Lilia Salazar, 95, has lived in her house in Long Beach, California, since 1953. To Lilia, suburban homeownership represented the achievement of the American dream. In her house on Lees Avenue, she raised five daughters (including my mother-in-law); from it, she commuted by car to the café she owned seven miles away, in Paramount. But since her husband died in 2005, she has lived in the house alone.
Carson Park, the Long Beach neighborhood that Lilia has called home since the Eisenhower years, retains mid-century sensibilities. A Levittown of the West, the neighborhood was built in a postwar tract style at the edge of Los Angeles County. Its ranch houses mostly range between 1,200 and 1,700 square feet, each with a patch of Bermuda grass out front. While Carson Park’s houses may appear modest compared with those in tonier Belmont Heights and Orange County, the neighborhood’s grid has aged gracefully. Houses routinely sell for more than $900,000.
For longtime residents like Lilia, however, Carson Park now represents constraint as much as it once did opportunity. Bounded to the east by the San Gabriel River and sliced through by three high-speed arterials, the neighborhood does not present a friendly environment for a nonagenarian who no longer drives. Long Beach zoning permits no grocery stores in Carson Park. The nearest one to Lilia’s house is a Walmart on the other side of the river, a mile-long walk on Los Coyotes Diagonal and Carson Boulevard, both roads with speed limits of 40 miles per hour. Transit is not much more inviting: to take a bus from Lilia’s house toward the heart of Long Beach would require walking across eight lanes of high-speed traffic to reach an unshaded stop on its opposite side. Hardly pleasant for anyone, the route is especially hazardous for an old woman. For Lilia and her elderly neighbors, participating in Long Beach community life means hitching a ride in a car with someone else.
“I wish that I didn’t have to count on other people to get around so much,” Lilia tells me. Though all her daughters still live in Southern California, none is near enough to visit daily. They stop by on the weekends and pick her up or arrange a ride-hail for her to join family gatherings. And while, happily, she has a younger friend, 75, who takes her out regularly, Lilia would prefer to have more independence outside the home.
The Carson Park neighborhood isn’t changing anytime soon. While the approval of new accessory dwelling units may boost its density, it will remain a residential area where nondrivers are isolated. Technology, however, is changing. Lilia can no longer drive herself, but she and millions of other Americans may soon see their communities reopened to them by the self-driving car—if public policy doesn’t erect roadblocks.
Autonomous vehicles (AVs) work, and work well. First, an extensive roster of instruments collects sensory data. Then, an onboard computing system plots the safest, most efficient path to a desired destination. The sensors that tend to do the heaviest lifting are 360-degree lidar (light-wave detection and ranging), radar (radio-wave detection and ranging), infrared tech, ultrasonic tech, and cameras. Artificial intelligence does the rest, taking the place of a human decision-maker in the driver’s seat.
Companies like Tesla are rolling out software updates that give car owners the chance to use autonomous features in certain settings, but the most exciting developments in self-driving for enhancing elderly mobility are happening on the fleet-owned side of the industry. Soon to be a household name, the Google spinoff Waymo now offers autonomous ride-hailing under the industry designation of Level 4 automation in Phoenix, San Francisco, and Los Angeles, with Austin on deck.
The utopian idea of a personal car that can take you anywhere and anytime at the push of a button may yet come, but Level 4 automation represents a revolutionary innovation in its own right, as I explained in a recent Manhattan Institute report. It means that the vehicle can operate entirely on its own within mapped, well-defined parameters. After years of testing with humans on board as a precaution, such vehicles now run in true autonomous fashion. Waymo has logged millions of driverless miles, tested a new ride-hailing model, and demonstrated the viability of the technology that enables it.
Government approval has come in fits and starts, and Americans largely remain wary. As Manhattan Institute senior fellow James B. Meigs has argued, the biggest remaining hurdle for AVs to clear may be psychological, not technical. According to a Pew survey, barely one in four Americans thinks that widespread AV use would be a good idea. The American Automobile Association reached similar results in a 2023 poll, finding that 68 percent of Americans fear AVs. AAA’s annual sentiment tracking shows that this fear is rising.
Such concern is natural, for technological change can be dangerous. But autonomous ride-hailing vehicles are proving to be a safe addition to the transportation mix. As technology journalist Timothy B. Lee has meticulously documented, the AVs operating commercially are already causing harm at lower rates than would be expected from human drivers. Lee’s analysis concludes that vehicles operated by Waymo and Cruise (GM’s autonomous-vehicle unit) were involved in 102 collisions in San Francisco over the course of 6 million driverless miles on the road through August 2023. The overwhelming majority of these collisions, the records of which Lee gathered from a California state database, occurred at low speeds; most were the fault of the other vehicle’s (human) driver.
Self-reporting indicates that Waymo and Cruise are involved in fewer collisions than human-driven ride-hails in similar conditions. Estimates generated by Cruise in conjunction with researchers at the University of Michigan and Virginia Tech concluded that its cars are 50 percent less likely to be involved in collisions and 70 percent less likely to get into injury-causing collisions in like circumstances. Swiss Re, an insurer, has now reached a similar positive conclusion regarding Waymo. “[I]n over 3.8 million miles driven without a human being behind the steering wheel,” Swiss Re announced in October, “[Waymo vehicles] incurred zero bodily injury claims in comparison with the human driver baseline of 1.11 claims per million miles . . . [and] significantly reduced property damage claims to 0.78 claims per million miles in comparison with the human driver baseline of 3.26 claims per million miles.”
In recent months, though, high-profile tech mishaps and poor human leadership have cost Cruise heavily and have embarrassed the wider industry. In August, Cruise vehicles navigating San Francisco’s North Beach nightlife district decided that the scene was too chaotic and froze in place, causing a traffic jam. Days later, a Cruise car plunged into freshly poured cement. Worst of all, however, was a terrible October incident in which a Cruise vehicle rolled over a woman and dragged her 20 feet. The incident was catalyzed by a human driver striking the woman, propelling her into the Cruise’s path, but the Cruise vehicle’s subsequent actions are concerning, nevertheless. Compounding matters, Cruise withheld portions of video evidence, according to the California Department of Motor Vehicles. In response, the agency suspended Cruise’s driverless permit, and the company has since put its operations on hold nationwide.
Also damaging for the reputation of autonomous vehicles have been fatal collisions involving Tesla’s “Autopilot” and “Full Self-Driving” features, the lens through which many Americans are likely to view AVs. The obvious problem is that while Tesla’s marketing materials imply that a human driver can kick back and relax, the fine print says otherwise, requiring a person to spring into action if things go awry. Despite its marketing monikers, Tesla’s features are not considered to meet industry-standard Level 4 automation that removes responsibility from the human on board. The ambiguity invites risk: a Washington Post analysis of data from the National Highway Traffic Safety Administration found that Tesla’s Autopilot has been a factor in more than 700 collisions since 2019, killing 17 people. In December, declaring that Tesla’s guardrails “may not be sufficient to prevent driver misuse” of its technology, the National Highway Traffic Safety Administration pressured the company to update the software on 2 million vehicles—a process that Tesla has completed over-the-air, rather than requiring visits to service sites. Neither Waymo nor Cruise has caused any fatalities.
Despite these setbacks, the market is responding to the prospect of autonomous ride-hailing in American cities. Waymo reports a waiting list of 100,000 San Franciscans eager to hop in to its AVs.
This dynamic of early adopters in major cities prefiguring mass pickup echoes historical patterns. Machines that we find perfectly normal today were once objects of uncertainty. “Humans often converge around massive technological shifts—around any change, really—with a flurry of anxieties,” Adrienne LaFrance wrote in 2015 for The Atlantic. That article reflected on the brief period—from the early 1980s through the mid-1990s—when a wave of “computerphobia” swept the country. Needless to say, the wave passed.
As with other technologies, older people tend to be less receptive to autonomous driving. A 2021 paper published in Geriatrics found that 21 percent of survey respondents aged 65 and older were receptive to the idea of hitching rides in AVs. Researchers from Georgia Tech reported similarly in a 2022 paper that 19 percent of people 65 and older are interested in riding in driverless ride-hails, compared with 25 percent of 55- to 64-year-olds, 36 percent of 45- to 54-year-olds, 46 percent of 35- to 44-year-olds, and 52 percent of 18- to 34-year-olds.
Ride-hailing of the human variety shows the same pattern. Today’s elderly are the least likely to use services like Uber. In a 2020 study, researchers at San Jose State University examined the effect of age on ride-hailing and found that Californians 55–64 years old were 14 percent more likely to have used ride-hailing than Californians 75 and older. Of the 55–64-year-olds, more than half reported having experience with a ride-hailing app, and more than a third reported having booked rides themselves.
But while members of the Silent Generation may not be eager to hail rides, the technology could be tremendously empowering. And for older baby boomers more accustomed to ride-sharing services, it could one day herald a revolution in getting around. The number of Americans 85 and older is expected to pass 12 million by 2035 and reach 19 million by 2060. As Americans living in suburban locales age, they will be farther from amenities and have fewer children to drive them around town than their parents have today. The social conundrum, then, is that millions of Americans living in areas that demand driving will soon age out of the safe range to do so.
For older Americans, suburban neighborhoods like Carson Park have always been home. Coming of age in the 1970s, Lilia’s youngest daughter, Margarita, saw the personal car as her ticket to independence, fun, and status. True to generational norms, the car still holds a central place in her daily life in suburban San Diego. It could do so for decades to come.
The great challenge for the companies offering AV ride-hailing is to show the merit of the technology before regulation arrests its development. In California, various government agencies are working at cross-purposes. In August, the California Public Utilities Commission granted Waymo and Cruise permission to operate in San Francisco around the clock. Yet, even prior to the October incident, the Department of Motor Vehicles halved the number of cars that Cruise can have on the road from 400 to 200 just days later. Meantime, in an adjacent industry, trucking, the California labor movement worked with legislators on a bill—vetoed by the governor, fortunately—that would all but have excluded the state from autonomous freight adoption. Though California has the advantage of its resident tech industry, Arizona and Texas have better regulatory environments for companies operating in the space. Texas, in particular, stands out for its Connected and Autonomous Vehicles Task Force, an agency that serves as a statewide clearinghouse of sorts for AV issues. In freight trucking, Texas is also ahead of the game. The autonomous freight company Aurora will begin running trucks back and forth between its terminals outside Dallas and Houston this year.
Skeptical regulators have tended to ignore the safety benefits of AVs in an era of demographic change. According to data reported by the states and aggregated by the Federal Highway Administration, the share of Americans aged 70 and older with driver’s licenses rose from 73 percent in 1997 to 87 percent in 2021. While advanced age alone does not disqualify anyone from operating a motor vehicle, it comes with elevated risk, deriving from weakening vision, slowing reaction times, and general health decline. Statistics on the roadway safety of older and elderly drivers show that drivers aged 65–69 have the lowest rates of property-damage liability claims and collision claims per insured vehicle annually; but beyond age 70, claim rates rise, reflecting higher collision rates.
The Catch-22 is that mobility, enabled by car for those not living in dense cities, is often crucial to well-being. Remaining home can worsen an elderly person’s health. Research on aging and mobility indicates that driving cessation is associated not only with loneliness and depression but also increased mortality risk. Hard though it is to unspool the feedback loops involving physical limitation, driving risk, driving cessation, cognitive decline, and health outcomes, expanding access to getting around is imperative.
If America’s changing demography presents a compelling argument for promoting ride-hailing, the economics of serving older customers will ensure a solid market. Ride-hailing companies like Uber and Lyft are restricted by labor costs, but the business model that Waymo and other firms are pioneering will take ownership of vehicles and avoid the cost of paying drivers. Drivers appear to hoover up the majority of Uber fares, with the Economic Policy Institute deducing in 2018 that their share amounts to about two-thirds. A year earlier, UBS put the portion going to drivers at 70 percent or higher. According to 2023 fare data self-reporting from Uber, drivers have gotten 71 percent to 76 percent of pay in the most recent two years.
Despite the enormous R&D investments needed to jumpstart autonomous driving, then, the AV industry has an opportunity to undercut Uber and Lyft in short order. In 2022, McKinsey projected that by the end of this decade, the autonomous ride-hailing model would drop fares 60 percent below existing rates. While today, riding in an Uber costs about 200 percent more per mile than personal car ownership, McKinsey anticipates that by 2030 an autonomous ride-hail will cost just 20 percent more than personal car ownership. Riding in autonomous shuttles for several people, which companies are already rolling out, will cost 40 percent less than driving one’s own car. Bypassing driver pay will enable autonomous ride-hailing to outcompete existing offerings on price, while delivering a vital service in the process. Perhaps unsurprisingly, Uber is now partnering with Waymo in Phoenix, offering Waymo’s AV rides to its larger customer base.
Our modern autocentric status quo comes with some clear benefits, but elderly people bear many unseen costs. As the nation ages, this problem will become more acute: a growing number of people will be stuck between driving at elevated risk levels or languishing in isolation. Autonomous ride-hailing vehicles have thus emerged just in time. They can ameliorate this problem and keep people connected to their communities in their golden years. As Americans get older in the suburbs, expanding the frontier of autonomous ride-hailing beyond the confines of major-city municipal boundaries is a worthy policy objective. Lilia Salazar and the Silent Generation may get only a small bite of the AV apple—but generations to come will get much more.
Top Photo: Despite public doubts, research suggests that Waymo and Cruise vehicles are involved in fewer collisions than human-driven ride-hails in similar conditions. (Allen J. Schaben/Los Angeles Times/Getty Images)
