" report on the Klamath that narrowly examined the scientific basis of decisions to restrict water used for irrigation, and found them inadequate. Department of the Interior scientists, the committee found, relied partly on judgment instead of field data to reserve water for flow downstream. The interim report didn't say the judgment was wrong, simply that it was not based on tested evidence. Still, partly in response to that report, the Department of the Interior opened the gates for irrigation, lowering the amount of water going downstream.

The decision killed one-quarter of the salmon downstream last fall, according to the California Department of Fish and Game. "It's a nice, fairly concrete science controversy standing by itself, but you bring in the politics and legal aspects and you have a situation that has blown up," says Kristen Boyle, an attorney with the environmental group Earthjustice in Seattle.

Things were expected to be dry on the Klamath this year, too, so when the irrigation canals were opened again on April 1, the water levels were kept at 75%, according to Jeff McCracken of the Bureau of Reclamation. Fortunately, it's a wet year, but, even so, Department of Interior reports on the fish kill missed their deadlines to have any impact on irrigation decisions.

A "Best Science" document by the American Fisheries Society calls the Klamath River an example of the downside of sound-science rules: While regulators wait for peer-reviewed studies, endangered species may expire. "The concern is that policymakers are defining science rather than scientists," the society contends.

Advocates and critics of sound science agree that scientific method can only provide estimates of risk, not infallible predictions, for most situations. Even with the best information, policymakers still face tough decisions. "Science is important, but no substitute in many cases for human judgment," Graham says. "Science serves to inform that judgment"