Under ordinary circumstances, the death of a homeless woman trying to cross a road in Arizona would barely arouse notice. After all, about 16 pedestrians are killed every day on U.S. roads, and Arizona is one of the deadliest states.
But the March 18 death of Elaine Herzberg, 49, in Tempe, was an international news event. Herzberg — or “Miss Elle,” as she was known in Tempe’s homeless community — was struck by a Volvo SUV steered by Uber’s self-driving car technology.
She was the first pedestrian killed by a self-driving car.
Until recently, self-driving cars were an abstraction, a dreamy, far-off idea. Now they are literally crashing into us. Twenty-four U.S. cities are either running autonomous vehicle (AV) test
programs or have committed to doing so, according to the Bloomberg Aspen Initiative on Cities and Autonomous Vehicles. Uber is one of more than a dozen firms racing to bring the technology to the market; some experts believe AVs could fully replace human-driven cars by 2040.
AVs are already influencing public policy. This spring, the Koch brothers’ group Americans for Prosperity helped defeat a $5.4 billion mass transit initiative in Nashville, in part by suggesting AVs would make such transit obsolete. In short, we are on the precipice of a major technological change, one that reshapes the way we live and travel in ways we can’t yet fully grasp.
In an oft-cited thought experiment, Robin Chase, co-founder of the car-sharing company Zipcar, offered two scenarios for how the self-driving revolution could unfold. She calls them “heaven” and “hell.”
In “heaven,” drawing on and expanding Chase’s vision, AVs usher in a new age of personal and civic well-being. Traffic deaths are eliminated; an 8‑year-old child can bike to school in a major
city without fear. As outdoor activity becomes safe and popular, public health outcomes improve. Urban dwellers rely on shared AVs and expanded public transit to get around and are largely freed from the financial burden of individual car ownership. Cities repair some of the damage done in the highway era. Massive surface parking lots are no longer needed, for example, and that space is repurposed for parks or desperately needed affordable housing. In a boon for the climate, AVs run on zero-carbon electricity. While some workers lose their jobs in the transition — such as bus drivers, truckers and delivery people — social policies like universal basic income help ensure a safety net.
hell,” however, AVs further entrench Americans’ already dysfunctional relationship with cars. Social isolation, congestion and inequality increase. Less concerned about long commutes, people sprawl across the countryside, eliminating farmland and establishing elite and exclusive communities far from the masses. The total number of miles driven increases exponentially, along with urban congestion. AV owners order their empty cars to cruise the city all day to avoid paying for parking. Buses are mired in heavy traffic or simply discontinued as public support for them erodes, leaving lower-income people stranded. The transition to AVs doesn’t coincide with a transition away from internal combustion engines. Passenger cars and trucks, already the source of roughly 15
percent of U.S. greenhouse gas emissions, increase their footprint dramatically, broiling the planet.
Right now, it’s unclear which direction we’re headed. But we are at an inflection point. The auto industry is pushing Congress to greenlight not only the testing of self-driving cars, but also sale to the public, currently lobbying a handful of Democratic holdouts in the Senate on a bill that would do just that. What’s lacking is what’s needed most: public oversight.
AVs could introduce us to a new, even more wasteful possibility: zero-occupancy vehicles. Cities might be full of empty cars that are traveling around to pick up kids, or pick up dinner.
THE NEW MOTORDOM
According to University of Virginia historian Peter Norton, who studies the early automotive era, we’re on the brink of the most sweeping technological change in transportation since the transition from the horse and carriage to the car about a century ago.
We’re just starting to fully recognize what was lost in that transition: Huge portions of downtowns turned over to parking lots; thriving communities cut through by highways; the death of legacy transit like streetcars; white flight and urban depopulation. The auto revolution even caused political polarization, as suburbanization created increasingly homogenous communities with increasingly divergent interests.
The popular myth is that Americans in the early 20th century immediately embraced cars. But Norton argues that the American “love affair with the automobile,” as the auto industry presents it, was carefully manufactured. The introduction of cars into cities was met at first with skepticism, even horror. The public was aghast when hundreds of people—disproportionately children—were run over and killed in streets where just a few years earlier they had been free to roam and play. The front page of the New York Times in 1924 pictured a car with a human skull mounted on the hood. “Nation Roused Against Motor Killings,” read the headline.
In response, car companies and other auto interests—a set of forces that Norton calls “motordom”—launched a lobbying and publicity blitz. One key battle was the struggle for a resource that remains hotly contested: street space.
Streets had been gathering places where merchants sold goods and children played. Urbanites were reluctant to cede this territory to noisy, dangerous and noxious cars.
A turning point came when motordom introduced the term “jaywalker,” branding pedestrians as intruders in the rightful domain of drivers. In 1921, for example, auto clubs recruited Boy Scouts to distribute cards to pedestrians in Hartford, Conn., that explained “jay walking” and instructed them how to properly cross the street. In 1952, a more explicit card distributed in Cincinnati read “I am a jaywalker” and gave the recipient a choice of hospitals to be delivered to when she was struck.
There was continued resistance from engineers, planners and ordinary citizens, most notably New York City’s Jane Jacobs, the mother of contemporary urban planning, who fought expressways and advocated for mixed-use public space. But many urban planners, such as Jacobs’ nemesis Robert Moses, enthusiastically embraced the redesign of cities around cars. Zoning rules were overhauled to require a minimum number of parking spaces for most new development. Black and brown neighborhoods were torn down to make way for highways.
Norton sees echoes of motordom in the sunny depictions of an AV future. Ford’s website, for example, presents a “City of Tomorrow” with an enormous highway right through a city center. Norton points to a joint 2010 presentation by SAIC (a Chinese car manufacturer) and GM in Shanghai. The companies, he says, offered a vision of total auto dependence, imagining a world where “the minute you leave the door you’re in a pod of some kind looking at screens the whole time, and you never see anybody doing anything else.”
Tech and car companies developing self-driving vehicles, including Uber, Waymo, Ford and Volvo, established a lobbying group called the Self-Driving Coalition for Safer Streets in 2016. They hired the former head of the National Highway Traffic Safety Administration, David Strickland, to lead the effort. The group has lobbied Congress—so far successfully—to keep collision data secret.
They are also pushing against state and local efforts to create their own laws governing AVs. That’s especially significant because if AVs are regulated to protect urban interests, it will most likely happen at the local level: Cities traditionally serve as policy laboratories, and have the most at stake in this transition.
THE PRESENT PURGATORY
Right now, there aren’t a whole lot of voices pushing back against auto companies, beyond a few consumer and safety groups who warn that the U.S. public is serving as “crash test dummies.” They’ve won some state-level safety regulations, but federal regulators have mostly stayed out of AV companies’ way. Only in the last two years has the U.S. Department of Transportation issued “voluntary guidelines” for autonomous vehicles.
Elaine Herzberg’s death suggests that safety advocates were right to sound alarms. We won’t know all the details until the National Transportation Safety Board’s investigation is complete. But the organization’s preliminary findings are disturbing. The agency reports that the Volvo self-driving system detected Herzberg 6 seconds before the collision and called for an emergency brake 1.3 seconds before. But the car didn’t respond, because Uber had programmed the car not to brake in those situations. The reason? Money.
Anonymous sources with Uber told the tech news site The Information that the company had wanted to start offering paid rides in driverless cars in Arizona by year end. But while frequent braking for unknown objects is safer, it can make the ride uncomfortable and jerky for passengers. So Uber had programmed the cars to avoid braking. NTSB has not issued a final report, or
named either party at fault.
Part of the reason federal regulators have been hesitant to intervene is that most experts expect AVs to become safer than their predecessors—eventually. The goal is achievable in large part because humans are such terrible drivers. About 40,000 Americans were killed in traffic last year, and millions more were seriously injured.
Costa Samaras, a civil engineering professor at Carnegie Mellon who studies energy system transitions, estimates that if the automated features already available in cars—for example, blind spot monitoring, lane departure warning and forward collision-warning systems—were to become standard, the net benefit to society in reduced social cost of crashes could be as much as $200 billion annually, including up to 10,000 fewer fatal crashes.
With the right planning, AVs offer other potential benefits to society. Right now, the average American family spends 16 percent of its income on transportation, and many cars spend 95 percent of their time parked. AVs could bring these costs down dramatically—but whether they do so depends on shared ownership.
Shared ownership is a big open question, however, and perhaps one of the most important for social equity and the health of cities. It’s in the interests of auto companies like Tesla to continue with individual ownership, the status quo, in which every working adult is more or less forced to own an expensive vehicle for their own private use. Uber, on the other hand, is more invested in the idea of f leets of self-driving cars that sell mobility as a service on a per-ride basis.
Another potential benefit is increased freedom and quality of life for people who, because of the way we’ve set up our society around driving, have been isolated and dependent: elderly people, children, people with disabilities and some low-income people. “A whole big portion of the population doesn’t drive,” notes Samaras. “If [with AVs] elderly people are traveling more to reach doctors, that’s a huge increase in social welfare.”
This will especially apply to rural and suburban areas that lack efficient mass transit because people and destinations are so far apart. With government investment, a sort of public Uber service that uses automated cars, vans or mini-buses to move small groups of people could serve as an affordable, efficient alternative to a city bus in places where there isn’t the population to fill one.
Most of the potential negative consequences of AVs—the “hell scenario”—stem from the logic of the market. The technology lowers the “costs” of driving, costs that are already far too low compared to the societal costs. For generations now, Americans have been sprawling across urban areas, building low-density housing on former farmlands and using highways to commute into urban centers.
The consequences for the environment and for social equity have been tremendous and are well chronicled. According to the American Farmland Trust, we lost three acres of farmland a minute between 1992 and 2012, about 60 percent due to sprawl-type development. Cars and urban sprawl have also helped enable massive geographic inequality: In the Cleveland metro area, there is a 24-year difference in life expectancy between inner-city Hough (a black neighborhood) and the middle-class suburb of Lyndhurst, just eight miles away.
Despite that, the one thing that has tenuously held metro areas together is the time it takes to commute. Most people are unwilling to commute much longer than 45 minutes. Self-driving cars will upend that calculus. If people are free to read or sleep or work while they are traveling in a car, they may accept much longer commutes. That means more cars on highways, more sprawl and more prime farmland disappearing. The result could be gridlock, especially in already congested growing areas.
While cars can be relatively efficient if they’re full of people, most of the time they’re not. Right now, their average occupancy is 1.5 people. AVs could introduce us to a new, even more wasteful possibility: zero-occupancy vehicles. Cities might be full of empty cars that are traveling around to pick up kids, or pick up dinner.
Those cars require road space, and space is at a premium in booming areas like Manhattan, San Francisco and Boston. AVs could harm high-capacity transit in the urban areas that depend on it.
Automation also threatens to eliminate millions of decent middle-class jobs. Between truck drivers and delivery drivers alone, there are 3.2 million jobs that will likely disappear. There are another half million people working as school bus drivers, 160,000 as city bus drivers and almost 200,000 taxi drivers, not including Uber. Progressive government policies—perhaps universal basic income, jobs programs or a stronger safety net—will be necessary to cushion the transition.
Right now it’s unclear whether AVs can have environmental benefits even if they don’t lead to a huge increase in driving. A lot depends on whether they join a broader transition to electric vehicles. University of Sydney civil engineering professor David Levinson thinks they will, based on the declining cost and improving batteries. Electric vehicles may also lend themselves better to shared fleets than to individual ownership, because more cars can take advantage of a single charging station. Samaras, however, thinks theecological effects are an open question. Much depends on public policy. “Design it from the start to be environmentally and ethically better,” Samaras says.
But regulators’ laissez-faire attitude toward the road-testing of AVs does not bode well when it comes to shaping our autonomous future. The stakes are especially high in cities, which are just, after a century, beginning to correct for the ravages of the auto transition.
THE CITY OF THE FUTURE
One of the most authoritative—and one of the only organized—voices for strong cities in the age of AVs is the National Association of City Transportation Officials (NACTO), which represents officials in more than 60 North American cities. NACTO envisions cities of the future that center pedestrians, not cars, with much of the road space repurposed for bicyclists, walkers and public transit. AVs can assist in this transition if they are programmed to always yield to pedestrians, or to never exceed certain speeds in pedestrian-dense areas.
That vision blurs the lines between what we today consider space for cars and public space. We may see more “shared spaces,” like the plazas of Europe that allow vehicles to traverse cautiously while bicyclists and pedestrians roam freely. Since we can worry less about cars careening into sidewalks, streets could be redesigned with no curbs, removing an accessibility obstacle for people in wheelchairs.
With the looming possibility that AVs will make car use more attractive, it’s important for cities to prioritize the kind of transportation that is going to improve the health and well-being of residents—biking and walking and mass transit—right now. Without any deliberate planning, we could repeat the same mistakes we made a century ago. “We spent the last 100 years making it cheap and easy to own and operate your personal car,” says Robin Chase. “We need to rebalance.”
That sea change has already, slowly, begun. The most progressive U.S. cities are restoring their streets to how they were before motordom. New York City, where only about half of households have a car, has led the way in making streets more democratic and just. Dedicated bus lanes, for example, have been built in cities such as New York, San Francisco and Cleveland. These help reapportion road space to offer the greatest benefit to the largest number of people. Why should crowded, lower-income bus riders get mired in traffic caused by rich people’s one-passenger cars?
Bike infrastructure is another powerful tool for improving equity and reducing congestion. Thanks to dogged activism and some progressive leadership, New York has built 98 miles of protected bike lanes over the last decade. Other cities are following suit.
Some cities are also investing in mass transit. Seattle, where voters approved an additional $54 billion in light rail investment, has seen solo car commuting into downtown decline, even as it added an additional 60,000 jobs.
Ideally, the AV transition would reinforce rather than reverse those healthy patterns. But these are difficult political fights, made even more difficult by the narrative—like the one pushed by the Koch brothers in Nashville—that we can simply sit back and wait for a glittering AV future.
In Los Angeles, a noisy group of local residents have launched fierce resistance to Mayor Eric Garcetti’s series of “road diet” projects to repurpose some of the street width for bikes, buses or sidewalks. The projects are part of the city’s “Vision Zero” efforts, which seek to eliminate traffic deaths by 2025. Naysayers fear the projects will reduce parking or cause congestion. Despite Los Angeles’ high rate of traffic deaths, residents launched a formal effort to recall City Councilmember Mike Bonin because of his support for road diets.
Sometimes resistance is so intense, it’s almost comical. In Minneapolis, residents took to the streets to protest the installation of a bike lane. One of the protest signs decried “Nazi lanes.”
Car companies generally don’t involve themselves in these local fights. They don’t need to—urban car owners zealously guard their parking space. The Koch brothers, however, have a history of astroturf opposition to transit measures, including Nashville’s recent failed ballot initiative, which would have launched several new rail and bus lines.
Conservative and libertarian pushback is likely to be even greater if cities take the next step and introduce the kind of market-based policy changes that have successfully reduced car use in Europe, such as higher gas taxes, less subsidized parking and congestion pricing, which charges tolls to discourage driving during peak hours. Both Seattle and New York are considering congestion pricing.
Robin Chase notes that these measures could provide a key source of income in the automated future. AVs will reduce or eliminate two major sources of municipal revenue: parking fees and traffic fines.
Critics of congestion pricing and high gas taxes often argue that these measures are regressive. That may indeed be true in rural areas, where cars are a necessity, but these measures are typically aimed at highways and dense urban areas like Manhattan, where drivers are disproportionately higher-income professionals. These concerns could be further assuaged if the revenues went to services like buses that benefit lower-income people. Chase believes these tolls and taxes ultimately make driving more fair by making the price better reflect the social costs.
As AVs increase the temptation to rely on cars, Chase thinks more aggressive policies are needed. Without a big deliberate change in direction, she says, “We’re going toward the hell version, because the status quo produces hell.
“The passive resistance is hell.”