So much of what we think and do about risk does not make sense. In a 1990 paper, researchers George Loewenstein and Jane Mather compared people’s levels of concern about nine risks—including AIDS, crime, and teen suicide—with objective measures of those risks. The results can only be described as scrambled. In some cases, concern rose and fell as the risk rose and fell. In others, there was “wild fluctuation” in levels of concern that had absolutely no connection to the real risk. “There is no generally applicable dynamic relationship between perceived and actual risk,” the researchers politely concluded.

There are countless illustrations of our confused and confusing relationship with risk. The single greatest risk factor for breast cancer is age— the older the woman, the greater the risk—but when a 2007 survey by Oxford University researchers asked British women when a woman is most likely to get breast cancer, more than half said, “Age doesn’t matter.” One in five thought the risk is highest when a woman “is in her 50s”; 9.3 percent said the risk is highest “in her 40s”; and 1.3 percent said “in her 70s.” A grand total of 0.7 percent of women chose the correct answer: “80 and older.” Breast cancer has been a major public concern and topic of discussion since at least the early 1990s, and yet the survey revealed that the vast majority of women still know nothing about the most important risk factor. How is that possible?

In Europe, where there are more cell phones than people and sales keep climbing, a survey found that more than 50 percent of Europeans believe the dubious claims that cell phones are a serious threat to health. And then there’s the striking contrast between Europeans’ smoking habits and their aversion to foods containing genetically modified organisms. Surely one of the great riddles to be answered by science is how the same person who doesn’t think twice about lighting up a Gauloise will march in the streets demanding a ban on products that have never been proven to have caused so much as a single case of indigestion.

In Europe and elsewhere, people tremble at the sight of a nuclear reactor but shrug at the thought of having an X-ray—even though X-rays expose them to the radiation they are terrified might leak from a nuclear plant. Stranger still, they pay thousands of dollars for the opportunity to fly somewhere distant, lie on a beach, and soak up the radiation emitted by the sun—even though the estimated death toll from the Chernobyl meltdown (9,000) is actually quite modest compared to the number of Americans diagnosed with skin cancer each year (more than one million) and the number killed (more than 10,000).

Or compare attitudes about two popular forms of entertainment: watching car races and smoking pot. Over a five-year period, NASCAR drivers crashed more than three thousand times. Dale Earnhardt’s death in 2001 was the seventh fatal smashup in seven years. Governments permit NASCAR drivers to take these risks, and the public sees NASCAR as wholesome family entertainment. But if a NASCAR driver were to relieve post-race stress by smoking marijuana, he would be subject to arrest and imprisonment for possession of a banned substance that governments worldwide have deemed to be so risky that not even consenting adults are allowed to consume it—even though it is impossible for someone to consume enough to cause a fatal overdose.

The same logic applies to steroids and other forms of doping: One of the reasons that these substances are banned in sports is the belief that they are so dangerous that not even athletes who know the risks should be allowed to take them. But in many cases, the sports those athletes compete in are far more dangerous than doping. Aerial skiing—to take only one example—requires a competitor to race down a hill, hurtle off a jump, soar through the air, twist, turn, spin, and return to earth safely. The slightest mistake can mean a headfirst landing and serious injury, even a broken neck. But aerial skiing isn’t banned. It’s celebrated. In the 2006 Olympics, a Canadian skier who had broken her neck only months before was lionized when she and the metal plate holding her vertebrae together returned to the slopes to once again risk paralysis and death. “I would prefer my child take anabolic steroids and growth hormone than play rugby,” a British scientist who studies doping told the Financial Times. “I don’t know of any cases of quadriplegia caused by growth hormone.” The same is all the more true of American football, a beloved game that snaps the occasional teenage neck and routinely turns the stars of the National Football League into shambling, pain-racked, middle-aged wrecks.

Handguns are scary, but driving to work? It’s just a boring part of the daily routine. So it’s no surprise that handgun killings grab headlines and dominate elections while traffic accidents are dismissed as nothing more than the unpleasant background noise of modern life. But in country after country—including the United States—cars kill far more people than do handguns. In Canada, twenty-six people die in car crashes for every one life taken by a handgun. And if you are not a drug dealer or the friend of a drug dealer, and you don’t hang out in places patronized by drug dealers and their friends, your chance of being murdered with a handgun shrinks almost to invisibility—unlike the risk of dying in a car crash, which applies to anyone who pulls out of a driveway.

Daniel Gardner, The Science of Fear