WHAT SCARES YOU more -- flu or mad cows? If we were being rational about it, we would do some serious fretting about runny noses but keep munching on that steak. (Ratio of flu deaths to mad-cow-disease deaths in the United States last year: 36,000 to 0.) Nuclear power vs. sunburn? While Chernobyl caused 2,000 cases of cancer, mostly nonfatal, among Russians, melanoma kills 7,800 Americans annually.

These are the kinds of arguments brandished by economists who specialize in risk analysis, a field that's moved from obscurity to the corridors of power in the last two decades. The discipline truly arrived in 2001, when John D. Graham, founder of the Harvard Center for Risk Analysis, at the Harvard School of Public Health, moved to Washington to become the nation's ''regulatory czar.'' From his perch in the Office of Information and Regulatory Affairs (OIRA) in the Office of Management and Budget, he now has the power to veto rules issued by agencies like the Environmental Protection Agency that he and his staff deem insufficiently scientific.

Environmentalists dislike Graham because he has long argued that regulations to, say, eliminate the last traces of toxin in drinking water often aren't worth the cost -- as much as billions of dollars per life saved, according to his calculations. In contrast, regulations requiring collapsible steering columns in cars save thousands of people at minimal cost. Lately, Graham has taken to arguing that it is time to apply his methods to the endeavors of the Department of Homeland Security.

But now Tufts economist Frank Ackerman and Georgetown law professor Lisa Heinzerling have mounted an assault on risk analysis as currently practiced, placing Graham and his center squarely in their sights. Their new book, ''Priceless: On Knowing the Price of Everything and the Value of Nothing'' (New Press), is an unwelcome-to-office present for the new director of the Harvard center, James K. Hammitt, promoted from within three weeks ago. (Among other things, Hammitt is known for his research on air pollution regulation in Mexico City.)

Cost-benefit analysis begins by examining how much ordinary people typically pay to avoid risk in their own lives: A miner takes a $5,000 pay cut to avoid a dangerous assignment, for example, while a yuppie skips the Volvo and buys a less safe econobox in order to save $8,000. Economists extrapolate from these figures to estimate how much someone would be willing to pay to reduce, say, a one-in-10,000 chance of dying from a carcinogen to one-in-a-million. The theory is that government shouldn't spend millions of dollars per person to eliminate risks people happily tolerate daily.

But if a person drives a Honda instead of a Volvo, critics like Ackerman and Heinzerling ask, does that really explain whether she will tolerate pollutants in her water? Whatever economists say, it is not irrational to feel heightened ''dread,'' the authors argue, in the face of chemicals and toxins, especially since the long-term genetic effects are unknown.

And despite economists' claims of greater rigor, Ackerman says, ''you actually lose information'' when putting things like cancer and loss of species into monetary terms: ''You lose clarity about how many deaths, how many diseases, how much damage to the ecosystem.'' Besides, he argues, current processes are antidemocratic: Rather than having Graham's office and the EPA bicker over Byzantine economic models, why not just put the issues to a vote -- and see what real live people actually think?

Cass R. Sunstein, a law professor at the University of Chicago whose 2002 book ''Risk and Reason'' defends a liberal version of what he calls the ''cost-benefit state,'' says he thinks the authors score some points when they criticize the fuzzy numbers in the field. But to suggest, as they do, that people who study environmental costs and benefits are mere shills of industry is ''beneath the dignity of the book's better arguments.'' Ackerman and Heinzerling also ''don't adequately acknowledge that cost-benefit analysis has sometimes been a tool for more aggressive regulation.'' Under Graham, for example, OIRA is pushing the Food and Drug Administration to order labels placed on foods that contain dangerous trans fatty acids -- a potential boon for public health.

As an alternative to economic models, the authors hold up the somewhat fuzzy ''precautionary principle'' as an ideal. But the executive director of the Harvard center, George M. Gray, counters that while ''the precautionary principle tells us how we want to behave -- we want to be careful -- it doesn't tell us what to do'' about mad cow or dirty water. The only way forward, he says, is to measure the dangers, put price tags on the different options (ban all beef? do nothing?) and -- yes -- start crunching the numbers.

Christopher Shea's column appears in Ideas biweekly. E-mail: critical.faculties@verizon.net.

This story ran on page C5 of the Boston Globe on 1/4/2004.
© Copyright 2003 Globe Newspaper Company.