Sarah Holder is a staff writer at CityLab covering local policy, housing, labor, and technology.
Pedestrians fatalities are rising sharply as Americans spend more time behind the wheel. And self-driving technology isn’t likely to be the fix we need.
In some ways, the crash that killed pedestrian Elaine Herzberg in Tempe, Arizona, last year, was a typical distracted-driving incident, with a cruel high-tech twist: As Herzberg walked her bike across the road in the dark of night, the driver of the Volvo SUV hurtling toward her was streaming an episode of The Voice on her cell phone.
But the driver wasn’t the only operator that was distracted: The car was part of Uber’s fleet of self-driving test vehicles, racking up miles in computer control mode. Its many sensors should have recognized the pedestrian obstacle in its path and avoided the collision. Instead, the SUV’s operating system kept right on driving; and the human driver failed to intervene. Herzberg was fatally injured, and died in the hospital.
This week, at least one chapter of the long legal battle that ensued against Uber—which made the self-driving technology that powered the car—closed, when an Arizona prosecutor ruled that the company was not criminally liable for Herzberg’s death. The driver may still face manslaughter charges.
The Tempe case was so high-profile in part because it was historic—the first recorded case of a pedestrian killed by an autonomous vehicle, a long-dreaded industry milestone that threatened to confirm the public’s worst fears about self-driving technology.
But the coverage of the incident may have obscured a larger tragedy: That every day in the U.S., pedestrians like Herzberg are being killed by regular drivers at a staggering rate. And though autonomous vehicles promise to eventually replace humans at the wheel—and eventually, promoters of this technology insist, make the streets safer—right now pedestrians are being killed and injured by motorists at the highest rate in decades.
Last week, the Governors Highway Safety Association (GHSA) dropped preliminary data on 2018’s U.S. pedestrian fatalities, finding that American motorists killed 5,977 walkers and cyclists in 2017. After 2018 data is fully analyzed, the U.S. is on track to report the highest number of pedestrian fatalities since 1990: 6,277.
“There was a 30-year decline starting in 1979 in the number of pedestrian fatalities,” said Richard Retting, who researched and wrote the report. Now, the U.S. is reaching the peak of a decade-long surge. “Something’s gone terribly wrong in the last ten years.”
U.S. Pedestrian Fatalities: 1990-2018
What went so wrong? Another recently released statistic offers one clue: Americans are simply spending a growing amount of time driving their cars. According to the AAA Foundation for Traffic Safety’s annual American Driving Survey (ADS), which was also released last week, 87 percent of the U.S. population age 16 and up are driving, and collectively, they logged 70.1 billion hours behind the wheel between 2016 and 2017. That’s a new record high, at least since the AAA started collecting this data in May 2013.
Heightened activity on the roads isn’t the only factor making them more dangerous. The rise of smartphones has introduced major distractions for both drivers and walkers, as the number of devices in use more than quintupled from 2010 to 2017. The shape of the vehicles on the roads has also changed, as popular heavy-duty SUVs and pickup trucks have proliferated. Though passenger cars killed more people, the number of deaths involving SUVs increased 20 percent faster than that of passenger cars between 2013 and 2017, as retail sales of light trucks like them increased dramatically. With their greater mass and limited driver visibility, SUVs have proved to be more lethal than their smaller cousins. In June, a damning Detroit Free Press investigation concluded that “the SUV revolution is a key, leading cause of escalating pedestrian deaths nationwide, which are up 46 percent since 2009.”
Retail sales (in thousands) of passenger cars and light trucks, 2008-2017
Meanwhile, the dangers that have long flustered or incapacitated human drivers—like darkness and alcohol—remain stubbornly pervasive: 75 percent of all pedestrian fatalities occurred at night, and 17 percent of drivers were intoxicated when involved in a fatal crash (interestingly, compared to 32 percent of pedestrians).
Still, it’s hard to avoid connecting the dots. The number of pedestrian deaths increased by 35 percent from 2008 to 2017, while the number of all other traffic deaths dropped by 6 percent. Even with more cars on the streets, more frequently—collectively driving a total of 183 billion trips, according to the ADS—people in cars aren’t getting into deadlier crashes with other driver peers. Thanks to increasingly advanced airbags, crumple zones, and other government-mandated safety features, the people inside America’s cars and trucks have never been better protected.
Those on the other side of the windshield, however? Not so much.
Numbers of U.S. Traffic Deaths in 2008 and 2017
“People in cars are safer than they had been in the past, and people outside of cars are less safe than they’ve been in the past,” said Retting.
While vehicles have become more resilient, the report reads, “pedestrians remain just as susceptible to sustaining serious or fatal injuries when struck by a motor vehicle.” This survivability gap isn’t inevitable, the report argues. It reflects our ongoing inability to enact nationwide pedestrian safety measures with the same attention that policymakers and car manufacturers have lavished on the welfare of those inside vehicles.
Indeed, the sense of invulnerability people feel within their increasingly vault-like machines could be part of the reason people spend so much time there, at least in some places. According to the ADS, Americans aren’t just driving more because commutes are longer (which they are), or because they’re living farther from jobs (which is especially true for lower-income households). There’s also a more emotional factor at play: A car can be a refuge. When IKEA recently asked 22,000 people worldwide what “home” meant to them, researchers found that almost half of the Americans they asked go to their car to “have a private moment to themselves.”
That primal desire to insulate oneself from the rages of the road might help explain why Americans are choosing to roll up in ever-more-hulking machines. On Jalopnik, Jason Torchinsky recently pondered the latest design trends involving pickup trucks, which, like full-size SUVs, now account for a growing share of vehicles on America’s roads. “Truck grilles are growing at alarming rates, and becoming more and more intricate, Baroque, and confrontational,” he wrote. He concludes:
Walking by trucks like these feels more like walking past a building sometimes, confronted with vast curtain walls of vents and meshes and perforated, vertical walls. They’re starting to feel less and less like vehicles.
It’s all starting to feel sort of insecure and crazy, if we’re honest. It’s not like any of these trucks actually looks all that good, really. It could just be me, but these massive, over-complicated grilles feel desperate and attention-hungry, like showing up at a barbecue slathered in blood and with your sleeve on fire so there will be no doubt as to what a badass you are.
Pickup trucks are among the best-selling vehicles in America. I’m not exactly sure what the effect on our national mindset will be if we’re populating our roads with these colossal, wrathful gaping maws, but I don’t feel like it’s entirely healthy.
Retting, however, doesn’t buy the notion that there could be some deeper psychological forces lurking behind the growing pedestrian carnage. “I’m not concerned about people’s feelings,” he said. “I’m just interested in outcomes. And right now, the outcomes are dreadful.”
Cities and vehicle manufacturers should focus on evidence-based solutions, he says, by installing pedestrian automatic braking in more cars, and prioritizing their deployment in SUVs. The report also recommends increasing street lighting around pedestrian corridors, deploying nighttime enforcement patrols, enforcing speed limits with cameras, and implementing curb extensions and pedestrian refuge islands.
The pedestrian fatality picture isn’t entirely bleak. At least 18 U.S. cities have adopted Vision Zero plans, with the stated goal of eliminating all pedestrian deaths by 2040. Those efforts have not been futile: The 10 largest cities reported a combined 15 percent decline in pedestrian fatalities in 2017; New York City, whose Vision Zero plan is one of the most extensive, reported the sharpest decrease, along with Los Angeles, San Antonio, and San Diego. Since Vision Zero actions are concentrated in cities, however, they ignore the periphery of the urban core, where vehicle speeds are higher and a growing number of fatal crashes are happening.
Almost half of all the reported pedestrian deaths happened in just five states—Arizona, California, Florida, Georgia, and Texas—partly because of their higher populations, but also because of their car-centric planning cultures and spatial idiosyncrasies. Texas, for example, is one of the fastest-growing states, and is home to four of the 10 fastest-growing cities (San Antonio, Dallas, Fort Worth, and Frisco). “But unfortunately, growth isn’t even,” said Retting. “A lot of the population growth is not in the downtown core where speeds are low and pedestrians have sidewalks. A lot of the growth is in the suburbs and exurbs, which are often hostile and not friendly to pedestrian access.”
It’s these sprawling and car-dominated suburban spaces—places a lot like the suburban Tempe roadway that Elaine Herzberg was trying to cross—where the prospect of self-driving might hold the greatest life-saving promise. Once the technology to deploy more autonomous vehicles is fully perfected, we have often been told, AVs will make streets exponentially safer for pedestrians, as they replace human error with expert calculation.
But that day hasn’t come yet—and, indeed, may never arrive, at least not for every driving situation. In the meantime, AVs appear to be just one more thing for human drivers to be afraid of: In 2017, AAA found that three-quarters of survey respondents felt unsafe sharing the road with fully autonomous cars.