Traffic (35 page)

To return to the frenetic streets of Delhi: My impression was that many drivers did not seem to be particularly qualified for a license. There’s a good reason for this. A study conducted by a team of researchers for the U.S. National Bureau of Economic Research looked at the process of getting a driver’s license in Delhi. The group tracked 822 individuals in three groups: a “bonus” group, whose members would get a financial reward if they could obtain a license in the fastest time legally possible; a “lesson” group, whose members were given free driving lessons before they attempted to get the license; and a “comparison” group, which was given no special instructions.

The researchers found that those who wanted the license soonest—that is, the members of the bonus group—got it more often, and faster, than people in the other groups. The reason, it turned out, was that like many drivers in Delhi, they used an “agent” to speed the process. But when the researchers later gave all the survey participants a driving test, 69 percent of the bonus group failed, compared to just 11 percent of the drivers who had taken lessons. But learning to drive properly clearly did not pay off: The people who had the best driving skills were 29 percent less likely to get a license than the people with the worst driving skills. Corruption did indeed grease the wheels, but at the expense of the quality of those behind the wheel. “Corruption,” the authors wrote, “appears to substitute for actual driving skill.”

This study provides a hint about how the norms discussed in the previous section evolve and flourish. The scores of new drivers who land on Delhi streets each month learn the norms of a system made up of the collective experience of all the previous drivers who bribed their way through the Regional Transport Office. No small wonder this traffic system isn’t marked by scrupulous attention to formal rules. In the writer Pavan Varma’s description of what motivates corruption in India, it is not hard to see a metaphor for the country’s traffic behavior: “In a cut-throat world, the immediate task is to get on with the job, to reach a desired goal, to finesse an obstacle. The premium is on pragmatism and agility, the capacity to seize an opportunity when it comes, and to profit when possible. What matters is not fixity of principle but clarity of purpose.”

What
is
surprising is how strong these corruption norms can be, even in a different context. In one study, the economists Ray Fisman and Edward Miguel looked at the number of parking tickets issued to diplomats in New York City between 1997 and 2002. During this time, diplomats could be given parking tickets, but there was no enforceable punishment for not paying them. Thus empowered, diplomats racked up some 150,000 tickets.

The tickets were not acquired randomly. The diplomats who got the most tickets tended to be from the countries deemed to be more corrupt by the Transparency International index (those countries also got more “egregious” tickets, such as for blocking fire hydrants). The countries whose diplomats received
no
tickets included Sweden, Norway, Japan, and Denmark—judged among the least corrupt countries. These countries were scrupulous in following the law, even when it was clearly not necessary. India, in case you were wondering, was roughly halfway down the list, just as it is on the corruption index. Lest you think I am singling out India, I might add that the United States embassy in London, as of 2007, owed the highest amount (ahead of even corruption-plagued Nigeria) of unpaid traffic congestion-pricing fees to the city of London. The United States, which claims that its diplomats are exempt from the congestion-pricing “tax,” is not one of the ten least corrupt countries (it was ranked twentieth in 2007). (The least-corrupt country, Finland, whose diplomats are also exempt from taxes, pays the charge.)

In traffic, laws are only as good as the norms regarding them. This may be why, as I discussed in Chapter 7, the engineer Hans Monderman could strip the signs from a roundabout and Dutch drivers would still act in a responsible, safe manner; and why, in other countries, a roundabout can be filled with signs and drivers will still act in an irresponsible, dangerous manner. Which brings us back to two questions: Are developing countries fated to have a disproportionate share of traffic fatalities? And how many of these fatalities come from lack of money, how many from laws or norms weakened by corruption? The passengers crowded into unsafely overloaded buses may be there because it’s the only transportation they can afford or because there is no one to stop the bus from being overloaded—perhaps because the government thinks it can’t afford to
not
let people ride the overcrowded bus.

The vexing, intertwined nature of this dilemma is reflected in a piece of Hindi slang I learned while in Delhi:
jugad.
The word has a shifting palette of meanings, mostly arrayed around the central idea of “creative improvisation.” It can refer, on the one hand, to the jury-rigged vehicles one finds in India, especially in rural areas. Lacking money for a car, say, a farmer will craft a functioning vehicle out of an old motorcycle, a car axle, and a diesel engine. That this
jugad
vehicle might not be safe, at least when it’s sharing the road with newer cars, is one of the clear kinds of traffic risks that come with lack of money.

But
jugad
is also used as a kind of surrogate for “bribe” here it refers to doing whatever needs to be done to get something accomplished. The case of the Delhi drivers who acquired licenses quickly is a form of
jugad
in practice. Would-be drivers know that corrupt bureaucrats respond more to money than driving skills. Is this kind of corruption, which has a ripple effect that translates into the myriad traffic violations that occur in Delhi every day—and studies suggest that the more traffic laws are violated, the more casualties there will be—purely an effect of lack of resources? Or is it, as many would argue, precisely the sort of thing that holds up the development of a country? If GDP and traffic fatalities are somewhat related, and GDP and corruption are somewhat related, and traffic fatalities and corruption seem to be the most clearly related, then fighting corruption may be the best way to lower traffic fatalities and raise GDP.

There are, after all, creative ways of combating corruption that do not require huge amounts of money. In Mexico City, Alfredo Hernández García, the city’s traffic czar, described a novel plan to fight corruption and improve traffic safety. In 2007, he noted, the last of the city’s male traffic officers had been phased out, replaced entirely by women (known as
cisnes,
or “swans”). Why? “Because women are less likely to be corrupted,” he explained in his office in the Secretaría de Seguridad Pública. Previously, Mexico City traffic cops were famous for soliciting
refrescos
or “soft drinks”—that is, bribes in lieu of a ticket. According to Hernández García, the
cisnes
have increased the number of tickets written on the order of 300 percent. They have been given handheld units to issue tickets and ensure payment—drivers can use credit cards—and take photographs. “People do not accept they are breaking the law,” he said. “We have to provide evidence.”

The theory of women as less corruptible may be based on more than the hunches of a few higher-ups in the police department. A study by a group of U.S. economists found that women were less likely to engage in hypothetical corruption, that female managers in one country they studied were less likely to engage in actual corruption, and that the countries that rank as least corrupt on the global indices tend to have more women in government. Indeed, they may be onto something: Finland, ranked as the least-corrupt country in the world, set the record in 2007 for having the government with the most women in cabinet-level positions. As you will recall, they do not mess about with their traffic tickets.

Why You Shouldn’t Drive with a Beer-Drinking Divorced Doctor Named Fred on Super Bowl Sunday in a Pickup Truck in Rural Montana: What’s Risky on the Road and Why

In a basement laboratory in the looming red-brick Henry Ford Hospital in Detroit, Michigan, a team of researchers has, for the past few years, been looking at what happens to our brains as we drive. The device that measures the faint magnetic fields the brain emits is too massive to fit inside of a car, so research subjects are instead studied in the hospital’s Neuromagnetism Laboratory, where they watch film clips of a car navigating through traffic. As I lay back on the cozy bed inside the magnetically shielded lab to get a feel for the procedure, Richard Young, a scientist with General Motors who leads the research team, told me, “Our biggest problem is people falling asleep in the bed.”

To keep people awake as they play passenger to the filmed driving, subjects are given a simple “event-detection task.” When a red light near the screen goes on, the subject, attached to a neuromagnetometer, presses a simulated brake pedal. This simple habit of braking in response to a red light (i.e, brake lights), something drivers do an estimated fifty thousand times a year, triggers a burst of activity in the brain. The visual cortex lights up about 80 to 110 milliseconds after the red signal comes on. This indicates that you have seen the signal. The left prefrontal lobe, an area of the brain linked to decision making, begins to buzz with activity. This is the microinstant during which you’re deciding what to do with the information you have acquired—here, the rather simple response of simply pressing the brake. It comes about 300 milliseconds before you actually do it. About 180 milliseconds before braking begins, the motor cortex sees action—your foot is about to be told to move. About 80 milliseconds after you have pressed the brake, the visual cortex is again activated. You’re registering that the red signal has been turned off.

The scientists are probing the neural pathways involved in what they call the “mind on the drive,” in part to learn what cell phone conversations and other activities do to our brains as we drive. But sometimes, as they watch these real-time movies of people’s brains in traffic, there are strange and unanticipated plot twists.

Once, while watching the real-time fMRI (functional magnetic resonance imaging) readings of a subject, Young noticed a burst of brain activity, not during the braking event but during “normal” driving. “There was a spike. There were brain areas lighting up in the emotional cortex, the amygdala, the limbic cortex, the lower brain,” Young recalled. This hinted at more complex responses than what usually showed up in the fairly well-conditioned responses to braking or keeping the vehicle on the road at a certain speed. What was going on? Young compared the activity to the actual video of the drive. At the moment his brain went on the boil, the driver was passing a semitrailer. After the trial, Young asked the subject if he had noticed “anything unusual during the last run.” He had. According to Young, “The person said, ‘Oh yes, I was passing that eighteen-wheeler and every time I pass one of those things I get real nervous.’”

That small peek into the brain of the driver revealed a simple, if underappreciated truth about driving: When we are in traffic, we all become on-the-fly risk analysts. We are endlessly having to make snap decisions in fragments of moments, about whether it is safe to turn in front of an oncoming car, about the right speed to travel on a curve, about how soon we should apply the brakes when we see a cluster of brake lights in the distance. We make these decisions not with some kind of mathematical probability in the back of our heads—I have a 97.5 percent chance of passing this car successfully—but with a complicated set of human tools. These could be cobbled from the most primeval instincts lurking in the ancient brain, the experience from a lifetime of driving, or something we heard yesterday on the television news.

On the one hand, it was perfectly natural, normal, and wise for the driver in Detroit to show fear in the face of an eighteen-wheeler. Large trucks, from the point of view of a car, are dangerous. Because of the staggering differences in mass—trucks weigh twenty to thirty times more than a car—the simple physics of a collision are horrifically skewed against the car. When trucks and cars collide, nearly nine of ten times it’s the truck driver who walks away alive.

As the driver’s brain activity would seem to indicate, we know this on some instinctual level, as if our discomfort in driving next to a looming truck on a highway is some modern version of the moment our prehistoric ancestor felt the hairs on the back of his neck raise when confronted with a large predator. Indeed, the amygdala, one of the areas that lit up in the Detroit driver, is thought to be linked with fear. It can be activated even before the cognitive regions kick in—neuroscientists have described the amygdala as a kind of alarm that triggers our attention to things we should probably fear. And we all likely have proof of the dangerous nature of trucks. We have seen cars crumpled on the roadside. We’ve heard news stories of truck drivers, wired on stimulants, forced to drive the deregulated trucking industry’s increasingly long shifts. We can easily recall being tailgated or cut off by some crazy trucker.

Just one thing complicates this image of trucks as the biggest hazard on the road today: In most cases, when cars and trucks collide, the car bears the greater share of what are called “contributory factors.” This was the surprising conclusion that Daniel Blower, a researcher at the University of Michigan Transport Research Institute, came to after sifting through two years’ worth of federal crash data.

It was a controversial finding. Blower, to begin with, had to determine that it did not simply stem from “survivor bias”: “The truck driver is the only one that survives these eighty-five percent of the time,” he explained. “He’s the one who gets to tell the story. That’s what’s reflected in the police report.” So he dug deeper into the records, analyzing the relative position and motion of the vehicles before a crash. Instead of relying on drivers’ accounts, he looked at “unmistakable” physical evidence. “In certain crash types like head-ons, the vehicle that crosses the center much more likely contributed to the crash than the vehicle that didn’t cross the center line,” he said. “Similarly, in rear-end crashes, the striking vehicle in the crash is much more likely to have contributed to the crash in a major way than the vehicle that was struck.” After examining more than five thousand fatal truck-car crashes, Blower found that in 70 percent of cases, the driver of the car had the
sole
contributing responsibility in the crash.

This hardly means trucks are not dangerous. But the reason trucks are dangerous seems to have more to do with the actions of car drivers combined with the physical characteristics of trucks (in head-on collisions, for example, they are obviously less able to get out of the way) and less to do with the actions of truck drivers. “The caricature that we have that the highways are thronged with fatigued, drug-addled truck drivers is, I think, just wrong,” Blower said. Certainly there are aggressive truck drivers and truckers jacked up on methamphetamine, but the more pressing problem, the evidence tells us, seems to be that car drivers do not fully understand the risk of heavy trucks as they drive in their presence. This is not something we are necessarily taught when we learn to drive. “In a light vehicle you are correct to be afraid of them, but it’s not because the drivers are disproportionately aggressive or bad drivers,” Blower said. “It’s because of physics, truck design, the different performance characteristics. You can make a mistake around a Geo Metro and live to tell about it. You make that same mistake around a truck and you could easily be dead.”

What all this seems to suggest is that car drivers have less to fear from trucks than from what they themselves do around trucks. I had a glimpse of this a few years back when I rode in an eighteen-wheel tractor-trailer for the first time, watching in horror as cars darted in front of the truck with dangerous proximity, sometimes disappearing from sight beneath the truck’s long, high hood. So why does it seem that virtually everyone, like my Latin-teacher friend in the Prologue, has some horror story about crazy truckers?

One possible answer goes back to the spike in brain activity of the Detroit driver. He was afraid, probably before he even knew why. The size of trucks makes most of us nervous—and rightfully so. When we have a close brush with a truck or we see the horrific results of a crash between a car and a truck, it undoubtedly leaves a greater impression on our consciousness, which can skew our view of the world. “Being tailgated by a big truck is worth getting tailgated by fifty Geo Metros,” as Blower put it. “It stays with you, and you generalize with that.” (Studies have suggested that people think there are more trucks on the road than is actually the case.)

Here’s the conundrum: If, on both an instinctual level and a more intellectual level, the drivers of cars fear trucks, why do car drivers, in so many cases, act so dangerously around them? The answer, as we are about to see, is that on the road we make imperfect guesses as to exactly what is risky and why, and we act on those biases in ways we may not even be aware of.

Psychologists have suggested that we generally think about risk in two different ways. One way, called “risk as analysis,” involves reason, logic, and careful consideration about the consequences of choices. This is what we do when we tell ourselves, on the way to the airport with a nervous stomach, “Statistically, flying is much safer than driving.”

The second way has been called “risk as feelings.” This is why you have the nervous stomach in the first place. Perhaps it’s the act of leaving the ground: Flying just
seems
more dangerous than driving, even though you keep telling yourself it isn’t. Studies have suggested that we tend to lean more on “risk as feelings” when we have less time to make a decision, which seems like a survival instinct. It was smart of the Detroit driver to feel risk from the truck next to him, but the instinctual fear response doesn’t always help us. In collisions between cars and deer, for example, the greatest risk to the driver comes in trying to avoid hitting the animal. No one with a conscience wants to hit a deer, but we may also be fooled into thinking that the deer itself presents the greatest hazard. Hence the traffic signs that say
DON’T VEER WHEN YOU SEE A DEER.

One good reason why we rely on our feelings in thinking about risk is that “risk as analysis” is an incredibly complex and daunting process, more familiar to mathematicians and actuaries than the average driver. Even when we’re given actual probabilities of risk on the road, often the picture just gets muddier. Take the simple question of whether driving is safe or dangerous. Consider two sets of statistics: For every 100 million miles that are driven in vehicles in the United States, there are 1.3 deaths. One hundred million miles is a massive distance, the rough equivalent of crisscrossing the country more than thirty thousand times. Now consider another number: If you drive an average of 15,500 miles per year, as many Americans do, there is a roughly 1 in 100 chance you’ll die in a fatal car crash over a lifetime of 50 years of driving.

To most people, the first statistic sounds a whole lot better than the second. Each trip taken is incredibly safe. On an average drive to work or the mall, you’d have a 1 in 100 million chance of dying in a car crash. Over a lifetime of trips, however, it doesn’t sound as good: 1 in 100. How do you know if this one trip is going to be
the
trip? Psychologists, as you may suspect, have found that we are more sensitive to the latter sorts of statistics. When subjects in one study were given odds, similar to the aforementioned ones, of dying in a car crash on a “per trip” versus a “per lifetime” basis, more people said they were in favor of seat-belt laws when given the lifetime probability.

This is why, it has been argued, it has long been difficult to convince people to drive in a safer manner. Each safe trip we take reinforces the image of a safe trip. It sometimes hardly seems worth the bother to wear a seat belt for a short trip to a local store, given that the odds are so low. But events that the odds say will almost certainly never happen have a strange way of happening sometimes (risk scholars call these moments “black swans”). Or, perhaps more accurately, when they do happen we are utterly unprepared for them—suddenly, there’s a train at the always empty railroad crossing.

The risk of driving can be framed in several ways. One way is that most people get through a lifetime without a fatal car crash. Another way, as described by one study, is that “traffic fatalities are by far the most important contributor to the danger of leaving home.” If you considered only the first line of thinking, you might drive without much of a sense of risk. If you listened to only the second, you might never again get in a car. There is a built-in dilemma to how societies think about the risk of driving; driving
is
relatively safe, considering how much it is done, but it could be much safer. How much safer? If the number of deaths on the road were held to the acceptable-risk standards that the U.S. Occupational Safety and Health Administration maintains for service-industry fatalities, it has been estimated, there would be just under four thousand deaths a year; instead, the number is eleven times that. Does telling people it is dangerous make it safer?