Traffic (39 page)

The average person, the criticism goes, is hardly aware of what their chances actually would be of surviving a severe crash while wearing a seat belt or protected by the unseen air bag lurking inside the steering wheel. Then again, as any trip to Las Vegas will demonstrate, we seem quite capable of making confident choices based on imperfect information of risk and odds. The loud, and occasionally vicious, debate over “risk compensation” and its various offshoots seems less about whether it can happen and more about whether it always happens, or exactly why.

Most researchers agree that behavioral adaptation seems more robust in response to direct feedback. When you can actually feel something, it’s easier to change your behavior in response to it. We cannot feel air bags and seat belts at work, and we do not regularly test their capabilities—if they make us feel safer, that sense comes from something besides the devices themselves. Driving in snow, on the other hand, we don’t have to rely on internalized risk calculations: One can feel how dangerous or safe it is through the act of driving. (Some studies have shown that drivers with studded winter tires drive faster than those without them.)

A classic way we sense feedback as drivers is through the size of the vehicle we are driving. The feedback is felt in various ways, from our closeness to the ground to the amount of road noise. Studies have suggested that drivers of small cars take fewer risks (as judged by speed, distance to the vehicle ahead of them, and seat-belt wearing) than drivers of larger cars. Many drivers, particularly in the United States, drive sportutility vehicles for their perceived safety benefits from increased weight and visibility. There is evidence, however, that SUV drivers trade these advantages for more aggressive driving behavior. The result, studies have argued, is that SUVs are, overall, no safer than medium or large passenger cars, and less safe than minivans.

Studies have also shown that SUV drivers drive faster, which may be a result of feeling safer. They seem to behave differently in other ways as well. A study in New Zealand observed the position of passing drivers’ hands on their steering wheels. This positioning has been suggested as a measure of perceived risk—research has found, for instance, that more people are likely to have their hands on the top half of the steering wheel when they’re driving on roads with higher speeds and more lanes. The study found that SUV drivers, more than car drivers, tended to drive either with only one hand or with both hands on the bottom half of the steering wheel, positions that seemed to indicate lower feelings of risk. Another study looked at several locations in London. After observing more than forty thousand vehicles, researchers found that SUV drivers were more likely to be talking on a cell phone than car drivers, more likely not to be wearing a seat belt, and—no surprise—more likely not to be wearing a seat belt
while
talking on a cell phone.

It could just be that the types of people who talk on cell phones and disdain seat belts while driving also like to drive SUVs. But do they like to drive an SUV because they think it’s a safer vehicle or because it gives them license to act more adventurously on the road? To return to the mythical Fred, pickup drivers are less likely than other drivers to wear their seat belts. Under risk-compensation theory, he is doing this because he feels safer in the large pickup truck. But could he not drive in an even more risky fashion yet lower the “cost” of that risky driving by buckling up? It all leads to questions of where we get our information about what is risky and safe, and how we act upon it. Since relatively few of us have firsthand experience with severe crashes in which the air bags deployed, can we really have an accurate sense of how safe we are in a car with air bags versus one without—enough to get us to change our behavior?

Risk is never as simple as it seems. One might think the safest course of action on the road would be to drive the newest car possible, one filled with the latest safety improvements and stuffed full of technological wonders. This car must be safer than your previous model. But, as a study in Norway found,
new cars crash most.
It’s not simply that there are more new cars on the road—the
rate
is higher. After studying the records of more than two hundred thousand cars, the researchers concluded: “If you drive a newer car, the probability of both damage and injury is higher than if you drive an older car.”

Given that a newer car would seem to offer more protection in a crash, the researchers suggested that the most likely explanation was drivers changing the way they drive in response to the new car. “When using an older car which may not feel very safe,” they argued, “a driver probably drives more slowly and is more concentrated and cautious, possibly keeping a greater distance to the car in front.” The finding that new cars crash most has shown up elsewhere, including in the United States, although another explanation has been offered: When people buy new cars, they drive them more than old cars. This in itself, however, may be a subtle form of risk compensation: I feel safer in my new car, thus I am going to drive it more often.

Studying risk is not rocket science; it’s more complicated. Cars keep getting objectively safer, but the challenge is to design a car that can overcome the inherent risks of human nature.

In most places in the world, there are more suicides than homicides. Globally, more people take their own lives in an average year—roughly a million—than the total murdered
and
killed in war. We always find these sorts of statistics surprising, even if we are simultaneously aware of one of the major reasons for our misconception: Homicides and war receive much more media coverage than suicides, so they seem more prevalent.

A similar bias helps explain why, in countries like the United States, the annual death toll from car crashes does not elicit more attention. If the media can be taken as some version of the authentic voice of public concern, one might assume that, over the last few years, the biggest threat to life in this country is terrorism. This is reinforced all the time. We hear constant talk about “suspicious packages” left in public buildings. We’re searched at airports and we watch other people being searched. We live under coded warnings from the Department of Homeland Security. The occasional terrorist cell is broken up, even if it often seems to be a hapless group of wannabes.

Grimly tally the number of people who have been killed by terrorism in the United States since the State Department began keeping records in the 1960s, and you’ll get a total of less than 5,000—roughly the same number, it has been pointed out, as those who have been struck by lightning. But each year, with some fluctuation, the number of people killed in car crashes in the United States tops 40,000. More people are killed on the roads each month than were killed in the September 11 attacks. In the wake of those attacks, polls found that many citizens thought it was acceptable to curtail civil liberties to help counter the threat of terrorism, to help preserve our “way of life.” Those same citizens, meanwhile, in polls and in personal behavior, have routinely resisted traffic measures designed to reduce the annual death toll (e.g., lowering speed limits, introducing more red-light cameras, stiffer blood alcohol limits, stricter cell phone laws).

Ironically, the normal business of life that we are so dedicated to preserving is actually more dangerous to the average person than the threats against it. Road deaths in the three months after 9/11, for example, were 9 percent higher than those during the similar periods in the two years before. Given that airline passenger numbers dropped during this same period, it can be assumed some people chose to drive rather than fly. It might be precisely because of all the vigilance that no further deaths due to terrorism have occurred in the United States since 9/11—even as more than two hundred thousand people have died on the roads. This raises the question of why we do not mount a similarly concerted effort to improve the “security” of the nation’s roads; instead, in the wake of 9/11, newspapers have been filled with stories of traffic police being taken off the roads and assigned to counterterrorism.

In the 1990s, the United Kingdom dropped its road fatalities by 34 percent. The United States managed a 6.5 percent reduction. Why the difference? Better air bags, safer cars? It was mostly speed, one study concluded (although U.S. drivers also rack up many more miles each year). While the United Kingdom was introducing speed cameras, the United States was resisting cameras and raising speed limits. Had the United States pulled off what the United Kingdom did, it is suggested, 10,000 fewer people would have been killed.

Why doesn’t the annual road death toll elicit the proportionate amount of concern? One reason may simply be the trouble we have in making sense of large numbers, because of what has been called “psychophysical numbing.” Studies have shown that people think it’s more important to save the same number of lives in a small refugee camp than a large refugee camp: Saving ten lives in a fifty-person camp seems more desirable than saving ten lives in a two-hundred-person camp, even though ten lives is ten lives. We seem less sensitive to changes when the numbers are larger.

By contrast, in what is called the “identifiable victim effect,” we can be quite sensitive to the suffering of one person, like the victim of a terrible disease. We are, in fact, so sensitive to the suffering of one person that, as work by the American psychologist and risk-analysis expert Paul Slovic has shown, people are more likely to give more money to charity campaigns that feature one child rather than those that show multiple children—even when the appeal features only
one
more child.

Numbers, rather than commanding more attention for a problem, just seem to push us toward paralysis. (Perhaps this goes back to that evolutionary small-group hypothesis.) Traffic deaths present a further problem: Whereas a person in jeopardy can possibly be saved, we cannot know with certainty ahead of time who will be a crash victim—even most legally drunk drivers, after all, make it home safely. In fatal crashes, victims usually die instantly, out of sight. Their deaths are dispersed in space and time, with no regular accumulated reporting of all who died. There are no vigils or pledge drives for fatal car-crash victims, just eulogies, condolences, and thoughts about how “it can happen to anyone,” even if fatal car crashes are not as statistically random as we might think.

Psychologists have argued that our fears tend to be amplified by “dread” and “novelty.” A bioterrorism attack is a new threat that we dread because it seems beyond our control. People have been dying in cars, on the other hand, for more than a century, often by factors presumably within their control. We also seem to think things are somehow less risky when we can feel a personal benefit they provide (like cars) than when we cannot (like nuclear power). Still, even within the realm of traffic, risks seem to be misperceived. Take so-called road rage. The number of people shot and killed on the road every year, even in gun-happy America, unofficially numbers around a dozen (far fewer than those killed by lightning). Fatigue, meanwhile, contributes to some 12 percent of crashes. We are better advised to watch out for yawning drivers than pistol-packing drivers.

Our feelings about which risks we should fear, as the English risk expert John Adams argues, are colored by several important factors. Is something voluntary or not? Do we feel that something is in our control or beyond our control? What is the potential reward? Some risks are voluntary, in our control (we think), and there is a reward. “A pure self-imposed, self-controlled voluntary risk might be something like rock climbing,” Adams said. “The risk is the reward.” No one forces a rock climber to take risks, and when rock climbers die, no one else feels threatened. (The same might be said of suicide versus murder.) Other risks are voluntary but we cede control—for example, taking a cross-country bus trip. We have no sway over the situation. Imagine that you are at the bus station and see a driver drinking a beer at the bar. Then imagine you see the same driver at the wheel as you board your bus. How would you feel? Nervous, I would guess.

Now imagine yourself at a bar having a beer. Then imagine yourself getting in your car to drive home. Did you envision the same dread and panic? Probably not, because you were, at least in your own mind, in control. You’re the manager of your own risk. This is why people think they have a better chance of winning the lottery if they pick the numbers (it is also, admittedly, more fun that way). We get nervous about ceding control over risk to other people. Not surprisingly, we tend to inflate risk most dramatically for things that are involuntary, out of our control, and offer no reward. “The July 7 bombings here in London killed six days’ worth of death on the road,” Adams said. “After this event, ten thousand people gathered in Trafalgar Square. You don’t get ten thousand people in Trafalgar Square lamenting last week’s road death toll.”

Why
is
there no outrage? Driving is voluntary, it’s in our control, and there’s a reward. And so we fail to recognize the real danger cars present. Research in the United States has shown, for example, that exurban areas—the sprawling regions beyond the old inner-ring suburbs—pose greater risks to their inhabitants than central cities as a whole. This despite a cultural preconception that the opposite is true. The key culprit? Traffic fatalities. The less dense the environment, the more dangerous it is. If we wanted dramatically safer roads overnight—virtually fatality-free—it wouldn’t actually be difficult. We could simply lower the speed limit to ten miles per hour (as in those Dutch
woonerven
). Does that seem absurd? In the early 1900s that
was
the speed limit. In Bermuda, very few people die in cars each year. The island-wide speed limit is 35 kilometers per hour (roughly 22 miles per hour). In the United States, to take one example, Sanibel Island, Florida, which like Bermuda has a 35 mph maximum, has not seen a traffic fatality this century, despite a heavy volume of cars and cyclists. But merely lowering mean speeds as little as one mile per hour, as Australian researchers have found, lowers crash risks.