Updated at 2:00 p.m. ET on October 10, 2019.
When my father was 18 years old, he fell asleep at the wheel while driving home late. That’s never good, but it was particularly bad then, in 1954, the year he crashed his car on the early-morning Milwaukee streets. Seat belts weren’t common until the 1960s, and federal law didn’t mandate them until 1968. Dad’s unharnessed body was thrown out the windshield, through the quiet night, and onto the pavement. He survived—otherwise I wouldn’t be writing this—but not without permanent disability.
The road to mandatory seat belts was a long one, involving decades of medical and military research, legislative intervention, and corporate acquiescence. But today, according to the National Highway Traffic Safety Administration (NHTSA), 90 percent of Americans use seat belts, which, the agency claims, save some 15,000 lives a year. They are almost automatic for most drivers and passengers.
Except in one case. Even people who wear seat belts religiously tend not to do so in taxis. New York City cab passengers over age 16 are exempt from seat-belt laws (for now), and a 2014 NYC Taxi and Limousine Commission survey of such riders reported that only 38 percent of them wear seat belts. Why? In a 2008 study of 100,000 car trips, two psychologists concluded that riders might assume a lower risk of calamity in cars for hire: The trips are shorter, the drivers are seen as professionals with lots of road experience, and the passengers (wrongly) perceive a lower risk of harm in the back seat.
For years, this particular threat to public safety didn’t affect people who didn’t take taxis. But now, thanks to Uber and Lyft, trips by hired car have increased dramatically all over the country. Americans took 3.2 billion ride-share trips in 2018, and they were almost as likely to forgo a seat belt on those trips as New York City taxi patrons. One survey found that 43 percent of car-hire customers reported not always wearing a seat belt, and 80 percent don’t buckle up for short trips.
Sixty-five years later, my oldest kids are now about the age my father was when he was hurled from a vehicle, absent a seat belt, to permanent effect. Because of Uber and Lyft, they are far more likely than I was to risk a similar fate.
Once you start looking, examples like this appear everywhere. Technology services are systematically, if invisibly, eroding long-won victories in public safety.
Take vaping. Much like Uber transformed the cab ride with smartphones, e-cigarettes have reinvented cigarette culture with a discreet new technology: a battery-powered stick that heats and aerosolizes nicotine (or sometimes cannabis). They were invented in the early 2000s by Hon Lik, a 52-year-old Chinese pharmacist (and smoker), who wanted to find a way to scratch the same itch that a cigarette could, but without all the harmful chemicals—smoking without the smoke. Makers marketed the products as therapeutic devices, beneficial to smokers hoping to quit. But health officials expressed concern about that claim from the start. In 2008, the World Health Organization declared that it “does not consider [vaping] to be a legitimate therapy for smokers trying to quit.” Ruyan Holdings, the Chinese company that brought Lik’s invention to market, responded by funding a report arguing that e-cigs are substantially safer than traditional cigarettes, partly because “each puff contains one third to one half the nicotine in a tobacco cigarette’s puff.”
But as my colleague Amanda Mull wrote, “To market a product as less harmful than cigarettes is to damn it with faint praise.” Juul Labs, the San Francisco-based company that makes the most popular brand of e-cigarette in the U.S., sells pods at 3 and 5 percent nicotine, which is stronger than earlier vaping products. A single Juul pod delivers about 200 puffs, making each pod roughly equivalent to a pack of cigarettes in nicotine content. As Juuling became synonymous with vaping, the company’s stronger, more capacious nicotine products became the norm.
Meanwhile, e-cigarettes still deliver particulate matter and toxins into the lungs (just not as much as cigarettes do). Secondhand aerosol poses some risk to people in the proximity of e-cigarettes, which are often used indoors because the odorless vapor doesn’t seem harmful, like tobacco smoke.
Juul has made vapers of all ages, but it is especially popular among teens—who largely appear to have no idea that their mango-flavored e-cigarettes even contain nicotine at all, and who are already more susceptible to the drug’s effects. In 2015, 74 percent fewer American high-school students had recently smoked than in 1997. But 60 percent more had used electronic cigarettes than traditional ones. And on top of it, adolescents who use e-cigarettes are much more likely to start smoking traditional cigarettes, too. Teen tobacco use has fallen as the regulation and stigmatization of cigarettes have risen. But vaping reversed that—a pattern as old as smoking itself—providing a new path toward nicotine addiction among young people.
Vaping’s safe reputation, never really earned, is now quickly deteriorating. Recently the Food and Drug Administration issued a warning to Juul about the company’s marketing of vaping products as healthier than cigarettes. A respiratory “vaping illness” has struck more than 1,000 Americans in 48 states, most with otherwise healthy lungs; by early October, it had resulted in at least 18 deaths. A contaminant (perhaps hydrogen cyanide) in marijuana vaping products may be the cause. But even before the scare issued a reminder of how little is known about the health consequences of vaping, e-cigarettes had already set in motion social changes that upended decades of smoking-cessation advances.
The technology sector’s global conquest rests partly on its total unconcern for the real-world impact of its products and services. Tesla beta-tests its “autopilot” semiautonomous mode on public roads because it can get away with doing so. Google Maps and Waze optimize for the fastest trip, no matter what effect those routes might have on traffic or safety. Amazon has evaded responsibility for damages, injuries, and deaths caused by its delivery network by presenting it as arm’s-length contractors, even as the retail giant imposes pressure to increase the speed and volume of distribution. Technology firms cover their exposure to risk the same way they create it: through willful ignorance.
Public safety protects the public—a collectivity worth shielding. It is a shared effort that begets a shared benefit. Some technological innovations, such as automobiles with lane-departure warnings or automatic emergency braking, help owners of the equipment along with those who surround them. But tech in the sense of Silicon Valley Big Tech has subtly rewritten our conception of the public. The industry has undoubtedly improved people’s individual, private lives—that’s the business model. But it has not necessarily benefited their communal ones. The vision of the future that firms such as Uber, Amazon, and Facebook have grown rich selling is a decidedly individualist one: Get a ride just for you, wherever you are, via Uber. Receive almost any product tomorrow, without leaving home, from Amazon. Hear from only the people and groups you choose on Facebook. Technology products can improve health and safety, but largely at the personal level: carrying a cellphone for emergencies, or wearing a fitness tracker to motivate regular exercise. Solving one’s own problems can de-escalate interest in solving communal ones. This libertarian individualism also grips the big-tech companies themselves, which pursue their private aspirations no matter the public cost.
You and me, your kids, your neighbors, your co-workers—all the members of the general public—are caught in the middle. On the one hand, we like and desire the tech industry’s offerings, personally. But on the other hand, we can’t always see the consequences of those offerings at a social scale. When those effects do become visible, we look for answers that eradicate all the bad stuff about tech without changing the good stuff.
It’s never so simple. For example, ride-share drivers have been accused of attacking, kidnapping, and sexually assaulting passengers; some victims have alleged that the companies knew about these assaults, yet failed to stop them. But people are also violently attacked in traditional taxis and car hires; limited data on all sides have made it difficult to determine whether ride-shares really are more dangerous than taxi and livery drivers.
More certain is that exposure to such risk has risen substantially as a result of ride-sharing. More drivers convey more passengers more often. Over the past decade, people have completed billions of new rides in new contexts—trips that never would have happened at all before Uber and Lyft came on the scene.
Whether they mean to or not, companies such as these are changing the standards of public safety. Those effects come, in part, from globalizing and automating previously local, human processes. But they also arise when innovation imports the undesirable features of earlier, neighboring technologies and then intensifies them at scale.
Here’s another one: New York City medallion taxis are exempt from state laws that otherwise require children under the age of 4 to use child-safety seats. Passengers can bring their own child seat, and taxi drivers must allow riders to install and use one if they do so. But instead, most folks rationalize the risk in much the same way they do with seat belts: It’s a short trip, and What are you gonna do, really?
Ride-sharing apps have inherited this quandary and exported it to cities everywhere. Uber and Lyft do let riders request a car with a child-safety seat—but only in a few cities, and for an extra fee (and probably a longer wait). Elsewhere, both companies encourage riders to bring their own as required by local law, but neither one appears to have a policy that requires drivers to accept such fares. And even if they did, bringing a car seat to a concert or ball game undoes the convenience of ride-sharing in the first place.
Car seats are the life-saving result of decades of incremental progress on child safety. Some automakers offered them in the 1960s, and the NHTSA first adopted standards in 1971. Over time, the requirements became stricter and more complex (rear-facing for infants, boosters for preschoolers, tethered anchors, the LATCH installation standard). It’s not uncommon for states to require specialized equipment for children up to age 8.
Motor-vehicle crashes are the leading cause of death among children in the United States. Passengers used to take the risk in taxicabs only under duress. But ride-sharing has given that exception a shortcut to convention, by exporting the false sense of comfort from taxi trips to rides of all kinds.
Uber and Lyft allowed drivers to operate, with little oversight, as they gunned for global growth at all costs. Like other “sharing economy” companies, they exploited lax regulation and ambiguities and gaps in local laws to gain advantage. When California recently passed a bill demanding that companies give gig workers more benefits, Uber responded that it does not intend to reclassify its drivers as employees, because “drivers’ work is outside the usual course of Uber’s business.” The company has always claimed to be “just” a technology business, providing a set of software tools that form a marketplace on which drivers and passengers can connect. Even safety becomes a worry that a suite of technological solutions can remedy. This detached approach breeds indifference; the consequences become someone else’s problem.
Even with your seat belt safely fastened in your own car, with your kid properly harnessed in the back seat, technology has made driving more dangerous, because everyone is distracted by smartphones all the time.
Drivers have gotten sidetracked at the wheel for as long as cars have been on the road. In the early 20th century, the roads themselves were sometimes the cause, lulling drivers to diversion in a condition that came to be known as “highway hypnotism.” Later, as cars added more devices, such as radios, CD players, and navigation systems, the car itself started to sidetrack drivers.
But equipment is just part of the picture. The NHTSA tracks three kinds of distraction: visual, manual, and cognitive. Visual diversions take your eyes off the road. That could mean kids arguing in the back seat, or scintillating neon signs along the route. Manual distractions take your hands off the wheel. Radios and controls do that, as do takeout cups and Eggs McMuffin. Cognitive distractions take your mind off driving. Those are innumerable, covering everything from struggles at home to anger at traffic to worries about the workday coming or gone.
Smartphone use is particularly insidious for drivers because it fuses all three kinds of distraction: You look away from the road, take your hands off the wheel, and shift your thoughts to the world behind the screen. Moreover, smartphones distract drivers frequently, perhaps even constantly, as vehicle operators might pick up the phone at every stoplight. The NHTSA calls it “among the worst of all driver distractions.”
The effects can be substantial. From 2012 to 2015, an average of 283,000 injuries a year could be traced to texting and smartphones. More recently, a combination of hands-free laws and distracted-driving awareness has reduced driver use of cellphones, but the handheld manipulation of these devices has fallen less rapidly. By 2017, car crashes had risen by 11 percent, compared with just a few years earlier, and distraction-caused collisions involving cellphones remained relatively steady.
Smartphones didn’t cause these problems on their own; widespread reliance on them did. Adults, especially younger ones, feel pressure to respond to work messages while driving. Social-media services can inspire people to catch up on their feed, post photos, or bang out quips on the road.
And a 2018 law requiring new cars to include backup cameras to prevent injury when moving in reverse might have made things worse. Installing cameras meant adding a big screen to the dashboard, a change that happened to arrive alongside the proliferation of smartphone touch-screen interfaces. Many automakers redesigned the whole dashboard to match them. Now it’s common for a single touch-screen display to operate the radio, the climate control, the navigation, and, of course, the backup camera. They are known as “in-vehicle information systems,” or IVIS, in auto-industry lingo.
As you might expect, IVIS screens only distract drivers more. Now they can’t operate the radio preset or the blower control by feel, without looking. The options are more complex, too: satellite radio, playback controls, map panning. It’s almost like using a smartphone.
Then Apple CarPlay and Android Auto arrived. Their familiar appearance does seem to reduce distraction compared with native IVIS equipment, but according to an American Automobile Association study, the software still produces substantial cognitive and visual diversion. Responding to texts (even by dictation), scrolling Spotify playlists, and managing the quirks of these systems still generate increased demands on driver attention compared with driving without them.
In the same urban spaces where more cars are creating more traffic, with more passengers skipping seat belts and diverting their attention to IVIS and CarPlay screens while they inhale aerosolized tobacco, people are also using new kinds of transportation devices to get around. Electric scooter shares, like Bird and Lime, are among the “micromobility” services that proponents hope will help reduce city dwellers’ reliance on cars, especially for short trips.
Bird, Lime, and other services followed Uber’s model, expanding rapidly to new cities to grow their market and stimulate consumer demand—often without permits. In the cities where they took off, scooters have become staples of urban transport; they allow professionals to get to work without rumpling clothes or breaking a sweat. “I cannot imagine life without them,” my colleague Robinson Meyer wrote of the scooters last year.
But according to a study published earlier this year, over 12 months two California emergency rooms admitted 249 patients with injuries from scooter mishaps. Most common were head injuries—largely caused by riders’ failure to wear helmets—although fractures were also common. A Centers for Disease Control and Prevention (CDC) study conducted in Austin, Texas, counted 20 injuries presented to medical-emergency services per 100,000 e-scooter trips over a three-month period. Almost half were head injuries, and 15 percent were traumatic brain injuries. Of 190 injured riders, only one had worn a helmet.
Motorbike helmets were commercialized in the 1950s, and the U.S. Department of Transportation issued standards for manufacturers in 1964. In 1975, models for cycling appeared in earnest, and adoption increased in the decade that followed. Starting in 1987, some states began enacting laws requiring helmets, especially for children.
Over time, a culture of helmet use became entrenched among casual and commuter cyclists alike. In 2012, 29 percent of bike-riding adults and 42 percent of children reported always wearing helmets; half said they wore helmets at least some of the time. Ninety percent of kids wore helmets when adults in their orbit also did; they were also more likely to do so in states with laws mandating helmet use.
The problem is, helmets ruin the appeal of shared e-scooters almost completely. As Michael Keating, the CEO of the micromobility company Scoot, told me last year, scooters such as Birds are most useful for making trips that would require a long walk easier without a car. Donning a helmet means carrying one, which could erase the benefits in convenience.
That’s the same problem bicycles have faced. Detractors of bicycle-helmet laws often complain that they discourage ridership or apologize for the ongoing reign of the car. They have a point. It’s easy to wag a finger at riders who forgo helmets, and serious e-scooter injuries are relatively infrequent—about 0.2 percent of trips, according to the Austin data. But there, the CDC found that only 10 percent of e-scooter riders sustained their injuries from vehicles, making auto-centrism an incomplete counterpoint to concerns about scooter safety.
Helmet critics want to juggle a series of risks and rewards at urban scale. And the meaning of public safety should evolve, as the writer Sam Bloch has put it. But that’s not what’s happening when uninformed and inexperienced e-scooter riders injure themselves straightaway. If the vehicles’ popularity continues to rise, that trend will only continue: The CDC studies suggest that peer influence—riding with other people and seeing people ride safely—has the greatest impact on cyclists’ choices.
In the end, it might be worthwhile to make certain safety compromises to spur new urban design and mobility, not to mention carbon use. But that’s not how the scooter rollout has happened. Instead, it has shifted public safety absent the public’s understanding that it is happening at all.
There’s an irony in the technological erosion of general welfare, because innovations so often promise to benefit our lives—sometimes by improving public safety explicitly. It’s one of the promises of autonomous vehicles, for example: that they might reduce traffic injuries and fatalities like those caused by distracted drivers or motorists who fail to yield to scooters. That promise is speculative and possibly wrong, but its commercial value has inspired some governments to make substantial compromises in public safety in the present in order to chase improvements in the future. Arizona Governor Doug Ducey, for example, issued executive orders allowing experimental use of driverless cars on Arizona roads, hoping to spur investment and job creation in the desert state.
When an Uber operating in autonomous mode struck and killed a pedestrian in Tempe, critics blamed lax regulation, in part, for allowing Uber to take unnecessary risks with its equipment. But as the scooter advocates had done, some robocar proponents scoffed, seeing selective concern about the fatal consequences of traffic in America. With more than 40,000 automobile-related fatalities in the United States each year, why single out one autonomous-vehicle death for special attention?
The total number of deaths from shirking seat belts or car seats in ride-shares, or even from the unknown pathogens afflicting vaping devices, is small compared with those from other trends. Griping about Uber or Juul can seem like missing the forest for the trees—or come across as knee-jerk traditionalism that unfairly demonizes youthful trends. Imagine being worried about a dozen vaping deaths or e-scooter injuries, some complain, instead of almost 40,000 deaths each year from guns in America, or 480,000 from smoking.
Responding to that criticism about autonomous-car deaths specifically last year, I argued that focusing on the quantity of deaths or injuries, via thought experiments such as the trolley problem, misses the point. A utilitarian obsession with outcomes overlooks the flaws in the social, political, and economic processes that aspire to realize those outcomes. Even when the numbers affected are small, the way the problems arise can be significant. Likewise, victories can also be ambiguous. Ridesharing might have reduced drunk driving in some cities, but not others, for example.
Eroding widespread successes—such as buckling seat belts or shunning nicotine—should worry you because these changes represent reversals of public-health victories so complete that they had been forgotten. Juul marketed flavor-masked e-cigarettes to young people, who had been less likely to use nicotine products year after year. The convenience of e-scooters has inspired riders to embrace the risk of traumatic injuries that they’d be unlikely to take in other circumstances. Likewise seat belts or child restraints in the back of car shares, or smartphones at the wheel.
If Americans could choose only a single path to mutual benefit, then maybe it would make sense to narrow attention to smoking or gun deaths alone. But the world isn’t monolithic. Earlier adoption of smoking cessation wouldn’t have helped my dad avoid being hurled through his windshield. Gun control won’t improve the regulatory oversight of autonomous cars. Just because people die from lung cancer or firearms discharge doesn’t obviate the need for traffic planning or building-construction codes or occupational-safety practices. When it comes to public safety, citizens shouldn’t want redress of just one or two risks with the greatest real or perceived impact. They should demand widespread improvements in as many aspects of public life as possible.
When those victories start to walk themselves back, they introduce new exposures that might no longer be understood in modern contexts. And once the changes become commonplace, it’s difficult to quell them back to rarity. Instead, when a new technology is still truly new, those who will live in its future wake must look for and mitigate its future risks more deliberately. Past improvements in public safety required regulatory intervention and social transformation; the present circumstances likely will as well. Where possible, it is necessary to minimize technology’s bad effects while preserving its good ones—which is much harder after widespread adoption. But in addition, both private and public agents should learn from the frequent and obvious lessons of the past, and the present, as new technologies appear.
The conditions at the fringes of a social practice get amplified as it becomes more predominant. Seat-belt and child-seat use is almost universal—except in taxis, and ride-sharing has proliferated that exception everywhere. Cigarettes are mostly a memory, unless a company finds a new way of delivering nicotine that evades their decline, as Juul did. Only important executives or government officials would find it necessary to text or email at a stoplight, until smartphones and apps made everyone’s work and social lives seem urgent. Most of these consequences were foreseeable, but nobody was looking for them. Private benefit took the wheel, and public protection jumped in the back seat, untethered, and ready to roll.