Authors: Tom Vanderbilt
The fact that Fred is divorced puts him in a riskier pool. A French study that looked into the experiences of some thirteen thousand company employees over eight years found that a recent divorce or separation was linked to a fourfold increase in the risk of a crash that could be at least partially attributed to the driver. One could hypothesize many reasons: There’s the emotional stress (as John Hiatt once sang in a breakup song, “Don’t think about her while you’re trying to drive”), and perhaps more drinking. Or there may be lifestyle changes, like driving more to visit the kids on weekends. Perhaps people who get divorced are simply the type of people who take more risks. Fred might take some comfort, however, from a New Zealand study that found that people who have never been married have even a higher crash risk than those who are divorced. (The study took into account age and gender differences.)
Fred may not have a life partner, but he should be glad if you chose to join him in his truck: Passengers seem to be a life-saving device. Studies from Spain to California have come to the conclusion that a driver has a lower chance of being in a fatal crash if there’s a passenger. This holds particularly true for middle-aged drivers—especially when the passenger is a woman and the driver is a man. (Whether this stems from men looking out for women or women telling men to drive more safely is open to debate.)
The exception here is teenage drivers. Teens are less likely to be wearing seat belts and more likely to be drinking when driving when there
are
passengers in the car. Many studies have found that teen drivers are more likely to crash with passengers onboard, which is why, in many places, teens are restricted from carrying passengers of their own age during their first few years of driving.
Researchers are beginning to uncover fascinating things about how that risk plays out. A study that looked at the drivers exiting the parking lot at ten different high schools found that teenage drivers seemed to drive faster and follow cars at closer distances than other drivers did. Males drove more riskily than females. This is common knowledge, verified by insurance rates. But their risk-taking varied: Male drivers drove faster and followed closer when they had a male riding shotgun. When they had a female in the front seat, they actually behaved
less
riskily, and they were safer still when they drove by themselves (a pattern that also held for female drivers).
What seems to be a need to impress in the presence of males turns into a protective impulse when a female passenger (possibly a girlfriend) is in the car—or it could be that the female passenger serves as the voice of reason. This “girlfriend effect” seems to take root early and persist through later life. It need not be a romantic partner: The Israel Defense Forces, in an effort to reduce road deaths for soldiers on leave, trains female soldiers (dubbed “angels”) to act as a “calming” influence on their male comrades.
Now consider
where
Fred is driving. What’s the matter with Montana? In 2005, 205 people were killed on Montana’s roads, roughly one-third the number that were killed in New Jersey. But Montana has just under one-tenth the population of New Jersey. People clearly drive more in Montana, but even adjusting for what is known as VMT (or “vehicle miles traveled”), Montana drivers are still twice as likely as New Jersey drivers to die on the roads. The big culprit is alcohol: Montana drivers were nearly three times as likely as New Jersey drivers to be involved in an alcohol-related fatal crash. Montana also has higher speed limits than New Jersey, and fewer chances to get caught violating traffic laws. And, most importantly, most Montana roads are rural.
There is, in theory, nothing nicer than a drive in the country, away from the “crazy traffic” of the city. But there is also nothing more dangerous. We would all do well to heed what the sign says:
IT’S GOD COUNTRY, DON’T DRIVE LIKE HELL THROUGH IT.
Rural, noninterstate roads have a death rate more than two and half times higher than all other roads—even after adjusting for the fewer vehicles found on rural roads. Taking a curve on a rural, noninterstate road is more than six times as dangerous as doing so on any other road. Most crashes involve single cars leaving the roadway, which suggests poorly marked roads, high speeds, fatigue or falling asleep, or alcohol—or some combination of any or all of these. When crashes do happen, medical help is often far away.
In Fred’s case, he
is
the medical assistance. But what of the fact that he is a doctor? Why should that be a risk? Doctors are usually well-educated, affluent, upstanding members of the community; they drive expensive cars in good condition. But a study by Quality Planning Corporation, a San Francisco–based insurance research firm, found doctors to have the second-highest crash risk in an eight-month sample of a million drivers, just after students (whose risk is largely influenced by their young age). Why is that? Are doctors overconfident, type A drivers racing from open-heart surgery to the golf course?
One simple contributing factor may be that, in the United States at least, many doctors are male (nearly 75 percent in 2005). But firefighters and pilots are usually male as well, and those two professions were at the bottom of the risk list. Firefighters spend a lot of time in fire stations, not on the road, and pilots spend much of their time in the air. Exposure matters, which is seemingly why real estate agents, always driving from house to house, showed up high on the list. (Architects ranked high as well, prompting QPC’s vice president to speculate that they’re often distracted by looking at buildings!) Doctors drive a lot, often in urban settings, often with a certain urgency, perhaps dispensing advice via cell phone. Most important, they may also be tired. A report in the
New England Journal of Medicine
suggested that every time in a given month interns at Harvard Medical School pulled an extended shift, their crash risk rose by 9.1 percent. The more shifts they worked, the greater the risk that they would fall asleep while stopped in traffic, or even while driving.
Now let’s talk about Dr. Fred’s vehicle of choice, the pickup truck. It’s an increasingly popular vehicle in the United States. The number of households owning pickups rose by nearly 50 percent from 1977 to 1990, and pickup registrations continue to rise every year. It is also the most dangerous vehicle on the road: More people in the United States die in pickups per 100 million vehicles registered than in any other kind of vehicle.
Pickups also impose the most risk on drivers of other vehicles. One study showed that the Ford F-350 presents nearly seven times the risk to other cars as the Dodge Caravan, a minivan. From a vehicular point of view, pickups are high, heavy, and have very stiff front ends—meaning other vehicles have to absorb more energy in a crash. When drivers of pickups crash into other cars, they die at a lower rate than the drivers of smaller cars. Because of simple physics, larger vehicles, with larger crush zones and, often, higher-quality materials, are better able to sustain a collision.
Though not always. As some crash tests have shown, weight is often no help at all when a vehicle hits a fixed object like a wall or a large tree. Marc Ross, a physicist at the University of Michigan, told me that “mass sort of drops out of the calculation for a fixed barrier.” The car’s design—its ability to absorb its own kinetic energy—is as important as its size. In crash testing done several years ago by the Insurance Institute for Highway Safety, vehicles with crash-test dummies were sent into a barrier at 40 miles per hour. Consider two vehicles: the big and brawny Ford F-150 pickup truck, weighing in at nearly 5,000 pounds, and the tiny Mini Cooper, at just under 2,500 pounds. Which would you have rather been in? The test photos make the answer clear: the Mini Cooper. The Ford, despite having more space between the obstacle and the driver, saw a “major collapse of the occupant compartment” that “left little survival space for the driver.” In the Mini, meanwhile, “the dummy’s position in relation to the steering wheel and instrument panel after the crash test indicates that the driver’s survival space was maintained very well.”
As Malcolm Gladwell argued in the
New Yorker,
larger, heavier vehicles, which are more difficult to maneuver and slower to stop, may also make it harder for a driver to avoid a crash in the first place. What complicates this is the finding that, in the United States, small cars are involved in more
single-car
fatal crashes than large cars—and it’s single-car crashes that the greater maneuverability of smaller, lighter cars should help prevent. Smaller cars may be more maneuverable, but they also tend to be driven by riskier younger drivers, while sports cars that handle well may be “self-selected” by more adventurous drivers. Researchers with the National Highway Traffic Safety Administration raised another question: Would the higher maneuverability of smaller cars lead drivers to take more risks? “The quicker response of light vehicles,” they argued, “may give the average driver yet more opportunity to blunder.”
Risk can be deceiving. The answer to “What are the riskiest vehicles on the road?” is more complicated than it seems. Assigning risk based purely on “vehicle factors” is limiting, because it neglects the idea of who is driving the vehicle and how it is being driven. Leonard Evans, the former GM researcher, notes that crash rates are higher for two-door cars than four-door cars (up to a certain weight, where the rates become equal). “The believers in vehicle factors would say, ‘We’ve got it, you’ve just got to weld another couple doors on the vehicle and you’ve got a safe car.’”
Those two doors are often not an engineering distinction, but a lifestyle distinction: the difference, say, between a two-door Acura RSX and a four-door Toyota Corolla. In the United States from 2002 through 2005, the death rate in the “fast and furious” Acura was more than twice as high as that in the sleepy Corolla. In terms of weight, the two vehicles are virtually identical. The different crash rate owes more to the drivers of four-doors and two-doors than to the cars themselves.
The idea that who is driving (and how) affects the risk of what is being driven is well depicted in the case of the Ford Crown Victoria and the Mercury Marquis, as Marc Ross and Tom Wenzel have pointed out. The Crown Vic and its corporate cousin the Marquis, large, staid V-8 sedans both, are basically the same car—one repair manual covers both models. They both pose the same relative risk to their drivers, which is no surprise given their similarities. The Crown Vic, however, statistically poses more risk to others. Why is that? The Crown Victoria is a popular police car, meaning that it’s involved in a lot more dangerous high-speed pursuits than the Marquis. (Crown Vics, it must be said, are also the taxi of choice in New York City.)
There are “safer” cars in the hands of dangerous drivers, and “more dangerous” cars in the hands of safe drivers. Small cars such as subcompacts do pose a greater risk for their occupants if involved in a crash—although more-expensive subcompacts are less risky than cheaper subcompacts—but subcompacts also tend to be driven by people (e.g., younger drivers) with higher risks of getting into a crash, because of “behavioral factors.” Still, age is just one behavioral factor, and it interacts with the type of car being driven. As I will discuss in the next section, the drivers of small cars may actually act in safer ways because of the size of the car. Are large passenger cars statistically the safest because they pose less of a rollover risk than SUVs or because they weigh more than small cars? Or is it because they tend to be driven by the statistically safest demographic?
Returning to Fred and his pickup truck: It’s hard to tell where the risks of one end and those of the other begin. Men tend to drive pickup trucks more than women, men tend to wear seat belts less often, men who live in rural areas are more likely to drive pickup trucks without seat belts, and, after motorcyclists, the drivers of pickup trucks are the most likely to have been drinking when involved in a fatal crash. These would be only a handful of the potential risk factors—an Australian study, for example, found that black cars were more likely to crash than white cars. Is it visibility, or the types of people who drive black cars versus white ones? We all know no one washes a rental car, but are rental cars driven more recklessly? (There is some evidence to suggest so.) A study in Israel found that fewer drivers died on the roads in the first and second days after a suicide-bombing attack but then tracked an
increase
in danger on the third day. Are people simply staying off the roads in the aftermath, then rejoining them en masse? (Or does the aftereffect of terror cause people to act with less regard for life?)
As the risk expert John Adams likes to say, understanding risk is not rocket science—it’s more complicated. Looking at statistics from the United Kingdom, he notes that a young man is 100 times more likely than a middle-aged woman to be killed in traffic. Someone driving on Sunday morning at three a.m. has a risk
134 times
greater than someone driving at ten a.m. on Sunday. Someone with a personality disorder is 10 times more likely to have a serious crash, while someone 2.9 times over the BAC limit would be 20 times more likely than a sober driver to crash.
“So if these were independent variables,” he told me, “you could multiply them and come to the conclusion that a disturbed, drunken young man on the road on a Sunday morning was about 2.5 million times more likely to have a serious accident than a normal, sober middle-aged woman driving to church seven hours later.” They are, however,
not
independent. “There are proportionally more disturbed, drunken young men on the road at three o’clock on a Sunday morning,” Adams noted. Now add other factors. Were the car’s tires in good shape? Was it foggy? Was the driver tired or awake? “Once you start trying to imagine all the factors,” Adams said, “that might not be an exaggeration of the disparity between one person’s risk and another person’s risk.” He used this example to “have a go” at what he calls the “Richter scales” of risk, which show, for instance, that a person has a 1 in 8,000 chance of dying or being seriously injured in a car crash, and a 1 in 25,000 chance of the same thing happening while playing soccer. “The purveyors of these tables say they produce them to guide the lay public in making risks. The lay public is hopeless at making use of numbers like this.”