At the top of every ski slope, economist Lester Lave faces a choice. He can loosen the screw on each of his ski bindings so they will release more easily, reducing his risk of serious injury but possibly cutting short an exhilarating run. Or he can throw caution to the wind and tighten the screw, knowing that the binding will release only under great force.

It's only natural that Lave, an expert in risk assessment and president of the Society for Risk Analysis, should see his favorite sport as an exercise in risk management -- a classic tradeoff between fun and safety.

"Every time, at the top of the hill, I make that choice," said Lave, professor of economics at Carnegie-Mellon University in Pittsburgh. "A quarter turn to the right or left."

Usually the method works. But twice he took spectacular falls that resulted in serious injury -- a broken shoulder and a cracked vertebra.

"If any regulator tried to impose on me the magnitude of risk I was imposing on myself at the top of the hill," Lave said, "I would have shot him."

Love Canal, Bhopal, Tylenol, Qaddafi, Challenger, Chernobyl. Cancer, AIDS, toxic shock. Saccharine, asbestos, vinyl chloride. Acid rain, yellow rain, red tide, Agent Orange. VDTs, PCBs, DDT, EDB. This is the age of risk.

Or is it?

"We live in a less and less risky society, and yet we're more and more anxious about it," said Edward Burger, director of Georgetown University's Institute for Health Policy Analysis.

A Harris poll found that more than three of four Americans believed life was riskier in 1980 than 20 years earlier.

"How extraordinary!" wrote Aaron Wildavsky, political scientist at the University of California at Berkeley and coauthor of "Risk and Culture." "The richest, longest-lived, best-protected, most resourceful civilization, with the highest degree of insight into its own technology, is on its way to becoming the most frightened. Has there ever been, one wonders, a society that produced more uncertainty more often about everyday life?

"It isn't much, really, in dispute -- only the land we live on, the water we drink, the air we breathe, the food we eat, the energy that supports us. Chicken Little is alive and well in America."

Nearly 2 million Americans planning trips abroad this year changed their minds in February, apparently out of fear of terrorist attacks, according to the U.S. Travel Data Center. And that was before the U.S. bombing of Libya and the Soviet nuclear accident at Chernobyl.

Statistics suggest they're overreacting -- more Americans drowned in their bathtubs last year than were killed or injured in terrorist attacks. But perception of risk is more than a numbers game. After all, this is a society where an estimated 1,000 Americans die each day from smoking-related causes and about 10,000 die each year because their seat belts weren't buckled -- but the artificial sweetener cyclamate was banned in 1969 as a potential human carcinogen.

What matters to most individuals is not the size of a risk, but whether or not it is seen as acceptable, said Chris G. Whipple, an expert in risk assessment with the Electric Power Research Institute.

"You don't know whether they think the risk is one in two or one in a billion," Whipple said. "You just know they're canceling reservations."

The risk Lave takes on a ski slope is statistically much greater than his risk of being caught in a terrorist attack -- he's planning a family trip to Italy in June -- or being exposed to radiation from a nuclear accident. But skiing is a different kind of risk.

It's voluntary, well known and fun. Though the risk of breaking a bone is considerable, the risk of dying in a ski accident is almost nil. And the skier has at least the illusion of being in control.

"Most people I know feel safer when they're driving an automobile than when they're sitting in the back of an airplane," Whipple said. "The numbers say they're wrong."

The risk per mile of dying in a car is about 10 times that of dying in a commercial airplane.

"After six days in Greece, I'm alive," deadpanned newspaper columnist Colman McCarthy in a recent column from Athens. "So are other Americans here. And Greeks. I have survived danger, let's not minimize it. I was on the fourth hole of the Glyfada golf course when a wild drive from the opposite fairway sliced over my head.

"Terror? It was present whenever I crossed an Athens street and the cars and mopeds surged at me in tidal frenzy."

New York averages 117 murders a month, about twice the annual total for all of Greece, McCarthy pointed out.

There's no such thing, to paraphrase economist Milton Friedman, as a risk-free lunch. More than 3,000 Americans choked to death on food last year; 15 died in elevator accidents.

"Take calculated risks," Gen. George Patton wrote in a letter to his grandson on D-Day. "That is quite different from being rash."

The effort to define that difference is called risk analysis, which puts percentages and dollar signs on chancy human behavior and "acts of God." A couple of decades ago, risk analysis was an arcane concept. Now it's a growth industry (there are 1,100 members of the Society for Risk Analysis) called upon by businesses, utilities, labor unions, insurance companies, hospitals and government agencies for help in estimating and controlling risks from toxic chemicals, accidents and other hazards.

In part it's an offshoot of the consumer and environmental movements of the 1960s and 1970s, signaled by two influential books, Ralph Nader's "Unsafe at Any Speed" and Rachel Carson's "Silent Spring." These movements spawned unprecedented government regulation of automobile safety, air and water pollution, food additives, chemicals and consumer product safety, which in turn prompted demands for better estimates of risk and benefit.

Public interest groups claimed that government and industry failed to protect consumers from a myriad of risks from smoggy air to flammable sleepwear. Regulated industries complained that Congress passed environmental laws with popular but impossible goals, without regard to the costs of cleaning up air, water and land. Regulators were caught in the middle.

"We have tended to do our risk management in an insanely simplistic fashion," Lave said. " We said, 'My God, it's a carcinogen -- ban it!' "

"Basically the lawyers for both sides called up the most extreme scientists they could find on both sides, and they fought it out in adversarial proceedings," said Elizabeth Anderson, former director of the Environmental Protection Agency's office of health and environmental assessment and now president of ICF-Clement, a Washington consulting firm.

When Anderson joined EPA in 1971, "there was no such thing as quantitative risk assessment of chemicals. It just wasn't done."

Any chemical that caused tumors in laboratory animals was by definition a carcinogen, she said, and for any carcinogen the only tolerable goal was "zero risk." Differences in dose, exposure and potency were ignored.

But in 1976 EPA set up a risk assessment office, with Anderson as director. The agency committed itself to reviewing the scientific evidence and applying numbers -- the best estimates available -- to the risks of toxic chemicals.

"For a long time we were out on a limb alone," Anderson said. "Now risk assessment is used everywhere all the time -- used and misused, understood and misunderstood."

Everyone, amateur or professional, is a risk assessor -- the tourist planning a trip to Europe, the motorist deciding whether to buckle up, EPA officials testing the safety of some 60,000 commonly used chemicals. No matter what the context, the risk assessor's question is the same: How safe is safe enough?

Rarely is the answer simple.

"Thinking about uncertainty is very difficult for people -- for experts as well as laypeople," said Paul Slovic, a psychologist and expert in risk analysis at Decision Research in Eugene, Ore.

Experts know a lot about the technical aspects of, say, nuclear power or food chemistry. But they tend to define risk narrowly, in terms of factors that can be reduced to numbers, such as how many people died last year from a given hazard.

The public, on the other hand, often cares less about numbers than about other factors, such as whether a hazard is new or old, known or unknown, controllable or uncontrollable, voluntary or involuntary, and whether or not it is potentially catastrophic. Fear of nuclear radiation, for example, is heightened by each of these considerations.

In short, the experts worry mainly about quantity of risk, while laypeople worry about quality of risk.

"There's a kind of wisdom and foolishness on both sides," Slovic said.

Experts can offer the public numbers and estimates to put risks into perspective. But public perceptions can warn the experts against using sheer statistics to compare risks as different as living near a nuclear reactor and drinking diet sodas.

Several years ago Slovic and two other psychologists at Decision Research, Baruch Fischhoff and Sarah Lichtenstein, asked four groups of people to rank 30 technologies and behaviors in order of their riskiness. The groups were college students, League of Women Voters members, business and professional leaders, and experts in risk assessment. Box, Page 15.

There was some agreement -- all four groups ranked motor vehicles, smoking and handguns among the top five risks. But there were also many differences.

Nuclear power, for example, was ranked first by the college students and the League of Women Voters members, eighth by the business and professional leaders and 20th by the risk experts. Swimming was ranked 10th by the experts but 30th by the college students. Skiing was ranked 16th by the civic club members and 30th by the risk experts. X-rays were ranked seventh by the experts, 22nd by the League members.

In general, according to another study by Slovic and his colleagues, people overestimate rare but highly publicized causes of death -- such as botulism and tornadoes -- and underestimate common causes that kill one person at a time -- such as asthma, emphysema and stroke.

Homicides and stroke were seen as about equally common causes of death, when in fact stroke kills 11 times as many. People estimated that tornadoes and asthma kill about 500 Americans a year, when in fact tornadoes kill about 100 a year and asthma kills about 3,000.

The way to size up a risk is to compare it with other, more familiar risks, said Bernard Cohen, a physicist at the University of Pittsburgh. He has compiled a life expectancy reduction (LER) scale that ranks hazards according to how much they shorten the average American's life.

Heart disease tops the list by lopping 2,100 days, or nearly six years, off the average American's life. Cigarette smoking, cancer, being overweight and being unmarried all have an LER of several years. By contrast, tornadoes shorten the average life span by a single day.

Cohen's scale ranks nuclear power much safer than coal, because no American has died in an accident at a commercial nuclear power plant but thousands die every year from coal-related air pollution and mining accidents.

A fervent advocate of nuclear power, Cohen blames the news media for "grossly exaggerating" the risks of radiation.

By the LER scale, the average radiation exposure to nearby residents from the Three Mile Island nuclear accident in 1979 -- about one millirem -- is as risky as three street crossings or three extra miles driven in a car.

Such comparisons are considered seriously flawed by many risk analysts and ridiculed by some as "kindergarten lists."

"It's comparing apples and oranges," Slovic said. Purely statistical comparisons between a nuclear plant meltdown and walking across the street fail to account for important differences in the two types of uncertainty.

Although more people died last year crossing the street than in nuclear accidents, the public views nuclear accidents as much more frightening, because of what Slovic calls the "dread factor." The risks of a nuclear accident are seen as unknown, uncontrollable and potentially catastrophic.

Human beings are pretty good at learning from experience, Slovic said, but less good at dealing with abstract concepts such as risk.

"Our traditional way of dealing with risk," he said, "is trial and error. If the fire's too hot, you back off. If you need to get warm, you get closer. But our conceptual abilities are much more limited."

People -- especially voters and sports fans -- also have a hard time distinguishing between a bad decision and bad outcome. (Winning the lottery is a good outcome from a bad decision; Lester Lave's skiing accidents were bad outcomes of good decisions.)

Intuition often conflicts with probability. Most people are astonished to learn that if 50 people are in a room, there is a 97 percent chance that two of them have the same birthday. And studies suggest that it makes little difference to the public whether a given risk is one-in-a-thousand or one-in-a-billion.

Scientists who try to emphasize the smallness of a risk by comparing it to one crouton in a five-ton salad or a drop in a railroad tank car, Slovic said, only succeed in making the risk more easily imaginable. "We find it very hard to think about probability," he said. "It's not wired into our nervous system."

Studies show that the vast majority of people not only believe they are better-than-average drivers but also are unrealistically optimistic about their chances of living past 80 and avoiding a heart attack.

Habits also influence perceptions of risk. A British study in 1979 found that only half of all smokers, compared with 89 percent of nonsmokers, believed smoking was "really as dangerous as people say." Nonsmokers were twice as likely as smokers to realize that smoking causes more deaths than traffic accidents.

And those who asserted that people have a right to take health risks, as opposed to a moral responsibility not to risk their health, were more likely to smoke and less likely to wear seat belts.

Risk assessment is at best an imperfect science. Estimating the risk of a potentially cancer-causing pesticide or food additive, for example, depends on a myriad of factors, some of which are unmeasurable or even unknown.

Risk estimates themselves run the risk of losing credibility when they fail to acknowledge the uncertainty by including a possible range of error, cautioned Georgetown's Burger.

"There are a lot of pressures for simple numerical answers -- from regulators, lawyers, economists, the public itself," Burger said. "There's a strong push to put numbers on risk, and yet those numbers often are so fuzzy that they can give a misleading impression of accuracy and confidence."

The Journal of Irreproducible Results, a tongue-in-cheek publication that spoofs statistics and modern science, has warned that if everyone keeps stacking National Geographic magazines in garages and attics instead of throwing them away, the weight will soon sink the continent 100 feet, causing extensive flooding. The journal also reported that 80 percent of the U.S. coastline will disappear in 10 years if beachgoers keep returning home with sand clinging to their bodies. Many risk analysts see the public's preoccupation with risk as ironic. Some sociologists suggest that leisure-oriented baby-boomers have the "luxury" of time to worry about risk. Whatever the reason, Americans today seem less willing than their forbears to accept risk as a price of living.

And efforts to manage risk will always be up against two formidable obstacles: the limitations of scientific data and the idiosyncrasies of the human psyche.

"The best we can hope for," concluded psychologist Fischhoff, "is some muddling through intelligently."