One in every three Americans will get cancer and as many as one in five will die from it. While few people would oppose a full-scale effort to eradicate this scourge, legitimate questions need to be asked about some of the federal government's methods in waging war against it.

An important task force report released in Michigan by The Mackinac Center calls into question the standards and procedures of the federal Environmental Protection Agency in evaluating certain public health risks. For example, despite the fact that industrial products and food additives cause less than 3 percent of all cancers (by EPA's own estimates), the agency's flawed approach is imposing billions of dollars of costs on the American public through regulations of dubious value.

Consider EPA's process of calculating potential risks from exposure to a suspected air pollutant, which is done by testing the chemical for toxicity in laboratory animals.

The substance is administered to rats and mice in massive daily doses just below the amount that would kill them immediately. The EPA then extrapolates from rodents to humans and estimates the human risk of cancer from exposure to the chemical. One problem with this is that at these high levels of exposure, one out of every two chemicals ever tested (both natural and man-made) eventually causes cancer in at least one species of rodent.

Another problem is that about one-quarter of the tests that produced cancer in mice failed to do so in rats, and vice-versa. Since rats and mice are biologically similar and both are dissimilar from humans, extrapolating to statements about risks for people is, as the task force report puts it modestly, "a considerable leap."

To calculate the "risk" to human populations, the EPA postulates an imaginary "Most Exposed Individual" (MEI) who lives right on the property line of the emissions source and breathes the highest level of emissions from that source for 70 years, 24 hours each day. The agency then assumes that everyone is an MEI.

Even with these almost ridiculously pessimistic assumptions, the EPA figures that only 1,700 to 2,700 cancers are caused each year by exposure to a wide range of 90 potentially hazardous air pollutants, a small fraction of the nearly one million cancer cases occurring each year in America.

Assuming the generous, that EPA's risk assessments are accurate, it's quite apparent that the cost of preventing cancer through EPA regulations is extremely high. The toxics section of the amended Clean Air Act, for instance, will cost between $20 billion and $30 billion, 10 to 15 times the entire budget of the National Cancer Institute. That translates into a cost of between $40 million and $86 million per cancer case avoided!

Realistic calculations indicate that the EPA's regulations on coke emissions will cost $682 million for each instance of cancer they prevent. Cost seems to be no object at the EPA. Its new benzene regulations, by its own admission, will impose a cost of $200 million a year to prevent just 3.4 cases of cancer. Perhaps this is what an EPA executive had in mind when he said that the most effective way to combat cancer would be to give the entire EPA budget to the American Cancer Society.

Most Americans would be surprised to learn that nature itself is a virtual carcinogen factory. About 99.99 percent of the pesticides we eat are natural (10,000 times the level of synthetic pesticide residues) and half of the natural pesticides tested in rodents have proven to be carcinogenic at very high dosage levels, as our recent study points out.

The amount of carcinogens produced by cooking--in the browned and burned foods we eat in a day--is several hundred times greater than the amount we inhale when we breathe severely polluted air. Indeed, the numbers show that virtually all dietary cancer risks are trivial compared to the actual risk the average American takes just driving to work each day.

Cancer risks are not to be laughed at, but neither should they be exaggerated beyond reason. Failure to put them in proper perspective can actually be counterproductive: regulations which unduly restrict the use of pesticides, for instance, can raise the prices of fruits and vegetables, which in turn leads to reduced consumption and increased cancer rates. Little or no regard for the costs of regulation is an invitation for expensive abuse by regulatory bodies.

Perhaps we are overdue for a realistic assessment of how the federal government assesses cancer risks.